{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.5, "eval_steps": 500, "global_step": 40000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.609763741493225, "learning_rate": 2.25e-07, "loss": 1.3278, "step": 10 }, { "grad_norm": 1.8155934810638428, "learning_rate": 4.75e-07, "loss": 1.3246, "step": 20 }, { "grad_norm": 1.6316994428634644, "learning_rate": 7.25e-07, "loss": 1.3257, "step": 30 }, { "grad_norm": 1.2802839279174805, "learning_rate": 9.75e-07, "loss": 1.3205, "step": 40 }, { "grad_norm": 1.161996841430664, "learning_rate": 1.2250000000000001e-06, "loss": 1.3011, "step": 50 }, { "grad_norm": 0.9059290885925293, "learning_rate": 1.475e-06, "loss": 1.2693, "step": 60 }, { "grad_norm": 0.5230932831764221, "learning_rate": 1.7250000000000002e-06, "loss": 1.2504, "step": 70 }, { "grad_norm": 0.508309543132782, "learning_rate": 1.975e-06, "loss": 1.2321, "step": 80 }, { "grad_norm": 0.3696719706058502, "learning_rate": 2.225e-06, "loss": 1.2228, "step": 90 }, { "grad_norm": 0.4355830252170563, "learning_rate": 2.4750000000000004e-06, "loss": 1.2341, "step": 100 }, { "grad_norm": 0.5072057843208313, "learning_rate": 2.725e-06, "loss": 1.2302, "step": 110 }, { "grad_norm": 0.4295227527618408, "learning_rate": 2.975e-06, "loss": 1.22, "step": 120 }, { "grad_norm": 0.5316113233566284, "learning_rate": 3.225e-06, "loss": 1.2066, "step": 130 }, { "grad_norm": 0.5699491500854492, "learning_rate": 3.4750000000000006e-06, "loss": 1.1932, "step": 140 }, { "grad_norm": 0.566224992275238, "learning_rate": 3.725e-06, "loss": 1.1591, "step": 150 }, { "grad_norm": 0.5540301203727722, "learning_rate": 3.975e-06, "loss": 1.1348, "step": 160 }, { "grad_norm": 0.6491700410842896, "learning_rate": 4.225e-06, "loss": 1.1199, "step": 170 }, { "grad_norm": 0.575451672077179, "learning_rate": 4.475e-06, "loss": 1.1075, "step": 180 }, { "grad_norm": 0.5091739892959595, "learning_rate": 4.7250000000000005e-06, "loss": 1.0908, "step": 190 }, { "grad_norm": 0.6332139372825623, "learning_rate": 4.975000000000001e-06, "loss": 1.0752, "step": 200 }, { "grad_norm": 0.5818896293640137, "learning_rate": 5.225e-06, "loss": 1.0673, "step": 210 }, { "grad_norm": 0.6217289566993713, "learning_rate": 5.475e-06, "loss": 1.0631, "step": 220 }, { "grad_norm": 0.7273525595664978, "learning_rate": 5.725e-06, "loss": 1.0495, "step": 230 }, { "grad_norm": 0.676984429359436, "learning_rate": 5.975e-06, "loss": 1.0425, "step": 240 }, { "grad_norm": 0.8840965628623962, "learning_rate": 6.2250000000000005e-06, "loss": 1.0455, "step": 250 }, { "grad_norm": 0.744970440864563, "learning_rate": 6.475000000000001e-06, "loss": 1.049, "step": 260 }, { "grad_norm": 0.5394951105117798, "learning_rate": 6.725000000000001e-06, "loss": 1.0428, "step": 270 }, { "grad_norm": 0.7758359909057617, "learning_rate": 6.975000000000001e-06, "loss": 1.044, "step": 280 }, { "grad_norm": 0.5587018728256226, "learning_rate": 7.2249999999999994e-06, "loss": 1.0333, "step": 290 }, { "grad_norm": 0.6326748132705688, "learning_rate": 7.4750000000000004e-06, "loss": 1.0309, "step": 300 }, { "grad_norm": 0.5205458998680115, "learning_rate": 7.725e-06, "loss": 1.0339, "step": 310 }, { "grad_norm": 0.8577150702476501, "learning_rate": 7.975e-06, "loss": 1.0298, "step": 320 }, { "grad_norm": 0.4417423605918884, "learning_rate": 8.225e-06, "loss": 1.0214, "step": 330 }, { "grad_norm": 0.6328150629997253, "learning_rate": 8.475000000000001e-06, "loss": 1.0114, "step": 340 }, { "grad_norm": 1.269781231880188, "learning_rate": 8.725e-06, "loss": 1.0071, "step": 350 }, { "grad_norm": 0.6686986088752747, "learning_rate": 8.975e-06, "loss": 0.9842, "step": 360 }, { "grad_norm": 0.9036184549331665, "learning_rate": 9.225e-06, "loss": 0.9658, "step": 370 }, { "grad_norm": 0.8397446870803833, "learning_rate": 9.475e-06, "loss": 0.9622, "step": 380 }, { "grad_norm": 0.7943445444107056, "learning_rate": 9.725000000000001e-06, "loss": 0.9473, "step": 390 }, { "grad_norm": 1.1855570077896118, "learning_rate": 9.975e-06, "loss": 0.9278, "step": 400 }, { "grad_norm": 1.4037675857543945, "learning_rate": 1.0225e-05, "loss": 0.8986, "step": 410 }, { "grad_norm": 1.5672457218170166, "learning_rate": 1.0475e-05, "loss": 0.8606, "step": 420 }, { "grad_norm": 1.395384669303894, "learning_rate": 1.0725e-05, "loss": 0.8412, "step": 430 }, { "grad_norm": 1.5197160243988037, "learning_rate": 1.0975e-05, "loss": 0.8096, "step": 440 }, { "grad_norm": 1.6633671522140503, "learning_rate": 1.1225e-05, "loss": 0.7844, "step": 450 }, { "grad_norm": 1.8633966445922852, "learning_rate": 1.1475000000000001e-05, "loss": 0.7505, "step": 460 }, { "grad_norm": 1.7362576723098755, "learning_rate": 1.1725e-05, "loss": 0.7229, "step": 470 }, { "grad_norm": 1.9630100727081299, "learning_rate": 1.1975e-05, "loss": 0.6907, "step": 480 }, { "grad_norm": 1.9028385877609253, "learning_rate": 1.2225e-05, "loss": 0.6546, "step": 490 }, { "grad_norm": 1.4572972059249878, "learning_rate": 1.2475e-05, "loss": 0.638, "step": 500 }, { "grad_norm": 1.7094491720199585, "learning_rate": 1.2725000000000001e-05, "loss": 0.602, "step": 510 }, { "grad_norm": 2.0004892349243164, "learning_rate": 1.2975e-05, "loss": 0.5794, "step": 520 }, { "grad_norm": 2.27589750289917, "learning_rate": 1.3225000000000001e-05, "loss": 0.5495, "step": 530 }, { "grad_norm": 1.547186255455017, "learning_rate": 1.3475000000000002e-05, "loss": 0.5267, "step": 540 }, { "grad_norm": 1.7012335062026978, "learning_rate": 1.3725000000000002e-05, "loss": 0.4957, "step": 550 }, { "grad_norm": 1.6557270288467407, "learning_rate": 1.3975000000000003e-05, "loss": 0.4732, "step": 560 }, { "grad_norm": 2.296168804168701, "learning_rate": 1.4225e-05, "loss": 0.4433, "step": 570 }, { "grad_norm": 2.4349303245544434, "learning_rate": 1.4475e-05, "loss": 0.4235, "step": 580 }, { "grad_norm": 2.012977123260498, "learning_rate": 1.4725e-05, "loss": 0.3829, "step": 590 }, { "grad_norm": 1.7090418338775635, "learning_rate": 1.4975e-05, "loss": 0.3705, "step": 600 }, { "grad_norm": 1.9706214666366577, "learning_rate": 1.5225e-05, "loss": 0.3423, "step": 610 }, { "grad_norm": 2.054928779602051, "learning_rate": 1.5475e-05, "loss": 0.333, "step": 620 }, { "grad_norm": 2.8221399784088135, "learning_rate": 1.5725e-05, "loss": 0.3013, "step": 630 }, { "grad_norm": 2.3137879371643066, "learning_rate": 1.5975000000000002e-05, "loss": 0.2767, "step": 640 }, { "grad_norm": 3.165996789932251, "learning_rate": 1.6225e-05, "loss": 0.2711, "step": 650 }, { "grad_norm": 1.8318530321121216, "learning_rate": 1.6475e-05, "loss": 0.2514, "step": 660 }, { "grad_norm": 1.818946123123169, "learning_rate": 1.6725000000000003e-05, "loss": 0.2228, "step": 670 }, { "grad_norm": 2.0168988704681396, "learning_rate": 1.6975000000000003e-05, "loss": 0.2225, "step": 680 }, { "grad_norm": 2.2375540733337402, "learning_rate": 1.7225e-05, "loss": 0.2187, "step": 690 }, { "grad_norm": 1.8472027778625488, "learning_rate": 1.7475e-05, "loss": 0.2037, "step": 700 }, { "grad_norm": 2.772651433944702, "learning_rate": 1.7725e-05, "loss": 0.208, "step": 710 }, { "grad_norm": 1.7465503215789795, "learning_rate": 1.7975e-05, "loss": 0.1779, "step": 720 }, { "grad_norm": 2.0525588989257812, "learning_rate": 1.8225e-05, "loss": 0.1522, "step": 730 }, { "grad_norm": 2.1394004821777344, "learning_rate": 1.8475000000000002e-05, "loss": 0.1616, "step": 740 }, { "grad_norm": 1.8793686628341675, "learning_rate": 1.8725e-05, "loss": 0.157, "step": 750 }, { "grad_norm": 1.463911771774292, "learning_rate": 1.8975e-05, "loss": 0.1545, "step": 760 }, { "grad_norm": 1.6199036836624146, "learning_rate": 1.9225e-05, "loss": 0.1482, "step": 770 }, { "grad_norm": 2.1191611289978027, "learning_rate": 1.9475000000000002e-05, "loss": 0.1484, "step": 780 }, { "grad_norm": 1.5473886728286743, "learning_rate": 1.9725000000000002e-05, "loss": 0.1413, "step": 790 }, { "grad_norm": 1.7293376922607422, "learning_rate": 1.9975e-05, "loss": 0.138, "step": 800 }, { "grad_norm": 1.5900216102600098, "learning_rate": 2.0225000000000004e-05, "loss": 0.1331, "step": 810 }, { "grad_norm": 1.8217607736587524, "learning_rate": 2.0475e-05, "loss": 0.1273, "step": 820 }, { "grad_norm": 1.5571990013122559, "learning_rate": 2.0725e-05, "loss": 0.1291, "step": 830 }, { "grad_norm": 1.7084885835647583, "learning_rate": 2.0975e-05, "loss": 0.1242, "step": 840 }, { "grad_norm": 1.9380112886428833, "learning_rate": 2.1225e-05, "loss": 0.1243, "step": 850 }, { "grad_norm": 1.853886604309082, "learning_rate": 2.1475e-05, "loss": 0.133, "step": 860 }, { "grad_norm": 1.817953109741211, "learning_rate": 2.1725e-05, "loss": 0.1212, "step": 870 }, { "grad_norm": 1.8641302585601807, "learning_rate": 2.1975000000000002e-05, "loss": 0.1183, "step": 880 }, { "grad_norm": 1.7489721775054932, "learning_rate": 2.2225e-05, "loss": 0.1116, "step": 890 }, { "grad_norm": 1.757407307624817, "learning_rate": 2.2475e-05, "loss": 0.1039, "step": 900 }, { "grad_norm": 1.301628828048706, "learning_rate": 2.2725000000000003e-05, "loss": 0.1079, "step": 910 }, { "grad_norm": 2.164989948272705, "learning_rate": 2.2975000000000003e-05, "loss": 0.114, "step": 920 }, { "grad_norm": 2.140270471572876, "learning_rate": 2.3225000000000002e-05, "loss": 0.1256, "step": 930 }, { "grad_norm": 1.2337141036987305, "learning_rate": 2.3475e-05, "loss": 0.106, "step": 940 }, { "grad_norm": 1.3270549774169922, "learning_rate": 2.3725e-05, "loss": 0.1114, "step": 950 }, { "grad_norm": 1.2835454940795898, "learning_rate": 2.3975e-05, "loss": 0.1065, "step": 960 }, { "grad_norm": 1.5953614711761475, "learning_rate": 2.4225e-05, "loss": 0.105, "step": 970 }, { "grad_norm": 1.8449679613113403, "learning_rate": 2.4475000000000002e-05, "loss": 0.1085, "step": 980 }, { "grad_norm": 1.5764297246932983, "learning_rate": 2.4725e-05, "loss": 0.1076, "step": 990 }, { "grad_norm": 1.545379877090454, "learning_rate": 2.4975e-05, "loss": 0.114, "step": 1000 }, { "grad_norm": 1.3646624088287354, "learning_rate": 2.5225e-05, "loss": 0.107, "step": 1010 }, { "grad_norm": 1.3966397047042847, "learning_rate": 2.5475e-05, "loss": 0.1123, "step": 1020 }, { "grad_norm": 1.4768576622009277, "learning_rate": 2.5725e-05, "loss": 0.1128, "step": 1030 }, { "grad_norm": 1.7385600805282593, "learning_rate": 2.5974999999999998e-05, "loss": 0.1031, "step": 1040 }, { "grad_norm": 1.5427240133285522, "learning_rate": 2.6225e-05, "loss": 0.0942, "step": 1050 }, { "grad_norm": 1.282538890838623, "learning_rate": 2.6475e-05, "loss": 0.0964, "step": 1060 }, { "grad_norm": 1.0972903966903687, "learning_rate": 2.6725e-05, "loss": 0.0927, "step": 1070 }, { "grad_norm": 1.185531497001648, "learning_rate": 2.6975000000000002e-05, "loss": 0.0829, "step": 1080 }, { "grad_norm": 1.318634033203125, "learning_rate": 2.7225e-05, "loss": 0.092, "step": 1090 }, { "grad_norm": 1.6268579959869385, "learning_rate": 2.7475e-05, "loss": 0.0967, "step": 1100 }, { "grad_norm": 1.5126841068267822, "learning_rate": 2.7725e-05, "loss": 0.0861, "step": 1110 }, { "grad_norm": 1.484811544418335, "learning_rate": 2.7975000000000002e-05, "loss": 0.0976, "step": 1120 }, { "grad_norm": 1.3723887205123901, "learning_rate": 2.8225e-05, "loss": 0.0951, "step": 1130 }, { "grad_norm": 1.444368839263916, "learning_rate": 2.8475e-05, "loss": 0.0923, "step": 1140 }, { "grad_norm": 1.5154602527618408, "learning_rate": 2.8725e-05, "loss": 0.0963, "step": 1150 }, { "grad_norm": 1.3796039819717407, "learning_rate": 2.8975000000000003e-05, "loss": 0.0862, "step": 1160 }, { "grad_norm": 1.149346113204956, "learning_rate": 2.9225000000000002e-05, "loss": 0.0878, "step": 1170 }, { "grad_norm": 1.90553617477417, "learning_rate": 2.9475e-05, "loss": 0.0928, "step": 1180 }, { "grad_norm": 1.2240837812423706, "learning_rate": 2.9725000000000004e-05, "loss": 0.0881, "step": 1190 }, { "grad_norm": 1.1540358066558838, "learning_rate": 2.9975000000000004e-05, "loss": 0.0944, "step": 1200 }, { "grad_norm": 1.4626652002334595, "learning_rate": 3.0225000000000003e-05, "loss": 0.0872, "step": 1210 }, { "grad_norm": 1.3133652210235596, "learning_rate": 3.0475000000000002e-05, "loss": 0.0787, "step": 1220 }, { "grad_norm": 1.1019482612609863, "learning_rate": 3.0725e-05, "loss": 0.0819, "step": 1230 }, { "grad_norm": 1.3753159046173096, "learning_rate": 3.0975e-05, "loss": 0.0881, "step": 1240 }, { "grad_norm": 1.2031173706054688, "learning_rate": 3.122500000000001e-05, "loss": 0.0847, "step": 1250 }, { "grad_norm": 1.1128253936767578, "learning_rate": 3.1475e-05, "loss": 0.0736, "step": 1260 }, { "grad_norm": 1.6023499965667725, "learning_rate": 3.1725e-05, "loss": 0.092, "step": 1270 }, { "grad_norm": 1.4361248016357422, "learning_rate": 3.1975e-05, "loss": 0.0817, "step": 1280 }, { "grad_norm": 1.3706294298171997, "learning_rate": 3.2225e-05, "loss": 0.0838, "step": 1290 }, { "grad_norm": 1.1677745580673218, "learning_rate": 3.2474999999999997e-05, "loss": 0.085, "step": 1300 }, { "grad_norm": 1.2363678216934204, "learning_rate": 3.2725e-05, "loss": 0.0755, "step": 1310 }, { "grad_norm": 1.321259617805481, "learning_rate": 3.2975e-05, "loss": 0.0951, "step": 1320 }, { "grad_norm": 1.1303952932357788, "learning_rate": 3.3225e-05, "loss": 0.0792, "step": 1330 }, { "grad_norm": 1.2076300382614136, "learning_rate": 3.3475e-05, "loss": 0.0695, "step": 1340 }, { "grad_norm": 1.1507999897003174, "learning_rate": 3.3725e-05, "loss": 0.0814, "step": 1350 }, { "grad_norm": 1.3013414144515991, "learning_rate": 3.3975e-05, "loss": 0.0871, "step": 1360 }, { "grad_norm": 0.9929667115211487, "learning_rate": 3.4225e-05, "loss": 0.0756, "step": 1370 }, { "grad_norm": 1.2897142171859741, "learning_rate": 3.4475000000000005e-05, "loss": 0.0727, "step": 1380 }, { "grad_norm": 1.294641375541687, "learning_rate": 3.4725000000000004e-05, "loss": 0.0711, "step": 1390 }, { "grad_norm": 1.2155081033706665, "learning_rate": 3.4975e-05, "loss": 0.0725, "step": 1400 }, { "grad_norm": 1.1517822742462158, "learning_rate": 3.5225e-05, "loss": 0.0793, "step": 1410 }, { "grad_norm": 0.8690961599349976, "learning_rate": 3.5475e-05, "loss": 0.0761, "step": 1420 }, { "grad_norm": 1.2019202709197998, "learning_rate": 3.5725e-05, "loss": 0.0876, "step": 1430 }, { "grad_norm": 1.1133569478988647, "learning_rate": 3.5975e-05, "loss": 0.075, "step": 1440 }, { "grad_norm": 0.9797796010971069, "learning_rate": 3.6225000000000006e-05, "loss": 0.0685, "step": 1450 }, { "grad_norm": 0.9948344826698303, "learning_rate": 3.6475000000000006e-05, "loss": 0.0795, "step": 1460 }, { "grad_norm": 1.219495415687561, "learning_rate": 3.6725000000000005e-05, "loss": 0.0703, "step": 1470 }, { "grad_norm": 1.1820889711380005, "learning_rate": 3.6975000000000004e-05, "loss": 0.0701, "step": 1480 }, { "grad_norm": 1.3867719173431396, "learning_rate": 3.7225000000000004e-05, "loss": 0.0785, "step": 1490 }, { "grad_norm": 1.126526951789856, "learning_rate": 3.7475e-05, "loss": 0.0757, "step": 1500 }, { "grad_norm": 0.7901089787483215, "learning_rate": 3.7725e-05, "loss": 0.0722, "step": 1510 }, { "grad_norm": 1.2209901809692383, "learning_rate": 3.7975e-05, "loss": 0.0727, "step": 1520 }, { "grad_norm": 0.9143607020378113, "learning_rate": 3.8225e-05, "loss": 0.0682, "step": 1530 }, { "grad_norm": 0.9692726731300354, "learning_rate": 3.8475e-05, "loss": 0.0707, "step": 1540 }, { "grad_norm": 0.8995283246040344, "learning_rate": 3.8725e-05, "loss": 0.071, "step": 1550 }, { "grad_norm": 1.0713919401168823, "learning_rate": 3.8975e-05, "loss": 0.0723, "step": 1560 }, { "grad_norm": 1.1776140928268433, "learning_rate": 3.9225e-05, "loss": 0.0719, "step": 1570 }, { "grad_norm": 1.028486967086792, "learning_rate": 3.9475000000000004e-05, "loss": 0.0644, "step": 1580 }, { "grad_norm": 0.8951988816261292, "learning_rate": 3.9725e-05, "loss": 0.0665, "step": 1590 }, { "grad_norm": 0.8805592060089111, "learning_rate": 3.9975e-05, "loss": 0.075, "step": 1600 }, { "grad_norm": 0.8125795722007751, "learning_rate": 4.0225e-05, "loss": 0.0686, "step": 1610 }, { "grad_norm": 1.135894775390625, "learning_rate": 4.0475e-05, "loss": 0.0658, "step": 1620 }, { "grad_norm": 1.2084405422210693, "learning_rate": 4.0725e-05, "loss": 0.0718, "step": 1630 }, { "grad_norm": 1.2208480834960938, "learning_rate": 4.0975e-05, "loss": 0.0715, "step": 1640 }, { "grad_norm": 0.885971188545227, "learning_rate": 4.1225e-05, "loss": 0.0641, "step": 1650 }, { "grad_norm": 0.8390306830406189, "learning_rate": 4.1475000000000005e-05, "loss": 0.0636, "step": 1660 }, { "grad_norm": 1.002145528793335, "learning_rate": 4.1725000000000005e-05, "loss": 0.0714, "step": 1670 }, { "grad_norm": 1.0860170125961304, "learning_rate": 4.1975000000000004e-05, "loss": 0.0631, "step": 1680 }, { "grad_norm": 0.8816429376602173, "learning_rate": 4.2225e-05, "loss": 0.0609, "step": 1690 }, { "grad_norm": 0.7317923307418823, "learning_rate": 4.2475e-05, "loss": 0.0731, "step": 1700 }, { "grad_norm": 0.7520487904548645, "learning_rate": 4.2725e-05, "loss": 0.06, "step": 1710 }, { "grad_norm": 0.7894096374511719, "learning_rate": 4.2975e-05, "loss": 0.0674, "step": 1720 }, { "grad_norm": 0.8585862517356873, "learning_rate": 4.322500000000001e-05, "loss": 0.0651, "step": 1730 }, { "grad_norm": 1.1899316310882568, "learning_rate": 4.3475000000000006e-05, "loss": 0.0642, "step": 1740 }, { "grad_norm": 0.9894418716430664, "learning_rate": 4.3725000000000006e-05, "loss": 0.0606, "step": 1750 }, { "grad_norm": 0.8882922530174255, "learning_rate": 4.3975e-05, "loss": 0.0572, "step": 1760 }, { "grad_norm": 0.8644171357154846, "learning_rate": 4.4225e-05, "loss": 0.0754, "step": 1770 }, { "grad_norm": 0.9033655524253845, "learning_rate": 4.4475e-05, "loss": 0.0813, "step": 1780 }, { "grad_norm": 0.9285604357719421, "learning_rate": 4.4725e-05, "loss": 0.0657, "step": 1790 }, { "grad_norm": 0.9482817053794861, "learning_rate": 4.4975e-05, "loss": 0.0786, "step": 1800 }, { "grad_norm": 0.9869294166564941, "learning_rate": 4.5225e-05, "loss": 0.0699, "step": 1810 }, { "grad_norm": 0.9271327257156372, "learning_rate": 4.5475e-05, "loss": 0.0699, "step": 1820 }, { "grad_norm": 0.680384635925293, "learning_rate": 4.5725e-05, "loss": 0.063, "step": 1830 }, { "grad_norm": 0.8649093508720398, "learning_rate": 4.5975e-05, "loss": 0.0689, "step": 1840 }, { "grad_norm": 0.837262749671936, "learning_rate": 4.6225e-05, "loss": 0.0662, "step": 1850 }, { "grad_norm": 0.7167726159095764, "learning_rate": 4.6475000000000005e-05, "loss": 0.0598, "step": 1860 }, { "grad_norm": 0.8097646832466125, "learning_rate": 4.6725000000000004e-05, "loss": 0.0619, "step": 1870 }, { "grad_norm": 0.8313556909561157, "learning_rate": 4.6975000000000003e-05, "loss": 0.0597, "step": 1880 }, { "grad_norm": 0.7028264403343201, "learning_rate": 4.7225e-05, "loss": 0.06, "step": 1890 }, { "grad_norm": 1.247399926185608, "learning_rate": 4.7475e-05, "loss": 0.0654, "step": 1900 }, { "grad_norm": 0.7923406958580017, "learning_rate": 4.7725e-05, "loss": 0.0666, "step": 1910 }, { "grad_norm": 0.7047503590583801, "learning_rate": 4.7975e-05, "loss": 0.0628, "step": 1920 }, { "grad_norm": 0.6851572394371033, "learning_rate": 4.822500000000001e-05, "loss": 0.0606, "step": 1930 }, { "grad_norm": 0.8303053975105286, "learning_rate": 4.8475000000000006e-05, "loss": 0.0721, "step": 1940 }, { "grad_norm": 0.8470669984817505, "learning_rate": 4.8725000000000005e-05, "loss": 0.062, "step": 1950 }, { "grad_norm": 1.1616108417510986, "learning_rate": 4.8975000000000005e-05, "loss": 0.0597, "step": 1960 }, { "grad_norm": 0.5819072723388672, "learning_rate": 4.9225000000000004e-05, "loss": 0.061, "step": 1970 }, { "grad_norm": 0.8464227914810181, "learning_rate": 4.9475e-05, "loss": 0.0648, "step": 1980 }, { "grad_norm": 0.7199952006340027, "learning_rate": 4.9725e-05, "loss": 0.0567, "step": 1990 }, { "grad_norm": 0.8480494618415833, "learning_rate": 4.9975e-05, "loss": 0.0602, "step": 2000 }, { "grad_norm": 0.658765971660614, "learning_rate": 5.0225e-05, "loss": 0.0557, "step": 2010 }, { "grad_norm": 0.7687562108039856, "learning_rate": 5.047500000000001e-05, "loss": 0.0615, "step": 2020 }, { "grad_norm": 0.7604524493217468, "learning_rate": 5.0725e-05, "loss": 0.0546, "step": 2030 }, { "grad_norm": 0.7438364028930664, "learning_rate": 5.0975000000000006e-05, "loss": 0.0588, "step": 2040 }, { "grad_norm": 0.8400558829307556, "learning_rate": 5.1225e-05, "loss": 0.0585, "step": 2050 }, { "grad_norm": 0.9639374613761902, "learning_rate": 5.1475000000000004e-05, "loss": 0.0578, "step": 2060 }, { "grad_norm": 0.8440175652503967, "learning_rate": 5.1725000000000004e-05, "loss": 0.0583, "step": 2070 }, { "grad_norm": 0.6768313646316528, "learning_rate": 5.197500000000001e-05, "loss": 0.0595, "step": 2080 }, { "grad_norm": 0.9968610405921936, "learning_rate": 5.2225e-05, "loss": 0.0649, "step": 2090 }, { "grad_norm": 0.7608867883682251, "learning_rate": 5.247500000000001e-05, "loss": 0.0576, "step": 2100 }, { "grad_norm": 0.7837508320808411, "learning_rate": 5.2725e-05, "loss": 0.0564, "step": 2110 }, { "grad_norm": 0.6221339702606201, "learning_rate": 5.297500000000001e-05, "loss": 0.0575, "step": 2120 }, { "grad_norm": 0.5899678468704224, "learning_rate": 5.3225e-05, "loss": 0.0597, "step": 2130 }, { "grad_norm": 0.9251871705055237, "learning_rate": 5.3475e-05, "loss": 0.0573, "step": 2140 }, { "grad_norm": 0.7509172558784485, "learning_rate": 5.3725000000000005e-05, "loss": 0.0595, "step": 2150 }, { "grad_norm": 0.7112311720848083, "learning_rate": 5.3975e-05, "loss": 0.0478, "step": 2160 }, { "grad_norm": 0.7615765333175659, "learning_rate": 5.4225000000000003e-05, "loss": 0.0568, "step": 2170 }, { "grad_norm": 0.7463403344154358, "learning_rate": 5.4474999999999996e-05, "loss": 0.0586, "step": 2180 }, { "grad_norm": 0.5818749666213989, "learning_rate": 5.4725e-05, "loss": 0.0559, "step": 2190 }, { "grad_norm": 0.7068076729774475, "learning_rate": 5.4975e-05, "loss": 0.0512, "step": 2200 }, { "grad_norm": 0.6013463139533997, "learning_rate": 5.522500000000001e-05, "loss": 0.0494, "step": 2210 }, { "grad_norm": 0.7436593174934387, "learning_rate": 5.5475e-05, "loss": 0.0509, "step": 2220 }, { "grad_norm": 0.5676515698432922, "learning_rate": 5.5725000000000006e-05, "loss": 0.0526, "step": 2230 }, { "grad_norm": 0.812030017375946, "learning_rate": 5.5975e-05, "loss": 0.0588, "step": 2240 }, { "grad_norm": 0.9136979579925537, "learning_rate": 5.6225000000000005e-05, "loss": 0.0638, "step": 2250 }, { "grad_norm": 0.7437300682067871, "learning_rate": 5.6475e-05, "loss": 0.0639, "step": 2260 }, { "grad_norm": 0.7953630089759827, "learning_rate": 5.6725e-05, "loss": 0.0625, "step": 2270 }, { "grad_norm": 0.6982801556587219, "learning_rate": 5.6975e-05, "loss": 0.0554, "step": 2280 }, { "grad_norm": 0.7675284147262573, "learning_rate": 5.722500000000001e-05, "loss": 0.0519, "step": 2290 }, { "grad_norm": 0.7702083587646484, "learning_rate": 5.7475e-05, "loss": 0.0561, "step": 2300 }, { "grad_norm": 0.7242898344993591, "learning_rate": 5.772500000000001e-05, "loss": 0.0591, "step": 2310 }, { "grad_norm": 0.7835424542427063, "learning_rate": 5.7975e-05, "loss": 0.0552, "step": 2320 }, { "grad_norm": 0.9267889857292175, "learning_rate": 5.8225000000000006e-05, "loss": 0.0641, "step": 2330 }, { "grad_norm": 0.6127759218215942, "learning_rate": 5.8475000000000005e-05, "loss": 0.0593, "step": 2340 }, { "grad_norm": 1.0412110090255737, "learning_rate": 5.8725000000000004e-05, "loss": 0.061, "step": 2350 }, { "grad_norm": 0.7494506239891052, "learning_rate": 5.8975000000000004e-05, "loss": 0.0739, "step": 2360 }, { "grad_norm": 0.7000832557678223, "learning_rate": 5.922500000000001e-05, "loss": 0.0656, "step": 2370 }, { "grad_norm": 0.5600836277008057, "learning_rate": 5.9475e-05, "loss": 0.0614, "step": 2380 }, { "grad_norm": 0.7076703310012817, "learning_rate": 5.9724999999999995e-05, "loss": 0.056, "step": 2390 }, { "grad_norm": 0.6326077580451965, "learning_rate": 5.9975e-05, "loss": 0.0579, "step": 2400 }, { "grad_norm": 0.8572502136230469, "learning_rate": 6.0225e-05, "loss": 0.0515, "step": 2410 }, { "grad_norm": 0.672612726688385, "learning_rate": 6.0475000000000006e-05, "loss": 0.0555, "step": 2420 }, { "grad_norm": 0.46678104996681213, "learning_rate": 6.0725e-05, "loss": 0.0543, "step": 2430 }, { "grad_norm": 0.5345669388771057, "learning_rate": 6.0975000000000005e-05, "loss": 0.0526, "step": 2440 }, { "grad_norm": 0.6343130469322205, "learning_rate": 6.1225e-05, "loss": 0.0533, "step": 2450 }, { "grad_norm": 0.6937117576599121, "learning_rate": 6.1475e-05, "loss": 0.0607, "step": 2460 }, { "grad_norm": 0.623308002948761, "learning_rate": 6.1725e-05, "loss": 0.0507, "step": 2470 }, { "grad_norm": 0.6295056343078613, "learning_rate": 6.1975e-05, "loss": 0.0553, "step": 2480 }, { "grad_norm": 0.6251871585845947, "learning_rate": 6.2225e-05, "loss": 0.0511, "step": 2490 }, { "grad_norm": 0.761501669883728, "learning_rate": 6.2475e-05, "loss": 0.0538, "step": 2500 }, { "grad_norm": 0.5880475640296936, "learning_rate": 6.2725e-05, "loss": 0.0556, "step": 2510 }, { "grad_norm": 0.7141260504722595, "learning_rate": 6.297500000000001e-05, "loss": 0.0512, "step": 2520 }, { "grad_norm": 0.6467633247375488, "learning_rate": 6.3225e-05, "loss": 0.0517, "step": 2530 }, { "grad_norm": 0.5360629558563232, "learning_rate": 6.347500000000001e-05, "loss": 0.0499, "step": 2540 }, { "grad_norm": 0.6864625811576843, "learning_rate": 6.3725e-05, "loss": 0.0586, "step": 2550 }, { "grad_norm": 0.7282418012619019, "learning_rate": 6.397500000000001e-05, "loss": 0.0489, "step": 2560 }, { "grad_norm": 0.7517862319946289, "learning_rate": 6.4225e-05, "loss": 0.054, "step": 2570 }, { "grad_norm": 0.5441966652870178, "learning_rate": 6.447500000000001e-05, "loss": 0.0574, "step": 2580 }, { "grad_norm": 0.6316747665405273, "learning_rate": 6.4725e-05, "loss": 0.0504, "step": 2590 }, { "grad_norm": 0.7340215444564819, "learning_rate": 6.497500000000001e-05, "loss": 0.0527, "step": 2600 }, { "grad_norm": 0.6639565229415894, "learning_rate": 6.5225e-05, "loss": 0.0518, "step": 2610 }, { "grad_norm": 0.970940351486206, "learning_rate": 6.5475e-05, "loss": 0.0544, "step": 2620 }, { "grad_norm": 0.5749973654747009, "learning_rate": 6.5725e-05, "loss": 0.0606, "step": 2630 }, { "grad_norm": 0.6175587773323059, "learning_rate": 6.5975e-05, "loss": 0.0554, "step": 2640 }, { "grad_norm": 0.5030772686004639, "learning_rate": 6.6225e-05, "loss": 0.0531, "step": 2650 }, { "grad_norm": 0.44821229577064514, "learning_rate": 6.6475e-05, "loss": 0.053, "step": 2660 }, { "grad_norm": 1.306839108467102, "learning_rate": 6.672500000000001e-05, "loss": 0.0564, "step": 2670 }, { "grad_norm": 1.2493832111358643, "learning_rate": 6.6975e-05, "loss": 0.0548, "step": 2680 }, { "grad_norm": 1.0581810474395752, "learning_rate": 6.722500000000001e-05, "loss": 0.0613, "step": 2690 }, { "grad_norm": 0.7844956517219543, "learning_rate": 6.7475e-05, "loss": 0.057, "step": 2700 }, { "grad_norm": 0.6968123316764832, "learning_rate": 6.7725e-05, "loss": 0.058, "step": 2710 }, { "grad_norm": 0.4456035792827606, "learning_rate": 6.7975e-05, "loss": 0.0506, "step": 2720 }, { "grad_norm": 0.5860215425491333, "learning_rate": 6.8225e-05, "loss": 0.0529, "step": 2730 }, { "grad_norm": 0.6344910264015198, "learning_rate": 6.8475e-05, "loss": 0.0538, "step": 2740 }, { "grad_norm": 0.6461586356163025, "learning_rate": 6.8725e-05, "loss": 0.0516, "step": 2750 }, { "grad_norm": 0.7884969115257263, "learning_rate": 6.8975e-05, "loss": 0.0545, "step": 2760 }, { "grad_norm": 0.6607843041419983, "learning_rate": 6.9225e-05, "loss": 0.0465, "step": 2770 }, { "grad_norm": 0.5909495949745178, "learning_rate": 6.9475e-05, "loss": 0.0512, "step": 2780 }, { "grad_norm": 0.537689208984375, "learning_rate": 6.9725e-05, "loss": 0.056, "step": 2790 }, { "grad_norm": 0.7072675824165344, "learning_rate": 6.997500000000001e-05, "loss": 0.0502, "step": 2800 }, { "grad_norm": 0.7460647821426392, "learning_rate": 7.022500000000001e-05, "loss": 0.0479, "step": 2810 }, { "grad_norm": 0.39364323019981384, "learning_rate": 7.0475e-05, "loss": 0.0442, "step": 2820 }, { "grad_norm": 0.5745301842689514, "learning_rate": 7.072500000000001e-05, "loss": 0.0516, "step": 2830 }, { "grad_norm": 0.4373280704021454, "learning_rate": 7.0975e-05, "loss": 0.0527, "step": 2840 }, { "grad_norm": 0.44675442576408386, "learning_rate": 7.122500000000001e-05, "loss": 0.047, "step": 2850 }, { "grad_norm": 0.7202121615409851, "learning_rate": 7.1475e-05, "loss": 0.0449, "step": 2860 }, { "grad_norm": 0.4915679097175598, "learning_rate": 7.172500000000001e-05, "loss": 0.0468, "step": 2870 }, { "grad_norm": 0.43736374378204346, "learning_rate": 7.1975e-05, "loss": 0.0451, "step": 2880 }, { "grad_norm": 0.4668284058570862, "learning_rate": 7.2225e-05, "loss": 0.042, "step": 2890 }, { "grad_norm": 0.682341992855072, "learning_rate": 7.2475e-05, "loss": 0.0448, "step": 2900 }, { "grad_norm": 0.5675418376922607, "learning_rate": 7.272499999999999e-05, "loss": 0.0466, "step": 2910 }, { "grad_norm": 0.5599114298820496, "learning_rate": 7.2975e-05, "loss": 0.0453, "step": 2920 }, { "grad_norm": 0.5984386205673218, "learning_rate": 7.3225e-05, "loss": 0.0513, "step": 2930 }, { "grad_norm": 0.6827234029769897, "learning_rate": 7.347500000000001e-05, "loss": 0.0536, "step": 2940 }, { "grad_norm": 0.49290385842323303, "learning_rate": 7.3725e-05, "loss": 0.0498, "step": 2950 }, { "grad_norm": 0.6331766843795776, "learning_rate": 7.397500000000001e-05, "loss": 0.0535, "step": 2960 }, { "grad_norm": 0.46048396825790405, "learning_rate": 7.4225e-05, "loss": 0.0446, "step": 2970 }, { "grad_norm": 0.5611748695373535, "learning_rate": 7.447500000000001e-05, "loss": 0.0446, "step": 2980 }, { "grad_norm": 0.5913515090942383, "learning_rate": 7.4725e-05, "loss": 0.0437, "step": 2990 }, { "grad_norm": 0.544611394405365, "learning_rate": 7.4975e-05, "loss": 0.0475, "step": 3000 }, { "grad_norm": 0.40595072507858276, "learning_rate": 7.5225e-05, "loss": 0.0516, "step": 3010 }, { "grad_norm": 0.6375641822814941, "learning_rate": 7.5475e-05, "loss": 0.0461, "step": 3020 }, { "grad_norm": 0.3940492570400238, "learning_rate": 7.5725e-05, "loss": 0.0438, "step": 3030 }, { "grad_norm": 0.6022951602935791, "learning_rate": 7.5975e-05, "loss": 0.0423, "step": 3040 }, { "grad_norm": 0.6211546063423157, "learning_rate": 7.6225e-05, "loss": 0.046, "step": 3050 }, { "grad_norm": 0.6333584189414978, "learning_rate": 7.6475e-05, "loss": 0.0468, "step": 3060 }, { "grad_norm": 0.5094603300094604, "learning_rate": 7.672500000000001e-05, "loss": 0.0467, "step": 3070 }, { "grad_norm": 0.5822936296463013, "learning_rate": 7.697500000000001e-05, "loss": 0.0454, "step": 3080 }, { "grad_norm": 0.5196226239204407, "learning_rate": 7.722500000000001e-05, "loss": 0.0454, "step": 3090 }, { "grad_norm": 0.4697386920452118, "learning_rate": 7.747500000000001e-05, "loss": 0.0499, "step": 3100 }, { "grad_norm": 0.6816082000732422, "learning_rate": 7.7725e-05, "loss": 0.0509, "step": 3110 }, { "grad_norm": 0.46242374181747437, "learning_rate": 7.797500000000001e-05, "loss": 0.0399, "step": 3120 }, { "grad_norm": 0.5932023525238037, "learning_rate": 7.8225e-05, "loss": 0.0545, "step": 3130 }, { "grad_norm": 0.6690108776092529, "learning_rate": 7.8475e-05, "loss": 0.0472, "step": 3140 }, { "grad_norm": 0.5768377184867859, "learning_rate": 7.8725e-05, "loss": 0.0484, "step": 3150 }, { "grad_norm": 0.5579317808151245, "learning_rate": 7.8975e-05, "loss": 0.0427, "step": 3160 }, { "grad_norm": 0.4241048991680145, "learning_rate": 7.9225e-05, "loss": 0.0476, "step": 3170 }, { "grad_norm": 0.5672839879989624, "learning_rate": 7.9475e-05, "loss": 0.0466, "step": 3180 }, { "grad_norm": 0.6738792061805725, "learning_rate": 7.9725e-05, "loss": 0.0504, "step": 3190 }, { "grad_norm": 0.5103583931922913, "learning_rate": 7.9975e-05, "loss": 0.0448, "step": 3200 }, { "grad_norm": 0.4033074975013733, "learning_rate": 8.022500000000001e-05, "loss": 0.0417, "step": 3210 }, { "grad_norm": 0.7127251625061035, "learning_rate": 8.0475e-05, "loss": 0.0444, "step": 3220 }, { "grad_norm": 0.47499921917915344, "learning_rate": 8.072500000000001e-05, "loss": 0.0549, "step": 3230 }, { "grad_norm": 0.4665334224700928, "learning_rate": 8.0975e-05, "loss": 0.0451, "step": 3240 }, { "grad_norm": 0.49986517429351807, "learning_rate": 8.122500000000001e-05, "loss": 0.0465, "step": 3250 }, { "grad_norm": 0.5379824638366699, "learning_rate": 8.1475e-05, "loss": 0.0475, "step": 3260 }, { "grad_norm": 0.5345013737678528, "learning_rate": 8.172500000000001e-05, "loss": 0.0468, "step": 3270 }, { "grad_norm": 0.639836847782135, "learning_rate": 8.1975e-05, "loss": 0.0409, "step": 3280 }, { "grad_norm": 0.505789041519165, "learning_rate": 8.2225e-05, "loss": 0.0479, "step": 3290 }, { "grad_norm": 0.4632497727870941, "learning_rate": 8.2475e-05, "loss": 0.0472, "step": 3300 }, { "grad_norm": 0.48713013529777527, "learning_rate": 8.2725e-05, "loss": 0.046, "step": 3310 }, { "grad_norm": 0.49235033988952637, "learning_rate": 8.2975e-05, "loss": 0.0501, "step": 3320 }, { "grad_norm": 0.5878191590309143, "learning_rate": 8.3225e-05, "loss": 0.0446, "step": 3330 }, { "grad_norm": 0.45359861850738525, "learning_rate": 8.347500000000001e-05, "loss": 0.0447, "step": 3340 }, { "grad_norm": 0.5364097952842712, "learning_rate": 8.3725e-05, "loss": 0.0472, "step": 3350 }, { "grad_norm": 0.49855002760887146, "learning_rate": 8.397500000000001e-05, "loss": 0.0427, "step": 3360 }, { "grad_norm": 0.4520539343357086, "learning_rate": 8.422500000000001e-05, "loss": 0.0454, "step": 3370 }, { "grad_norm": 0.46096259355545044, "learning_rate": 8.447500000000001e-05, "loss": 0.0436, "step": 3380 }, { "grad_norm": 0.49285227060317993, "learning_rate": 8.4725e-05, "loss": 0.0463, "step": 3390 }, { "grad_norm": 0.45933452248573303, "learning_rate": 8.4975e-05, "loss": 0.0483, "step": 3400 }, { "grad_norm": 0.5284451842308044, "learning_rate": 8.5225e-05, "loss": 0.0486, "step": 3410 }, { "grad_norm": 0.39977461099624634, "learning_rate": 8.5475e-05, "loss": 0.0425, "step": 3420 }, { "grad_norm": 0.713625967502594, "learning_rate": 8.5725e-05, "loss": 0.0418, "step": 3430 }, { "grad_norm": 0.6533679366111755, "learning_rate": 8.5975e-05, "loss": 0.0462, "step": 3440 }, { "grad_norm": 0.6241810917854309, "learning_rate": 8.6225e-05, "loss": 0.0386, "step": 3450 }, { "grad_norm": 0.5186618566513062, "learning_rate": 8.6475e-05, "loss": 0.0437, "step": 3460 }, { "grad_norm": 0.5300339460372925, "learning_rate": 8.672500000000001e-05, "loss": 0.0471, "step": 3470 }, { "grad_norm": 0.44799116253852844, "learning_rate": 8.6975e-05, "loss": 0.0441, "step": 3480 }, { "grad_norm": 0.509917676448822, "learning_rate": 8.7225e-05, "loss": 0.0425, "step": 3490 }, { "grad_norm": 0.5340568423271179, "learning_rate": 8.747500000000001e-05, "loss": 0.0426, "step": 3500 }, { "grad_norm": 0.6597117185592651, "learning_rate": 8.7725e-05, "loss": 0.0496, "step": 3510 }, { "grad_norm": 0.4959251284599304, "learning_rate": 8.797500000000001e-05, "loss": 0.0435, "step": 3520 }, { "grad_norm": 0.5126045346260071, "learning_rate": 8.8225e-05, "loss": 0.0464, "step": 3530 }, { "grad_norm": 0.6432322263717651, "learning_rate": 8.847500000000001e-05, "loss": 0.0463, "step": 3540 }, { "grad_norm": 0.4597359597682953, "learning_rate": 8.8725e-05, "loss": 0.0441, "step": 3550 }, { "grad_norm": 0.4878421425819397, "learning_rate": 8.897500000000001e-05, "loss": 0.0445, "step": 3560 }, { "grad_norm": 0.4870072603225708, "learning_rate": 8.9225e-05, "loss": 0.047, "step": 3570 }, { "grad_norm": 0.4989506006240845, "learning_rate": 8.9475e-05, "loss": 0.0683, "step": 3580 }, { "grad_norm": 0.4251370131969452, "learning_rate": 8.9725e-05, "loss": 0.0555, "step": 3590 }, { "grad_norm": 0.47309866547584534, "learning_rate": 8.9975e-05, "loss": 0.053, "step": 3600 }, { "grad_norm": 0.5568931698799133, "learning_rate": 9.0225e-05, "loss": 0.0505, "step": 3610 }, { "grad_norm": 0.5528845191001892, "learning_rate": 9.0475e-05, "loss": 0.0475, "step": 3620 }, { "grad_norm": 0.40993499755859375, "learning_rate": 9.072500000000001e-05, "loss": 0.048, "step": 3630 }, { "grad_norm": 0.5121254324913025, "learning_rate": 9.0975e-05, "loss": 0.044, "step": 3640 }, { "grad_norm": 0.4579777121543884, "learning_rate": 9.122500000000001e-05, "loss": 0.0515, "step": 3650 }, { "grad_norm": 0.5501810312271118, "learning_rate": 9.1475e-05, "loss": 0.0538, "step": 3660 }, { "grad_norm": 0.5352097749710083, "learning_rate": 9.172500000000001e-05, "loss": 0.0456, "step": 3670 }, { "grad_norm": 0.46208709478378296, "learning_rate": 9.1975e-05, "loss": 0.0431, "step": 3680 }, { "grad_norm": 0.4856385886669159, "learning_rate": 9.2225e-05, "loss": 0.0501, "step": 3690 }, { "grad_norm": 0.5076271891593933, "learning_rate": 9.2475e-05, "loss": 0.0401, "step": 3700 }, { "grad_norm": 0.44316086173057556, "learning_rate": 9.2725e-05, "loss": 0.0438, "step": 3710 }, { "grad_norm": 0.45971542596817017, "learning_rate": 9.2975e-05, "loss": 0.0437, "step": 3720 }, { "grad_norm": 0.38122907280921936, "learning_rate": 9.3225e-05, "loss": 0.0435, "step": 3730 }, { "grad_norm": 0.4904205799102783, "learning_rate": 9.3475e-05, "loss": 0.0454, "step": 3740 }, { "grad_norm": 0.4196482300758362, "learning_rate": 9.3725e-05, "loss": 0.0419, "step": 3750 }, { "grad_norm": 0.4237809181213379, "learning_rate": 9.397500000000001e-05, "loss": 0.0457, "step": 3760 }, { "grad_norm": 0.42059576511383057, "learning_rate": 9.422500000000001e-05, "loss": 0.0384, "step": 3770 }, { "grad_norm": 0.47593939304351807, "learning_rate": 9.4475e-05, "loss": 0.0446, "step": 3780 }, { "grad_norm": 0.33894872665405273, "learning_rate": 9.472500000000001e-05, "loss": 0.041, "step": 3790 }, { "grad_norm": 0.425527960062027, "learning_rate": 9.4975e-05, "loss": 0.0421, "step": 3800 }, { "grad_norm": 0.5598286390304565, "learning_rate": 9.522500000000001e-05, "loss": 0.0508, "step": 3810 }, { "grad_norm": 0.42299482226371765, "learning_rate": 9.5475e-05, "loss": 0.0501, "step": 3820 }, { "grad_norm": 0.35498201847076416, "learning_rate": 9.572500000000001e-05, "loss": 0.041, "step": 3830 }, { "grad_norm": 0.3442164957523346, "learning_rate": 9.5975e-05, "loss": 0.0411, "step": 3840 }, { "grad_norm": 0.48438239097595215, "learning_rate": 9.622500000000001e-05, "loss": 0.0424, "step": 3850 }, { "grad_norm": 0.3291952311992645, "learning_rate": 9.6475e-05, "loss": 0.0453, "step": 3860 }, { "grad_norm": 0.4471481740474701, "learning_rate": 9.6725e-05, "loss": 0.0394, "step": 3870 }, { "grad_norm": 0.5023987293243408, "learning_rate": 9.6975e-05, "loss": 0.0418, "step": 3880 }, { "grad_norm": 0.44215789437294006, "learning_rate": 9.7225e-05, "loss": 0.0448, "step": 3890 }, { "grad_norm": 0.4624996781349182, "learning_rate": 9.747500000000001e-05, "loss": 0.0474, "step": 3900 }, { "grad_norm": 0.6112748384475708, "learning_rate": 9.7725e-05, "loss": 0.0473, "step": 3910 }, { "grad_norm": 0.5663880705833435, "learning_rate": 9.797500000000001e-05, "loss": 0.0518, "step": 3920 }, { "grad_norm": 0.7093642950057983, "learning_rate": 9.8225e-05, "loss": 0.0421, "step": 3930 }, { "grad_norm": 0.4678977131843567, "learning_rate": 9.847500000000001e-05, "loss": 0.0476, "step": 3940 }, { "grad_norm": 0.42080679535865784, "learning_rate": 9.8725e-05, "loss": 0.0488, "step": 3950 }, { "grad_norm": 0.45515406131744385, "learning_rate": 9.897500000000001e-05, "loss": 0.0454, "step": 3960 }, { "grad_norm": 0.5549961924552917, "learning_rate": 9.9225e-05, "loss": 0.043, "step": 3970 }, { "grad_norm": 0.49888551235198975, "learning_rate": 9.9475e-05, "loss": 0.0412, "step": 3980 }, { "grad_norm": 0.3849007189273834, "learning_rate": 9.9725e-05, "loss": 0.0412, "step": 3990 }, { "grad_norm": 0.4961939752101898, "learning_rate": 9.9975e-05, "loss": 0.0431, "step": 4000 }, { "grad_norm": 0.44645360112190247, "learning_rate": 9.999999653982884e-05, "loss": 0.0405, "step": 4010 }, { "grad_norm": 0.39994311332702637, "learning_rate": 9.999998457874392e-05, "loss": 0.0398, "step": 4020 }, { "grad_norm": 0.37800687551498413, "learning_rate": 9.999996407402913e-05, "loss": 0.0441, "step": 4030 }, { "grad_norm": 0.5396924614906311, "learning_rate": 9.999993502568801e-05, "loss": 0.0469, "step": 4040 }, { "grad_norm": 0.5175265073776245, "learning_rate": 9.999989743372548e-05, "loss": 0.0494, "step": 4050 }, { "grad_norm": 0.41543933749198914, "learning_rate": 9.999985129814798e-05, "loss": 0.0452, "step": 4060 }, { "grad_norm": 0.39991307258605957, "learning_rate": 9.99997966189634e-05, "loss": 0.0405, "step": 4070 }, { "grad_norm": 0.41280728578567505, "learning_rate": 9.999973339618107e-05, "loss": 0.0422, "step": 4080 }, { "grad_norm": 0.3433506488800049, "learning_rate": 9.999966162981179e-05, "loss": 0.0381, "step": 4090 }, { "grad_norm": 0.39202168583869934, "learning_rate": 9.999958131986784e-05, "loss": 0.0383, "step": 4100 }, { "grad_norm": 0.40492185950279236, "learning_rate": 9.999949246636293e-05, "loss": 0.0558, "step": 4110 }, { "grad_norm": 0.42412662506103516, "learning_rate": 9.999939506931224e-05, "loss": 0.0533, "step": 4120 }, { "grad_norm": 0.5615144968032837, "learning_rate": 9.999928912873243e-05, "loss": 0.0536, "step": 4130 }, { "grad_norm": 0.4418800473213196, "learning_rate": 9.999917464464159e-05, "loss": 0.0457, "step": 4140 }, { "grad_norm": 0.3550400733947754, "learning_rate": 9.999905161705929e-05, "loss": 0.0456, "step": 4150 }, { "grad_norm": 0.4374850392341614, "learning_rate": 9.999892004600653e-05, "loss": 0.0423, "step": 4160 }, { "grad_norm": 0.388772189617157, "learning_rate": 9.999877993150581e-05, "loss": 0.0477, "step": 4170 }, { "grad_norm": 0.4890954792499542, "learning_rate": 9.999863127358108e-05, "loss": 0.0473, "step": 4180 }, { "grad_norm": 0.4368743300437927, "learning_rate": 9.999847407225773e-05, "loss": 0.0491, "step": 4190 }, { "grad_norm": 0.3310664892196655, "learning_rate": 9.999830832756262e-05, "loss": 0.0394, "step": 4200 }, { "grad_norm": 0.4448356628417969, "learning_rate": 9.999813403952407e-05, "loss": 0.0393, "step": 4210 }, { "grad_norm": 0.5338175892829895, "learning_rate": 9.999795120817187e-05, "loss": 0.0405, "step": 4220 }, { "grad_norm": 0.3958755135536194, "learning_rate": 9.999775983353725e-05, "loss": 0.041, "step": 4230 }, { "grad_norm": 0.40772688388824463, "learning_rate": 9.999755991565292e-05, "loss": 0.0399, "step": 4240 }, { "grad_norm": 0.42285946011543274, "learning_rate": 9.999735145455303e-05, "loss": 0.043, "step": 4250 }, { "grad_norm": 0.39227426052093506, "learning_rate": 9.99971344502732e-05, "loss": 0.0431, "step": 4260 }, { "grad_norm": 0.4272048771381378, "learning_rate": 9.999690890285053e-05, "loss": 0.0413, "step": 4270 }, { "grad_norm": 0.5415866374969482, "learning_rate": 9.999667481232356e-05, "loss": 0.0468, "step": 4280 }, { "grad_norm": 0.31098106503486633, "learning_rate": 9.999643217873225e-05, "loss": 0.0388, "step": 4290 }, { "grad_norm": 0.3921031653881073, "learning_rate": 9.999618100211809e-05, "loss": 0.0369, "step": 4300 }, { "grad_norm": 0.524545431137085, "learning_rate": 9.999592128252402e-05, "loss": 0.038, "step": 4310 }, { "grad_norm": 0.4104313552379608, "learning_rate": 9.999565301999437e-05, "loss": 0.0406, "step": 4320 }, { "grad_norm": 0.4053952693939209, "learning_rate": 9.999537621457502e-05, "loss": 0.037, "step": 4330 }, { "grad_norm": 0.3362249433994293, "learning_rate": 9.999509086631323e-05, "loss": 0.0402, "step": 4340 }, { "grad_norm": 0.5196506381034851, "learning_rate": 9.99947969752578e-05, "loss": 0.0413, "step": 4350 }, { "grad_norm": 0.3526574373245239, "learning_rate": 9.999449454145891e-05, "loss": 0.0369, "step": 4360 }, { "grad_norm": 0.3929980993270874, "learning_rate": 9.999418356496827e-05, "loss": 0.0417, "step": 4370 }, { "grad_norm": 0.5054603219032288, "learning_rate": 9.999386404583899e-05, "loss": 0.0404, "step": 4380 }, { "grad_norm": 0.47113654017448425, "learning_rate": 9.999353598412568e-05, "loss": 0.0462, "step": 4390 }, { "grad_norm": 0.43987777829170227, "learning_rate": 9.999319937988442e-05, "loss": 0.0401, "step": 4400 }, { "grad_norm": 0.35857242345809937, "learning_rate": 9.999285423317268e-05, "loss": 0.0395, "step": 4410 }, { "grad_norm": 0.44648808240890503, "learning_rate": 9.999250054404947e-05, "loss": 0.0389, "step": 4420 }, { "grad_norm": 0.30834338068962097, "learning_rate": 9.99921383125752e-05, "loss": 0.0373, "step": 4430 }, { "grad_norm": 0.44708752632141113, "learning_rate": 9.99917675388118e-05, "loss": 0.0403, "step": 4440 }, { "grad_norm": 0.3997916877269745, "learning_rate": 9.99913882228226e-05, "loss": 0.0406, "step": 4450 }, { "grad_norm": 0.3929179608821869, "learning_rate": 9.999100036467242e-05, "loss": 0.0361, "step": 4460 }, { "grad_norm": 0.3173561096191406, "learning_rate": 9.999060396442753e-05, "loss": 0.0352, "step": 4470 }, { "grad_norm": 0.3632104992866516, "learning_rate": 9.999019902215566e-05, "loss": 0.0368, "step": 4480 }, { "grad_norm": 0.4156198799610138, "learning_rate": 9.998978553792602e-05, "loss": 0.0416, "step": 4490 }, { "grad_norm": 0.4332118332386017, "learning_rate": 9.998936351180926e-05, "loss": 0.0355, "step": 4500 }, { "grad_norm": 0.3885791003704071, "learning_rate": 9.998893294387747e-05, "loss": 0.0339, "step": 4510 }, { "grad_norm": 0.47569289803504944, "learning_rate": 9.998849383420426e-05, "loss": 0.035, "step": 4520 }, { "grad_norm": 0.4389174282550812, "learning_rate": 9.998804618286465e-05, "loss": 0.0397, "step": 4530 }, { "grad_norm": 0.42283549904823303, "learning_rate": 9.99875899899351e-05, "loss": 0.0361, "step": 4540 }, { "grad_norm": 0.3404773473739624, "learning_rate": 9.99871252554936e-05, "loss": 0.0318, "step": 4550 }, { "grad_norm": 0.3923811614513397, "learning_rate": 9.998665197961955e-05, "loss": 0.0363, "step": 4560 }, { "grad_norm": 0.3850061595439911, "learning_rate": 9.998617016239379e-05, "loss": 0.0352, "step": 4570 }, { "grad_norm": 0.37737271189689636, "learning_rate": 9.998567980389869e-05, "loss": 0.0428, "step": 4580 }, { "grad_norm": 0.478417307138443, "learning_rate": 9.998518090421802e-05, "loss": 0.0438, "step": 4590 }, { "grad_norm": 0.4358876645565033, "learning_rate": 9.998467346343703e-05, "loss": 0.039, "step": 4600 }, { "grad_norm": 0.38017603754997253, "learning_rate": 9.998415748164243e-05, "loss": 0.0429, "step": 4610 }, { "grad_norm": 0.3091076910495758, "learning_rate": 9.998363295892238e-05, "loss": 0.0357, "step": 4620 }, { "grad_norm": 0.271602600812912, "learning_rate": 9.998309989536652e-05, "loss": 0.0364, "step": 4630 }, { "grad_norm": 0.3934313654899597, "learning_rate": 9.998255829106593e-05, "loss": 0.0371, "step": 4640 }, { "grad_norm": 0.32232555747032166, "learning_rate": 9.998200814611316e-05, "loss": 0.0359, "step": 4650 }, { "grad_norm": 0.5098556280136108, "learning_rate": 9.998144946060219e-05, "loss": 0.0376, "step": 4660 }, { "grad_norm": 0.47436389327049255, "learning_rate": 9.998088223462852e-05, "loss": 0.0359, "step": 4670 }, { "grad_norm": 0.35391613841056824, "learning_rate": 9.998030646828905e-05, "loss": 0.0323, "step": 4680 }, { "grad_norm": 0.40620648860931396, "learning_rate": 9.997972216168217e-05, "loss": 0.0369, "step": 4690 }, { "grad_norm": 0.3488348722457886, "learning_rate": 9.997912931490771e-05, "loss": 0.0463, "step": 4700 }, { "grad_norm": 0.45770204067230225, "learning_rate": 9.9978527928067e-05, "loss": 0.0405, "step": 4710 }, { "grad_norm": 0.3162718117237091, "learning_rate": 9.997791800126277e-05, "loss": 0.0343, "step": 4720 }, { "grad_norm": 0.3393849730491638, "learning_rate": 9.997729953459927e-05, "loss": 0.0372, "step": 4730 }, { "grad_norm": 0.3142169117927551, "learning_rate": 9.997667252818214e-05, "loss": 0.0377, "step": 4740 }, { "grad_norm": 0.455417662858963, "learning_rate": 9.997603698211855e-05, "loss": 0.0355, "step": 4750 }, { "grad_norm": 0.4108242094516754, "learning_rate": 9.99753928965171e-05, "loss": 0.0413, "step": 4760 }, { "grad_norm": 0.4228082299232483, "learning_rate": 9.997474027148781e-05, "loss": 0.0449, "step": 4770 }, { "grad_norm": 0.3937317430973053, "learning_rate": 9.997407910714223e-05, "loss": 0.0342, "step": 4780 }, { "grad_norm": 0.2753814160823822, "learning_rate": 9.997340940359332e-05, "loss": 0.0395, "step": 4790 }, { "grad_norm": 0.40873220562934875, "learning_rate": 9.997273116095552e-05, "loss": 0.0364, "step": 4800 }, { "grad_norm": 0.4110976457595825, "learning_rate": 9.997204437934473e-05, "loss": 0.0322, "step": 4810 }, { "grad_norm": 0.354056179523468, "learning_rate": 9.997134905887829e-05, "loss": 0.0346, "step": 4820 }, { "grad_norm": 0.3258845806121826, "learning_rate": 9.997064519967501e-05, "loss": 0.0331, "step": 4830 }, { "grad_norm": 0.37447667121887207, "learning_rate": 9.996993280185517e-05, "loss": 0.0326, "step": 4840 }, { "grad_norm": 0.359535276889801, "learning_rate": 9.99692118655405e-05, "loss": 0.0332, "step": 4850 }, { "grad_norm": 0.38163861632347107, "learning_rate": 9.996848239085417e-05, "loss": 0.0316, "step": 4860 }, { "grad_norm": 0.5058581829071045, "learning_rate": 9.996774437792085e-05, "loss": 0.0378, "step": 4870 }, { "grad_norm": 0.3757554292678833, "learning_rate": 9.996699782686664e-05, "loss": 0.0352, "step": 4880 }, { "grad_norm": 0.34641340374946594, "learning_rate": 9.996624273781909e-05, "loss": 0.0374, "step": 4890 }, { "grad_norm": 0.32302024960517883, "learning_rate": 9.996547911090725e-05, "loss": 0.0353, "step": 4900 }, { "grad_norm": 0.3208027184009552, "learning_rate": 9.996470694626157e-05, "loss": 0.0325, "step": 4910 }, { "grad_norm": 0.30068525671958923, "learning_rate": 9.996392624401403e-05, "loss": 0.0386, "step": 4920 }, { "grad_norm": 0.347551167011261, "learning_rate": 9.996313700429801e-05, "loss": 0.0384, "step": 4930 }, { "grad_norm": 0.32836684584617615, "learning_rate": 9.996233922724836e-05, "loss": 0.0317, "step": 4940 }, { "grad_norm": 0.2974022626876831, "learning_rate": 9.996153291300141e-05, "loss": 0.0365, "step": 4950 }, { "grad_norm": 0.42923039197921753, "learning_rate": 9.996071806169494e-05, "loss": 0.0384, "step": 4960 }, { "grad_norm": 0.3056226074695587, "learning_rate": 9.995989467346817e-05, "loss": 0.0309, "step": 4970 }, { "grad_norm": 0.36673858761787415, "learning_rate": 9.995906274846183e-05, "loss": 0.038, "step": 4980 }, { "grad_norm": 0.4666679799556732, "learning_rate": 9.995822228681803e-05, "loss": 0.0409, "step": 4990 }, { "grad_norm": 0.365070104598999, "learning_rate": 9.99573732886804e-05, "loss": 0.0379, "step": 5000 }, { "grad_norm": 0.26574277877807617, "learning_rate": 9.995651575419402e-05, "loss": 0.0355, "step": 5010 }, { "grad_norm": 0.39820677042007446, "learning_rate": 9.995564968350541e-05, "loss": 0.0387, "step": 5020 }, { "grad_norm": 0.3052152693271637, "learning_rate": 9.995477507676256e-05, "loss": 0.0333, "step": 5030 }, { "grad_norm": 0.3384539484977722, "learning_rate": 9.995389193411493e-05, "loss": 0.032, "step": 5040 }, { "grad_norm": 0.33887213468551636, "learning_rate": 9.995300025571339e-05, "loss": 0.0355, "step": 5050 }, { "grad_norm": 0.3464163541793823, "learning_rate": 9.995210004171034e-05, "loss": 0.0376, "step": 5060 }, { "grad_norm": 0.3302582800388336, "learning_rate": 9.995119129225956e-05, "loss": 0.0348, "step": 5070 }, { "grad_norm": 0.35970214009284973, "learning_rate": 9.995027400751637e-05, "loss": 0.0315, "step": 5080 }, { "grad_norm": 0.2884008288383484, "learning_rate": 9.994934818763751e-05, "loss": 0.0348, "step": 5090 }, { "grad_norm": 0.5651416182518005, "learning_rate": 9.994841383278115e-05, "loss": 0.0366, "step": 5100 }, { "grad_norm": 0.3183937668800354, "learning_rate": 9.994747094310695e-05, "loss": 0.0371, "step": 5110 }, { "grad_norm": 0.2873990535736084, "learning_rate": 9.994651951877604e-05, "loss": 0.0336, "step": 5120 }, { "grad_norm": 0.3406108319759369, "learning_rate": 9.994555955995099e-05, "loss": 0.031, "step": 5130 }, { "grad_norm": 0.3606358766555786, "learning_rate": 9.994459106679581e-05, "loss": 0.0332, "step": 5140 }, { "grad_norm": 0.3142997622489929, "learning_rate": 9.994361403947603e-05, "loss": 0.0336, "step": 5150 }, { "grad_norm": 0.3760536313056946, "learning_rate": 9.994262847815854e-05, "loss": 0.0346, "step": 5160 }, { "grad_norm": 0.3122391700744629, "learning_rate": 9.99416343830118e-05, "loss": 0.031, "step": 5170 }, { "grad_norm": 0.3252725899219513, "learning_rate": 9.994063175420565e-05, "loss": 0.0311, "step": 5180 }, { "grad_norm": 0.3924965560436249, "learning_rate": 9.99396205919114e-05, "loss": 0.0347, "step": 5190 }, { "grad_norm": 0.3171039819717407, "learning_rate": 9.993860089630185e-05, "loss": 0.032, "step": 5200 }, { "grad_norm": 0.3668302297592163, "learning_rate": 9.993757266755123e-05, "loss": 0.0372, "step": 5210 }, { "grad_norm": 0.3569835126399994, "learning_rate": 9.993653590583522e-05, "loss": 0.0403, "step": 5220 }, { "grad_norm": 0.33074331283569336, "learning_rate": 9.993549061133102e-05, "loss": 0.0338, "step": 5230 }, { "grad_norm": 0.33517077565193176, "learning_rate": 9.993443678421719e-05, "loss": 0.0298, "step": 5240 }, { "grad_norm": 0.33033502101898193, "learning_rate": 9.993337442467384e-05, "loss": 0.0334, "step": 5250 }, { "grad_norm": 0.33219602704048157, "learning_rate": 9.993230353288248e-05, "loss": 0.0336, "step": 5260 }, { "grad_norm": 0.49784958362579346, "learning_rate": 9.993122410902608e-05, "loss": 0.0333, "step": 5270 }, { "grad_norm": 0.5342609882354736, "learning_rate": 9.993013615328912e-05, "loss": 0.0465, "step": 5280 }, { "grad_norm": 0.3908643424510956, "learning_rate": 9.992903966585747e-05, "loss": 0.0433, "step": 5290 }, { "grad_norm": 0.3100164532661438, "learning_rate": 9.992793464691852e-05, "loss": 0.0431, "step": 5300 }, { "grad_norm": 0.38851526379585266, "learning_rate": 9.992682109666105e-05, "loss": 0.0437, "step": 5310 }, { "grad_norm": 0.3801441490650177, "learning_rate": 9.992569901527538e-05, "loss": 0.0439, "step": 5320 }, { "grad_norm": 0.336963951587677, "learning_rate": 9.99245684029532e-05, "loss": 0.0347, "step": 5330 }, { "grad_norm": 0.33969762921333313, "learning_rate": 9.992342925988774e-05, "loss": 0.0376, "step": 5340 }, { "grad_norm": 0.3124370872974396, "learning_rate": 9.992228158627361e-05, "loss": 0.0361, "step": 5350 }, { "grad_norm": 0.46610331535339355, "learning_rate": 9.992112538230693e-05, "loss": 0.0327, "step": 5360 }, { "grad_norm": 0.3500784635543823, "learning_rate": 9.991996064818527e-05, "loss": 0.0327, "step": 5370 }, { "grad_norm": 0.27614477276802063, "learning_rate": 9.991878738410768e-05, "loss": 0.0326, "step": 5380 }, { "grad_norm": 0.311614990234375, "learning_rate": 9.991760559027457e-05, "loss": 0.0316, "step": 5390 }, { "grad_norm": 0.3738064169883728, "learning_rate": 9.991641526688793e-05, "loss": 0.0308, "step": 5400 }, { "grad_norm": 0.539320170879364, "learning_rate": 9.991521641415113e-05, "loss": 0.0422, "step": 5410 }, { "grad_norm": 0.3526660203933716, "learning_rate": 9.991400903226904e-05, "loss": 0.0409, "step": 5420 }, { "grad_norm": 0.4708975553512573, "learning_rate": 9.991279312144794e-05, "loss": 0.0455, "step": 5430 }, { "grad_norm": 0.34634849429130554, "learning_rate": 9.991156868189564e-05, "loss": 0.0527, "step": 5440 }, { "grad_norm": 0.42472490668296814, "learning_rate": 9.991033571382131e-05, "loss": 0.0508, "step": 5450 }, { "grad_norm": 0.3765963613986969, "learning_rate": 9.990909421743569e-05, "loss": 0.0398, "step": 5460 }, { "grad_norm": 0.37628206610679626, "learning_rate": 9.990784419295085e-05, "loss": 0.0376, "step": 5470 }, { "grad_norm": 0.37842443585395813, "learning_rate": 9.990658564058044e-05, "loss": 0.0418, "step": 5480 }, { "grad_norm": 0.270518958568573, "learning_rate": 9.990531856053948e-05, "loss": 0.0396, "step": 5490 }, { "grad_norm": 0.40595611929893494, "learning_rate": 9.99040429530445e-05, "loss": 0.0345, "step": 5500 }, { "grad_norm": 0.38575878739356995, "learning_rate": 9.990275881831346e-05, "loss": 0.04, "step": 5510 }, { "grad_norm": 0.43990910053253174, "learning_rate": 9.990146615656577e-05, "loss": 0.0408, "step": 5520 }, { "grad_norm": 0.2804604172706604, "learning_rate": 9.990016496802233e-05, "loss": 0.0362, "step": 5530 }, { "grad_norm": 0.37825530767440796, "learning_rate": 9.989885525290548e-05, "loss": 0.037, "step": 5540 }, { "grad_norm": 0.39570584893226624, "learning_rate": 9.989753701143897e-05, "loss": 0.047, "step": 5550 }, { "grad_norm": 0.4060802757740021, "learning_rate": 9.989621024384812e-05, "loss": 0.0453, "step": 5560 }, { "grad_norm": 0.44655120372772217, "learning_rate": 9.989487495035959e-05, "loss": 0.0407, "step": 5570 }, { "grad_norm": 0.3579506278038025, "learning_rate": 9.989353113120156e-05, "loss": 0.0334, "step": 5580 }, { "grad_norm": 0.3092242181301117, "learning_rate": 9.989217878660366e-05, "loss": 0.0354, "step": 5590 }, { "grad_norm": 0.7281389832496643, "learning_rate": 9.989081791679695e-05, "loss": 0.0381, "step": 5600 }, { "grad_norm": 0.39230021834373474, "learning_rate": 9.988944852201397e-05, "loss": 0.0434, "step": 5610 }, { "grad_norm": 0.3048754036426544, "learning_rate": 9.988807060248873e-05, "loss": 0.0391, "step": 5620 }, { "grad_norm": 0.33723023533821106, "learning_rate": 9.988668415845665e-05, "loss": 0.032, "step": 5630 }, { "grad_norm": 0.37975576519966125, "learning_rate": 9.988528919015466e-05, "loss": 0.037, "step": 5640 }, { "grad_norm": 0.23965099453926086, "learning_rate": 9.988388569782112e-05, "loss": 0.036, "step": 5650 }, { "grad_norm": 0.3485466241836548, "learning_rate": 9.988247368169583e-05, "loss": 0.0408, "step": 5660 }, { "grad_norm": 0.4039449989795685, "learning_rate": 9.988105314202007e-05, "loss": 0.0367, "step": 5670 }, { "grad_norm": 0.3235609233379364, "learning_rate": 9.987962407903659e-05, "loss": 0.0373, "step": 5680 }, { "grad_norm": 0.27321696281433105, "learning_rate": 9.987818649298957e-05, "loss": 0.0371, "step": 5690 }, { "grad_norm": 0.32204267382621765, "learning_rate": 9.987674038412465e-05, "loss": 0.0363, "step": 5700 }, { "grad_norm": 0.5109764337539673, "learning_rate": 9.987528575268891e-05, "loss": 0.0342, "step": 5710 }, { "grad_norm": 0.330104261636734, "learning_rate": 9.987382259893095e-05, "loss": 0.0347, "step": 5720 }, { "grad_norm": 0.3546832203865051, "learning_rate": 9.987235092310074e-05, "loss": 0.0372, "step": 5730 }, { "grad_norm": 0.26066920161247253, "learning_rate": 9.987087072544978e-05, "loss": 0.0386, "step": 5740 }, { "grad_norm": 0.27211469411849976, "learning_rate": 9.9869382006231e-05, "loss": 0.0341, "step": 5750 }, { "grad_norm": 0.32011687755584717, "learning_rate": 9.986788476569875e-05, "loss": 0.0314, "step": 5760 }, { "grad_norm": 0.2572726607322693, "learning_rate": 9.986637900410887e-05, "loss": 0.0346, "step": 5770 }, { "grad_norm": 0.3319950997829437, "learning_rate": 9.986486472171869e-05, "loss": 0.0328, "step": 5780 }, { "grad_norm": 0.32415324449539185, "learning_rate": 9.986334191878692e-05, "loss": 0.0355, "step": 5790 }, { "grad_norm": 0.3183695673942566, "learning_rate": 9.986181059557378e-05, "loss": 0.032, "step": 5800 }, { "grad_norm": 0.34653428196907043, "learning_rate": 9.986027075234094e-05, "loss": 0.0314, "step": 5810 }, { "grad_norm": 0.3007928133010864, "learning_rate": 9.985872238935152e-05, "loss": 0.0324, "step": 5820 }, { "grad_norm": 0.3183874487876892, "learning_rate": 9.985716550687008e-05, "loss": 0.036, "step": 5830 }, { "grad_norm": 0.34281739592552185, "learning_rate": 9.985560010516264e-05, "loss": 0.0328, "step": 5840 }, { "grad_norm": 0.36475664377212524, "learning_rate": 9.985402618449668e-05, "loss": 0.0325, "step": 5850 }, { "grad_norm": 0.33416691422462463, "learning_rate": 9.985244374514118e-05, "loss": 0.0358, "step": 5860 }, { "grad_norm": 0.33051103353500366, "learning_rate": 9.985085278736651e-05, "loss": 0.035, "step": 5870 }, { "grad_norm": 0.37805914878845215, "learning_rate": 9.984925331144452e-05, "loss": 0.0337, "step": 5880 }, { "grad_norm": 0.2656953036785126, "learning_rate": 9.984764531764851e-05, "loss": 0.0327, "step": 5890 }, { "grad_norm": 0.38959717750549316, "learning_rate": 9.984602880625326e-05, "loss": 0.0329, "step": 5900 }, { "grad_norm": 0.4199771285057068, "learning_rate": 9.9844403777535e-05, "loss": 0.0351, "step": 5910 }, { "grad_norm": 0.40378522872924805, "learning_rate": 9.984277023177135e-05, "loss": 0.0321, "step": 5920 }, { "grad_norm": 0.2689759135246277, "learning_rate": 9.984112816924148e-05, "loss": 0.0288, "step": 5930 }, { "grad_norm": 0.3292727470397949, "learning_rate": 9.983947759022596e-05, "loss": 0.0327, "step": 5940 }, { "grad_norm": 0.33387938141822815, "learning_rate": 9.983781849500682e-05, "loss": 0.0297, "step": 5950 }, { "grad_norm": 0.35696136951446533, "learning_rate": 9.98361508838676e-05, "loss": 0.0309, "step": 5960 }, { "grad_norm": 0.27351662516593933, "learning_rate": 9.98344747570932e-05, "loss": 0.0334, "step": 5970 }, { "grad_norm": 0.2618231475353241, "learning_rate": 9.983279011497004e-05, "loss": 0.0282, "step": 5980 }, { "grad_norm": 0.3550776243209839, "learning_rate": 9.983109695778596e-05, "loss": 0.0317, "step": 5990 }, { "grad_norm": 0.35351601243019104, "learning_rate": 9.982939528583032e-05, "loss": 0.0325, "step": 6000 }, { "grad_norm": 0.28928929567337036, "learning_rate": 9.982768509939385e-05, "loss": 0.0295, "step": 6010 }, { "grad_norm": 0.3197455108165741, "learning_rate": 9.982596639876879e-05, "loss": 0.0288, "step": 6020 }, { "grad_norm": 0.29459190368652344, "learning_rate": 9.982423918424881e-05, "loss": 0.0275, "step": 6030 }, { "grad_norm": 0.24794085323810577, "learning_rate": 9.982250345612908e-05, "loss": 0.0305, "step": 6040 }, { "grad_norm": 0.2776229679584503, "learning_rate": 9.982075921470611e-05, "loss": 0.0315, "step": 6050 }, { "grad_norm": 0.294268399477005, "learning_rate": 9.981900646027802e-05, "loss": 0.0281, "step": 6060 }, { "grad_norm": 0.3230022192001343, "learning_rate": 9.981724519314425e-05, "loss": 0.0324, "step": 6070 }, { "grad_norm": 0.3841678500175476, "learning_rate": 9.981547541360581e-05, "loss": 0.0347, "step": 6080 }, { "grad_norm": 0.3651130497455597, "learning_rate": 9.981369712196508e-05, "loss": 0.0276, "step": 6090 }, { "grad_norm": 0.30709022283554077, "learning_rate": 9.981191031852592e-05, "loss": 0.0349, "step": 6100 }, { "grad_norm": 0.2331516146659851, "learning_rate": 9.981011500359362e-05, "loss": 0.0327, "step": 6110 }, { "grad_norm": 0.25019049644470215, "learning_rate": 9.9808311177475e-05, "loss": 0.0354, "step": 6120 }, { "grad_norm": 0.30261775851249695, "learning_rate": 9.980649884047826e-05, "loss": 0.0304, "step": 6130 }, { "grad_norm": 0.32881611585617065, "learning_rate": 9.980467799291307e-05, "loss": 0.0297, "step": 6140 }, { "grad_norm": 0.2921956479549408, "learning_rate": 9.980284863509058e-05, "loss": 0.0324, "step": 6150 }, { "grad_norm": 0.30418339371681213, "learning_rate": 9.980101076732334e-05, "loss": 0.03, "step": 6160 }, { "grad_norm": 0.3600499629974365, "learning_rate": 9.979916438992544e-05, "loss": 0.034, "step": 6170 }, { "grad_norm": 0.2922988831996918, "learning_rate": 9.979730950321237e-05, "loss": 0.0321, "step": 6180 }, { "grad_norm": 0.3097001314163208, "learning_rate": 9.979544610750104e-05, "loss": 0.027, "step": 6190 }, { "grad_norm": 0.3137573301792145, "learning_rate": 9.97935742031099e-05, "loss": 0.0301, "step": 6200 }, { "grad_norm": 0.29859405755996704, "learning_rate": 9.979169379035878e-05, "loss": 0.0309, "step": 6210 }, { "grad_norm": 0.3536825180053711, "learning_rate": 9.978980486956899e-05, "loss": 0.0328, "step": 6220 }, { "grad_norm": 0.29734253883361816, "learning_rate": 9.978790744106332e-05, "loss": 0.0299, "step": 6230 }, { "grad_norm": 0.3068651854991913, "learning_rate": 9.978600150516594e-05, "loss": 0.0297, "step": 6240 }, { "grad_norm": 0.2734467089176178, "learning_rate": 9.978408706220259e-05, "loss": 0.0343, "step": 6250 }, { "grad_norm": 0.31229695677757263, "learning_rate": 9.978216411250032e-05, "loss": 0.0337, "step": 6260 }, { "grad_norm": 0.31601840257644653, "learning_rate": 9.978023265638778e-05, "loss": 0.0264, "step": 6270 }, { "grad_norm": 0.22858934104442596, "learning_rate": 9.977829269419495e-05, "loss": 0.0279, "step": 6280 }, { "grad_norm": 0.3669416010379791, "learning_rate": 9.977634422625335e-05, "loss": 0.0282, "step": 6290 }, { "grad_norm": 0.2506018877029419, "learning_rate": 9.97743872528959e-05, "loss": 0.0337, "step": 6300 }, { "grad_norm": 0.24194996058940887, "learning_rate": 9.9772421774457e-05, "loss": 0.0334, "step": 6310 }, { "grad_norm": 0.28431206941604614, "learning_rate": 9.977044779127252e-05, "loss": 0.0311, "step": 6320 }, { "grad_norm": 0.32780182361602783, "learning_rate": 9.976846530367971e-05, "loss": 0.0296, "step": 6330 }, { "grad_norm": 0.4018631875514984, "learning_rate": 9.976647431201735e-05, "loss": 0.0326, "step": 6340 }, { "grad_norm": 0.3479234576225281, "learning_rate": 9.976447481662568e-05, "loss": 0.0341, "step": 6350 }, { "grad_norm": 0.3231974244117737, "learning_rate": 9.976246681784629e-05, "loss": 0.0338, "step": 6360 }, { "grad_norm": 0.2708033621311188, "learning_rate": 9.976045031602234e-05, "loss": 0.036, "step": 6370 }, { "grad_norm": 0.350852370262146, "learning_rate": 9.975842531149837e-05, "loss": 0.0361, "step": 6380 }, { "grad_norm": 0.2712596654891968, "learning_rate": 9.975639180462043e-05, "loss": 0.0342, "step": 6390 }, { "grad_norm": 0.3553732931613922, "learning_rate": 9.975434979573596e-05, "loss": 0.0292, "step": 6400 }, { "grad_norm": 0.35133787989616394, "learning_rate": 9.97522992851939e-05, "loss": 0.0367, "step": 6410 }, { "grad_norm": 0.2975817918777466, "learning_rate": 9.975024027334461e-05, "loss": 0.0343, "step": 6420 }, { "grad_norm": 0.29427701234817505, "learning_rate": 9.974817276053993e-05, "loss": 0.0302, "step": 6430 }, { "grad_norm": 0.24519965052604675, "learning_rate": 9.974609674713315e-05, "loss": 0.0315, "step": 6440 }, { "grad_norm": 0.34085026383399963, "learning_rate": 9.9744012233479e-05, "loss": 0.0314, "step": 6450 }, { "grad_norm": 0.37959110736846924, "learning_rate": 9.974191921993366e-05, "loss": 0.0309, "step": 6460 }, { "grad_norm": 0.31928691267967224, "learning_rate": 9.973981770685474e-05, "loss": 0.0298, "step": 6470 }, { "grad_norm": 0.2971026301383972, "learning_rate": 9.97377076946014e-05, "loss": 0.0332, "step": 6480 }, { "grad_norm": 0.25844523310661316, "learning_rate": 9.973558918353412e-05, "loss": 0.0331, "step": 6490 }, { "grad_norm": 0.32475146651268005, "learning_rate": 9.973346217401494e-05, "loss": 0.0303, "step": 6500 }, { "grad_norm": 0.2619224190711975, "learning_rate": 9.973132666640726e-05, "loss": 0.0292, "step": 6510 }, { "grad_norm": 0.3639601767063141, "learning_rate": 9.972918266107602e-05, "loss": 0.0278, "step": 6520 }, { "grad_norm": 0.35530954599380493, "learning_rate": 9.972703015838756e-05, "loss": 0.0286, "step": 6530 }, { "grad_norm": 0.2640044093132019, "learning_rate": 9.97248691587097e-05, "loss": 0.0308, "step": 6540 }, { "grad_norm": 0.33494481444358826, "learning_rate": 9.972269966241166e-05, "loss": 0.0306, "step": 6550 }, { "grad_norm": 0.3184661567211151, "learning_rate": 9.972052166986417e-05, "loss": 0.0328, "step": 6560 }, { "grad_norm": 0.2899337112903595, "learning_rate": 9.971833518143938e-05, "loss": 0.0359, "step": 6570 }, { "grad_norm": 0.2623591125011444, "learning_rate": 9.971614019751093e-05, "loss": 0.0284, "step": 6580 }, { "grad_norm": 0.28621000051498413, "learning_rate": 9.971393671845383e-05, "loss": 0.03, "step": 6590 }, { "grad_norm": 0.29390376806259155, "learning_rate": 9.971172474464464e-05, "loss": 0.0347, "step": 6600 }, { "grad_norm": 0.2554798722267151, "learning_rate": 9.97095042764613e-05, "loss": 0.036, "step": 6610 }, { "grad_norm": 0.37677836418151855, "learning_rate": 9.970727531428324e-05, "loss": 0.0254, "step": 6620 }, { "grad_norm": 0.28950953483581543, "learning_rate": 9.970503785849132e-05, "loss": 0.0344, "step": 6630 }, { "grad_norm": 0.2514253556728363, "learning_rate": 9.970279190946788e-05, "loss": 0.0317, "step": 6640 }, { "grad_norm": 0.3144059479236603, "learning_rate": 9.970053746759667e-05, "loss": 0.0297, "step": 6650 }, { "grad_norm": 0.3406069278717041, "learning_rate": 9.969827453326292e-05, "loss": 0.0316, "step": 6660 }, { "grad_norm": 0.3117373287677765, "learning_rate": 9.969600310685332e-05, "loss": 0.0328, "step": 6670 }, { "grad_norm": 0.26987820863723755, "learning_rate": 9.969372318875596e-05, "loss": 0.0298, "step": 6680 }, { "grad_norm": 0.27957627177238464, "learning_rate": 9.969143477936043e-05, "loss": 0.0323, "step": 6690 }, { "grad_norm": 0.2414153516292572, "learning_rate": 9.968913787905775e-05, "loss": 0.0288, "step": 6700 }, { "grad_norm": 0.4204576909542084, "learning_rate": 9.968683248824045e-05, "loss": 0.0314, "step": 6710 }, { "grad_norm": 0.31073111295700073, "learning_rate": 9.968451860730238e-05, "loss": 0.0303, "step": 6720 }, { "grad_norm": 0.2517956793308258, "learning_rate": 9.968219623663896e-05, "loss": 0.0277, "step": 6730 }, { "grad_norm": 0.3040815591812134, "learning_rate": 9.967986537664702e-05, "loss": 0.0301, "step": 6740 }, { "grad_norm": 0.3620397448539734, "learning_rate": 9.967752602772483e-05, "loss": 0.028, "step": 6750 }, { "grad_norm": 0.313873291015625, "learning_rate": 9.967517819027212e-05, "loss": 0.0275, "step": 6760 }, { "grad_norm": 0.259196013212204, "learning_rate": 9.967282186469009e-05, "loss": 0.027, "step": 6770 }, { "grad_norm": 0.31170669198036194, "learning_rate": 9.967045705138135e-05, "loss": 0.0277, "step": 6780 }, { "grad_norm": 0.26722973585128784, "learning_rate": 9.966808375074998e-05, "loss": 0.0282, "step": 6790 }, { "grad_norm": 0.38341024518013, "learning_rate": 9.966570196320154e-05, "loss": 0.0294, "step": 6800 }, { "grad_norm": 0.30353736877441406, "learning_rate": 9.966331168914299e-05, "loss": 0.0295, "step": 6810 }, { "grad_norm": 0.33119818568229675, "learning_rate": 9.966091292898277e-05, "loss": 0.0325, "step": 6820 }, { "grad_norm": 0.262693852186203, "learning_rate": 9.965850568313076e-05, "loss": 0.0268, "step": 6830 }, { "grad_norm": 0.34299370646476746, "learning_rate": 9.965608995199827e-05, "loss": 0.0286, "step": 6840 }, { "grad_norm": 0.2869609296321869, "learning_rate": 9.965366573599812e-05, "loss": 0.0306, "step": 6850 }, { "grad_norm": 0.31411245465278625, "learning_rate": 9.965123303554453e-05, "loss": 0.0313, "step": 6860 }, { "grad_norm": 0.28400278091430664, "learning_rate": 9.964879185105317e-05, "loss": 0.026, "step": 6870 }, { "grad_norm": 0.38364431262016296, "learning_rate": 9.964634218294119e-05, "loss": 0.0325, "step": 6880 }, { "grad_norm": 0.27736568450927734, "learning_rate": 9.964388403162714e-05, "loss": 0.027, "step": 6890 }, { "grad_norm": 0.285224974155426, "learning_rate": 9.96414173975311e-05, "loss": 0.0289, "step": 6900 }, { "grad_norm": 0.3073465824127197, "learning_rate": 9.963894228107451e-05, "loss": 0.0267, "step": 6910 }, { "grad_norm": 0.35965272784233093, "learning_rate": 9.963645868268032e-05, "loss": 0.0304, "step": 6920 }, { "grad_norm": 0.2485080361366272, "learning_rate": 9.963396660277289e-05, "loss": 0.0315, "step": 6930 }, { "grad_norm": 0.2486938089132309, "learning_rate": 9.963146604177807e-05, "loss": 0.0256, "step": 6940 }, { "grad_norm": 0.31976431608200073, "learning_rate": 9.962895700012311e-05, "loss": 0.0278, "step": 6950 }, { "grad_norm": 0.36299586296081543, "learning_rate": 9.962643947823677e-05, "loss": 0.0308, "step": 6960 }, { "grad_norm": 0.23235608637332916, "learning_rate": 9.962391347654921e-05, "loss": 0.0253, "step": 6970 }, { "grad_norm": 0.3062026798725128, "learning_rate": 9.962137899549204e-05, "loss": 0.0274, "step": 6980 }, { "grad_norm": 0.24825993180274963, "learning_rate": 9.961883603549835e-05, "loss": 0.0266, "step": 6990 }, { "grad_norm": 0.30879825353622437, "learning_rate": 9.961628459700267e-05, "loss": 0.0287, "step": 7000 }, { "grad_norm": 0.29447034001350403, "learning_rate": 9.961372468044095e-05, "loss": 0.0331, "step": 7010 }, { "grad_norm": 0.30963701009750366, "learning_rate": 9.961115628625062e-05, "loss": 0.0303, "step": 7020 }, { "grad_norm": 0.28570908308029175, "learning_rate": 9.960857941487056e-05, "loss": 0.0273, "step": 7030 }, { "grad_norm": 0.26770713925361633, "learning_rate": 9.960599406674106e-05, "loss": 0.0264, "step": 7040 }, { "grad_norm": 0.26823270320892334, "learning_rate": 9.960340024230393e-05, "loss": 0.0272, "step": 7050 }, { "grad_norm": 0.2758004367351532, "learning_rate": 9.960079794200232e-05, "loss": 0.0281, "step": 7060 }, { "grad_norm": 0.3007809817790985, "learning_rate": 9.959818716628096e-05, "loss": 0.026, "step": 7070 }, { "grad_norm": 0.21521875262260437, "learning_rate": 9.95955679155859e-05, "loss": 0.0256, "step": 7080 }, { "grad_norm": 0.3185882568359375, "learning_rate": 9.959294019036472e-05, "loss": 0.0284, "step": 7090 }, { "grad_norm": 0.25881436467170715, "learning_rate": 9.959030399106646e-05, "loss": 0.0232, "step": 7100 }, { "grad_norm": 0.36083823442459106, "learning_rate": 9.958765931814153e-05, "loss": 0.029, "step": 7110 }, { "grad_norm": 0.27585265040397644, "learning_rate": 9.958500617204184e-05, "loss": 0.0295, "step": 7120 }, { "grad_norm": 0.2656874358654022, "learning_rate": 9.958234455322075e-05, "loss": 0.0259, "step": 7130 }, { "grad_norm": 0.2779974043369293, "learning_rate": 9.957967446213308e-05, "loss": 0.0295, "step": 7140 }, { "grad_norm": 0.24876749515533447, "learning_rate": 9.957699589923501e-05, "loss": 0.0245, "step": 7150 }, { "grad_norm": 0.32461634278297424, "learning_rate": 9.957430886498431e-05, "loss": 0.0275, "step": 7160 }, { "grad_norm": 0.24160906672477722, "learning_rate": 9.957161335984008e-05, "loss": 0.0253, "step": 7170 }, { "grad_norm": 0.2900542616844177, "learning_rate": 9.956890938426291e-05, "loss": 0.0301, "step": 7180 }, { "grad_norm": 0.2683820426464081, "learning_rate": 9.956619693871482e-05, "loss": 0.026, "step": 7190 }, { "grad_norm": 0.2961862087249756, "learning_rate": 9.956347602365934e-05, "loss": 0.0269, "step": 7200 }, { "grad_norm": 0.28537508845329285, "learning_rate": 9.956074663956135e-05, "loss": 0.0295, "step": 7210 }, { "grad_norm": 0.32538455724716187, "learning_rate": 9.955800878688726e-05, "loss": 0.0283, "step": 7220 }, { "grad_norm": 0.46540752053260803, "learning_rate": 9.955526246610489e-05, "loss": 0.0311, "step": 7230 }, { "grad_norm": 0.2894623279571533, "learning_rate": 9.955250767768349e-05, "loss": 0.0276, "step": 7240 }, { "grad_norm": 0.29666247963905334, "learning_rate": 9.95497444220938e-05, "loss": 0.0281, "step": 7250 }, { "grad_norm": 0.40074822306632996, "learning_rate": 9.954697269980797e-05, "loss": 0.0285, "step": 7260 }, { "grad_norm": 0.2923004925251007, "learning_rate": 9.954419251129962e-05, "loss": 0.0268, "step": 7270 }, { "grad_norm": 0.30607151985168457, "learning_rate": 9.95414038570438e-05, "loss": 0.0238, "step": 7280 }, { "grad_norm": 0.2913547158241272, "learning_rate": 9.953860673751703e-05, "loss": 0.0292, "step": 7290 }, { "grad_norm": 0.3000546991825104, "learning_rate": 9.953580115319725e-05, "loss": 0.0288, "step": 7300 }, { "grad_norm": 0.3146362006664276, "learning_rate": 9.953298710456387e-05, "loss": 0.0281, "step": 7310 }, { "grad_norm": 0.2595890164375305, "learning_rate": 9.953016459209771e-05, "loss": 0.0294, "step": 7320 }, { "grad_norm": 0.3815585672855377, "learning_rate": 9.952733361628108e-05, "loss": 0.0312, "step": 7330 }, { "grad_norm": 0.3215445876121521, "learning_rate": 9.952449417759772e-05, "loss": 0.0294, "step": 7340 }, { "grad_norm": 0.28534165024757385, "learning_rate": 9.952164627653279e-05, "loss": 0.0317, "step": 7350 }, { "grad_norm": 0.27830877900123596, "learning_rate": 9.951878991357292e-05, "loss": 0.0267, "step": 7360 }, { "grad_norm": 0.24574284255504608, "learning_rate": 9.951592508920622e-05, "loss": 0.0261, "step": 7370 }, { "grad_norm": 0.2512884736061096, "learning_rate": 9.951305180392219e-05, "loss": 0.0279, "step": 7380 }, { "grad_norm": 0.3274686932563782, "learning_rate": 9.951017005821178e-05, "loss": 0.0284, "step": 7390 }, { "grad_norm": 0.27718251943588257, "learning_rate": 9.95072798525674e-05, "loss": 0.0256, "step": 7400 }, { "grad_norm": 0.34561827778816223, "learning_rate": 9.950438118748293e-05, "loss": 0.0303, "step": 7410 }, { "grad_norm": 0.1997852474451065, "learning_rate": 9.950147406345366e-05, "loss": 0.0271, "step": 7420 }, { "grad_norm": 0.22861960530281067, "learning_rate": 9.949855848097635e-05, "loss": 0.034, "step": 7430 }, { "grad_norm": 0.2882862389087677, "learning_rate": 9.949563444054916e-05, "loss": 0.0304, "step": 7440 }, { "grad_norm": 0.27287787199020386, "learning_rate": 9.949270194267178e-05, "loss": 0.0278, "step": 7450 }, { "grad_norm": 0.26233503222465515, "learning_rate": 9.948976098784526e-05, "loss": 0.0294, "step": 7460 }, { "grad_norm": 0.2923811674118042, "learning_rate": 9.948681157657213e-05, "loss": 0.0317, "step": 7470 }, { "grad_norm": 0.3178064525127411, "learning_rate": 9.948385370935638e-05, "loss": 0.0285, "step": 7480 }, { "grad_norm": 0.2969828248023987, "learning_rate": 9.94808873867034e-05, "loss": 0.0301, "step": 7490 }, { "grad_norm": 0.2780396640300751, "learning_rate": 9.947791260912009e-05, "loss": 0.0287, "step": 7500 }, { "grad_norm": 0.27495118975639343, "learning_rate": 9.947492937711474e-05, "loss": 0.0253, "step": 7510 }, { "grad_norm": 0.21375812590122223, "learning_rate": 9.947193769119707e-05, "loss": 0.0247, "step": 7520 }, { "grad_norm": 0.3567165732383728, "learning_rate": 9.946893755187834e-05, "loss": 0.0278, "step": 7530 }, { "grad_norm": 0.2643527090549469, "learning_rate": 9.946592895967115e-05, "loss": 0.0275, "step": 7540 }, { "grad_norm": 0.28180697560310364, "learning_rate": 9.94629119150896e-05, "loss": 0.0271, "step": 7550 }, { "grad_norm": 0.3120986521244049, "learning_rate": 9.94598864186492e-05, "loss": 0.0288, "step": 7560 }, { "grad_norm": 0.29780784249305725, "learning_rate": 9.945685247086696e-05, "loss": 0.0253, "step": 7570 }, { "grad_norm": 0.367081880569458, "learning_rate": 9.945381007226129e-05, "loss": 0.0296, "step": 7580 }, { "grad_norm": 0.23549754917621613, "learning_rate": 9.945075922335203e-05, "loss": 0.0254, "step": 7590 }, { "grad_norm": 0.25413188338279724, "learning_rate": 9.944769992466049e-05, "loss": 0.0266, "step": 7600 }, { "grad_norm": 0.2546578347682953, "learning_rate": 9.944463217670945e-05, "loss": 0.0265, "step": 7610 }, { "grad_norm": 0.31086114048957825, "learning_rate": 9.944155598002307e-05, "loss": 0.0277, "step": 7620 }, { "grad_norm": 0.2706267535686493, "learning_rate": 9.943847133512701e-05, "loss": 0.025, "step": 7630 }, { "grad_norm": 0.33484533429145813, "learning_rate": 9.943537824254834e-05, "loss": 0.0322, "step": 7640 }, { "grad_norm": 0.3511607348918915, "learning_rate": 9.943227670281559e-05, "loss": 0.0289, "step": 7650 }, { "grad_norm": 0.2645726203918457, "learning_rate": 9.942916671645873e-05, "loss": 0.0236, "step": 7660 }, { "grad_norm": 0.2771812081336975, "learning_rate": 9.942604828400916e-05, "loss": 0.0276, "step": 7670 }, { "grad_norm": 0.2716110348701477, "learning_rate": 9.942292140599975e-05, "loss": 0.0286, "step": 7680 }, { "grad_norm": 0.2959499955177307, "learning_rate": 9.94197860829648e-05, "loss": 0.0271, "step": 7690 }, { "grad_norm": 0.3206130564212799, "learning_rate": 9.941664231544004e-05, "loss": 0.0271, "step": 7700 }, { "grad_norm": 0.23999905586242676, "learning_rate": 9.941349010396264e-05, "loss": 0.029, "step": 7710 }, { "grad_norm": 0.36157453060150146, "learning_rate": 9.941032944907125e-05, "loss": 0.0235, "step": 7720 }, { "grad_norm": 0.3076336681842804, "learning_rate": 9.940716035130596e-05, "loss": 0.0268, "step": 7730 }, { "grad_norm": 0.31218400597572327, "learning_rate": 9.940398281120821e-05, "loss": 0.025, "step": 7740 }, { "grad_norm": 0.23467285931110382, "learning_rate": 9.940079682932102e-05, "loss": 0.027, "step": 7750 }, { "grad_norm": 0.2895459532737732, "learning_rate": 9.939760240618877e-05, "loss": 0.0283, "step": 7760 }, { "grad_norm": 0.24530920386314392, "learning_rate": 9.939439954235729e-05, "loss": 0.0251, "step": 7770 }, { "grad_norm": 0.28225305676460266, "learning_rate": 9.939118823837387e-05, "loss": 0.0266, "step": 7780 }, { "grad_norm": 0.30382809042930603, "learning_rate": 9.938796849478725e-05, "loss": 0.0244, "step": 7790 }, { "grad_norm": 0.36795222759246826, "learning_rate": 9.938474031214755e-05, "loss": 0.0263, "step": 7800 }, { "grad_norm": 0.303459107875824, "learning_rate": 9.938150369100643e-05, "loss": 0.0276, "step": 7810 }, { "grad_norm": 0.2890806794166565, "learning_rate": 9.93782586319169e-05, "loss": 0.0252, "step": 7820 }, { "grad_norm": 0.23668508231639862, "learning_rate": 9.937500513543348e-05, "loss": 0.0251, "step": 7830 }, { "grad_norm": 0.28377673029899597, "learning_rate": 9.937174320211207e-05, "loss": 0.0275, "step": 7840 }, { "grad_norm": 0.27537116408348083, "learning_rate": 9.936847283251009e-05, "loss": 0.0234, "step": 7850 }, { "grad_norm": 0.2807295322418213, "learning_rate": 9.936519402718632e-05, "loss": 0.026, "step": 7860 }, { "grad_norm": 0.27625954151153564, "learning_rate": 9.936190678670102e-05, "loss": 0.0249, "step": 7870 }, { "grad_norm": 0.29720914363861084, "learning_rate": 9.935861111161593e-05, "loss": 0.0283, "step": 7880 }, { "grad_norm": 0.24064810574054718, "learning_rate": 9.935530700249416e-05, "loss": 0.0244, "step": 7890 }, { "grad_norm": 0.2646888196468353, "learning_rate": 9.935199445990028e-05, "loss": 0.0273, "step": 7900 }, { "grad_norm": 0.3217192590236664, "learning_rate": 9.934867348440033e-05, "loss": 0.0313, "step": 7910 }, { "grad_norm": 0.26002615690231323, "learning_rate": 9.934534407656176e-05, "loss": 0.0264, "step": 7920 }, { "grad_norm": 0.35019269585609436, "learning_rate": 9.93420062369535e-05, "loss": 0.027, "step": 7930 }, { "grad_norm": 0.30773407220840454, "learning_rate": 9.933865996614589e-05, "loss": 0.0304, "step": 7940 }, { "grad_norm": 0.24067196249961853, "learning_rate": 9.933530526471068e-05, "loss": 0.0245, "step": 7950 }, { "grad_norm": 0.26752278208732605, "learning_rate": 9.933194213322114e-05, "loss": 0.0267, "step": 7960 }, { "grad_norm": 0.26547080278396606, "learning_rate": 9.932857057225192e-05, "loss": 0.0254, "step": 7970 }, { "grad_norm": 0.24315033853054047, "learning_rate": 9.932519058237912e-05, "loss": 0.027, "step": 7980 }, { "grad_norm": 0.25267869234085083, "learning_rate": 9.932180216418032e-05, "loss": 0.0232, "step": 7990 }, { "grad_norm": 0.30445393919944763, "learning_rate": 9.931840531823446e-05, "loss": 0.0243, "step": 8000 }, { "grad_norm": 0.30330565571784973, "learning_rate": 9.9315000045122e-05, "loss": 0.0268, "step": 8010 }, { "grad_norm": 0.32466721534729004, "learning_rate": 9.931158634542481e-05, "loss": 0.0296, "step": 8020 }, { "grad_norm": 0.22164176404476166, "learning_rate": 9.930816421972617e-05, "loss": 0.0243, "step": 8030 }, { "grad_norm": 0.36087286472320557, "learning_rate": 9.930473366861086e-05, "loss": 0.0271, "step": 8040 }, { "grad_norm": 0.2838979959487915, "learning_rate": 9.930129469266505e-05, "loss": 0.0247, "step": 8050 }, { "grad_norm": 0.29753413796424866, "learning_rate": 9.929784729247638e-05, "loss": 0.0234, "step": 8060 }, { "grad_norm": 0.2416101098060608, "learning_rate": 9.929439146863389e-05, "loss": 0.0262, "step": 8070 }, { "grad_norm": 0.3213789165019989, "learning_rate": 9.92909272217281e-05, "loss": 0.0269, "step": 8080 }, { "grad_norm": 0.25728368759155273, "learning_rate": 9.928745455235097e-05, "loss": 0.0248, "step": 8090 }, { "grad_norm": 0.3163979649543762, "learning_rate": 9.928397346109588e-05, "loss": 0.0274, "step": 8100 }, { "grad_norm": 0.22557896375656128, "learning_rate": 9.928048394855762e-05, "loss": 0.0324, "step": 8110 }, { "grad_norm": 0.23745207488536835, "learning_rate": 9.92769860153325e-05, "loss": 0.0279, "step": 8120 }, { "grad_norm": 0.2884468734264374, "learning_rate": 9.927347966201819e-05, "loss": 0.0272, "step": 8130 }, { "grad_norm": 0.308135986328125, "learning_rate": 9.926996488921383e-05, "loss": 0.0277, "step": 8140 }, { "grad_norm": 0.2116638571023941, "learning_rate": 9.926644169752001e-05, "loss": 0.027, "step": 8150 }, { "grad_norm": 0.24709001183509827, "learning_rate": 9.926291008753875e-05, "loss": 0.0272, "step": 8160 }, { "grad_norm": 0.2528913617134094, "learning_rate": 9.92593700598735e-05, "loss": 0.0257, "step": 8170 }, { "grad_norm": 0.3068314492702484, "learning_rate": 9.925582161512915e-05, "loss": 0.0279, "step": 8180 }, { "grad_norm": 0.2585375905036926, "learning_rate": 9.925226475391205e-05, "loss": 0.0266, "step": 8190 }, { "grad_norm": 0.25951501727104187, "learning_rate": 9.924869947682993e-05, "loss": 0.0221, "step": 8200 }, { "grad_norm": 0.27614569664001465, "learning_rate": 9.924512578449204e-05, "loss": 0.0259, "step": 8210 }, { "grad_norm": 0.2889922559261322, "learning_rate": 9.924154367750901e-05, "loss": 0.0248, "step": 8220 }, { "grad_norm": 0.3857645094394684, "learning_rate": 9.923795315649293e-05, "loss": 0.0253, "step": 8230 }, { "grad_norm": 0.24734961986541748, "learning_rate": 9.92343542220573e-05, "loss": 0.0237, "step": 8240 }, { "grad_norm": 0.2392263114452362, "learning_rate": 9.92307468748171e-05, "loss": 0.0235, "step": 8250 }, { "grad_norm": 0.2578246593475342, "learning_rate": 9.922713111538873e-05, "loss": 0.0246, "step": 8260 }, { "grad_norm": 0.2863660156726837, "learning_rate": 9.922350694439003e-05, "loss": 0.0291, "step": 8270 }, { "grad_norm": 0.28136977553367615, "learning_rate": 9.921987436244024e-05, "loss": 0.0276, "step": 8280 }, { "grad_norm": 0.3421774208545685, "learning_rate": 9.921623337016008e-05, "loss": 0.0266, "step": 8290 }, { "grad_norm": 0.3205151855945587, "learning_rate": 9.921258396817172e-05, "loss": 0.0247, "step": 8300 }, { "grad_norm": 0.2364971935749054, "learning_rate": 9.920892615709874e-05, "loss": 0.0265, "step": 8310 }, { "grad_norm": 0.31118547916412354, "learning_rate": 9.920525993756612e-05, "loss": 0.0259, "step": 8320 }, { "grad_norm": 0.2455354481935501, "learning_rate": 9.920158531020036e-05, "loss": 0.0239, "step": 8330 }, { "grad_norm": 0.2753344774246216, "learning_rate": 9.919790227562933e-05, "loss": 0.0229, "step": 8340 }, { "grad_norm": 0.19379620254039764, "learning_rate": 9.919421083448237e-05, "loss": 0.0234, "step": 8350 }, { "grad_norm": 0.2797226309776306, "learning_rate": 9.919051098739022e-05, "loss": 0.0223, "step": 8360 }, { "grad_norm": 0.3285495936870575, "learning_rate": 9.918680273498514e-05, "loss": 0.0234, "step": 8370 }, { "grad_norm": 0.21963045001029968, "learning_rate": 9.918308607790072e-05, "loss": 0.0254, "step": 8380 }, { "grad_norm": 0.24825063347816467, "learning_rate": 9.917936101677205e-05, "loss": 0.023, "step": 8390 }, { "grad_norm": 0.23198597133159637, "learning_rate": 9.917562755223564e-05, "loss": 0.0213, "step": 8400 }, { "grad_norm": 0.29441413283348083, "learning_rate": 9.917188568492944e-05, "loss": 0.023, "step": 8410 }, { "grad_norm": 0.33521556854248047, "learning_rate": 9.916813541549283e-05, "loss": 0.0226, "step": 8420 }, { "grad_norm": 0.27306699752807617, "learning_rate": 9.916437674456663e-05, "loss": 0.0236, "step": 8430 }, { "grad_norm": 0.2572407126426697, "learning_rate": 9.916060967279308e-05, "loss": 0.025, "step": 8440 }, { "grad_norm": 0.2757345736026764, "learning_rate": 9.91568342008159e-05, "loss": 0.0256, "step": 8450 }, { "grad_norm": 0.27204757928848267, "learning_rate": 9.915305032928019e-05, "loss": 0.0255, "step": 8460 }, { "grad_norm": 0.2737986147403717, "learning_rate": 9.914925805883253e-05, "loss": 0.0256, "step": 8470 }, { "grad_norm": 0.2522967457771301, "learning_rate": 9.914545739012088e-05, "loss": 0.0222, "step": 8480 }, { "grad_norm": 0.3264315128326416, "learning_rate": 9.91416483237947e-05, "loss": 0.0228, "step": 8490 }, { "grad_norm": 0.18571530282497406, "learning_rate": 9.913783086050485e-05, "loss": 0.0236, "step": 8500 }, { "grad_norm": 0.2695249617099762, "learning_rate": 9.913400500090364e-05, "loss": 0.0228, "step": 8510 }, { "grad_norm": 0.31031712889671326, "learning_rate": 9.913017074564479e-05, "loss": 0.0266, "step": 8520 }, { "grad_norm": 0.31138724088668823, "learning_rate": 9.912632809538348e-05, "loss": 0.0282, "step": 8530 }, { "grad_norm": 0.27916982769966125, "learning_rate": 9.912247705077629e-05, "loss": 0.0279, "step": 8540 }, { "grad_norm": 0.2655140459537506, "learning_rate": 9.911861761248127e-05, "loss": 0.0254, "step": 8550 }, { "grad_norm": 0.30672046542167664, "learning_rate": 9.91147497811579e-05, "loss": 0.0267, "step": 8560 }, { "grad_norm": 0.28690674901008606, "learning_rate": 9.911087355746709e-05, "loss": 0.0226, "step": 8570 }, { "grad_norm": 0.30461496114730835, "learning_rate": 9.910698894207117e-05, "loss": 0.0277, "step": 8580 }, { "grad_norm": 0.2472553849220276, "learning_rate": 9.910309593563392e-05, "loss": 0.0241, "step": 8590 }, { "grad_norm": 0.22687065601348877, "learning_rate": 9.909919453882057e-05, "loss": 0.0233, "step": 8600 }, { "grad_norm": 0.3084118366241455, "learning_rate": 9.90952847522977e-05, "loss": 0.0229, "step": 8610 }, { "grad_norm": 0.2489946037530899, "learning_rate": 9.909136657673346e-05, "loss": 0.0245, "step": 8620 }, { "grad_norm": 0.2646307051181793, "learning_rate": 9.908744001279731e-05, "loss": 0.0214, "step": 8630 }, { "grad_norm": 0.27661463618278503, "learning_rate": 9.90835050611602e-05, "loss": 0.0225, "step": 8640 }, { "grad_norm": 0.24778582155704498, "learning_rate": 9.90795617224945e-05, "loss": 0.0239, "step": 8650 }, { "grad_norm": 0.24043375253677368, "learning_rate": 9.907560999747405e-05, "loss": 0.0257, "step": 8660 }, { "grad_norm": 0.24620282649993896, "learning_rate": 9.907164988677408e-05, "loss": 0.0258, "step": 8670 }, { "grad_norm": 0.2638263702392578, "learning_rate": 9.906768139107124e-05, "loss": 0.0268, "step": 8680 }, { "grad_norm": 0.18517261743545532, "learning_rate": 9.906370451104367e-05, "loss": 0.0248, "step": 8690 }, { "grad_norm": 0.20575949549674988, "learning_rate": 9.905971924737088e-05, "loss": 0.0223, "step": 8700 }, { "grad_norm": 0.2666829228401184, "learning_rate": 9.905572560073387e-05, "loss": 0.0236, "step": 8710 }, { "grad_norm": 0.23316988348960876, "learning_rate": 9.905172357181501e-05, "loss": 0.0206, "step": 8720 }, { "grad_norm": 0.2713091969490051, "learning_rate": 9.904771316129817e-05, "loss": 0.0256, "step": 8730 }, { "grad_norm": 0.23055215179920197, "learning_rate": 9.904369436986862e-05, "loss": 0.021, "step": 8740 }, { "grad_norm": 0.2421201765537262, "learning_rate": 9.903966719821303e-05, "loss": 0.0267, "step": 8750 }, { "grad_norm": 0.2748860716819763, "learning_rate": 9.903563164701956e-05, "loss": 0.0234, "step": 8760 }, { "grad_norm": 0.3484538197517395, "learning_rate": 9.903158771697778e-05, "loss": 0.0239, "step": 8770 }, { "grad_norm": 0.3013627529144287, "learning_rate": 9.902753540877867e-05, "loss": 0.0353, "step": 8780 }, { "grad_norm": 0.2657149136066437, "learning_rate": 9.902347472311466e-05, "loss": 0.0346, "step": 8790 }, { "grad_norm": 0.3137989342212677, "learning_rate": 9.901940566067962e-05, "loss": 0.0336, "step": 8800 }, { "grad_norm": 0.28809085488319397, "learning_rate": 9.901532822216883e-05, "loss": 0.027, "step": 8810 }, { "grad_norm": 0.2964083254337311, "learning_rate": 9.901124240827904e-05, "loss": 0.0278, "step": 8820 }, { "grad_norm": 0.26803508400917053, "learning_rate": 9.900714821970835e-05, "loss": 0.0326, "step": 8830 }, { "grad_norm": 0.23100896179676056, "learning_rate": 9.900304565715641e-05, "loss": 0.0295, "step": 8840 }, { "grad_norm": 0.25621575117111206, "learning_rate": 9.899893472132419e-05, "loss": 0.0298, "step": 8850 }, { "grad_norm": 0.28260454535484314, "learning_rate": 9.899481541291415e-05, "loss": 0.0272, "step": 8860 }, { "grad_norm": 0.30929091572761536, "learning_rate": 9.899068773263016e-05, "loss": 0.0306, "step": 8870 }, { "grad_norm": 0.2894105613231659, "learning_rate": 9.898655168117754e-05, "loss": 0.0279, "step": 8880 }, { "grad_norm": 0.2720106244087219, "learning_rate": 9.898240725926302e-05, "loss": 0.0289, "step": 8890 }, { "grad_norm": 0.28604769706726074, "learning_rate": 9.897825446759478e-05, "loss": 0.0261, "step": 8900 }, { "grad_norm": 0.29150834679603577, "learning_rate": 9.897409330688241e-05, "loss": 0.0285, "step": 8910 }, { "grad_norm": 0.29342901706695557, "learning_rate": 9.896992377783692e-05, "loss": 0.0292, "step": 8920 }, { "grad_norm": 0.916714072227478, "learning_rate": 9.89657458811708e-05, "loss": 0.0298, "step": 8930 }, { "grad_norm": 0.5137404799461365, "learning_rate": 9.896155961759792e-05, "loss": 0.0315, "step": 8940 }, { "grad_norm": 0.5105929374694824, "learning_rate": 9.895736498783361e-05, "loss": 0.0379, "step": 8950 }, { "grad_norm": 0.26307860016822815, "learning_rate": 9.895316199259462e-05, "loss": 0.0343, "step": 8960 }, { "grad_norm": 0.273598313331604, "learning_rate": 9.894895063259909e-05, "loss": 0.0305, "step": 8970 }, { "grad_norm": 0.2468273788690567, "learning_rate": 9.894473090856667e-05, "loss": 0.0287, "step": 8980 }, { "grad_norm": 0.23699112236499786, "learning_rate": 9.894050282121839e-05, "loss": 0.0286, "step": 8990 }, { "grad_norm": 0.19794486463069916, "learning_rate": 9.893626637127668e-05, "loss": 0.03, "step": 9000 }, { "grad_norm": 0.2348492294549942, "learning_rate": 9.893202155946546e-05, "loss": 0.0295, "step": 9010 }, { "grad_norm": 0.26428455114364624, "learning_rate": 9.892776838651006e-05, "loss": 0.0252, "step": 9020 }, { "grad_norm": 0.23352566361427307, "learning_rate": 9.892350685313722e-05, "loss": 0.0292, "step": 9030 }, { "grad_norm": 0.2168714851140976, "learning_rate": 9.891923696007513e-05, "loss": 0.0279, "step": 9040 }, { "grad_norm": 0.20620207488536835, "learning_rate": 9.891495870805336e-05, "loss": 0.0233, "step": 9050 }, { "grad_norm": 0.25342071056365967, "learning_rate": 9.891067209780298e-05, "loss": 0.0231, "step": 9060 }, { "grad_norm": 0.3405463397502899, "learning_rate": 9.890637713005646e-05, "loss": 0.026, "step": 9070 }, { "grad_norm": 0.27890944480895996, "learning_rate": 9.890207380554767e-05, "loss": 0.0252, "step": 9080 }, { "grad_norm": 0.3066551983356476, "learning_rate": 9.889776212501196e-05, "loss": 0.0253, "step": 9090 }, { "grad_norm": 0.2539692223072052, "learning_rate": 9.889344208918605e-05, "loss": 0.0237, "step": 9100 }, { "grad_norm": 0.30644699931144714, "learning_rate": 9.888911369880812e-05, "loss": 0.0256, "step": 9110 }, { "grad_norm": 0.30765172839164734, "learning_rate": 9.888477695461777e-05, "loss": 0.0258, "step": 9120 }, { "grad_norm": 0.27979224920272827, "learning_rate": 9.888043185735607e-05, "loss": 0.0242, "step": 9130 }, { "grad_norm": 0.22798892855644226, "learning_rate": 9.887607840776542e-05, "loss": 0.0253, "step": 9140 }, { "grad_norm": 0.3119165599346161, "learning_rate": 9.887171660658975e-05, "loss": 0.0231, "step": 9150 }, { "grad_norm": 0.23231817781925201, "learning_rate": 9.886734645457435e-05, "loss": 0.026, "step": 9160 }, { "grad_norm": 0.29335686564445496, "learning_rate": 9.886296795246597e-05, "loss": 0.0273, "step": 9170 }, { "grad_norm": 0.2420799732208252, "learning_rate": 9.885858110101276e-05, "loss": 0.0218, "step": 9180 }, { "grad_norm": 0.2603955566883087, "learning_rate": 9.885418590096434e-05, "loss": 0.0242, "step": 9190 }, { "grad_norm": 0.2752414345741272, "learning_rate": 9.88497823530717e-05, "loss": 0.0245, "step": 9200 }, { "grad_norm": 0.26370489597320557, "learning_rate": 9.884537045808732e-05, "loss": 0.0261, "step": 9210 }, { "grad_norm": 0.304301381111145, "learning_rate": 9.884095021676502e-05, "loss": 0.025, "step": 9220 }, { "grad_norm": 0.23307456076145172, "learning_rate": 9.883652162986017e-05, "loss": 0.0217, "step": 9230 }, { "grad_norm": 0.2704039216041565, "learning_rate": 9.883208469812943e-05, "loss": 0.0239, "step": 9240 }, { "grad_norm": 0.20825834572315216, "learning_rate": 9.882763942233098e-05, "loss": 0.0193, "step": 9250 }, { "grad_norm": 0.22832073271274567, "learning_rate": 9.882318580322441e-05, "loss": 0.0224, "step": 9260 }, { "grad_norm": 0.1604609340429306, "learning_rate": 9.881872384157067e-05, "loss": 0.0229, "step": 9270 }, { "grad_norm": 0.3027438223361969, "learning_rate": 9.881425353813225e-05, "loss": 0.0246, "step": 9280 }, { "grad_norm": 0.21989217400550842, "learning_rate": 9.880977489367296e-05, "loss": 0.022, "step": 9290 }, { "grad_norm": 0.2532685399055481, "learning_rate": 9.88052879089581e-05, "loss": 0.0231, "step": 9300 }, { "grad_norm": 0.24579070508480072, "learning_rate": 9.880079258475434e-05, "loss": 0.0203, "step": 9310 }, { "grad_norm": 0.19162872433662415, "learning_rate": 9.879628892182985e-05, "loss": 0.0211, "step": 9320 }, { "grad_norm": 0.2623721957206726, "learning_rate": 9.879177692095416e-05, "loss": 0.0227, "step": 9330 }, { "grad_norm": 0.19645681977272034, "learning_rate": 9.878725658289825e-05, "loss": 0.0208, "step": 9340 }, { "grad_norm": 0.20000097155570984, "learning_rate": 9.878272790843454e-05, "loss": 0.0238, "step": 9350 }, { "grad_norm": 0.24102862179279327, "learning_rate": 9.877819089833682e-05, "loss": 0.0209, "step": 9360 }, { "grad_norm": 0.2086973339319229, "learning_rate": 9.877364555338038e-05, "loss": 0.0195, "step": 9370 }, { "grad_norm": 0.21675536036491394, "learning_rate": 9.876909187434186e-05, "loss": 0.0216, "step": 9380 }, { "grad_norm": 0.23227964341640472, "learning_rate": 9.876452986199939e-05, "loss": 0.0186, "step": 9390 }, { "grad_norm": 0.24947337806224823, "learning_rate": 9.875995951713248e-05, "loss": 0.0217, "step": 9400 }, { "grad_norm": 0.280333548784256, "learning_rate": 9.875538084052207e-05, "loss": 0.0181, "step": 9410 }, { "grad_norm": 0.24935245513916016, "learning_rate": 9.875079383295053e-05, "loss": 0.0216, "step": 9420 }, { "grad_norm": 0.30426475405693054, "learning_rate": 9.874619849520167e-05, "loss": 0.0251, "step": 9430 }, { "grad_norm": 0.2473452091217041, "learning_rate": 9.874159482806069e-05, "loss": 0.0223, "step": 9440 }, { "grad_norm": 0.20086221396923065, "learning_rate": 9.873698283231426e-05, "loss": 0.0249, "step": 9450 }, { "grad_norm": 0.26358622312545776, "learning_rate": 9.87323625087504e-05, "loss": 0.0202, "step": 9460 }, { "grad_norm": 0.2208942472934723, "learning_rate": 9.872773385815863e-05, "loss": 0.0196, "step": 9470 }, { "grad_norm": 0.2879709005355835, "learning_rate": 9.872309688132986e-05, "loss": 0.0236, "step": 9480 }, { "grad_norm": 0.1933714598417282, "learning_rate": 9.871845157905639e-05, "loss": 0.0212, "step": 9490 }, { "grad_norm": 0.24055039882659912, "learning_rate": 9.871379795213201e-05, "loss": 0.0191, "step": 9500 }, { "grad_norm": 0.2551043927669525, "learning_rate": 9.87091360013519e-05, "loss": 0.0222, "step": 9510 }, { "grad_norm": 0.1939961463212967, "learning_rate": 9.870446572751262e-05, "loss": 0.0203, "step": 9520 }, { "grad_norm": 0.2842876613140106, "learning_rate": 9.869978713141224e-05, "loss": 0.0165, "step": 9530 }, { "grad_norm": 0.22121430933475494, "learning_rate": 9.869510021385016e-05, "loss": 0.0202, "step": 9540 }, { "grad_norm": 0.2658786177635193, "learning_rate": 9.869040497562727e-05, "loss": 0.0275, "step": 9550 }, { "grad_norm": 0.25379979610443115, "learning_rate": 9.868570141754587e-05, "loss": 0.0234, "step": 9560 }, { "grad_norm": 0.21855024993419647, "learning_rate": 9.868098954040965e-05, "loss": 0.022, "step": 9570 }, { "grad_norm": 0.2306457906961441, "learning_rate": 9.867626934502374e-05, "loss": 0.0246, "step": 9580 }, { "grad_norm": 0.27480506896972656, "learning_rate": 9.86715408321947e-05, "loss": 0.0237, "step": 9590 }, { "grad_norm": 0.2073504477739334, "learning_rate": 9.86668040027305e-05, "loss": 0.0194, "step": 9600 }, { "grad_norm": 0.43400704860687256, "learning_rate": 9.866205885744053e-05, "loss": 0.0224, "step": 9610 }, { "grad_norm": 0.285745769739151, "learning_rate": 9.865730539713563e-05, "loss": 0.0196, "step": 9620 }, { "grad_norm": 0.2310657650232315, "learning_rate": 9.8652543622628e-05, "loss": 0.0208, "step": 9630 }, { "grad_norm": 0.23762522637844086, "learning_rate": 9.864777353473132e-05, "loss": 0.0254, "step": 9640 }, { "grad_norm": 0.2312537282705307, "learning_rate": 9.864299513426068e-05, "loss": 0.0248, "step": 9650 }, { "grad_norm": 0.2112126350402832, "learning_rate": 9.863820842203254e-05, "loss": 0.0229, "step": 9660 }, { "grad_norm": 0.2804097831249237, "learning_rate": 9.863341339886483e-05, "loss": 0.0174, "step": 9670 }, { "grad_norm": 0.26027771830558777, "learning_rate": 9.86286100655769e-05, "loss": 0.0185, "step": 9680 }, { "grad_norm": 0.24180901050567627, "learning_rate": 9.862379842298953e-05, "loss": 0.0224, "step": 9690 }, { "grad_norm": 0.23606981337070465, "learning_rate": 9.861897847192485e-05, "loss": 0.0236, "step": 9700 }, { "grad_norm": 0.19816818833351135, "learning_rate": 9.86141502132065e-05, "loss": 0.0221, "step": 9710 }, { "grad_norm": 0.2033417969942093, "learning_rate": 9.860931364765946e-05, "loss": 0.0218, "step": 9720 }, { "grad_norm": 0.2882249653339386, "learning_rate": 9.860446877611021e-05, "loss": 0.0213, "step": 9730 }, { "grad_norm": 0.16273915767669678, "learning_rate": 9.859961559938655e-05, "loss": 0.0232, "step": 9740 }, { "grad_norm": 0.2762828469276428, "learning_rate": 9.85947541183178e-05, "loss": 0.0253, "step": 9750 }, { "grad_norm": 0.22010695934295654, "learning_rate": 9.858988433373463e-05, "loss": 0.0205, "step": 9760 }, { "grad_norm": 0.25412461161613464, "learning_rate": 9.858500624646918e-05, "loss": 0.0207, "step": 9770 }, { "grad_norm": 0.28532978892326355, "learning_rate": 9.858011985735497e-05, "loss": 0.0195, "step": 9780 }, { "grad_norm": 0.22160057723522186, "learning_rate": 9.857522516722693e-05, "loss": 0.0189, "step": 9790 }, { "grad_norm": 0.2851022779941559, "learning_rate": 9.857032217692145e-05, "loss": 0.0202, "step": 9800 }, { "grad_norm": 0.263643741607666, "learning_rate": 9.856541088727631e-05, "loss": 0.0283, "step": 9810 }, { "grad_norm": 0.23191837966442108, "learning_rate": 9.856049129913072e-05, "loss": 0.0238, "step": 9820 }, { "grad_norm": 0.2297893464565277, "learning_rate": 9.85555634133253e-05, "loss": 0.0234, "step": 9830 }, { "grad_norm": 0.15178857743740082, "learning_rate": 9.855062723070208e-05, "loss": 0.023, "step": 9840 }, { "grad_norm": 0.23723258078098297, "learning_rate": 9.854568275210454e-05, "loss": 0.0246, "step": 9850 }, { "grad_norm": 0.2019588202238083, "learning_rate": 9.854072997837754e-05, "loss": 0.0208, "step": 9860 }, { "grad_norm": 0.2040122151374817, "learning_rate": 9.853576891036737e-05, "loss": 0.0175, "step": 9870 }, { "grad_norm": 0.2566801607608795, "learning_rate": 9.853079954892177e-05, "loss": 0.0191, "step": 9880 }, { "grad_norm": 0.26289334893226624, "learning_rate": 9.852582189488983e-05, "loss": 0.0205, "step": 9890 }, { "grad_norm": 0.24759557843208313, "learning_rate": 9.852083594912212e-05, "loss": 0.0192, "step": 9900 }, { "grad_norm": 0.2343183010816574, "learning_rate": 9.851584171247058e-05, "loss": 0.0207, "step": 9910 }, { "grad_norm": 0.20029698312282562, "learning_rate": 9.851083918578863e-05, "loss": 0.02, "step": 9920 }, { "grad_norm": 0.19576194882392883, "learning_rate": 9.850582836993103e-05, "loss": 0.0213, "step": 9930 }, { "grad_norm": 0.2866073548793793, "learning_rate": 9.850080926575397e-05, "loss": 0.017, "step": 9940 }, { "grad_norm": 0.22228680551052094, "learning_rate": 9.849578187411515e-05, "loss": 0.0247, "step": 9950 }, { "grad_norm": 0.2259514033794403, "learning_rate": 9.849074619587354e-05, "loss": 0.0207, "step": 9960 }, { "grad_norm": 0.20802931487560272, "learning_rate": 9.848570223188964e-05, "loss": 0.02, "step": 9970 }, { "grad_norm": 0.2059260606765747, "learning_rate": 9.848064998302531e-05, "loss": 0.0173, "step": 9980 }, { "grad_norm": 0.2724813222885132, "learning_rate": 9.847558945014386e-05, "loss": 0.0223, "step": 9990 }, { "grad_norm": 0.2516973316669464, "learning_rate": 9.847052063410996e-05, "loss": 0.0213, "step": 10000 }, { "grad_norm": 0.1711120456457138, "learning_rate": 9.846544353578977e-05, "loss": 0.0191, "step": 10010 }, { "grad_norm": 0.21590742468833923, "learning_rate": 9.846035815605081e-05, "loss": 0.0209, "step": 10020 }, { "grad_norm": 0.23845860362052917, "learning_rate": 9.845526449576204e-05, "loss": 0.0216, "step": 10030 }, { "grad_norm": 0.179756298661232, "learning_rate": 9.845016255579383e-05, "loss": 0.0187, "step": 10040 }, { "grad_norm": 0.21693266928195953, "learning_rate": 9.844505233701794e-05, "loss": 0.0183, "step": 10050 }, { "grad_norm": 0.1711118519306183, "learning_rate": 9.843993384030757e-05, "loss": 0.0166, "step": 10060 }, { "grad_norm": 0.2330809086561203, "learning_rate": 9.843480706653737e-05, "loss": 0.018, "step": 10070 }, { "grad_norm": 0.22201579809188843, "learning_rate": 9.84296720165833e-05, "loss": 0.0252, "step": 10080 }, { "grad_norm": 0.25466352701187134, "learning_rate": 9.842452869132286e-05, "loss": 0.0206, "step": 10090 }, { "grad_norm": 0.18024522066116333, "learning_rate": 9.841937709163489e-05, "loss": 0.018, "step": 10100 }, { "grad_norm": 0.21271534264087677, "learning_rate": 9.841421721839962e-05, "loss": 0.0168, "step": 10110 }, { "grad_norm": 0.21224893629550934, "learning_rate": 9.840904907249879e-05, "loss": 0.0196, "step": 10120 }, { "grad_norm": 0.23806385695934296, "learning_rate": 9.840387265481545e-05, "loss": 0.0207, "step": 10130 }, { "grad_norm": 0.20288971066474915, "learning_rate": 9.839868796623411e-05, "loss": 0.0179, "step": 10140 }, { "grad_norm": 0.26396238803863525, "learning_rate": 9.839349500764072e-05, "loss": 0.0177, "step": 10150 }, { "grad_norm": 0.2912447452545166, "learning_rate": 9.83882937799226e-05, "loss": 0.0198, "step": 10160 }, { "grad_norm": 0.2866447865962982, "learning_rate": 9.838308428396849e-05, "loss": 0.0199, "step": 10170 }, { "grad_norm": 0.3966692388057709, "learning_rate": 9.837786652066854e-05, "loss": 0.0221, "step": 10180 }, { "grad_norm": 0.3079579174518585, "learning_rate": 9.837264049091437e-05, "loss": 0.0203, "step": 10190 }, { "grad_norm": 0.27611035108566284, "learning_rate": 9.836740619559893e-05, "loss": 0.0214, "step": 10200 }, { "grad_norm": 0.2958243191242218, "learning_rate": 9.836216363561659e-05, "loss": 0.0239, "step": 10210 }, { "grad_norm": 0.271915465593338, "learning_rate": 9.835691281186322e-05, "loss": 0.0249, "step": 10220 }, { "grad_norm": 0.2126733660697937, "learning_rate": 9.8351653725236e-05, "loss": 0.0192, "step": 10230 }, { "grad_norm": 0.22515985369682312, "learning_rate": 9.83463863766336e-05, "loss": 0.019, "step": 10240 }, { "grad_norm": 0.3273051381111145, "learning_rate": 9.834111076695602e-05, "loss": 0.0185, "step": 10250 }, { "grad_norm": 0.230808824300766, "learning_rate": 9.833582689710477e-05, "loss": 0.0241, "step": 10260 }, { "grad_norm": 0.19384004175662994, "learning_rate": 9.833053476798268e-05, "loss": 0.0215, "step": 10270 }, { "grad_norm": 0.2567552626132965, "learning_rate": 9.832523438049404e-05, "loss": 0.0229, "step": 10280 }, { "grad_norm": 0.23947404325008392, "learning_rate": 9.831992573554454e-05, "loss": 0.0188, "step": 10290 }, { "grad_norm": 0.19747987389564514, "learning_rate": 9.831460883404128e-05, "loss": 0.0214, "step": 10300 }, { "grad_norm": 0.25315743684768677, "learning_rate": 9.830928367689278e-05, "loss": 0.0195, "step": 10310 }, { "grad_norm": 0.26678869128227234, "learning_rate": 9.830395026500896e-05, "loss": 0.0192, "step": 10320 }, { "grad_norm": 0.2124910056591034, "learning_rate": 9.829860859930115e-05, "loss": 0.0209, "step": 10330 }, { "grad_norm": 0.21115513145923615, "learning_rate": 9.829325868068212e-05, "loss": 0.0195, "step": 10340 }, { "grad_norm": 0.2525855004787445, "learning_rate": 9.8287900510066e-05, "loss": 0.0257, "step": 10350 }, { "grad_norm": 0.2590568959712982, "learning_rate": 9.828253408836834e-05, "loss": 0.0199, "step": 10360 }, { "grad_norm": 0.2405785471200943, "learning_rate": 9.827715941650615e-05, "loss": 0.02, "step": 10370 }, { "grad_norm": 0.19365017116069794, "learning_rate": 9.82717764953978e-05, "loss": 0.0172, "step": 10380 }, { "grad_norm": 0.25648847222328186, "learning_rate": 9.826638532596308e-05, "loss": 0.0196, "step": 10390 }, { "grad_norm": 0.13857011497020721, "learning_rate": 9.82609859091232e-05, "loss": 0.0192, "step": 10400 }, { "grad_norm": 0.19567935168743134, "learning_rate": 9.825557824580076e-05, "loss": 0.0176, "step": 10410 }, { "grad_norm": 0.2667001783847809, "learning_rate": 9.82501623369198e-05, "loss": 0.017, "step": 10420 }, { "grad_norm": 0.3059537410736084, "learning_rate": 9.824473818340574e-05, "loss": 0.0206, "step": 10430 }, { "grad_norm": 0.1855013370513916, "learning_rate": 9.823930578618541e-05, "loss": 0.018, "step": 10440 }, { "grad_norm": 0.22627301514148712, "learning_rate": 9.823386514618709e-05, "loss": 0.0176, "step": 10450 }, { "grad_norm": 0.20103617012500763, "learning_rate": 9.82284162643404e-05, "loss": 0.0192, "step": 10460 }, { "grad_norm": 0.267746239900589, "learning_rate": 9.822295914157642e-05, "loss": 0.0199, "step": 10470 }, { "grad_norm": 0.18647514283657074, "learning_rate": 9.821749377882763e-05, "loss": 0.0187, "step": 10480 }, { "grad_norm": 0.18830452859401703, "learning_rate": 9.821202017702791e-05, "loss": 0.0178, "step": 10490 }, { "grad_norm": 0.3205047845840454, "learning_rate": 9.820653833711253e-05, "loss": 0.02, "step": 10500 }, { "grad_norm": 0.220184788107872, "learning_rate": 9.820104826001822e-05, "loss": 0.0206, "step": 10510 }, { "grad_norm": 0.24337656795978546, "learning_rate": 9.819554994668305e-05, "loss": 0.0217, "step": 10520 }, { "grad_norm": 0.15600962936878204, "learning_rate": 9.819004339804654e-05, "loss": 0.0198, "step": 10530 }, { "grad_norm": 0.19637763500213623, "learning_rate": 9.818452861504961e-05, "loss": 0.0196, "step": 10540 }, { "grad_norm": 0.24630466103553772, "learning_rate": 9.81790055986346e-05, "loss": 0.0196, "step": 10550 }, { "grad_norm": 0.18480509519577026, "learning_rate": 9.817347434974523e-05, "loss": 0.0166, "step": 10560 }, { "grad_norm": 0.25903111696243286, "learning_rate": 9.816793486932664e-05, "loss": 0.0202, "step": 10570 }, { "grad_norm": 0.20073233544826508, "learning_rate": 9.816238715832538e-05, "loss": 0.0168, "step": 10580 }, { "grad_norm": 0.20653970539569855, "learning_rate": 9.815683121768939e-05, "loss": 0.0224, "step": 10590 }, { "grad_norm": 0.22522598505020142, "learning_rate": 9.815126704836804e-05, "loss": 0.02, "step": 10600 }, { "grad_norm": 0.17565545439720154, "learning_rate": 9.81456946513121e-05, "loss": 0.0218, "step": 10610 }, { "grad_norm": 0.27188801765441895, "learning_rate": 9.814011402747373e-05, "loss": 0.0219, "step": 10620 }, { "grad_norm": 0.2824363708496094, "learning_rate": 9.813452517780651e-05, "loss": 0.0181, "step": 10630 }, { "grad_norm": 0.26387089490890503, "learning_rate": 9.81289281032654e-05, "loss": 0.0218, "step": 10640 }, { "grad_norm": 0.1932324320077896, "learning_rate": 9.812332280480683e-05, "loss": 0.0185, "step": 10650 }, { "grad_norm": 0.2855948507785797, "learning_rate": 9.811770928338854e-05, "loss": 0.0192, "step": 10660 }, { "grad_norm": 0.2228630632162094, "learning_rate": 9.811208753996979e-05, "loss": 0.0181, "step": 10670 }, { "grad_norm": 0.19863426685333252, "learning_rate": 9.810645757551113e-05, "loss": 0.0161, "step": 10680 }, { "grad_norm": 0.22575615346431732, "learning_rate": 9.810081939097459e-05, "loss": 0.0184, "step": 10690 }, { "grad_norm": 0.20164375007152557, "learning_rate": 9.809517298732356e-05, "loss": 0.0187, "step": 10700 }, { "grad_norm": 0.19213874638080597, "learning_rate": 9.80895183655229e-05, "loss": 0.0181, "step": 10710 }, { "grad_norm": 0.23266665637493134, "learning_rate": 9.808385552653877e-05, "loss": 0.0211, "step": 10720 }, { "grad_norm": 0.29015129804611206, "learning_rate": 9.807818447133886e-05, "loss": 0.0216, "step": 10730 }, { "grad_norm": 0.2869773209095001, "learning_rate": 9.807250520089215e-05, "loss": 0.017, "step": 10740 }, { "grad_norm": 0.16994121670722961, "learning_rate": 9.806681771616908e-05, "loss": 0.0183, "step": 10750 }, { "grad_norm": 0.22537025809288025, "learning_rate": 9.80611220181415e-05, "loss": 0.0208, "step": 10760 }, { "grad_norm": 0.2757217288017273, "learning_rate": 9.805541810778264e-05, "loss": 0.0211, "step": 10770 }, { "grad_norm": 0.23106302320957184, "learning_rate": 9.804970598606716e-05, "loss": 0.0198, "step": 10780 }, { "grad_norm": 0.3249918222427368, "learning_rate": 9.804398565397106e-05, "loss": 0.0238, "step": 10790 }, { "grad_norm": 0.3126215636730194, "learning_rate": 9.803825711247183e-05, "loss": 0.0227, "step": 10800 }, { "grad_norm": 0.19394293427467346, "learning_rate": 9.803252036254831e-05, "loss": 0.0167, "step": 10810 }, { "grad_norm": 0.2680874168872833, "learning_rate": 9.802677540518076e-05, "loss": 0.0205, "step": 10820 }, { "grad_norm": 0.21686530113220215, "learning_rate": 9.802102224135081e-05, "loss": 0.021, "step": 10830 }, { "grad_norm": 0.21341703832149506, "learning_rate": 9.801526087204155e-05, "loss": 0.0199, "step": 10840 }, { "grad_norm": 0.21616922318935394, "learning_rate": 9.800949129823743e-05, "loss": 0.0176, "step": 10850 }, { "grad_norm": 0.18263021111488342, "learning_rate": 9.80037135209243e-05, "loss": 0.0211, "step": 10860 }, { "grad_norm": 0.18213416635990143, "learning_rate": 9.799792754108946e-05, "loss": 0.0204, "step": 10870 }, { "grad_norm": 0.19786866009235382, "learning_rate": 9.799213335972152e-05, "loss": 0.0209, "step": 10880 }, { "grad_norm": 0.2660759687423706, "learning_rate": 9.798633097781058e-05, "loss": 0.0216, "step": 10890 }, { "grad_norm": 0.22050009667873383, "learning_rate": 9.79805203963481e-05, "loss": 0.0183, "step": 10900 }, { "grad_norm": 0.23821833729743958, "learning_rate": 9.797470161632697e-05, "loss": 0.0204, "step": 10910 }, { "grad_norm": 0.21293815970420837, "learning_rate": 9.796887463874145e-05, "loss": 0.0208, "step": 10920 }, { "grad_norm": 0.25037381052970886, "learning_rate": 9.796303946458718e-05, "loss": 0.0188, "step": 10930 }, { "grad_norm": 0.2365565001964569, "learning_rate": 9.795719609486127e-05, "loss": 0.0195, "step": 10940 }, { "grad_norm": 0.2157430797815323, "learning_rate": 9.795134453056219e-05, "loss": 0.0154, "step": 10950 }, { "grad_norm": 0.1998830884695053, "learning_rate": 9.794548477268979e-05, "loss": 0.0162, "step": 10960 }, { "grad_norm": 0.2125672698020935, "learning_rate": 9.793961682224537e-05, "loss": 0.0153, "step": 10970 }, { "grad_norm": 0.2017965465784073, "learning_rate": 9.793374068023156e-05, "loss": 0.02, "step": 10980 }, { "grad_norm": 0.20690003037452698, "learning_rate": 9.792785634765247e-05, "loss": 0.0223, "step": 10990 }, { "grad_norm": 0.225722536444664, "learning_rate": 9.792196382551357e-05, "loss": 0.0184, "step": 11000 }, { "grad_norm": 0.19263401627540588, "learning_rate": 9.791606311482171e-05, "loss": 0.0175, "step": 11010 }, { "grad_norm": 0.186325341463089, "learning_rate": 9.791015421658518e-05, "loss": 0.018, "step": 11020 }, { "grad_norm": 0.23567888140678406, "learning_rate": 9.790423713181362e-05, "loss": 0.02, "step": 11030 }, { "grad_norm": 0.17954254150390625, "learning_rate": 9.789831186151814e-05, "loss": 0.0181, "step": 11040 }, { "grad_norm": 0.36484140157699585, "learning_rate": 9.789237840671118e-05, "loss": 0.0196, "step": 11050 }, { "grad_norm": 0.22353847324848175, "learning_rate": 9.78864367684066e-05, "loss": 0.0193, "step": 11060 }, { "grad_norm": 0.30340078473091125, "learning_rate": 9.788048694761968e-05, "loss": 0.0186, "step": 11070 }, { "grad_norm": 0.23403975367546082, "learning_rate": 9.787452894536709e-05, "loss": 0.0204, "step": 11080 }, { "grad_norm": 0.23634079098701477, "learning_rate": 9.786856276266685e-05, "loss": 0.0195, "step": 11090 }, { "grad_norm": 0.21212653815746307, "learning_rate": 9.786258840053845e-05, "loss": 0.0217, "step": 11100 }, { "grad_norm": 0.25128939747810364, "learning_rate": 9.785660586000273e-05, "loss": 0.0185, "step": 11110 }, { "grad_norm": 0.1772269904613495, "learning_rate": 9.785061514208196e-05, "loss": 0.0183, "step": 11120 }, { "grad_norm": 0.16966281831264496, "learning_rate": 9.784461624779977e-05, "loss": 0.019, "step": 11130 }, { "grad_norm": 0.2666122615337372, "learning_rate": 9.783860917818123e-05, "loss": 0.0203, "step": 11140 }, { "grad_norm": 0.1933157742023468, "learning_rate": 9.783259393425277e-05, "loss": 0.0195, "step": 11150 }, { "grad_norm": 0.14423300325870514, "learning_rate": 9.782657051704221e-05, "loss": 0.0144, "step": 11160 }, { "grad_norm": 0.2017732709646225, "learning_rate": 9.782053892757883e-05, "loss": 0.016, "step": 11170 }, { "grad_norm": 0.21582989394664764, "learning_rate": 9.781449916689324e-05, "loss": 0.021, "step": 11180 }, { "grad_norm": 0.2940444052219391, "learning_rate": 9.780845123601746e-05, "loss": 0.0194, "step": 11190 }, { "grad_norm": 0.21112799644470215, "learning_rate": 9.780239513598492e-05, "loss": 0.0185, "step": 11200 }, { "grad_norm": 0.24813689291477203, "learning_rate": 9.779633086783047e-05, "loss": 0.0219, "step": 11210 }, { "grad_norm": 0.17511790990829468, "learning_rate": 9.779025843259031e-05, "loss": 0.0178, "step": 11220 }, { "grad_norm": 0.1952252984046936, "learning_rate": 9.778417783130204e-05, "loss": 0.0185, "step": 11230 }, { "grad_norm": 0.17818281054496765, "learning_rate": 9.777808906500468e-05, "loss": 0.0174, "step": 11240 }, { "grad_norm": 0.2445305585861206, "learning_rate": 9.777199213473862e-05, "loss": 0.0182, "step": 11250 }, { "grad_norm": 0.20802056789398193, "learning_rate": 9.77658870415457e-05, "loss": 0.0217, "step": 11260 }, { "grad_norm": 0.3132916986942291, "learning_rate": 9.775977378646906e-05, "loss": 0.0245, "step": 11270 }, { "grad_norm": 0.16427215933799744, "learning_rate": 9.775365237055331e-05, "loss": 0.018, "step": 11280 }, { "grad_norm": 0.20381340384483337, "learning_rate": 9.774752279484445e-05, "loss": 0.0176, "step": 11290 }, { "grad_norm": 0.23736369609832764, "learning_rate": 9.774138506038984e-05, "loss": 0.0179, "step": 11300 }, { "grad_norm": 0.3011503517627716, "learning_rate": 9.773523916823826e-05, "loss": 0.0208, "step": 11310 }, { "grad_norm": 0.21831920742988586, "learning_rate": 9.772908511943986e-05, "loss": 0.0164, "step": 11320 }, { "grad_norm": 0.23766176402568817, "learning_rate": 9.77229229150462e-05, "loss": 0.0187, "step": 11330 }, { "grad_norm": 0.27557918429374695, "learning_rate": 9.771675255611024e-05, "loss": 0.0194, "step": 11340 }, { "grad_norm": 0.165542870759964, "learning_rate": 9.771057404368632e-05, "loss": 0.0203, "step": 11350 }, { "grad_norm": 0.22856855392456055, "learning_rate": 9.770438737883018e-05, "loss": 0.0154, "step": 11360 }, { "grad_norm": 0.2658199071884155, "learning_rate": 9.769819256259898e-05, "loss": 0.0172, "step": 11370 }, { "grad_norm": 0.20054106414318085, "learning_rate": 9.769198959605119e-05, "loss": 0.0205, "step": 11380 }, { "grad_norm": 0.1841529756784439, "learning_rate": 9.768577848024678e-05, "loss": 0.0181, "step": 11390 }, { "grad_norm": 0.18670594692230225, "learning_rate": 9.767955921624702e-05, "loss": 0.0173, "step": 11400 }, { "grad_norm": 0.2592790424823761, "learning_rate": 9.767333180511465e-05, "loss": 0.0224, "step": 11410 }, { "grad_norm": 0.292689710855484, "learning_rate": 9.766709624791373e-05, "loss": 0.0214, "step": 11420 }, { "grad_norm": 0.32542991638183594, "learning_rate": 9.766085254570975e-05, "loss": 0.0216, "step": 11430 }, { "grad_norm": 0.2777533531188965, "learning_rate": 9.76546006995696e-05, "loss": 0.0212, "step": 11440 }, { "grad_norm": 0.2023884803056717, "learning_rate": 9.764834071056155e-05, "loss": 0.017, "step": 11450 }, { "grad_norm": 0.17225389182567596, "learning_rate": 9.764207257975526e-05, "loss": 0.0209, "step": 11460 }, { "grad_norm": 0.1542554497718811, "learning_rate": 9.763579630822179e-05, "loss": 0.0187, "step": 11470 }, { "grad_norm": 0.22374491393566132, "learning_rate": 9.762951189703356e-05, "loss": 0.0167, "step": 11480 }, { "grad_norm": 0.1513690948486328, "learning_rate": 9.762321934726442e-05, "loss": 0.0164, "step": 11490 }, { "grad_norm": 0.2065834105014801, "learning_rate": 9.761691865998959e-05, "loss": 0.0207, "step": 11500 }, { "grad_norm": 0.21070913970470428, "learning_rate": 9.76106098362857e-05, "loss": 0.0165, "step": 11510 }, { "grad_norm": 0.2510761320590973, "learning_rate": 9.760429287723072e-05, "loss": 0.0198, "step": 11520 }, { "grad_norm": 0.2339111715555191, "learning_rate": 9.759796778390406e-05, "loss": 0.0204, "step": 11530 }, { "grad_norm": 0.16769929230213165, "learning_rate": 9.759163455738653e-05, "loss": 0.0207, "step": 11540 }, { "grad_norm": 0.21335332095623016, "learning_rate": 9.75852931987603e-05, "loss": 0.0161, "step": 11550 }, { "grad_norm": 0.18650223314762115, "learning_rate": 9.757894370910891e-05, "loss": 0.0208, "step": 11560 }, { "grad_norm": 0.1993381828069687, "learning_rate": 9.757258608951733e-05, "loss": 0.0215, "step": 11570 }, { "grad_norm": 0.17651908099651337, "learning_rate": 9.75662203410719e-05, "loss": 0.0182, "step": 11580 }, { "grad_norm": 0.20196574926376343, "learning_rate": 9.755984646486034e-05, "loss": 0.0162, "step": 11590 }, { "grad_norm": 0.17715129256248474, "learning_rate": 9.75534644619718e-05, "loss": 0.0152, "step": 11600 }, { "grad_norm": 0.24702559411525726, "learning_rate": 9.754707433349676e-05, "loss": 0.0177, "step": 11610 }, { "grad_norm": 0.2433990091085434, "learning_rate": 9.754067608052715e-05, "loss": 0.0209, "step": 11620 }, { "grad_norm": 0.2590848505496979, "learning_rate": 9.753426970415622e-05, "loss": 0.018, "step": 11630 }, { "grad_norm": 0.21167349815368652, "learning_rate": 9.752785520547868e-05, "loss": 0.0183, "step": 11640 }, { "grad_norm": 0.19200752675533295, "learning_rate": 9.752143258559056e-05, "loss": 0.0196, "step": 11650 }, { "grad_norm": 0.18293334543704987, "learning_rate": 9.751500184558933e-05, "loss": 0.0183, "step": 11660 }, { "grad_norm": 0.21259409189224243, "learning_rate": 9.750856298657383e-05, "loss": 0.0151, "step": 11670 }, { "grad_norm": 0.173130601644516, "learning_rate": 9.750211600964428e-05, "loss": 0.0158, "step": 11680 }, { "grad_norm": 0.2667452394962311, "learning_rate": 9.749566091590226e-05, "loss": 0.0204, "step": 11690 }, { "grad_norm": 0.2527881860733032, "learning_rate": 9.748919770645083e-05, "loss": 0.021, "step": 11700 }, { "grad_norm": 0.25283315777778625, "learning_rate": 9.748272638239432e-05, "loss": 0.018, "step": 11710 }, { "grad_norm": 0.19260647892951965, "learning_rate": 9.747624694483855e-05, "loss": 0.0215, "step": 11720 }, { "grad_norm": 0.1664441078901291, "learning_rate": 9.746975939489065e-05, "loss": 0.0155, "step": 11730 }, { "grad_norm": 0.17266331613063812, "learning_rate": 9.746326373365918e-05, "loss": 0.0157, "step": 11740 }, { "grad_norm": 0.2845448851585388, "learning_rate": 9.745675996225403e-05, "loss": 0.0174, "step": 11750 }, { "grad_norm": 0.24508334696292877, "learning_rate": 9.745024808178657e-05, "loss": 0.0187, "step": 11760 }, { "grad_norm": 0.2768605947494507, "learning_rate": 9.744372809336947e-05, "loss": 0.0178, "step": 11770 }, { "grad_norm": 0.19388769567012787, "learning_rate": 9.743719999811682e-05, "loss": 0.0197, "step": 11780 }, { "grad_norm": 0.28225967288017273, "learning_rate": 9.743066379714412e-05, "loss": 0.0163, "step": 11790 }, { "grad_norm": 0.27953556180000305, "learning_rate": 9.74241194915682e-05, "loss": 0.0227, "step": 11800 }, { "grad_norm": 0.23703598976135254, "learning_rate": 9.741756708250731e-05, "loss": 0.0177, "step": 11810 }, { "grad_norm": 0.20163145661354065, "learning_rate": 9.741100657108109e-05, "loss": 0.0205, "step": 11820 }, { "grad_norm": 0.1795077621936798, "learning_rate": 9.740443795841054e-05, "loss": 0.0191, "step": 11830 }, { "grad_norm": 0.20922936499118805, "learning_rate": 9.739786124561805e-05, "loss": 0.017, "step": 11840 }, { "grad_norm": 0.21751584112644196, "learning_rate": 9.73912764338274e-05, "loss": 0.0197, "step": 11850 }, { "grad_norm": 0.159014031291008, "learning_rate": 9.738468352416377e-05, "loss": 0.0182, "step": 11860 }, { "grad_norm": 0.24216988682746887, "learning_rate": 9.737808251775369e-05, "loss": 0.0202, "step": 11870 }, { "grad_norm": 0.3710242211818695, "learning_rate": 9.737147341572512e-05, "loss": 0.018, "step": 11880 }, { "grad_norm": 0.16909940540790558, "learning_rate": 9.736485621920735e-05, "loss": 0.018, "step": 11890 }, { "grad_norm": 0.22526642680168152, "learning_rate": 9.735823092933108e-05, "loss": 0.0246, "step": 11900 }, { "grad_norm": 0.15906491875648499, "learning_rate": 9.735159754722838e-05, "loss": 0.0167, "step": 11910 }, { "grad_norm": 0.25703784823417664, "learning_rate": 9.734495607403275e-05, "loss": 0.0197, "step": 11920 }, { "grad_norm": 0.23905475437641144, "learning_rate": 9.733830651087901e-05, "loss": 0.0185, "step": 11930 }, { "grad_norm": 0.18944720923900604, "learning_rate": 9.733164885890338e-05, "loss": 0.0147, "step": 11940 }, { "grad_norm": 0.22467590868473053, "learning_rate": 9.732498311924349e-05, "loss": 0.0217, "step": 11950 }, { "grad_norm": 0.2179846614599228, "learning_rate": 9.731830929303833e-05, "loss": 0.0195, "step": 11960 }, { "grad_norm": 0.2253996878862381, "learning_rate": 9.731162738142827e-05, "loss": 0.0165, "step": 11970 }, { "grad_norm": 0.16181226074695587, "learning_rate": 9.730493738555506e-05, "loss": 0.0197, "step": 11980 }, { "grad_norm": 0.22060738503932953, "learning_rate": 9.729823930656186e-05, "loss": 0.0173, "step": 11990 }, { "grad_norm": 0.21419525146484375, "learning_rate": 9.729153314559316e-05, "loss": 0.0191, "step": 12000 }, { "grad_norm": 0.18322566151618958, "learning_rate": 9.728481890379486e-05, "loss": 0.0216, "step": 12010 }, { "grad_norm": 0.257799357175827, "learning_rate": 9.727809658231428e-05, "loss": 0.0198, "step": 12020 }, { "grad_norm": 0.2891811430454254, "learning_rate": 9.727136618230003e-05, "loss": 0.0186, "step": 12030 }, { "grad_norm": 0.17594243586063385, "learning_rate": 9.726462770490219e-05, "loss": 0.019, "step": 12040 }, { "grad_norm": 0.18036112189292908, "learning_rate": 9.725788115127214e-05, "loss": 0.017, "step": 12050 }, { "grad_norm": 0.20593859255313873, "learning_rate": 9.725112652256274e-05, "loss": 0.0172, "step": 12060 }, { "grad_norm": 0.17295965552330017, "learning_rate": 9.724436381992812e-05, "loss": 0.0179, "step": 12070 }, { "grad_norm": 0.23502680659294128, "learning_rate": 9.723759304452387e-05, "loss": 0.0184, "step": 12080 }, { "grad_norm": 0.13342241942882538, "learning_rate": 9.72308141975069e-05, "loss": 0.0159, "step": 12090 }, { "grad_norm": 0.18883174657821655, "learning_rate": 9.722402728003557e-05, "loss": 0.0167, "step": 12100 }, { "grad_norm": 0.20729127526283264, "learning_rate": 9.721723229326953e-05, "loss": 0.018, "step": 12110 }, { "grad_norm": 0.21123583614826202, "learning_rate": 9.721042923836992e-05, "loss": 0.0196, "step": 12120 }, { "grad_norm": 0.24871809780597687, "learning_rate": 9.720361811649914e-05, "loss": 0.0175, "step": 12130 }, { "grad_norm": 0.20436877012252808, "learning_rate": 9.719679892882106e-05, "loss": 0.0156, "step": 12140 }, { "grad_norm": 0.22756141424179077, "learning_rate": 9.718997167650085e-05, "loss": 0.0182, "step": 12150 }, { "grad_norm": 0.20609159767627716, "learning_rate": 9.718313636070515e-05, "loss": 0.0203, "step": 12160 }, { "grad_norm": 0.20333802700042725, "learning_rate": 9.717629298260192e-05, "loss": 0.0192, "step": 12170 }, { "grad_norm": 0.23279394209384918, "learning_rate": 9.716944154336047e-05, "loss": 0.016, "step": 12180 }, { "grad_norm": 0.26358604431152344, "learning_rate": 9.716258204415157e-05, "loss": 0.0159, "step": 12190 }, { "grad_norm": 0.21278487145900726, "learning_rate": 9.715571448614728e-05, "loss": 0.0204, "step": 12200 }, { "grad_norm": 0.23574970662593842, "learning_rate": 9.71488388705211e-05, "loss": 0.0159, "step": 12210 }, { "grad_norm": 0.233520969748497, "learning_rate": 9.714195519844788e-05, "loss": 0.0175, "step": 12220 }, { "grad_norm": 0.21153955161571503, "learning_rate": 9.713506347110386e-05, "loss": 0.0164, "step": 12230 }, { "grad_norm": 0.14170067012310028, "learning_rate": 9.712816368966662e-05, "loss": 0.0153, "step": 12240 }, { "grad_norm": 0.174855038523674, "learning_rate": 9.712125585531517e-05, "loss": 0.019, "step": 12250 }, { "grad_norm": 0.19380724430084229, "learning_rate": 9.711433996922988e-05, "loss": 0.0177, "step": 12260 }, { "grad_norm": 0.21681122481822968, "learning_rate": 9.710741603259245e-05, "loss": 0.0219, "step": 12270 }, { "grad_norm": 0.14100445806980133, "learning_rate": 9.710048404658603e-05, "loss": 0.0185, "step": 12280 }, { "grad_norm": 0.15663743019104004, "learning_rate": 9.709354401239508e-05, "loss": 0.0174, "step": 12290 }, { "grad_norm": 0.16564902663230896, "learning_rate": 9.708659593120546e-05, "loss": 0.0176, "step": 12300 }, { "grad_norm": 0.18388541042804718, "learning_rate": 9.707963980420443e-05, "loss": 0.0188, "step": 12310 }, { "grad_norm": 0.22790351510047913, "learning_rate": 9.707267563258058e-05, "loss": 0.015, "step": 12320 }, { "grad_norm": 0.20474064350128174, "learning_rate": 9.70657034175239e-05, "loss": 0.0162, "step": 12330 }, { "grad_norm": 0.15375617146492004, "learning_rate": 9.705872316022577e-05, "loss": 0.0209, "step": 12340 }, { "grad_norm": 0.2193571925163269, "learning_rate": 9.705173486187891e-05, "loss": 0.0168, "step": 12350 }, { "grad_norm": 0.2743431627750397, "learning_rate": 9.704473852367741e-05, "loss": 0.023, "step": 12360 }, { "grad_norm": 0.2301097959280014, "learning_rate": 9.70377341468168e-05, "loss": 0.0183, "step": 12370 }, { "grad_norm": 0.198034405708313, "learning_rate": 9.703072173249389e-05, "loss": 0.0194, "step": 12380 }, { "grad_norm": 0.22386199235916138, "learning_rate": 9.702370128190693e-05, "loss": 0.0181, "step": 12390 }, { "grad_norm": 0.19811466336250305, "learning_rate": 9.701667279625552e-05, "loss": 0.0204, "step": 12400 }, { "grad_norm": 0.1844702661037445, "learning_rate": 9.700963627674065e-05, "loss": 0.0181, "step": 12410 }, { "grad_norm": 0.16100110113620758, "learning_rate": 9.700259172456466e-05, "loss": 0.0162, "step": 12420 }, { "grad_norm": 0.21440571546554565, "learning_rate": 9.699553914093124e-05, "loss": 0.0152, "step": 12430 }, { "grad_norm": 0.28007206320762634, "learning_rate": 9.698847852704553e-05, "loss": 0.0185, "step": 12440 }, { "grad_norm": 0.22109797596931458, "learning_rate": 9.6981409884114e-05, "loss": 0.0238, "step": 12450 }, { "grad_norm": 0.23962804675102234, "learning_rate": 9.697433321334443e-05, "loss": 0.0163, "step": 12460 }, { "grad_norm": 0.19990967214107513, "learning_rate": 9.696724851594607e-05, "loss": 0.0219, "step": 12470 }, { "grad_norm": 0.21016570925712585, "learning_rate": 9.696015579312952e-05, "loss": 0.0164, "step": 12480 }, { "grad_norm": 0.19831646978855133, "learning_rate": 9.695305504610668e-05, "loss": 0.0159, "step": 12490 }, { "grad_norm": 0.22672967612743378, "learning_rate": 9.694594627609092e-05, "loss": 0.0142, "step": 12500 }, { "grad_norm": 0.2074083685874939, "learning_rate": 9.693882948429691e-05, "loss": 0.0164, "step": 12510 }, { "grad_norm": 0.21196043491363525, "learning_rate": 9.693170467194071e-05, "loss": 0.0175, "step": 12520 }, { "grad_norm": 0.23073795437812805, "learning_rate": 9.692457184023977e-05, "loss": 0.0206, "step": 12530 }, { "grad_norm": 0.22255592048168182, "learning_rate": 9.691743099041291e-05, "loss": 0.0184, "step": 12540 }, { "grad_norm": 0.20096434652805328, "learning_rate": 9.691028212368027e-05, "loss": 0.0188, "step": 12550 }, { "grad_norm": 0.21855144202709198, "learning_rate": 9.690312524126342e-05, "loss": 0.0155, "step": 12560 }, { "grad_norm": 0.20710411667823792, "learning_rate": 9.689596034438527e-05, "loss": 0.0159, "step": 12570 }, { "grad_norm": 0.3365044891834259, "learning_rate": 9.688878743427012e-05, "loss": 0.0155, "step": 12580 }, { "grad_norm": 0.33305373787879944, "learning_rate": 9.688160651214359e-05, "loss": 0.0168, "step": 12590 }, { "grad_norm": 0.22595250606536865, "learning_rate": 9.687441757923273e-05, "loss": 0.0167, "step": 12600 }, { "grad_norm": 0.2790733873844147, "learning_rate": 9.68672206367659e-05, "loss": 0.0172, "step": 12610 }, { "grad_norm": 0.269096314907074, "learning_rate": 9.686001568597291e-05, "loss": 0.0193, "step": 12620 }, { "grad_norm": 0.20516347885131836, "learning_rate": 9.685280272808486e-05, "loss": 0.017, "step": 12630 }, { "grad_norm": 0.2602628171443939, "learning_rate": 9.684558176433424e-05, "loss": 0.0176, "step": 12640 }, { "grad_norm": 0.221751868724823, "learning_rate": 9.683835279595495e-05, "loss": 0.0166, "step": 12650 }, { "grad_norm": 0.2194516956806183, "learning_rate": 9.683111582418216e-05, "loss": 0.0133, "step": 12660 }, { "grad_norm": 0.27125197649002075, "learning_rate": 9.682387085025254e-05, "loss": 0.0152, "step": 12670 }, { "grad_norm": 0.23843729496002197, "learning_rate": 9.681661787540401e-05, "loss": 0.0199, "step": 12680 }, { "grad_norm": 0.2157173603773117, "learning_rate": 9.680935690087593e-05, "loss": 0.0195, "step": 12690 }, { "grad_norm": 0.28277987241744995, "learning_rate": 9.680208792790901e-05, "loss": 0.0156, "step": 12700 }, { "grad_norm": 0.23129676282405853, "learning_rate": 9.679481095774529e-05, "loss": 0.0148, "step": 12710 }, { "grad_norm": 0.194877490401268, "learning_rate": 9.678752599162822e-05, "loss": 0.016, "step": 12720 }, { "grad_norm": 0.16914793848991394, "learning_rate": 9.678023303080259e-05, "loss": 0.0144, "step": 12730 }, { "grad_norm": 0.24200555682182312, "learning_rate": 9.677293207651459e-05, "loss": 0.0164, "step": 12740 }, { "grad_norm": 0.1947796493768692, "learning_rate": 9.676562313001173e-05, "loss": 0.0172, "step": 12750 }, { "grad_norm": 0.2753598988056183, "learning_rate": 9.675830619254293e-05, "loss": 0.0171, "step": 12760 }, { "grad_norm": 0.20487335324287415, "learning_rate": 9.675098126535843e-05, "loss": 0.0161, "step": 12770 }, { "grad_norm": 0.18296821415424347, "learning_rate": 9.674364834970988e-05, "loss": 0.0195, "step": 12780 }, { "grad_norm": 0.21957097947597504, "learning_rate": 9.673630744685028e-05, "loss": 0.015, "step": 12790 }, { "grad_norm": 0.18184109032154083, "learning_rate": 9.672895855803397e-05, "loss": 0.0162, "step": 12800 }, { "grad_norm": 0.20613276958465576, "learning_rate": 9.672160168451667e-05, "loss": 0.0142, "step": 12810 }, { "grad_norm": 0.24821284413337708, "learning_rate": 9.671423682755549e-05, "loss": 0.0172, "step": 12820 }, { "grad_norm": 0.2004411518573761, "learning_rate": 9.670686398840888e-05, "loss": 0.0158, "step": 12830 }, { "grad_norm": 0.21953096985816956, "learning_rate": 9.669948316833664e-05, "loss": 0.0151, "step": 12840 }, { "grad_norm": 0.28311654925346375, "learning_rate": 9.669209436859997e-05, "loss": 0.0156, "step": 12850 }, { "grad_norm": 0.23425641655921936, "learning_rate": 9.66846975904614e-05, "loss": 0.0171, "step": 12860 }, { "grad_norm": 0.2246619164943695, "learning_rate": 9.667729283518483e-05, "loss": 0.0149, "step": 12870 }, { "grad_norm": 0.18356527388095856, "learning_rate": 9.666988010403557e-05, "loss": 0.0148, "step": 12880 }, { "grad_norm": 0.21706371009349823, "learning_rate": 9.66624593982802e-05, "loss": 0.0167, "step": 12890 }, { "grad_norm": 0.27756163477897644, "learning_rate": 9.665503071918675e-05, "loss": 0.0175, "step": 12900 }, { "grad_norm": 0.25230517983436584, "learning_rate": 9.664759406802456e-05, "loss": 0.0173, "step": 12910 }, { "grad_norm": 0.1992475986480713, "learning_rate": 9.664014944606437e-05, "loss": 0.0171, "step": 12920 }, { "grad_norm": 0.2531428635120392, "learning_rate": 9.663269685457822e-05, "loss": 0.0174, "step": 12930 }, { "grad_norm": 0.21131007373332977, "learning_rate": 9.662523629483962e-05, "loss": 0.0179, "step": 12940 }, { "grad_norm": 0.22108909487724304, "learning_rate": 9.661776776812333e-05, "loss": 0.0147, "step": 12950 }, { "grad_norm": 0.24620850384235382, "learning_rate": 9.661029127570553e-05, "loss": 0.0164, "step": 12960 }, { "grad_norm": 0.23571152985095978, "learning_rate": 9.660280681886373e-05, "loss": 0.015, "step": 12970 }, { "grad_norm": 0.2615745961666107, "learning_rate": 9.659531439887685e-05, "loss": 0.016, "step": 12980 }, { "grad_norm": 0.272494375705719, "learning_rate": 9.658781401702511e-05, "loss": 0.0238, "step": 12990 }, { "grad_norm": 0.2600093185901642, "learning_rate": 9.658030567459015e-05, "loss": 0.0151, "step": 13000 }, { "grad_norm": 0.2021753191947937, "learning_rate": 9.65727893728549e-05, "loss": 0.0185, "step": 13010 }, { "grad_norm": 0.2433447688817978, "learning_rate": 9.656526511310375e-05, "loss": 0.0174, "step": 13020 }, { "grad_norm": 0.15607383847236633, "learning_rate": 9.655773289662233e-05, "loss": 0.0171, "step": 13030 }, { "grad_norm": 0.2107289433479309, "learning_rate": 9.655019272469772e-05, "loss": 0.0211, "step": 13040 }, { "grad_norm": 0.24004800617694855, "learning_rate": 9.654264459861832e-05, "loss": 0.0199, "step": 13050 }, { "grad_norm": 0.16790318489074707, "learning_rate": 9.653508851967391e-05, "loss": 0.0152, "step": 13060 }, { "grad_norm": 0.16406407952308655, "learning_rate": 9.65275244891556e-05, "loss": 0.0142, "step": 13070 }, { "grad_norm": 0.24400785565376282, "learning_rate": 9.651995250835591e-05, "loss": 0.0139, "step": 13080 }, { "grad_norm": 0.23925118148326874, "learning_rate": 9.651237257856862e-05, "loss": 0.017, "step": 13090 }, { "grad_norm": 0.1731141060590744, "learning_rate": 9.6504784701089e-05, "loss": 0.0135, "step": 13100 }, { "grad_norm": 0.21177230775356293, "learning_rate": 9.649718887721357e-05, "loss": 0.0182, "step": 13110 }, { "grad_norm": 0.13695217669010162, "learning_rate": 9.648958510824028e-05, "loss": 0.0148, "step": 13120 }, { "grad_norm": 0.26001688838005066, "learning_rate": 9.648197339546837e-05, "loss": 0.0188, "step": 13130 }, { "grad_norm": 0.17296287417411804, "learning_rate": 9.647435374019851e-05, "loss": 0.0182, "step": 13140 }, { "grad_norm": 0.221943199634552, "learning_rate": 9.646672614373266e-05, "loss": 0.0188, "step": 13150 }, { "grad_norm": 0.2299010008573532, "learning_rate": 9.645909060737418e-05, "loss": 0.0173, "step": 13160 }, { "grad_norm": 0.1939311921596527, "learning_rate": 9.645144713242778e-05, "loss": 0.0188, "step": 13170 }, { "grad_norm": 0.317006915807724, "learning_rate": 9.64437957201995e-05, "loss": 0.0164, "step": 13180 }, { "grad_norm": 0.23751774430274963, "learning_rate": 9.643613637199678e-05, "loss": 0.0182, "step": 13190 }, { "grad_norm": 0.2272309511899948, "learning_rate": 9.642846908912839e-05, "loss": 0.0175, "step": 13200 }, { "grad_norm": 0.20061230659484863, "learning_rate": 9.642079387290444e-05, "loss": 0.0169, "step": 13210 }, { "grad_norm": 0.27405184507369995, "learning_rate": 9.641311072463644e-05, "loss": 0.02, "step": 13220 }, { "grad_norm": 0.21924157440662384, "learning_rate": 9.640541964563722e-05, "loss": 0.0193, "step": 13230 }, { "grad_norm": 0.22439797222614288, "learning_rate": 9.639772063722096e-05, "loss": 0.0199, "step": 13240 }, { "grad_norm": 0.16740712523460388, "learning_rate": 9.639001370070324e-05, "loss": 0.0162, "step": 13250 }, { "grad_norm": 0.2848447263240814, "learning_rate": 9.638229883740095e-05, "loss": 0.0164, "step": 13260 }, { "grad_norm": 0.1655178964138031, "learning_rate": 9.637457604863233e-05, "loss": 0.0139, "step": 13270 }, { "grad_norm": 0.18055404722690582, "learning_rate": 9.636684533571703e-05, "loss": 0.0158, "step": 13280 }, { "grad_norm": 0.16907066106796265, "learning_rate": 9.635910669997599e-05, "loss": 0.0129, "step": 13290 }, { "grad_norm": 0.2086634486913681, "learning_rate": 9.635136014273154e-05, "loss": 0.0171, "step": 13300 }, { "grad_norm": 0.20570798218250275, "learning_rate": 9.634360566530735e-05, "loss": 0.0164, "step": 13310 }, { "grad_norm": 0.23277626931667328, "learning_rate": 9.633584326902845e-05, "loss": 0.0168, "step": 13320 }, { "grad_norm": 0.2055293470621109, "learning_rate": 9.632807295522124e-05, "loss": 0.0181, "step": 13330 }, { "grad_norm": 0.2299887239933014, "learning_rate": 9.632029472521342e-05, "loss": 0.0178, "step": 13340 }, { "grad_norm": 0.17513540387153625, "learning_rate": 9.631250858033409e-05, "loss": 0.0152, "step": 13350 }, { "grad_norm": 0.21192267537117004, "learning_rate": 9.630471452191371e-05, "loss": 0.0159, "step": 13360 }, { "grad_norm": 0.22488000988960266, "learning_rate": 9.629691255128405e-05, "loss": 0.0157, "step": 13370 }, { "grad_norm": 0.2267015129327774, "learning_rate": 9.628910266977825e-05, "loss": 0.0157, "step": 13380 }, { "grad_norm": 0.15946847200393677, "learning_rate": 9.628128487873083e-05, "loss": 0.0139, "step": 13390 }, { "grad_norm": 0.18564943969249725, "learning_rate": 9.627345917947761e-05, "loss": 0.0158, "step": 13400 }, { "grad_norm": 0.2158932089805603, "learning_rate": 9.626562557335579e-05, "loss": 0.0146, "step": 13410 }, { "grad_norm": 0.20434987545013428, "learning_rate": 9.625778406170393e-05, "loss": 0.0141, "step": 13420 }, { "grad_norm": 0.19679981470108032, "learning_rate": 9.624993464586193e-05, "loss": 0.0153, "step": 13430 }, { "grad_norm": 0.24700234830379486, "learning_rate": 9.624207732717105e-05, "loss": 0.018, "step": 13440 }, { "grad_norm": 0.28450003266334534, "learning_rate": 9.623421210697386e-05, "loss": 0.0172, "step": 13450 }, { "grad_norm": 0.23716723918914795, "learning_rate": 9.622633898661434e-05, "loss": 0.0176, "step": 13460 }, { "grad_norm": 0.15249769389629364, "learning_rate": 9.621845796743778e-05, "loss": 0.0208, "step": 13470 }, { "grad_norm": 0.239777609705925, "learning_rate": 9.621056905079082e-05, "loss": 0.0177, "step": 13480 }, { "grad_norm": 0.2301671802997589, "learning_rate": 9.620267223802149e-05, "loss": 0.0147, "step": 13490 }, { "grad_norm": 0.13154667615890503, "learning_rate": 9.619476753047911e-05, "loss": 0.0143, "step": 13500 }, { "grad_norm": 0.17437218129634857, "learning_rate": 9.618685492951438e-05, "loss": 0.0167, "step": 13510 }, { "grad_norm": 0.2238834649324417, "learning_rate": 9.617893443647938e-05, "loss": 0.0143, "step": 13520 }, { "grad_norm": 0.1587059050798416, "learning_rate": 9.617100605272746e-05, "loss": 0.014, "step": 13530 }, { "grad_norm": 0.19997438788414001, "learning_rate": 9.616306977961338e-05, "loss": 0.0128, "step": 13540 }, { "grad_norm": 0.1795455515384674, "learning_rate": 9.615512561849326e-05, "loss": 0.0141, "step": 13550 }, { "grad_norm": 0.18158526718616486, "learning_rate": 9.61471735707245e-05, "loss": 0.0186, "step": 13560 }, { "grad_norm": 0.2222793698310852, "learning_rate": 9.613921363766592e-05, "loss": 0.0164, "step": 13570 }, { "grad_norm": 0.20278845727443695, "learning_rate": 9.613124582067763e-05, "loss": 0.0144, "step": 13580 }, { "grad_norm": 0.15378710627555847, "learning_rate": 9.612327012112112e-05, "loss": 0.0147, "step": 13590 }, { "grad_norm": 0.14758604764938354, "learning_rate": 9.611528654035921e-05, "loss": 0.0128, "step": 13600 }, { "grad_norm": 0.2871081829071045, "learning_rate": 9.610729507975611e-05, "loss": 0.0171, "step": 13610 }, { "grad_norm": 0.22356505692005157, "learning_rate": 9.609929574067731e-05, "loss": 0.0151, "step": 13620 }, { "grad_norm": 0.16390575468540192, "learning_rate": 9.609128852448967e-05, "loss": 0.0155, "step": 13630 }, { "grad_norm": 0.17077480256557465, "learning_rate": 9.608327343256143e-05, "loss": 0.016, "step": 13640 }, { "grad_norm": 0.2534129023551941, "learning_rate": 9.607525046626216e-05, "loss": 0.0172, "step": 13650 }, { "grad_norm": 0.21574094891548157, "learning_rate": 9.606721962696272e-05, "loss": 0.0134, "step": 13660 }, { "grad_norm": 0.20160448551177979, "learning_rate": 9.60591809160354e-05, "loss": 0.0154, "step": 13670 }, { "grad_norm": 0.23362861573696136, "learning_rate": 9.605113433485378e-05, "loss": 0.0157, "step": 13680 }, { "grad_norm": 0.17621834576129913, "learning_rate": 9.604307988479279e-05, "loss": 0.0157, "step": 13690 }, { "grad_norm": 0.22728462517261505, "learning_rate": 9.603501756722876e-05, "loss": 0.0143, "step": 13700 }, { "grad_norm": 0.17789988219738007, "learning_rate": 9.602694738353927e-05, "loss": 0.0166, "step": 13710 }, { "grad_norm": 0.15810461342334747, "learning_rate": 9.601886933510331e-05, "loss": 0.0183, "step": 13720 }, { "grad_norm": 0.19786691665649414, "learning_rate": 9.60107834233012e-05, "loss": 0.0156, "step": 13730 }, { "grad_norm": 0.2015707790851593, "learning_rate": 9.60026896495146e-05, "loss": 0.0175, "step": 13740 }, { "grad_norm": 0.21483348309993744, "learning_rate": 9.599458801512652e-05, "loss": 0.0139, "step": 13750 }, { "grad_norm": 0.23020382225513458, "learning_rate": 9.598647852152129e-05, "loss": 0.015, "step": 13760 }, { "grad_norm": 0.21011820435523987, "learning_rate": 9.597836117008462e-05, "loss": 0.0174, "step": 13770 }, { "grad_norm": 0.2038397639989853, "learning_rate": 9.597023596220356e-05, "loss": 0.0186, "step": 13780 }, { "grad_norm": 0.20482175052165985, "learning_rate": 9.596210289926643e-05, "loss": 0.0142, "step": 13790 }, { "grad_norm": 0.23719441890716553, "learning_rate": 9.5953961982663e-05, "loss": 0.0156, "step": 13800 }, { "grad_norm": 0.17451229691505432, "learning_rate": 9.594581321378431e-05, "loss": 0.0147, "step": 13810 }, { "grad_norm": 0.21020199358463287, "learning_rate": 9.593765659402276e-05, "loss": 0.0162, "step": 13820 }, { "grad_norm": 0.21209287643432617, "learning_rate": 9.59294921247721e-05, "loss": 0.0151, "step": 13830 }, { "grad_norm": 0.18043021857738495, "learning_rate": 9.59213198074274e-05, "loss": 0.0151, "step": 13840 }, { "grad_norm": 0.2136881947517395, "learning_rate": 9.59131396433851e-05, "loss": 0.0159, "step": 13850 }, { "grad_norm": 0.21752209961414337, "learning_rate": 9.590495163404297e-05, "loss": 0.0179, "step": 13860 }, { "grad_norm": 0.17206427454948425, "learning_rate": 9.589675578080009e-05, "loss": 0.0158, "step": 13870 }, { "grad_norm": 0.2825464904308319, "learning_rate": 9.588855208505694e-05, "loss": 0.019, "step": 13880 }, { "grad_norm": 0.22640080749988556, "learning_rate": 9.588034054821529e-05, "loss": 0.0149, "step": 13890 }, { "grad_norm": 0.1546693742275238, "learning_rate": 9.587212117167826e-05, "loss": 0.0153, "step": 13900 }, { "grad_norm": 0.16966703534126282, "learning_rate": 9.586389395685033e-05, "loss": 0.0124, "step": 13910 }, { "grad_norm": 0.21907806396484375, "learning_rate": 9.585565890513733e-05, "loss": 0.014, "step": 13920 }, { "grad_norm": 0.1882125437259674, "learning_rate": 9.584741601794636e-05, "loss": 0.0143, "step": 13930 }, { "grad_norm": 0.14918777346611023, "learning_rate": 9.58391652966859e-05, "loss": 0.0141, "step": 13940 }, { "grad_norm": 0.1717442125082016, "learning_rate": 9.583090674276583e-05, "loss": 0.0153, "step": 13950 }, { "grad_norm": 0.22830110788345337, "learning_rate": 9.582264035759726e-05, "loss": 0.0146, "step": 13960 }, { "grad_norm": 0.20984764397144318, "learning_rate": 9.58143661425927e-05, "loss": 0.0151, "step": 13970 }, { "grad_norm": 0.1682412326335907, "learning_rate": 9.580608409916601e-05, "loss": 0.017, "step": 13980 }, { "grad_norm": 0.2540772259235382, "learning_rate": 9.579779422873233e-05, "loss": 0.0179, "step": 13990 }, { "grad_norm": 0.21260777115821838, "learning_rate": 9.578949653270819e-05, "loss": 0.0217, "step": 14000 }, { "grad_norm": 0.2095702737569809, "learning_rate": 9.578119101251144e-05, "loss": 0.014, "step": 14010 }, { "grad_norm": 0.20080415904521942, "learning_rate": 9.577287766956127e-05, "loss": 0.0155, "step": 14020 }, { "grad_norm": 0.2133273482322693, "learning_rate": 9.57645565052782e-05, "loss": 0.0139, "step": 14030 }, { "grad_norm": 0.2303811013698578, "learning_rate": 9.575622752108407e-05, "loss": 0.0166, "step": 14040 }, { "grad_norm": 0.18406140804290771, "learning_rate": 9.57478907184021e-05, "loss": 0.0171, "step": 14050 }, { "grad_norm": 0.22375258803367615, "learning_rate": 9.573954609865681e-05, "loss": 0.0148, "step": 14060 }, { "grad_norm": 0.19158223271369934, "learning_rate": 9.573119366327408e-05, "loss": 0.016, "step": 14070 }, { "grad_norm": 0.149745374917984, "learning_rate": 9.57228334136811e-05, "loss": 0.0152, "step": 14080 }, { "grad_norm": 0.2590252757072449, "learning_rate": 9.571446535130641e-05, "loss": 0.0182, "step": 14090 }, { "grad_norm": 0.22113032639026642, "learning_rate": 9.570608947757988e-05, "loss": 0.0164, "step": 14100 }, { "grad_norm": 0.22230006754398346, "learning_rate": 9.569770579393274e-05, "loss": 0.0171, "step": 14110 }, { "grad_norm": 0.14310018718242645, "learning_rate": 9.56893143017975e-05, "loss": 0.0154, "step": 14120 }, { "grad_norm": 0.2272331416606903, "learning_rate": 9.568091500260806e-05, "loss": 0.0145, "step": 14130 }, { "grad_norm": 0.1565503180027008, "learning_rate": 9.567250789779961e-05, "loss": 0.0134, "step": 14140 }, { "grad_norm": 0.18439382314682007, "learning_rate": 9.566409298880872e-05, "loss": 0.015, "step": 14150 }, { "grad_norm": 0.2059994488954544, "learning_rate": 9.565567027707326e-05, "loss": 0.0181, "step": 14160 }, { "grad_norm": 0.2621835470199585, "learning_rate": 9.56472397640324e-05, "loss": 0.0168, "step": 14170 }, { "grad_norm": 0.19555886089801788, "learning_rate": 9.563880145112675e-05, "loss": 0.0213, "step": 14180 }, { "grad_norm": 0.19541813433170319, "learning_rate": 9.563035533979814e-05, "loss": 0.0184, "step": 14190 }, { "grad_norm": 0.21775363385677338, "learning_rate": 9.562190143148981e-05, "loss": 0.0199, "step": 14200 }, { "grad_norm": 0.20640799403190613, "learning_rate": 9.561343972764627e-05, "loss": 0.0131, "step": 14210 }, { "grad_norm": 0.21432672441005707, "learning_rate": 9.560497022971343e-05, "loss": 0.0164, "step": 14220 }, { "grad_norm": 0.17763742804527283, "learning_rate": 9.559649293913847e-05, "loss": 0.0137, "step": 14230 }, { "grad_norm": 0.18220598995685577, "learning_rate": 9.558800785736993e-05, "loss": 0.0166, "step": 14240 }, { "grad_norm": 0.15277303755283356, "learning_rate": 9.557951498585767e-05, "loss": 0.0175, "step": 14250 }, { "grad_norm": 0.20066529512405396, "learning_rate": 9.557101432605293e-05, "loss": 0.0164, "step": 14260 }, { "grad_norm": 0.1933862417936325, "learning_rate": 9.556250587940818e-05, "loss": 0.0156, "step": 14270 }, { "grad_norm": 0.19193655252456665, "learning_rate": 9.555398964737734e-05, "loss": 0.0153, "step": 14280 }, { "grad_norm": 0.15999002754688263, "learning_rate": 9.554546563141555e-05, "loss": 0.0164, "step": 14290 }, { "grad_norm": 0.1810068041086197, "learning_rate": 9.553693383297937e-05, "loss": 0.0188, "step": 14300 }, { "grad_norm": 0.21018864214420319, "learning_rate": 9.552839425352663e-05, "loss": 0.0154, "step": 14310 }, { "grad_norm": 0.24803104996681213, "learning_rate": 9.551984689451652e-05, "loss": 0.0158, "step": 14320 }, { "grad_norm": 0.26540014147758484, "learning_rate": 9.551129175740953e-05, "loss": 0.0172, "step": 14330 }, { "grad_norm": 0.18048939108848572, "learning_rate": 9.550272884366754e-05, "loss": 0.015, "step": 14340 }, { "grad_norm": 0.15368875861167908, "learning_rate": 9.549415815475369e-05, "loss": 0.0159, "step": 14350 }, { "grad_norm": 0.15791326761245728, "learning_rate": 9.548557969213247e-05, "loss": 0.0149, "step": 14360 }, { "grad_norm": 0.15805822610855103, "learning_rate": 9.547699345726972e-05, "loss": 0.0135, "step": 14370 }, { "grad_norm": 0.1906329095363617, "learning_rate": 9.546839945163257e-05, "loss": 0.0135, "step": 14380 }, { "grad_norm": 0.24255239963531494, "learning_rate": 9.545979767668953e-05, "loss": 0.0182, "step": 14390 }, { "grad_norm": 0.20048584043979645, "learning_rate": 9.54511881339104e-05, "loss": 0.0178, "step": 14400 }, { "grad_norm": 0.17696967720985413, "learning_rate": 9.54425708247663e-05, "loss": 0.0182, "step": 14410 }, { "grad_norm": 0.16415774822235107, "learning_rate": 9.543394575072972e-05, "loss": 0.0159, "step": 14420 }, { "grad_norm": 0.1445925533771515, "learning_rate": 9.542531291327441e-05, "loss": 0.0148, "step": 14430 }, { "grad_norm": 0.180894136428833, "learning_rate": 9.541667231387552e-05, "loss": 0.0149, "step": 14440 }, { "grad_norm": 0.22182920575141907, "learning_rate": 9.540802395400949e-05, "loss": 0.015, "step": 14450 }, { "grad_norm": 0.19216981530189514, "learning_rate": 9.539936783515406e-05, "loss": 0.0146, "step": 14460 }, { "grad_norm": 0.16386307775974274, "learning_rate": 9.539070395878835e-05, "loss": 0.015, "step": 14470 }, { "grad_norm": 0.1994180679321289, "learning_rate": 9.538203232639277e-05, "loss": 0.0135, "step": 14480 }, { "grad_norm": 0.18411219120025635, "learning_rate": 9.537335293944907e-05, "loss": 0.0154, "step": 14490 }, { "grad_norm": 0.1792108416557312, "learning_rate": 9.536466579944032e-05, "loss": 0.0139, "step": 14500 }, { "grad_norm": 0.2048274576663971, "learning_rate": 9.535597090785091e-05, "loss": 0.0169, "step": 14510 }, { "grad_norm": 0.20736660063266754, "learning_rate": 9.534726826616656e-05, "loss": 0.0133, "step": 14520 }, { "grad_norm": 0.15596400201320648, "learning_rate": 9.53385578758743e-05, "loss": 0.0133, "step": 14530 }, { "grad_norm": 0.13792596757411957, "learning_rate": 9.532983973846252e-05, "loss": 0.0161, "step": 14540 }, { "grad_norm": 0.23419122397899628, "learning_rate": 9.53211138554209e-05, "loss": 0.0136, "step": 14550 }, { "grad_norm": 0.16986069083213806, "learning_rate": 9.531238022824047e-05, "loss": 0.0129, "step": 14560 }, { "grad_norm": 0.17294438183307648, "learning_rate": 9.530363885841355e-05, "loss": 0.015, "step": 14570 }, { "grad_norm": 0.17884811758995056, "learning_rate": 9.52948897474338e-05, "loss": 0.0153, "step": 14580 }, { "grad_norm": 0.21078158915042877, "learning_rate": 9.528613289679622e-05, "loss": 0.0139, "step": 14590 }, { "grad_norm": 0.18815600872039795, "learning_rate": 9.52773683079971e-05, "loss": 0.0128, "step": 14600 }, { "grad_norm": 0.14133097231388092, "learning_rate": 9.526859598253407e-05, "loss": 0.0146, "step": 14610 }, { "grad_norm": 0.21519437432289124, "learning_rate": 9.525981592190609e-05, "loss": 0.018, "step": 14620 }, { "grad_norm": 0.19289042055606842, "learning_rate": 9.525102812761342e-05, "loss": 0.0166, "step": 14630 }, { "grad_norm": 0.163262739777565, "learning_rate": 9.524223260115768e-05, "loss": 0.014, "step": 14640 }, { "grad_norm": 0.1338770091533661, "learning_rate": 9.523342934404175e-05, "loss": 0.0132, "step": 14650 }, { "grad_norm": 0.1701016128063202, "learning_rate": 9.522461835776989e-05, "loss": 0.0169, "step": 14660 }, { "grad_norm": 0.1888282299041748, "learning_rate": 9.521579964384764e-05, "loss": 0.0136, "step": 14670 }, { "grad_norm": 0.17827774584293365, "learning_rate": 9.52069732037819e-05, "loss": 0.0139, "step": 14680 }, { "grad_norm": 0.25751587748527527, "learning_rate": 9.519813903908083e-05, "loss": 0.0151, "step": 14690 }, { "grad_norm": 0.18763551115989685, "learning_rate": 9.5189297151254e-05, "loss": 0.0141, "step": 14700 }, { "grad_norm": 0.17686690390110016, "learning_rate": 9.518044754181218e-05, "loss": 0.0131, "step": 14710 }, { "grad_norm": 0.17783288657665253, "learning_rate": 9.51715902122676e-05, "loss": 0.0153, "step": 14720 }, { "grad_norm": 0.1623096764087677, "learning_rate": 9.516272516413368e-05, "loss": 0.0152, "step": 14730 }, { "grad_norm": 0.26056936383247375, "learning_rate": 9.515385239892525e-05, "loss": 0.0182, "step": 14740 }, { "grad_norm": 0.2277131825685501, "learning_rate": 9.514497191815839e-05, "loss": 0.0146, "step": 14750 }, { "grad_norm": 0.2023935765028, "learning_rate": 9.513608372335055e-05, "loss": 0.0129, "step": 14760 }, { "grad_norm": 0.168275848031044, "learning_rate": 9.512718781602045e-05, "loss": 0.015, "step": 14770 }, { "grad_norm": 0.18079715967178345, "learning_rate": 9.511828419768823e-05, "loss": 0.0141, "step": 14780 }, { "grad_norm": 0.18963082134723663, "learning_rate": 9.510937286987521e-05, "loss": 0.0152, "step": 14790 }, { "grad_norm": 0.20028680562973022, "learning_rate": 9.510045383410408e-05, "loss": 0.015, "step": 14800 }, { "grad_norm": 0.2639130651950836, "learning_rate": 9.509152709189892e-05, "loss": 0.016, "step": 14810 }, { "grad_norm": 0.20473183691501617, "learning_rate": 9.508259264478504e-05, "loss": 0.0168, "step": 14820 }, { "grad_norm": 0.174798846244812, "learning_rate": 9.507365049428909e-05, "loss": 0.0164, "step": 14830 }, { "grad_norm": 0.1652054339647293, "learning_rate": 9.506470064193902e-05, "loss": 0.0151, "step": 14840 }, { "grad_norm": 0.17699900269508362, "learning_rate": 9.505574308926414e-05, "loss": 0.0153, "step": 14850 }, { "grad_norm": 0.14409109950065613, "learning_rate": 9.504677783779505e-05, "loss": 0.0121, "step": 14860 }, { "grad_norm": 0.1685332953929901, "learning_rate": 9.503780488906365e-05, "loss": 0.0127, "step": 14870 }, { "grad_norm": 0.20687252283096313, "learning_rate": 9.502882424460319e-05, "loss": 0.016, "step": 14880 }, { "grad_norm": 0.2580704689025879, "learning_rate": 9.501983590594821e-05, "loss": 0.0157, "step": 14890 }, { "grad_norm": 0.1632561981678009, "learning_rate": 9.501083987463455e-05, "loss": 0.0148, "step": 14900 }, { "grad_norm": 0.17086398601531982, "learning_rate": 9.500183615219942e-05, "loss": 0.0125, "step": 14910 }, { "grad_norm": 0.1581311672925949, "learning_rate": 9.49928247401813e-05, "loss": 0.0157, "step": 14920 }, { "grad_norm": 0.1924428939819336, "learning_rate": 9.498380564011997e-05, "loss": 0.0153, "step": 14930 }, { "grad_norm": 0.2517300844192505, "learning_rate": 9.497477885355656e-05, "loss": 0.0146, "step": 14940 }, { "grad_norm": 0.2045097053050995, "learning_rate": 9.496574438203353e-05, "loss": 0.0126, "step": 14950 }, { "grad_norm": 0.26184746623039246, "learning_rate": 9.495670222709459e-05, "loss": 0.0138, "step": 14960 }, { "grad_norm": 0.20231293141841888, "learning_rate": 9.494765239028483e-05, "loss": 0.0193, "step": 14970 }, { "grad_norm": 0.19873714447021484, "learning_rate": 9.493859487315057e-05, "loss": 0.016, "step": 14980 }, { "grad_norm": 0.1350702941417694, "learning_rate": 9.492952967723953e-05, "loss": 0.0131, "step": 14990 }, { "grad_norm": 0.2625502943992615, "learning_rate": 9.492045680410068e-05, "loss": 0.0173, "step": 15000 }, { "grad_norm": 0.1932109147310257, "learning_rate": 9.491137625528436e-05, "loss": 0.0141, "step": 15010 }, { "grad_norm": 0.17516270279884338, "learning_rate": 9.490228803234215e-05, "loss": 0.0153, "step": 15020 }, { "grad_norm": 0.16629476845264435, "learning_rate": 9.489319213682701e-05, "loss": 0.0124, "step": 15030 }, { "grad_norm": 0.1784745752811432, "learning_rate": 9.488408857029316e-05, "loss": 0.0113, "step": 15040 }, { "grad_norm": 0.19523796439170837, "learning_rate": 9.487497733429616e-05, "loss": 0.0139, "step": 15050 }, { "grad_norm": 0.1644793450832367, "learning_rate": 9.486585843039286e-05, "loss": 0.0115, "step": 15060 }, { "grad_norm": 0.23531976342201233, "learning_rate": 9.485673186014143e-05, "loss": 0.014, "step": 15070 }, { "grad_norm": 0.14594687521457672, "learning_rate": 9.484759762510137e-05, "loss": 0.0151, "step": 15080 }, { "grad_norm": 0.20744961500167847, "learning_rate": 9.483845572683346e-05, "loss": 0.0139, "step": 15090 }, { "grad_norm": 0.20284146070480347, "learning_rate": 9.48293061668998e-05, "loss": 0.0163, "step": 15100 }, { "grad_norm": 0.2409936785697937, "learning_rate": 9.48201489468638e-05, "loss": 0.0178, "step": 15110 }, { "grad_norm": 0.19636066257953644, "learning_rate": 9.481098406829016e-05, "loss": 0.0135, "step": 15120 }, { "grad_norm": 0.17746473848819733, "learning_rate": 9.480181153274495e-05, "loss": 0.0154, "step": 15130 }, { "grad_norm": 0.21719495952129364, "learning_rate": 9.479263134179548e-05, "loss": 0.0176, "step": 15140 }, { "grad_norm": 0.22558645904064178, "learning_rate": 9.478344349701039e-05, "loss": 0.0176, "step": 15150 }, { "grad_norm": 0.18682080507278442, "learning_rate": 9.477424799995964e-05, "loss": 0.016, "step": 15160 }, { "grad_norm": 0.2008642703294754, "learning_rate": 9.476504485221448e-05, "loss": 0.0174, "step": 15170 }, { "grad_norm": 0.21191202104091644, "learning_rate": 9.475583405534748e-05, "loss": 0.0157, "step": 15180 }, { "grad_norm": 0.19362828135490417, "learning_rate": 9.474661561093251e-05, "loss": 0.0181, "step": 15190 }, { "grad_norm": 0.1843860149383545, "learning_rate": 9.473738952054478e-05, "loss": 0.0139, "step": 15200 }, { "grad_norm": 0.20040182769298553, "learning_rate": 9.472815578576073e-05, "loss": 0.0169, "step": 15210 }, { "grad_norm": 0.18519198894500732, "learning_rate": 9.471891440815817e-05, "loss": 0.0153, "step": 15220 }, { "grad_norm": 0.164746955037117, "learning_rate": 9.470966538931621e-05, "loss": 0.0133, "step": 15230 }, { "grad_norm": 0.15579853951931, "learning_rate": 9.470040873081525e-05, "loss": 0.0144, "step": 15240 }, { "grad_norm": 0.21090012788772583, "learning_rate": 9.469114443423698e-05, "loss": 0.014, "step": 15250 }, { "grad_norm": 0.15483099222183228, "learning_rate": 9.468187250116445e-05, "loss": 0.0146, "step": 15260 }, { "grad_norm": 0.15894703567028046, "learning_rate": 9.467259293318197e-05, "loss": 0.0136, "step": 15270 }, { "grad_norm": 0.20363810658454895, "learning_rate": 9.466330573187514e-05, "loss": 0.0164, "step": 15280 }, { "grad_norm": 0.1701412796974182, "learning_rate": 9.46540108988309e-05, "loss": 0.0146, "step": 15290 }, { "grad_norm": 0.2199689745903015, "learning_rate": 9.46447084356375e-05, "loss": 0.0128, "step": 15300 }, { "grad_norm": 0.19907130300998688, "learning_rate": 9.463539834388447e-05, "loss": 0.0129, "step": 15310 }, { "grad_norm": 0.21185387670993805, "learning_rate": 9.462608062516263e-05, "loss": 0.0149, "step": 15320 }, { "grad_norm": 0.1924649327993393, "learning_rate": 9.461675528106413e-05, "loss": 0.0133, "step": 15330 }, { "grad_norm": 0.19131815433502197, "learning_rate": 9.460742231318244e-05, "loss": 0.0145, "step": 15340 }, { "grad_norm": 0.25533297657966614, "learning_rate": 9.459808172311229e-05, "loss": 0.0153, "step": 15350 }, { "grad_norm": 0.17024080455303192, "learning_rate": 9.458873351244972e-05, "loss": 0.0144, "step": 15360 }, { "grad_norm": 0.20563574135303497, "learning_rate": 9.457937768279211e-05, "loss": 0.0173, "step": 15370 }, { "grad_norm": 0.2154892086982727, "learning_rate": 9.45700142357381e-05, "loss": 0.018, "step": 15380 }, { "grad_norm": 0.1552456021308899, "learning_rate": 9.456064317288765e-05, "loss": 0.0143, "step": 15390 }, { "grad_norm": 0.19954632222652435, "learning_rate": 9.455126449584201e-05, "loss": 0.016, "step": 15400 }, { "grad_norm": 0.2370849996805191, "learning_rate": 9.454187820620375e-05, "loss": 0.0157, "step": 15410 }, { "grad_norm": 0.2322322428226471, "learning_rate": 9.453248430557673e-05, "loss": 0.0159, "step": 15420 }, { "grad_norm": 0.1450721174478531, "learning_rate": 9.452308279556611e-05, "loss": 0.0162, "step": 15430 }, { "grad_norm": 0.13856881856918335, "learning_rate": 9.451367367777835e-05, "loss": 0.0177, "step": 15440 }, { "grad_norm": 0.16146190464496613, "learning_rate": 9.450425695382122e-05, "loss": 0.015, "step": 15450 }, { "grad_norm": 0.17454293370246887, "learning_rate": 9.449483262530375e-05, "loss": 0.0155, "step": 15460 }, { "grad_norm": 0.22213487327098846, "learning_rate": 9.448540069383633e-05, "loss": 0.0172, "step": 15470 }, { "grad_norm": 0.18047299981117249, "learning_rate": 9.447596116103061e-05, "loss": 0.0178, "step": 15480 }, { "grad_norm": 0.1796887069940567, "learning_rate": 9.446651402849955e-05, "loss": 0.0216, "step": 15490 }, { "grad_norm": 0.2187603861093521, "learning_rate": 9.44570592978574e-05, "loss": 0.0179, "step": 15500 }, { "grad_norm": 0.17622189223766327, "learning_rate": 9.444759697071972e-05, "loss": 0.0167, "step": 15510 }, { "grad_norm": 0.21582578122615814, "learning_rate": 9.443812704870336e-05, "loss": 0.0154, "step": 15520 }, { "grad_norm": 0.19551223516464233, "learning_rate": 9.442864953342649e-05, "loss": 0.0148, "step": 15530 }, { "grad_norm": 0.16564451158046722, "learning_rate": 9.441916442650852e-05, "loss": 0.0141, "step": 15540 }, { "grad_norm": 0.2685995101928711, "learning_rate": 9.440967172957023e-05, "loss": 0.0157, "step": 15550 }, { "grad_norm": 0.1757248044013977, "learning_rate": 9.440017144423364e-05, "loss": 0.0144, "step": 15560 }, { "grad_norm": 0.20240360498428345, "learning_rate": 9.439066357212209e-05, "loss": 0.0136, "step": 15570 }, { "grad_norm": 0.22323106229305267, "learning_rate": 9.438114811486022e-05, "loss": 0.0127, "step": 15580 }, { "grad_norm": 0.17867261171340942, "learning_rate": 9.4371625074074e-05, "loss": 0.0148, "step": 15590 }, { "grad_norm": 0.19632284343242645, "learning_rate": 9.436209445139059e-05, "loss": 0.016, "step": 15600 }, { "grad_norm": 0.15575653314590454, "learning_rate": 9.435255624843855e-05, "loss": 0.0148, "step": 15610 }, { "grad_norm": 0.18646714091300964, "learning_rate": 9.43430104668477e-05, "loss": 0.0135, "step": 15620 }, { "grad_norm": 0.1540815532207489, "learning_rate": 9.433345710824914e-05, "loss": 0.0145, "step": 15630 }, { "grad_norm": 0.19260448217391968, "learning_rate": 9.432389617427529e-05, "loss": 0.0151, "step": 15640 }, { "grad_norm": 0.12420589476823807, "learning_rate": 9.431432766655984e-05, "loss": 0.0146, "step": 15650 }, { "grad_norm": 0.22644519805908203, "learning_rate": 9.430475158673778e-05, "loss": 0.0152, "step": 15660 }, { "grad_norm": 0.21670815348625183, "learning_rate": 9.429516793644542e-05, "loss": 0.0151, "step": 15670 }, { "grad_norm": 0.16667963564395905, "learning_rate": 9.428557671732034e-05, "loss": 0.0162, "step": 15680 }, { "grad_norm": 0.17848919332027435, "learning_rate": 9.42759779310014e-05, "loss": 0.0123, "step": 15690 }, { "grad_norm": 0.17802955210208893, "learning_rate": 9.426637157912879e-05, "loss": 0.0115, "step": 15700 }, { "grad_norm": 0.24044370651245117, "learning_rate": 9.425675766334397e-05, "loss": 0.0143, "step": 15710 }, { "grad_norm": 0.2586868703365326, "learning_rate": 9.424713618528968e-05, "loss": 0.0135, "step": 15720 }, { "grad_norm": 0.17147377133369446, "learning_rate": 9.423750714661e-05, "loss": 0.0153, "step": 15730 }, { "grad_norm": 0.20213158428668976, "learning_rate": 9.422787054895022e-05, "loss": 0.0135, "step": 15740 }, { "grad_norm": 0.16428431868553162, "learning_rate": 9.4218226393957e-05, "loss": 0.0138, "step": 15750 }, { "grad_norm": 0.178126722574234, "learning_rate": 9.420857468327828e-05, "loss": 0.0169, "step": 15760 }, { "grad_norm": 0.16234837472438812, "learning_rate": 9.419891541856323e-05, "loss": 0.0155, "step": 15770 }, { "grad_norm": 0.1888536512851715, "learning_rate": 9.41892486014624e-05, "loss": 0.0149, "step": 15780 }, { "grad_norm": 0.19536668062210083, "learning_rate": 9.417957423362756e-05, "loss": 0.0149, "step": 15790 }, { "grad_norm": 0.19346582889556885, "learning_rate": 9.416989231671178e-05, "loss": 0.0148, "step": 15800 }, { "grad_norm": 0.1400822252035141, "learning_rate": 9.416020285236946e-05, "loss": 0.0132, "step": 15810 }, { "grad_norm": 0.17674557864665985, "learning_rate": 9.415050584225626e-05, "loss": 0.0164, "step": 15820 }, { "grad_norm": 0.1976936310529709, "learning_rate": 9.414080128802914e-05, "loss": 0.0109, "step": 15830 }, { "grad_norm": 0.14326395094394684, "learning_rate": 9.413108919134632e-05, "loss": 0.0175, "step": 15840 }, { "grad_norm": 0.20733998715877533, "learning_rate": 9.412136955386734e-05, "loss": 0.0141, "step": 15850 }, { "grad_norm": 0.18994253873825073, "learning_rate": 9.411164237725303e-05, "loss": 0.0151, "step": 15860 }, { "grad_norm": 0.11608327180147171, "learning_rate": 9.41019076631655e-05, "loss": 0.0135, "step": 15870 }, { "grad_norm": 0.19964349269866943, "learning_rate": 9.409216541326815e-05, "loss": 0.0163, "step": 15880 }, { "grad_norm": 0.16809974610805511, "learning_rate": 9.408241562922564e-05, "loss": 0.0161, "step": 15890 }, { "grad_norm": 0.19505825638771057, "learning_rate": 9.407265831270395e-05, "loss": 0.0183, "step": 15900 }, { "grad_norm": 0.17851653695106506, "learning_rate": 9.406289346537035e-05, "loss": 0.0169, "step": 15910 }, { "grad_norm": 0.21072518825531006, "learning_rate": 9.405312108889339e-05, "loss": 0.0186, "step": 15920 }, { "grad_norm": 0.27663278579711914, "learning_rate": 9.404334118494288e-05, "loss": 0.0146, "step": 15930 }, { "grad_norm": 0.23180313408374786, "learning_rate": 9.403355375518995e-05, "loss": 0.0136, "step": 15940 }, { "grad_norm": 0.1759326159954071, "learning_rate": 9.4023758801307e-05, "loss": 0.0175, "step": 15950 }, { "grad_norm": 0.23845942318439484, "learning_rate": 9.401395632496774e-05, "loss": 0.0152, "step": 15960 }, { "grad_norm": 0.1712338775396347, "learning_rate": 9.400414632784711e-05, "loss": 0.0154, "step": 15970 }, { "grad_norm": 0.17132152616977692, "learning_rate": 9.39943288116214e-05, "loss": 0.0143, "step": 15980 }, { "grad_norm": 0.1983083337545395, "learning_rate": 9.398450377796815e-05, "loss": 0.0141, "step": 15990 }, { "grad_norm": 0.16783158481121063, "learning_rate": 9.397467122856616e-05, "loss": 0.0183, "step": 16000 }, { "grad_norm": 0.20073743164539337, "learning_rate": 9.396483116509558e-05, "loss": 0.0149, "step": 16010 }, { "grad_norm": 0.15920765697956085, "learning_rate": 9.39549835892378e-05, "loss": 0.0157, "step": 16020 }, { "grad_norm": 0.17436183989048004, "learning_rate": 9.39451285026755e-05, "loss": 0.0137, "step": 16030 }, { "grad_norm": 0.1781500279903412, "learning_rate": 9.393526590709262e-05, "loss": 0.0141, "step": 16040 }, { "grad_norm": 0.19410814344882965, "learning_rate": 9.392539580417444e-05, "loss": 0.0155, "step": 16050 }, { "grad_norm": 0.2491885870695114, "learning_rate": 9.391551819560747e-05, "loss": 0.0169, "step": 16060 }, { "grad_norm": 0.17472004890441895, "learning_rate": 9.390563308307955e-05, "loss": 0.0138, "step": 16070 }, { "grad_norm": 0.17179718613624573, "learning_rate": 9.389574046827974e-05, "loss": 0.0152, "step": 16080 }, { "grad_norm": 0.18821652233600616, "learning_rate": 9.388584035289845e-05, "loss": 0.0127, "step": 16090 }, { "grad_norm": 0.17904329299926758, "learning_rate": 9.387593273862732e-05, "loss": 0.013, "step": 16100 }, { "grad_norm": 0.15666621923446655, "learning_rate": 9.386601762715929e-05, "loss": 0.0135, "step": 16110 }, { "grad_norm": 0.20845651626586914, "learning_rate": 9.38560950201886e-05, "loss": 0.0129, "step": 16120 }, { "grad_norm": 0.1606602668762207, "learning_rate": 9.384616491941071e-05, "loss": 0.0148, "step": 16130 }, { "grad_norm": 0.18379677832126617, "learning_rate": 9.383622732652245e-05, "loss": 0.014, "step": 16140 }, { "grad_norm": 0.18521882593631744, "learning_rate": 9.382628224322187e-05, "loss": 0.0137, "step": 16150 }, { "grad_norm": 0.1717977225780487, "learning_rate": 9.381632967120829e-05, "loss": 0.0146, "step": 16160 }, { "grad_norm": 0.13860207796096802, "learning_rate": 9.380636961218235e-05, "loss": 0.0133, "step": 16170 }, { "grad_norm": 0.2191903442144394, "learning_rate": 9.379640206784597e-05, "loss": 0.0138, "step": 16180 }, { "grad_norm": 0.22454378008842468, "learning_rate": 9.378642703990229e-05, "loss": 0.013, "step": 16190 }, { "grad_norm": 0.16459476947784424, "learning_rate": 9.37764445300558e-05, "loss": 0.0127, "step": 16200 }, { "grad_norm": 0.15716740489006042, "learning_rate": 9.376645454001222e-05, "loss": 0.0127, "step": 16210 }, { "grad_norm": 0.16975060105323792, "learning_rate": 9.375645707147858e-05, "loss": 0.015, "step": 16220 }, { "grad_norm": 0.1916893571615219, "learning_rate": 9.374645212616316e-05, "loss": 0.014, "step": 16230 }, { "grad_norm": 0.24321113526821136, "learning_rate": 9.373643970577555e-05, "loss": 0.0145, "step": 16240 }, { "grad_norm": 0.1236511692404747, "learning_rate": 9.372641981202659e-05, "loss": 0.0106, "step": 16250 }, { "grad_norm": 0.16878816485404968, "learning_rate": 9.37163924466284e-05, "loss": 0.0165, "step": 16260 }, { "grad_norm": 0.19582223892211914, "learning_rate": 9.370635761129438e-05, "loss": 0.0123, "step": 16270 }, { "grad_norm": 0.2326543629169464, "learning_rate": 9.36963153077392e-05, "loss": 0.0179, "step": 16280 }, { "grad_norm": 0.20633265376091003, "learning_rate": 9.368626553767888e-05, "loss": 0.0116, "step": 16290 }, { "grad_norm": 0.18039967119693756, "learning_rate": 9.367620830283057e-05, "loss": 0.0145, "step": 16300 }, { "grad_norm": 0.17162896692752838, "learning_rate": 9.366614360491281e-05, "loss": 0.0166, "step": 16310 }, { "grad_norm": 0.1642826795578003, "learning_rate": 9.365607144564539e-05, "loss": 0.0133, "step": 16320 }, { "grad_norm": 0.1602475345134735, "learning_rate": 9.364599182674934e-05, "loss": 0.0115, "step": 16330 }, { "grad_norm": 0.16306301951408386, "learning_rate": 9.3635904749947e-05, "loss": 0.0113, "step": 16340 }, { "grad_norm": 0.19244197010993958, "learning_rate": 9.362581021696202e-05, "loss": 0.015, "step": 16350 }, { "grad_norm": 0.2056296169757843, "learning_rate": 9.361570822951921e-05, "loss": 0.016, "step": 16360 }, { "grad_norm": 0.17592811584472656, "learning_rate": 9.360559878934476e-05, "loss": 0.0116, "step": 16370 }, { "grad_norm": 0.26363134384155273, "learning_rate": 9.359548189816611e-05, "loss": 0.0118, "step": 16380 }, { "grad_norm": 0.21175649762153625, "learning_rate": 9.358535755771193e-05, "loss": 0.0121, "step": 16390 }, { "grad_norm": 0.1548927128314972, "learning_rate": 9.357522576971221e-05, "loss": 0.0164, "step": 16400 }, { "grad_norm": 0.18949955701828003, "learning_rate": 9.356508653589819e-05, "loss": 0.0147, "step": 16410 }, { "grad_norm": 0.19280917942523956, "learning_rate": 9.355493985800237e-05, "loss": 0.0148, "step": 16420 }, { "grad_norm": 0.1824534833431244, "learning_rate": 9.354478573775857e-05, "loss": 0.0131, "step": 16430 }, { "grad_norm": 0.20356693863868713, "learning_rate": 9.353462417690186e-05, "loss": 0.013, "step": 16440 }, { "grad_norm": 0.15480750799179077, "learning_rate": 9.352445517716853e-05, "loss": 0.0144, "step": 16450 }, { "grad_norm": 0.20615136623382568, "learning_rate": 9.351427874029621e-05, "loss": 0.0158, "step": 16460 }, { "grad_norm": 0.1808253824710846, "learning_rate": 9.350409486802379e-05, "loss": 0.0133, "step": 16470 }, { "grad_norm": 0.18484821915626526, "learning_rate": 9.349390356209138e-05, "loss": 0.0145, "step": 16480 }, { "grad_norm": 0.1603071093559265, "learning_rate": 9.348370482424042e-05, "loss": 0.0133, "step": 16490 }, { "grad_norm": 0.17837700247764587, "learning_rate": 9.347349865621357e-05, "loss": 0.013, "step": 16500 }, { "grad_norm": 0.1708056926727295, "learning_rate": 9.346328505975481e-05, "loss": 0.0109, "step": 16510 }, { "grad_norm": 0.11983340233564377, "learning_rate": 9.345306403660936e-05, "loss": 0.0142, "step": 16520 }, { "grad_norm": 0.22298669815063477, "learning_rate": 9.344283558852371e-05, "loss": 0.0143, "step": 16530 }, { "grad_norm": 0.1603947877883911, "learning_rate": 9.343259971724563e-05, "loss": 0.0131, "step": 16540 }, { "grad_norm": 0.10918164253234863, "learning_rate": 9.342235642452413e-05, "loss": 0.0104, "step": 16550 }, { "grad_norm": 0.19090741872787476, "learning_rate": 9.341210571210954e-05, "loss": 0.0136, "step": 16560 }, { "grad_norm": 0.20543137192726135, "learning_rate": 9.340184758175338e-05, "loss": 0.0146, "step": 16570 }, { "grad_norm": 0.15830287337303162, "learning_rate": 9.339158203520854e-05, "loss": 0.0121, "step": 16580 }, { "grad_norm": 0.14357495307922363, "learning_rate": 9.338130907422908e-05, "loss": 0.0133, "step": 16590 }, { "grad_norm": 0.14726829528808594, "learning_rate": 9.337102870057037e-05, "loss": 0.0138, "step": 16600 }, { "grad_norm": 0.1165391281247139, "learning_rate": 9.336074091598907e-05, "loss": 0.0116, "step": 16610 }, { "grad_norm": 0.2256379872560501, "learning_rate": 9.335044572224306e-05, "loss": 0.0122, "step": 16620 }, { "grad_norm": 0.2186400443315506, "learning_rate": 9.334014312109151e-05, "loss": 0.0174, "step": 16630 }, { "grad_norm": 0.1693286895751953, "learning_rate": 9.332983311429486e-05, "loss": 0.0133, "step": 16640 }, { "grad_norm": 0.20366284251213074, "learning_rate": 9.33195157036148e-05, "loss": 0.0187, "step": 16650 }, { "grad_norm": 0.15364022552967072, "learning_rate": 9.330919089081432e-05, "loss": 0.0137, "step": 16660 }, { "grad_norm": 0.20763304829597473, "learning_rate": 9.32988586776576e-05, "loss": 0.0166, "step": 16670 }, { "grad_norm": 0.1571531593799591, "learning_rate": 9.328851906591016e-05, "loss": 0.0137, "step": 16680 }, { "grad_norm": 0.1444656401872635, "learning_rate": 9.327817205733875e-05, "loss": 0.0119, "step": 16690 }, { "grad_norm": 0.11683785915374756, "learning_rate": 9.326781765371142e-05, "loss": 0.0112, "step": 16700 }, { "grad_norm": 0.11307784169912338, "learning_rate": 9.325745585679741e-05, "loss": 0.0132, "step": 16710 }, { "grad_norm": 0.13569913804531097, "learning_rate": 9.32470866683673e-05, "loss": 0.0118, "step": 16720 }, { "grad_norm": 0.18258367478847504, "learning_rate": 9.323671009019288e-05, "loss": 0.0134, "step": 16730 }, { "grad_norm": 0.14784656465053558, "learning_rate": 9.322632612404725e-05, "loss": 0.0173, "step": 16740 }, { "grad_norm": 0.17753535509109497, "learning_rate": 9.321593477170471e-05, "loss": 0.0154, "step": 16750 }, { "grad_norm": 0.13057218492031097, "learning_rate": 9.320553603494088e-05, "loss": 0.014, "step": 16760 }, { "grad_norm": 0.2236151397228241, "learning_rate": 9.319512991553261e-05, "loss": 0.0148, "step": 16770 }, { "grad_norm": 0.17436997592449188, "learning_rate": 9.318471641525803e-05, "loss": 0.0144, "step": 16780 }, { "grad_norm": 0.2440842241048813, "learning_rate": 9.317429553589652e-05, "loss": 0.0119, "step": 16790 }, { "grad_norm": 0.18107903003692627, "learning_rate": 9.316386727922873e-05, "loss": 0.0132, "step": 16800 }, { "grad_norm": 0.20344282686710358, "learning_rate": 9.315343164703656e-05, "loss": 0.0106, "step": 16810 }, { "grad_norm": 0.21882730722427368, "learning_rate": 9.314298864110316e-05, "loss": 0.0123, "step": 16820 }, { "grad_norm": 0.19038590788841248, "learning_rate": 9.313253826321295e-05, "loss": 0.0138, "step": 16830 }, { "grad_norm": 0.21928557753562927, "learning_rate": 9.312208051515165e-05, "loss": 0.0125, "step": 16840 }, { "grad_norm": 0.1766219288110733, "learning_rate": 9.311161539870618e-05, "loss": 0.0135, "step": 16850 }, { "grad_norm": 0.19278235733509064, "learning_rate": 9.310114291566474e-05, "loss": 0.0124, "step": 16860 }, { "grad_norm": 0.15660931169986725, "learning_rate": 9.309066306781679e-05, "loss": 0.0136, "step": 16870 }, { "grad_norm": 0.19634799659252167, "learning_rate": 9.308017585695306e-05, "loss": 0.0137, "step": 16880 }, { "grad_norm": 0.2022879719734192, "learning_rate": 9.306968128486552e-05, "loss": 0.0132, "step": 16890 }, { "grad_norm": 0.19025133550167084, "learning_rate": 9.30591793533474e-05, "loss": 0.0143, "step": 16900 }, { "grad_norm": 0.13930177688598633, "learning_rate": 9.304867006419321e-05, "loss": 0.0124, "step": 16910 }, { "grad_norm": 0.19607365131378174, "learning_rate": 9.303815341919868e-05, "loss": 0.0163, "step": 16920 }, { "grad_norm": 0.17691941559314728, "learning_rate": 9.302762942016084e-05, "loss": 0.0122, "step": 16930 }, { "grad_norm": 0.18286816775798798, "learning_rate": 9.301709806887792e-05, "loss": 0.0125, "step": 16940 }, { "grad_norm": 0.19791291654109955, "learning_rate": 9.300655936714948e-05, "loss": 0.0146, "step": 16950 }, { "grad_norm": 0.25241580605506897, "learning_rate": 9.299601331677627e-05, "loss": 0.0178, "step": 16960 }, { "grad_norm": 0.2010880708694458, "learning_rate": 9.298545991956033e-05, "loss": 0.0172, "step": 16970 }, { "grad_norm": 0.18373385071754456, "learning_rate": 9.297489917730493e-05, "loss": 0.0139, "step": 16980 }, { "grad_norm": 0.13941553235054016, "learning_rate": 9.296433109181464e-05, "loss": 0.0121, "step": 16990 }, { "grad_norm": 0.19385574758052826, "learning_rate": 9.295375566489523e-05, "loss": 0.015, "step": 17000 }, { "grad_norm": 0.22171466052532196, "learning_rate": 9.294317289835379e-05, "loss": 0.0199, "step": 17010 }, { "grad_norm": 0.14527618885040283, "learning_rate": 9.293258279399859e-05, "loss": 0.0168, "step": 17020 }, { "grad_norm": 0.21428200602531433, "learning_rate": 9.292198535363919e-05, "loss": 0.0136, "step": 17030 }, { "grad_norm": 0.18029524385929108, "learning_rate": 9.291138057908641e-05, "loss": 0.0145, "step": 17040 }, { "grad_norm": 0.1980482041835785, "learning_rate": 9.290076847215234e-05, "loss": 0.0155, "step": 17050 }, { "grad_norm": 0.16983921825885773, "learning_rate": 9.289014903465025e-05, "loss": 0.0137, "step": 17060 }, { "grad_norm": 0.1708444505929947, "learning_rate": 9.287952226839475e-05, "loss": 0.0133, "step": 17070 }, { "grad_norm": 0.2168509066104889, "learning_rate": 9.286888817520164e-05, "loss": 0.0121, "step": 17080 }, { "grad_norm": 0.1962195485830307, "learning_rate": 9.285824675688803e-05, "loss": 0.0163, "step": 17090 }, { "grad_norm": 0.1718442291021347, "learning_rate": 9.28475980152722e-05, "loss": 0.0129, "step": 17100 }, { "grad_norm": 0.21013668179512024, "learning_rate": 9.283694195217379e-05, "loss": 0.0123, "step": 17110 }, { "grad_norm": 0.1782337725162506, "learning_rate": 9.282627856941356e-05, "loss": 0.0138, "step": 17120 }, { "grad_norm": 0.20914849638938904, "learning_rate": 9.281560786881363e-05, "loss": 0.013, "step": 17130 }, { "grad_norm": 0.15005385875701904, "learning_rate": 9.280492985219733e-05, "loss": 0.0147, "step": 17140 }, { "grad_norm": 0.15130697190761566, "learning_rate": 9.279424452138924e-05, "loss": 0.0128, "step": 17150 }, { "grad_norm": 0.18141500651836395, "learning_rate": 9.278355187821517e-05, "loss": 0.0146, "step": 17160 }, { "grad_norm": 0.14048554003238678, "learning_rate": 9.277285192450224e-05, "loss": 0.0109, "step": 17170 }, { "grad_norm": 0.20590706169605255, "learning_rate": 9.276214466207875e-05, "loss": 0.0124, "step": 17180 }, { "grad_norm": 0.20298773050308228, "learning_rate": 9.275143009277427e-05, "loss": 0.017, "step": 17190 }, { "grad_norm": 0.15305200219154358, "learning_rate": 9.274070821841964e-05, "loss": 0.011, "step": 17200 }, { "grad_norm": 0.12218886613845825, "learning_rate": 9.272997904084696e-05, "loss": 0.0099, "step": 17210 }, { "grad_norm": 0.16300274431705475, "learning_rate": 9.271924256188951e-05, "loss": 0.0128, "step": 17220 }, { "grad_norm": 0.2233513444662094, "learning_rate": 9.270849878338189e-05, "loss": 0.017, "step": 17230 }, { "grad_norm": 0.17043671011924744, "learning_rate": 9.269774770715991e-05, "loss": 0.0162, "step": 17240 }, { "grad_norm": 0.22314463555812836, "learning_rate": 9.268698933506061e-05, "loss": 0.0121, "step": 17250 }, { "grad_norm": 0.1601150631904602, "learning_rate": 9.267622366892235e-05, "loss": 0.0151, "step": 17260 }, { "grad_norm": 0.1750611513853073, "learning_rate": 9.266545071058465e-05, "loss": 0.0136, "step": 17270 }, { "grad_norm": 0.13885080814361572, "learning_rate": 9.265467046188833e-05, "loss": 0.0143, "step": 17280 }, { "grad_norm": 0.12124612182378769, "learning_rate": 9.264388292467543e-05, "loss": 0.0115, "step": 17290 }, { "grad_norm": 0.19760185480117798, "learning_rate": 9.263308810078926e-05, "loss": 0.0132, "step": 17300 }, { "grad_norm": 0.16168895363807678, "learning_rate": 9.262228599207434e-05, "loss": 0.0133, "step": 17310 }, { "grad_norm": 0.19271402060985565, "learning_rate": 9.261147660037647e-05, "loss": 0.0128, "step": 17320 }, { "grad_norm": 0.20263606309890747, "learning_rate": 9.26006599275427e-05, "loss": 0.0157, "step": 17330 }, { "grad_norm": 0.146996408700943, "learning_rate": 9.258983597542124e-05, "loss": 0.0134, "step": 17340 }, { "grad_norm": 0.19292062520980835, "learning_rate": 9.257900474586167e-05, "loss": 0.0157, "step": 17350 }, { "grad_norm": 0.1595739722251892, "learning_rate": 9.256816624071471e-05, "loss": 0.0111, "step": 17360 }, { "grad_norm": 0.19984070956707, "learning_rate": 9.25573204618324e-05, "loss": 0.0126, "step": 17370 }, { "grad_norm": 0.1823522299528122, "learning_rate": 9.254646741106796e-05, "loss": 0.0143, "step": 17380 }, { "grad_norm": 0.13905030488967896, "learning_rate": 9.253560709027589e-05, "loss": 0.0142, "step": 17390 }, { "grad_norm": 0.1720622181892395, "learning_rate": 9.252473950131192e-05, "loss": 0.0144, "step": 17400 }, { "grad_norm": 0.13292039930820465, "learning_rate": 9.251386464603302e-05, "loss": 0.013, "step": 17410 }, { "grad_norm": 0.17117644846439362, "learning_rate": 9.250298252629741e-05, "loss": 0.0132, "step": 17420 }, { "grad_norm": 0.19972409307956696, "learning_rate": 9.249209314396454e-05, "loss": 0.0122, "step": 17430 }, { "grad_norm": 0.13639935851097107, "learning_rate": 9.248119650089513e-05, "loss": 0.0133, "step": 17440 }, { "grad_norm": 0.19863411784172058, "learning_rate": 9.247029259895108e-05, "loss": 0.0157, "step": 17450 }, { "grad_norm": 0.14012299478054047, "learning_rate": 9.24593814399956e-05, "loss": 0.0117, "step": 17460 }, { "grad_norm": 0.20587247610092163, "learning_rate": 9.244846302589309e-05, "loss": 0.0146, "step": 17470 }, { "grad_norm": 0.19728651642799377, "learning_rate": 9.243753735850923e-05, "loss": 0.013, "step": 17480 }, { "grad_norm": 0.1738191395998001, "learning_rate": 9.24266044397109e-05, "loss": 0.0135, "step": 17490 }, { "grad_norm": 0.1762021780014038, "learning_rate": 9.241566427136624e-05, "loss": 0.0135, "step": 17500 }, { "grad_norm": 0.17677168548107147, "learning_rate": 9.240471685534463e-05, "loss": 0.0128, "step": 17510 }, { "grad_norm": 0.21114957332611084, "learning_rate": 9.239376219351667e-05, "loss": 0.011, "step": 17520 }, { "grad_norm": 0.16811013221740723, "learning_rate": 9.238280028775425e-05, "loss": 0.0134, "step": 17530 }, { "grad_norm": 0.1505614072084427, "learning_rate": 9.237183113993041e-05, "loss": 0.0127, "step": 17540 }, { "grad_norm": 0.1449541300535202, "learning_rate": 9.236085475191952e-05, "loss": 0.0114, "step": 17550 }, { "grad_norm": 0.16391035914421082, "learning_rate": 9.234987112559709e-05, "loss": 0.013, "step": 17560 }, { "grad_norm": 0.195335254073143, "learning_rate": 9.233888026283999e-05, "loss": 0.0132, "step": 17570 }, { "grad_norm": 0.12926065921783447, "learning_rate": 9.232788216552619e-05, "loss": 0.0125, "step": 17580 }, { "grad_norm": 0.1634429395198822, "learning_rate": 9.231687683553502e-05, "loss": 0.0138, "step": 17590 }, { "grad_norm": 0.24519561231136322, "learning_rate": 9.230586427474698e-05, "loss": 0.0124, "step": 17600 }, { "grad_norm": 0.17267078161239624, "learning_rate": 9.229484448504379e-05, "loss": 0.0132, "step": 17610 }, { "grad_norm": 0.13987067341804504, "learning_rate": 9.228381746830843e-05, "loss": 0.013, "step": 17620 }, { "grad_norm": 0.177252396941185, "learning_rate": 9.227278322642514e-05, "loss": 0.0142, "step": 17630 }, { "grad_norm": 0.18112032115459442, "learning_rate": 9.226174176127937e-05, "loss": 0.0126, "step": 17640 }, { "grad_norm": 0.2392246127128601, "learning_rate": 9.22506930747578e-05, "loss": 0.0111, "step": 17650 }, { "grad_norm": 0.17058934271335602, "learning_rate": 9.223963716874831e-05, "loss": 0.0104, "step": 17660 }, { "grad_norm": 0.12188567966222763, "learning_rate": 9.222857404514012e-05, "loss": 0.0126, "step": 17670 }, { "grad_norm": 0.16458967328071594, "learning_rate": 9.221750370582355e-05, "loss": 0.0152, "step": 17680 }, { "grad_norm": 0.2019423544406891, "learning_rate": 9.220642615269028e-05, "loss": 0.0133, "step": 17690 }, { "grad_norm": 0.1621701866388321, "learning_rate": 9.219534138763311e-05, "loss": 0.0135, "step": 17700 }, { "grad_norm": 0.15990027785301208, "learning_rate": 9.218424941254613e-05, "loss": 0.0109, "step": 17710 }, { "grad_norm": 0.17529930174350739, "learning_rate": 9.217315022932468e-05, "loss": 0.0156, "step": 17720 }, { "grad_norm": 0.13286730647087097, "learning_rate": 9.216204383986528e-05, "loss": 0.0126, "step": 17730 }, { "grad_norm": 0.2121018022298813, "learning_rate": 9.215093024606574e-05, "loss": 0.0162, "step": 17740 }, { "grad_norm": 0.21117661893367767, "learning_rate": 9.213980944982506e-05, "loss": 0.012, "step": 17750 }, { "grad_norm": 0.16284291446208954, "learning_rate": 9.212868145304346e-05, "loss": 0.0187, "step": 17760 }, { "grad_norm": 0.2497355192899704, "learning_rate": 9.211754625762241e-05, "loss": 0.0136, "step": 17770 }, { "grad_norm": 0.15524810552597046, "learning_rate": 9.210640386546463e-05, "loss": 0.0131, "step": 17780 }, { "grad_norm": 0.1321408450603485, "learning_rate": 9.209525427847405e-05, "loss": 0.0143, "step": 17790 }, { "grad_norm": 0.1712302267551422, "learning_rate": 9.208409749855583e-05, "loss": 0.0126, "step": 17800 }, { "grad_norm": 0.1872323751449585, "learning_rate": 9.207293352761633e-05, "loss": 0.0164, "step": 17810 }, { "grad_norm": 0.16004417836666107, "learning_rate": 9.206176236756319e-05, "loss": 0.0138, "step": 17820 }, { "grad_norm": 0.1323399692773819, "learning_rate": 9.205058402030525e-05, "loss": 0.0104, "step": 17830 }, { "grad_norm": 0.1916181445121765, "learning_rate": 9.203939848775259e-05, "loss": 0.0129, "step": 17840 }, { "grad_norm": 0.17880424857139587, "learning_rate": 9.202820577181652e-05, "loss": 0.0112, "step": 17850 }, { "grad_norm": 0.22304613888263702, "learning_rate": 9.201700587440953e-05, "loss": 0.0128, "step": 17860 }, { "grad_norm": 0.20500323176383972, "learning_rate": 9.200579879744544e-05, "loss": 0.016, "step": 17870 }, { "grad_norm": 0.17470784485340118, "learning_rate": 9.199458454283918e-05, "loss": 0.0111, "step": 17880 }, { "grad_norm": 0.19471287727355957, "learning_rate": 9.198336311250697e-05, "loss": 0.0157, "step": 17890 }, { "grad_norm": 0.15526719391345978, "learning_rate": 9.197213450836626e-05, "loss": 0.013, "step": 17900 }, { "grad_norm": 0.16366606950759888, "learning_rate": 9.19608987323357e-05, "loss": 0.0156, "step": 17910 }, { "grad_norm": 0.28706949949264526, "learning_rate": 9.194965578633517e-05, "loss": 0.0159, "step": 17920 }, { "grad_norm": 0.17415276169776917, "learning_rate": 9.193840567228582e-05, "loss": 0.012, "step": 17930 }, { "grad_norm": 0.21442747116088867, "learning_rate": 9.192714839210994e-05, "loss": 0.0148, "step": 17940 }, { "grad_norm": 0.21214967966079712, "learning_rate": 9.19158839477311e-05, "loss": 0.0116, "step": 17950 }, { "grad_norm": 0.19453856348991394, "learning_rate": 9.190461234107411e-05, "loss": 0.0149, "step": 17960 }, { "grad_norm": 0.17555773258209229, "learning_rate": 9.189333357406496e-05, "loss": 0.0131, "step": 17970 }, { "grad_norm": 0.22342373430728912, "learning_rate": 9.188204764863089e-05, "loss": 0.0131, "step": 17980 }, { "grad_norm": 0.17344820499420166, "learning_rate": 9.187075456670033e-05, "loss": 0.0134, "step": 17990 }, { "grad_norm": 0.19693416357040405, "learning_rate": 9.1859454330203e-05, "loss": 0.0125, "step": 18000 }, { "grad_norm": 0.24085097014904022, "learning_rate": 9.18481469410698e-05, "loss": 0.0151, "step": 18010 }, { "grad_norm": 0.19809448719024658, "learning_rate": 9.183683240123281e-05, "loss": 0.0122, "step": 18020 }, { "grad_norm": 0.16388729214668274, "learning_rate": 9.182551071262541e-05, "loss": 0.0113, "step": 18030 }, { "grad_norm": 0.17983710765838623, "learning_rate": 9.181418187718218e-05, "loss": 0.0119, "step": 18040 }, { "grad_norm": 0.13675357401371002, "learning_rate": 9.180284589683888e-05, "loss": 0.014, "step": 18050 }, { "grad_norm": 0.16914165019989014, "learning_rate": 9.17915027735325e-05, "loss": 0.0146, "step": 18060 }, { "grad_norm": 0.12255620211362839, "learning_rate": 9.178015250920133e-05, "loss": 0.0157, "step": 18070 }, { "grad_norm": 0.2018333375453949, "learning_rate": 9.176879510578477e-05, "loss": 0.0121, "step": 18080 }, { "grad_norm": 0.21872669458389282, "learning_rate": 9.17574305652235e-05, "loss": 0.0149, "step": 18090 }, { "grad_norm": 0.18386578559875488, "learning_rate": 9.174605888945942e-05, "loss": 0.013, "step": 18100 }, { "grad_norm": 0.1791485697031021, "learning_rate": 9.173468008043564e-05, "loss": 0.0127, "step": 18110 }, { "grad_norm": 0.21803049743175507, "learning_rate": 9.172329414009648e-05, "loss": 0.0176, "step": 18120 }, { "grad_norm": 0.19116617739200592, "learning_rate": 9.171190107038747e-05, "loss": 0.0137, "step": 18130 }, { "grad_norm": 0.1275191456079483, "learning_rate": 9.170050087325541e-05, "loss": 0.0105, "step": 18140 }, { "grad_norm": 0.20551888644695282, "learning_rate": 9.168909355064824e-05, "loss": 0.0107, "step": 18150 }, { "grad_norm": 0.18195761740207672, "learning_rate": 9.167767910451519e-05, "loss": 0.013, "step": 18160 }, { "grad_norm": 0.15501701831817627, "learning_rate": 9.166625753680669e-05, "loss": 0.0096, "step": 18170 }, { "grad_norm": 0.21460972726345062, "learning_rate": 9.165482884947431e-05, "loss": 0.0129, "step": 18180 }, { "grad_norm": 0.1837361752986908, "learning_rate": 9.164339304447098e-05, "loss": 0.0134, "step": 18190 }, { "grad_norm": 0.18966266512870789, "learning_rate": 9.163195012375072e-05, "loss": 0.0136, "step": 18200 }, { "grad_norm": 0.20267197489738464, "learning_rate": 9.16205000892688e-05, "loss": 0.0136, "step": 18210 }, { "grad_norm": 0.14384876191616058, "learning_rate": 9.160904294298175e-05, "loss": 0.0131, "step": 18220 }, { "grad_norm": 0.19484560191631317, "learning_rate": 9.159757868684727e-05, "loss": 0.015, "step": 18230 }, { "grad_norm": 0.20898985862731934, "learning_rate": 9.15861073228243e-05, "loss": 0.0148, "step": 18240 }, { "grad_norm": 0.15170450508594513, "learning_rate": 9.157462885287296e-05, "loss": 0.0145, "step": 18250 }, { "grad_norm": 0.19866938889026642, "learning_rate": 9.156314327895461e-05, "loss": 0.0141, "step": 18260 }, { "grad_norm": 0.16034775972366333, "learning_rate": 9.155165060303185e-05, "loss": 0.0152, "step": 18270 }, { "grad_norm": 0.15197300910949707, "learning_rate": 9.154015082706841e-05, "loss": 0.0137, "step": 18280 }, { "grad_norm": 0.221729576587677, "learning_rate": 9.152864395302936e-05, "loss": 0.0117, "step": 18290 }, { "grad_norm": 0.12433268129825592, "learning_rate": 9.151712998288085e-05, "loss": 0.0105, "step": 18300 }, { "grad_norm": 0.16438265144824982, "learning_rate": 9.150560891859031e-05, "loss": 0.0175, "step": 18310 }, { "grad_norm": 0.18073710799217224, "learning_rate": 9.14940807621264e-05, "loss": 0.0118, "step": 18320 }, { "grad_norm": 0.19025935232639313, "learning_rate": 9.148254551545894e-05, "loss": 0.0171, "step": 18330 }, { "grad_norm": 0.2336430698633194, "learning_rate": 9.147100318055901e-05, "loss": 0.0135, "step": 18340 }, { "grad_norm": 0.19145794212818146, "learning_rate": 9.145945375939888e-05, "loss": 0.0149, "step": 18350 }, { "grad_norm": 0.16894173622131348, "learning_rate": 9.144789725395203e-05, "loss": 0.0131, "step": 18360 }, { "grad_norm": 0.1864202916622162, "learning_rate": 9.14363336661931e-05, "loss": 0.012, "step": 18370 }, { "grad_norm": 0.14580422639846802, "learning_rate": 9.142476299809806e-05, "loss": 0.0139, "step": 18380 }, { "grad_norm": 0.1710616648197174, "learning_rate": 9.1413185251644e-05, "loss": 0.011, "step": 18390 }, { "grad_norm": 0.17863187193870544, "learning_rate": 9.140160042880923e-05, "loss": 0.012, "step": 18400 }, { "grad_norm": 0.1642799973487854, "learning_rate": 9.139000853157327e-05, "loss": 0.0123, "step": 18410 }, { "grad_norm": 0.2303670197725296, "learning_rate": 9.137840956191688e-05, "loss": 0.0124, "step": 18420 }, { "grad_norm": 0.19866491854190826, "learning_rate": 9.136680352182199e-05, "loss": 0.0144, "step": 18430 }, { "grad_norm": 0.1982426792383194, "learning_rate": 9.135519041327177e-05, "loss": 0.0142, "step": 18440 }, { "grad_norm": 0.1587243676185608, "learning_rate": 9.134357023825058e-05, "loss": 0.0137, "step": 18450 }, { "grad_norm": 0.16461136937141418, "learning_rate": 9.133194299874398e-05, "loss": 0.0118, "step": 18460 }, { "grad_norm": 0.18862298130989075, "learning_rate": 9.132030869673876e-05, "loss": 0.0111, "step": 18470 }, { "grad_norm": 0.14199349284172058, "learning_rate": 9.130866733422288e-05, "loss": 0.0108, "step": 18480 }, { "grad_norm": 0.24411459267139435, "learning_rate": 9.129701891318556e-05, "loss": 0.0149, "step": 18490 }, { "grad_norm": 0.13591380417346954, "learning_rate": 9.128536343561718e-05, "loss": 0.0135, "step": 18500 }, { "grad_norm": 0.1974743902683258, "learning_rate": 9.127370090350934e-05, "loss": 0.0131, "step": 18510 }, { "grad_norm": 0.16567419469356537, "learning_rate": 9.126203131885487e-05, "loss": 0.0117, "step": 18520 }, { "grad_norm": 0.14354778826236725, "learning_rate": 9.125035468364775e-05, "loss": 0.0111, "step": 18530 }, { "grad_norm": 0.2368907928466797, "learning_rate": 9.123867099988322e-05, "loss": 0.0137, "step": 18540 }, { "grad_norm": 0.16858448088169098, "learning_rate": 9.122698026955769e-05, "loss": 0.0146, "step": 18550 }, { "grad_norm": 0.1567610651254654, "learning_rate": 9.12152824946688e-05, "loss": 0.0117, "step": 18560 }, { "grad_norm": 0.17033196985721588, "learning_rate": 9.120357767721538e-05, "loss": 0.0124, "step": 18570 }, { "grad_norm": 0.14006425440311432, "learning_rate": 9.119186581919745e-05, "loss": 0.0106, "step": 18580 }, { "grad_norm": 0.17728213965892792, "learning_rate": 9.118014692261624e-05, "loss": 0.0134, "step": 18590 }, { "grad_norm": 0.3127535879611969, "learning_rate": 9.116842098947422e-05, "loss": 0.0131, "step": 18600 }, { "grad_norm": 0.16398152709007263, "learning_rate": 9.115668802177499e-05, "loss": 0.0136, "step": 18610 }, { "grad_norm": 0.1834898442029953, "learning_rate": 9.114494802152342e-05, "loss": 0.0149, "step": 18620 }, { "grad_norm": 0.17150543630123138, "learning_rate": 9.113320099072555e-05, "loss": 0.0149, "step": 18630 }, { "grad_norm": 0.13977530598640442, "learning_rate": 9.112144693138864e-05, "loss": 0.0107, "step": 18640 }, { "grad_norm": 0.1506539285182953, "learning_rate": 9.110968584552111e-05, "loss": 0.0111, "step": 18650 }, { "grad_norm": 0.15710872411727905, "learning_rate": 9.109791773513264e-05, "loss": 0.0122, "step": 18660 }, { "grad_norm": 0.16209104657173157, "learning_rate": 9.108614260223403e-05, "loss": 0.0104, "step": 18670 }, { "grad_norm": 0.15845796465873718, "learning_rate": 9.107436044883738e-05, "loss": 0.0126, "step": 18680 }, { "grad_norm": 0.1830548197031021, "learning_rate": 9.10625712769559e-05, "loss": 0.012, "step": 18690 }, { "grad_norm": 0.11877462267875671, "learning_rate": 9.105077508860406e-05, "loss": 0.0116, "step": 18700 }, { "grad_norm": 0.1643466055393219, "learning_rate": 9.103897188579751e-05, "loss": 0.0106, "step": 18710 }, { "grad_norm": 0.13980010151863098, "learning_rate": 9.102716167055308e-05, "loss": 0.01, "step": 18720 }, { "grad_norm": 0.15715600550174713, "learning_rate": 9.10153444448888e-05, "loss": 0.0136, "step": 18730 }, { "grad_norm": 0.19149385392665863, "learning_rate": 9.100352021082393e-05, "loss": 0.0118, "step": 18740 }, { "grad_norm": 0.20285263657569885, "learning_rate": 9.099168897037891e-05, "loss": 0.0109, "step": 18750 }, { "grad_norm": 0.12027908861637115, "learning_rate": 9.097985072557538e-05, "loss": 0.0093, "step": 18760 }, { "grad_norm": 0.2741585075855255, "learning_rate": 9.096800547843615e-05, "loss": 0.011, "step": 18770 }, { "grad_norm": 0.2337809056043625, "learning_rate": 9.095615323098526e-05, "loss": 0.0127, "step": 18780 }, { "grad_norm": 0.1510147601366043, "learning_rate": 9.094429398524795e-05, "loss": 0.0115, "step": 18790 }, { "grad_norm": 0.16867795586585999, "learning_rate": 9.093242774325061e-05, "loss": 0.0135, "step": 18800 }, { "grad_norm": 0.13693736493587494, "learning_rate": 9.092055450702088e-05, "loss": 0.0116, "step": 18810 }, { "grad_norm": 0.11439131945371628, "learning_rate": 9.090867427858756e-05, "loss": 0.0157, "step": 18820 }, { "grad_norm": 0.1754656285047531, "learning_rate": 9.089678705998066e-05, "loss": 0.011, "step": 18830 }, { "grad_norm": 0.14356540143489838, "learning_rate": 9.088489285323139e-05, "loss": 0.0114, "step": 18840 }, { "grad_norm": 0.16304275393486023, "learning_rate": 9.087299166037212e-05, "loss": 0.0119, "step": 18850 }, { "grad_norm": 0.17434343695640564, "learning_rate": 9.086108348343647e-05, "loss": 0.0108, "step": 18860 }, { "grad_norm": 0.23783473670482635, "learning_rate": 9.08491683244592e-05, "loss": 0.0105, "step": 18870 }, { "grad_norm": 0.20051024854183197, "learning_rate": 9.08372461854763e-05, "loss": 0.0129, "step": 18880 }, { "grad_norm": 0.16158638894557953, "learning_rate": 9.082531706852492e-05, "loss": 0.0148, "step": 18890 }, { "grad_norm": 0.14084669947624207, "learning_rate": 9.081338097564342e-05, "loss": 0.0114, "step": 18900 }, { "grad_norm": 0.15114234387874603, "learning_rate": 9.080143790887137e-05, "loss": 0.0137, "step": 18910 }, { "grad_norm": 0.2052232027053833, "learning_rate": 9.07894878702495e-05, "loss": 0.0121, "step": 18920 }, { "grad_norm": 0.14728987216949463, "learning_rate": 9.077753086181974e-05, "loss": 0.0118, "step": 18930 }, { "grad_norm": 0.18768039345741272, "learning_rate": 9.076556688562524e-05, "loss": 0.0112, "step": 18940 }, { "grad_norm": 0.15220823884010315, "learning_rate": 9.075359594371029e-05, "loss": 0.0128, "step": 18950 }, { "grad_norm": 0.180805504322052, "learning_rate": 9.07416180381204e-05, "loss": 0.0127, "step": 18960 }, { "grad_norm": 0.1838369518518448, "learning_rate": 9.072963317090228e-05, "loss": 0.0122, "step": 18970 }, { "grad_norm": 0.19434881210327148, "learning_rate": 9.071764134410382e-05, "loss": 0.0129, "step": 18980 }, { "grad_norm": 0.18265312910079956, "learning_rate": 9.070564255977407e-05, "loss": 0.0139, "step": 18990 }, { "grad_norm": 0.20922131836414337, "learning_rate": 9.06936368199633e-05, "loss": 0.0145, "step": 19000 }, { "grad_norm": 0.18900427222251892, "learning_rate": 9.0681624126723e-05, "loss": 0.0123, "step": 19010 }, { "grad_norm": 0.14426465332508087, "learning_rate": 9.066960448210576e-05, "loss": 0.0111, "step": 19020 }, { "grad_norm": 0.1535237729549408, "learning_rate": 9.065757788816543e-05, "loss": 0.0124, "step": 19030 }, { "grad_norm": 0.1547226756811142, "learning_rate": 9.064554434695705e-05, "loss": 0.0113, "step": 19040 }, { "grad_norm": 0.12569096684455872, "learning_rate": 9.063350386053677e-05, "loss": 0.0105, "step": 19050 }, { "grad_norm": 0.16701573133468628, "learning_rate": 9.062145643096202e-05, "loss": 0.0107, "step": 19060 }, { "grad_norm": 0.2407616674900055, "learning_rate": 9.060940206029136e-05, "loss": 0.014, "step": 19070 }, { "grad_norm": 0.32032665610313416, "learning_rate": 9.059734075058457e-05, "loss": 0.0211, "step": 19080 }, { "grad_norm": 0.42559319734573364, "learning_rate": 9.058527250390257e-05, "loss": 0.0318, "step": 19090 }, { "grad_norm": 0.3314374089241028, "learning_rate": 9.057319732230752e-05, "loss": 0.0322, "step": 19100 }, { "grad_norm": 0.25387147068977356, "learning_rate": 9.056111520786273e-05, "loss": 0.0275, "step": 19110 }, { "grad_norm": 0.6812101006507874, "learning_rate": 9.054902616263268e-05, "loss": 0.0249, "step": 19120 }, { "grad_norm": 0.27573585510253906, "learning_rate": 9.05369301886831e-05, "loss": 0.0319, "step": 19130 }, { "grad_norm": 0.16733458638191223, "learning_rate": 9.052482728808083e-05, "loss": 0.0232, "step": 19140 }, { "grad_norm": 0.204388827085495, "learning_rate": 9.051271746289391e-05, "loss": 0.0212, "step": 19150 }, { "grad_norm": 0.18034376204013824, "learning_rate": 9.050060071519162e-05, "loss": 0.0161, "step": 19160 }, { "grad_norm": 0.17209936678409576, "learning_rate": 9.048847704704437e-05, "loss": 0.0163, "step": 19170 }, { "grad_norm": 0.20197397470474243, "learning_rate": 9.047634646052376e-05, "loss": 0.0193, "step": 19180 }, { "grad_norm": 0.19694417715072632, "learning_rate": 9.046420895770256e-05, "loss": 0.0192, "step": 19190 }, { "grad_norm": 0.19813553988933563, "learning_rate": 9.045206454065473e-05, "loss": 0.0199, "step": 19200 }, { "grad_norm": 0.13599711656570435, "learning_rate": 9.043991321145546e-05, "loss": 0.0137, "step": 19210 }, { "grad_norm": 0.15034516155719757, "learning_rate": 9.042775497218105e-05, "loss": 0.0118, "step": 19220 }, { "grad_norm": 0.1521628051996231, "learning_rate": 9.041558982490901e-05, "loss": 0.0125, "step": 19230 }, { "grad_norm": 0.19234682619571686, "learning_rate": 9.040341777171805e-05, "loss": 0.0119, "step": 19240 }, { "grad_norm": 0.18777495622634888, "learning_rate": 9.039123881468802e-05, "loss": 0.0164, "step": 19250 }, { "grad_norm": 0.1492997109889984, "learning_rate": 9.037905295589998e-05, "loss": 0.0112, "step": 19260 }, { "grad_norm": 0.13330522179603577, "learning_rate": 9.036686019743617e-05, "loss": 0.0119, "step": 19270 }, { "grad_norm": 0.15894843637943268, "learning_rate": 9.035466054137997e-05, "loss": 0.0136, "step": 19280 }, { "grad_norm": 0.23440434038639069, "learning_rate": 9.0342453989816e-05, "loss": 0.0134, "step": 19290 }, { "grad_norm": 0.14977851510047913, "learning_rate": 9.033024054483e-05, "loss": 0.0099, "step": 19300 }, { "grad_norm": 0.19826918840408325, "learning_rate": 9.031802020850894e-05, "loss": 0.0128, "step": 19310 }, { "grad_norm": 0.19564476609230042, "learning_rate": 9.030579298294092e-05, "loss": 0.0168, "step": 19320 }, { "grad_norm": 0.17659598588943481, "learning_rate": 9.029355887021524e-05, "loss": 0.0193, "step": 19330 }, { "grad_norm": 0.14605483412742615, "learning_rate": 9.028131787242238e-05, "loss": 0.0159, "step": 19340 }, { "grad_norm": 0.1293816864490509, "learning_rate": 9.026906999165399e-05, "loss": 0.012, "step": 19350 }, { "grad_norm": 0.21232816576957703, "learning_rate": 9.025681523000291e-05, "loss": 0.0133, "step": 19360 }, { "grad_norm": 0.20331107079982758, "learning_rate": 9.024455358956315e-05, "loss": 0.0135, "step": 19370 }, { "grad_norm": 0.171175017952919, "learning_rate": 9.023228507242984e-05, "loss": 0.0132, "step": 19380 }, { "grad_norm": 0.14350931346416473, "learning_rate": 9.022000968069937e-05, "loss": 0.0123, "step": 19390 }, { "grad_norm": 0.1523503214120865, "learning_rate": 9.020772741646928e-05, "loss": 0.0129, "step": 19400 }, { "grad_norm": 0.1558396965265274, "learning_rate": 9.019543828183826e-05, "loss": 0.0147, "step": 19410 }, { "grad_norm": 0.14847339689731598, "learning_rate": 9.018314227890616e-05, "loss": 0.0105, "step": 19420 }, { "grad_norm": 0.1356242597103119, "learning_rate": 9.017083940977408e-05, "loss": 0.0109, "step": 19430 }, { "grad_norm": 0.12568265199661255, "learning_rate": 9.015852967654422e-05, "loss": 0.011, "step": 19440 }, { "grad_norm": 0.18345783650875092, "learning_rate": 9.014621308131996e-05, "loss": 0.011, "step": 19450 }, { "grad_norm": 0.18358319997787476, "learning_rate": 9.01338896262059e-05, "loss": 0.0103, "step": 19460 }, { "grad_norm": 0.15323877334594727, "learning_rate": 9.012155931330777e-05, "loss": 0.0126, "step": 19470 }, { "grad_norm": 0.10539256781339645, "learning_rate": 9.010922214473246e-05, "loss": 0.0126, "step": 19480 }, { "grad_norm": 0.18087179958820343, "learning_rate": 9.009687812258808e-05, "loss": 0.0106, "step": 19490 }, { "grad_norm": 0.1687115877866745, "learning_rate": 9.00845272489839e-05, "loss": 0.0088, "step": 19500 }, { "grad_norm": 0.172491192817688, "learning_rate": 9.007216952603031e-05, "loss": 0.0138, "step": 19510 }, { "grad_norm": 0.1786619871854782, "learning_rate": 9.005980495583894e-05, "loss": 0.0103, "step": 19520 }, { "grad_norm": 0.2040819525718689, "learning_rate": 9.004743354052252e-05, "loss": 0.0118, "step": 19530 }, { "grad_norm": 0.16783498227596283, "learning_rate": 9.003505528219503e-05, "loss": 0.0106, "step": 19540 }, { "grad_norm": 0.17846904695034027, "learning_rate": 9.002267018297154e-05, "loss": 0.0138, "step": 19550 }, { "grad_norm": 0.19575844705104828, "learning_rate": 9.001027824496834e-05, "loss": 0.0112, "step": 19560 }, { "grad_norm": 0.14792241156101227, "learning_rate": 8.999787947030287e-05, "loss": 0.0136, "step": 19570 }, { "grad_norm": 0.17597927153110504, "learning_rate": 8.998547386109376e-05, "loss": 0.0092, "step": 19580 }, { "grad_norm": 0.17238348722457886, "learning_rate": 8.997306141946073e-05, "loss": 0.0128, "step": 19590 }, { "grad_norm": 0.18497537076473236, "learning_rate": 8.996064214752481e-05, "loss": 0.0133, "step": 19600 }, { "grad_norm": 0.12539523839950562, "learning_rate": 8.994821604740806e-05, "loss": 0.0107, "step": 19610 }, { "grad_norm": 0.1905362755060196, "learning_rate": 8.993578312123377e-05, "loss": 0.0118, "step": 19620 }, { "grad_norm": 0.18882523477077484, "learning_rate": 8.992334337112639e-05, "loss": 0.0098, "step": 19630 }, { "grad_norm": 0.19525757431983948, "learning_rate": 8.991089679921154e-05, "loss": 0.0128, "step": 19640 }, { "grad_norm": 0.1252928078174591, "learning_rate": 8.989844340761599e-05, "loss": 0.0114, "step": 19650 }, { "grad_norm": 0.1993217170238495, "learning_rate": 8.988598319846768e-05, "loss": 0.0103, "step": 19660 }, { "grad_norm": 0.13667616248130798, "learning_rate": 8.987351617389574e-05, "loss": 0.0113, "step": 19670 }, { "grad_norm": 0.14851418137550354, "learning_rate": 8.98610423360304e-05, "loss": 0.0123, "step": 19680 }, { "grad_norm": 0.15290328860282898, "learning_rate": 8.984856168700317e-05, "loss": 0.0111, "step": 19690 }, { "grad_norm": 0.1520487666130066, "learning_rate": 8.983607422894658e-05, "loss": 0.01, "step": 19700 }, { "grad_norm": 0.11919601261615753, "learning_rate": 8.982357996399442e-05, "loss": 0.013, "step": 19710 }, { "grad_norm": 0.138382226228714, "learning_rate": 8.981107889428164e-05, "loss": 0.0093, "step": 19720 }, { "grad_norm": 0.15935742855072021, "learning_rate": 8.979857102194428e-05, "loss": 0.0152, "step": 19730 }, { "grad_norm": 0.18353845179080963, "learning_rate": 8.978605634911968e-05, "loss": 0.011, "step": 19740 }, { "grad_norm": 0.1532367318868637, "learning_rate": 8.977353487794616e-05, "loss": 0.0106, "step": 19750 }, { "grad_norm": 0.13072283565998077, "learning_rate": 8.976100661056334e-05, "loss": 0.0095, "step": 19760 }, { "grad_norm": 0.2257438749074936, "learning_rate": 8.974847154911197e-05, "loss": 0.0129, "step": 19770 }, { "grad_norm": 0.20684286952018738, "learning_rate": 8.973592969573393e-05, "loss": 0.0118, "step": 19780 }, { "grad_norm": 0.1826132833957672, "learning_rate": 8.972338105257228e-05, "loss": 0.0126, "step": 19790 }, { "grad_norm": 0.12871207296848297, "learning_rate": 8.971082562177125e-05, "loss": 0.0106, "step": 19800 }, { "grad_norm": 0.1639275699853897, "learning_rate": 8.96982634054762e-05, "loss": 0.0113, "step": 19810 }, { "grad_norm": 0.19785796105861664, "learning_rate": 8.96856944058337e-05, "loss": 0.0121, "step": 19820 }, { "grad_norm": 0.1522587686777115, "learning_rate": 8.967311862499144e-05, "loss": 0.0129, "step": 19830 }, { "grad_norm": 0.1691657304763794, "learning_rate": 8.966053606509825e-05, "loss": 0.0122, "step": 19840 }, { "grad_norm": 0.20835457742214203, "learning_rate": 8.964794672830417e-05, "loss": 0.0105, "step": 19850 }, { "grad_norm": 0.18518957495689392, "learning_rate": 8.963535061676038e-05, "loss": 0.0131, "step": 19860 }, { "grad_norm": 0.12648740410804749, "learning_rate": 8.962274773261918e-05, "loss": 0.0139, "step": 19870 }, { "grad_norm": 0.17288340628147125, "learning_rate": 8.961013807803409e-05, "loss": 0.0135, "step": 19880 }, { "grad_norm": 0.13717937469482422, "learning_rate": 8.959752165515973e-05, "loss": 0.012, "step": 19890 }, { "grad_norm": 0.16144733130931854, "learning_rate": 8.958489846615193e-05, "loss": 0.0117, "step": 19900 }, { "grad_norm": 0.2028270661830902, "learning_rate": 8.957226851316762e-05, "loss": 0.0134, "step": 19910 }, { "grad_norm": 0.15615488588809967, "learning_rate": 8.955963179836493e-05, "loss": 0.0118, "step": 19920 }, { "grad_norm": 0.15536747872829437, "learning_rate": 8.954698832390312e-05, "loss": 0.0115, "step": 19930 }, { "grad_norm": 0.15193496644496918, "learning_rate": 8.953433809194263e-05, "loss": 0.0122, "step": 19940 }, { "grad_norm": 0.12303940206766129, "learning_rate": 8.9521681104645e-05, "loss": 0.0095, "step": 19950 }, { "grad_norm": 0.16651859879493713, "learning_rate": 8.9509017364173e-05, "loss": 0.0118, "step": 19960 }, { "grad_norm": 0.22594094276428223, "learning_rate": 8.949634687269052e-05, "loss": 0.0127, "step": 19970 }, { "grad_norm": 0.20984500646591187, "learning_rate": 8.948366963236259e-05, "loss": 0.0108, "step": 19980 }, { "grad_norm": 0.16566535830497742, "learning_rate": 8.947098564535538e-05, "loss": 0.0099, "step": 19990 }, { "grad_norm": 0.23127159476280212, "learning_rate": 8.945829491383627e-05, "loss": 0.012, "step": 20000 }, { "grad_norm": 0.13534371554851532, "learning_rate": 8.944559743997374e-05, "loss": 0.0146, "step": 20010 }, { "grad_norm": 0.15512754023075104, "learning_rate": 8.943289322593746e-05, "loss": 0.0113, "step": 20020 }, { "grad_norm": 0.17603062093257904, "learning_rate": 8.942018227389821e-05, "loss": 0.0113, "step": 20030 }, { "grad_norm": 0.19580896198749542, "learning_rate": 8.940746458602795e-05, "loss": 0.011, "step": 20040 }, { "grad_norm": 0.11415717005729675, "learning_rate": 8.939474016449979e-05, "loss": 0.0106, "step": 20050 }, { "grad_norm": 0.1809801459312439, "learning_rate": 8.938200901148799e-05, "loss": 0.0124, "step": 20060 }, { "grad_norm": 0.19272084534168243, "learning_rate": 8.936927112916795e-05, "loss": 0.0129, "step": 20070 }, { "grad_norm": 0.12759433686733246, "learning_rate": 8.935652651971622e-05, "loss": 0.0102, "step": 20080 }, { "grad_norm": 0.15166404843330383, "learning_rate": 8.934377518531052e-05, "loss": 0.013, "step": 20090 }, { "grad_norm": 0.1725635826587677, "learning_rate": 8.933101712812967e-05, "loss": 0.0157, "step": 20100 }, { "grad_norm": 0.14532190561294556, "learning_rate": 8.931825235035374e-05, "loss": 0.0147, "step": 20110 }, { "grad_norm": 0.15284623205661774, "learning_rate": 8.930548085416382e-05, "loss": 0.0134, "step": 20120 }, { "grad_norm": 0.18277162313461304, "learning_rate": 8.92927026417422e-05, "loss": 0.0113, "step": 20130 }, { "grad_norm": 0.19049964845180511, "learning_rate": 8.92799177152724e-05, "loss": 0.0101, "step": 20140 }, { "grad_norm": 0.18714569509029388, "learning_rate": 8.926712607693895e-05, "loss": 0.0138, "step": 20150 }, { "grad_norm": 0.1318473517894745, "learning_rate": 8.925432772892762e-05, "loss": 0.0105, "step": 20160 }, { "grad_norm": 0.1570732295513153, "learning_rate": 8.924152267342529e-05, "loss": 0.0104, "step": 20170 }, { "grad_norm": 0.11657007783651352, "learning_rate": 8.922871091261998e-05, "loss": 0.0097, "step": 20180 }, { "grad_norm": 0.17866520583629608, "learning_rate": 8.92158924487009e-05, "loss": 0.0107, "step": 20190 }, { "grad_norm": 0.11104590445756912, "learning_rate": 8.920306728385834e-05, "loss": 0.0106, "step": 20200 }, { "grad_norm": 0.15956942737102509, "learning_rate": 8.919023542028379e-05, "loss": 0.0107, "step": 20210 }, { "grad_norm": 0.20126964151859283, "learning_rate": 8.917739686016988e-05, "loss": 0.0137, "step": 20220 }, { "grad_norm": 0.14119991660118103, "learning_rate": 8.916455160571033e-05, "loss": 0.0092, "step": 20230 }, { "grad_norm": 0.18157456815242767, "learning_rate": 8.915169965910008e-05, "loss": 0.0103, "step": 20240 }, { "grad_norm": 0.1740538328886032, "learning_rate": 8.913884102253514e-05, "loss": 0.0122, "step": 20250 }, { "grad_norm": 0.2010120451450348, "learning_rate": 8.912597569821273e-05, "loss": 0.0119, "step": 20260 }, { "grad_norm": 0.16883376240730286, "learning_rate": 8.911310368833118e-05, "loss": 0.011, "step": 20270 }, { "grad_norm": 0.17570234835147858, "learning_rate": 8.910022499508994e-05, "loss": 0.0118, "step": 20280 }, { "grad_norm": 0.16425827145576477, "learning_rate": 8.908733962068965e-05, "loss": 0.0126, "step": 20290 }, { "grad_norm": 0.1740826666355133, "learning_rate": 8.907444756733207e-05, "loss": 0.0106, "step": 20300 }, { "grad_norm": 0.16023127734661102, "learning_rate": 8.906154883722006e-05, "loss": 0.0143, "step": 20310 }, { "grad_norm": 0.16500811278820038, "learning_rate": 8.904864343255773e-05, "loss": 0.0111, "step": 20320 }, { "grad_norm": 0.15666130185127258, "learning_rate": 8.90357313555502e-05, "loss": 0.0115, "step": 20330 }, { "grad_norm": 0.14106708765029907, "learning_rate": 8.90228126084038e-05, "loss": 0.0117, "step": 20340 }, { "grad_norm": 0.17485077679157257, "learning_rate": 8.900988719332601e-05, "loss": 0.0116, "step": 20350 }, { "grad_norm": 0.1516212373971939, "learning_rate": 8.899695511252542e-05, "loss": 0.0116, "step": 20360 }, { "grad_norm": 0.14759787917137146, "learning_rate": 8.898401636821176e-05, "loss": 0.0124, "step": 20370 }, { "grad_norm": 0.17864905297756195, "learning_rate": 8.897107096259593e-05, "loss": 0.0129, "step": 20380 }, { "grad_norm": 0.10876105725765228, "learning_rate": 8.895811889788994e-05, "loss": 0.0131, "step": 20390 }, { "grad_norm": 0.11013015359640121, "learning_rate": 8.894516017630692e-05, "loss": 0.0116, "step": 20400 }, { "grad_norm": 0.13762463629245758, "learning_rate": 8.893219480006118e-05, "loss": 0.0113, "step": 20410 }, { "grad_norm": 0.18139494955539703, "learning_rate": 8.891922277136817e-05, "loss": 0.0078, "step": 20420 }, { "grad_norm": 0.09401977807283401, "learning_rate": 8.890624409244441e-05, "loss": 0.0078, "step": 20430 }, { "grad_norm": 0.1476939469575882, "learning_rate": 8.889325876550763e-05, "loss": 0.0115, "step": 20440 }, { "grad_norm": 0.11996317654848099, "learning_rate": 8.888026679277666e-05, "loss": 0.0098, "step": 20450 }, { "grad_norm": 0.14783059060573578, "learning_rate": 8.886726817647147e-05, "loss": 0.0111, "step": 20460 }, { "grad_norm": 0.19319452345371246, "learning_rate": 8.885426291881319e-05, "loss": 0.0107, "step": 20470 }, { "grad_norm": 0.16156621277332306, "learning_rate": 8.884125102202401e-05, "loss": 0.012, "step": 20480 }, { "grad_norm": 0.17089812457561493, "learning_rate": 8.882823248832736e-05, "loss": 0.0137, "step": 20490 }, { "grad_norm": 0.16721749305725098, "learning_rate": 8.881520731994772e-05, "loss": 0.0119, "step": 20500 }, { "grad_norm": 0.18288587033748627, "learning_rate": 8.880217551911077e-05, "loss": 0.0102, "step": 20510 }, { "grad_norm": 0.14604125916957855, "learning_rate": 8.878913708804323e-05, "loss": 0.0153, "step": 20520 }, { "grad_norm": 0.19096997380256653, "learning_rate": 8.877609202897308e-05, "loss": 0.0132, "step": 20530 }, { "grad_norm": 0.14733116328716278, "learning_rate": 8.876304034412933e-05, "loss": 0.0093, "step": 20540 }, { "grad_norm": 0.21957486867904663, "learning_rate": 8.874998203574214e-05, "loss": 0.0113, "step": 20550 }, { "grad_norm": 0.12235584110021591, "learning_rate": 8.873691710604284e-05, "loss": 0.0144, "step": 20560 }, { "grad_norm": 0.12717001140117645, "learning_rate": 8.872384555726387e-05, "loss": 0.0116, "step": 20570 }, { "grad_norm": 0.21380014717578888, "learning_rate": 8.871076739163878e-05, "loss": 0.013, "step": 20580 }, { "grad_norm": 0.14515310525894165, "learning_rate": 8.86976826114023e-05, "loss": 0.0104, "step": 20590 }, { "grad_norm": 0.11759300529956818, "learning_rate": 8.868459121879023e-05, "loss": 0.0099, "step": 20600 }, { "grad_norm": 0.1549883633852005, "learning_rate": 8.867149321603956e-05, "loss": 0.0117, "step": 20610 }, { "grad_norm": 0.1838006228208542, "learning_rate": 8.865838860538835e-05, "loss": 0.0092, "step": 20620 }, { "grad_norm": 0.20541423559188843, "learning_rate": 8.864527738907585e-05, "loss": 0.011, "step": 20630 }, { "grad_norm": 0.15350693464279175, "learning_rate": 8.863215956934239e-05, "loss": 0.0115, "step": 20640 }, { "grad_norm": 0.18222545087337494, "learning_rate": 8.861903514842947e-05, "loss": 0.0115, "step": 20650 }, { "grad_norm": 0.1464722752571106, "learning_rate": 8.860590412857966e-05, "loss": 0.0124, "step": 20660 }, { "grad_norm": 0.16799306869506836, "learning_rate": 8.85927665120367e-05, "loss": 0.0127, "step": 20670 }, { "grad_norm": 0.21342509984970093, "learning_rate": 8.857962230104546e-05, "loss": 0.0128, "step": 20680 }, { "grad_norm": 0.16992178559303284, "learning_rate": 8.856647149785193e-05, "loss": 0.0114, "step": 20690 }, { "grad_norm": 0.15025284886360168, "learning_rate": 8.855331410470322e-05, "loss": 0.0124, "step": 20700 }, { "grad_norm": 0.1687890738248825, "learning_rate": 8.854015012384756e-05, "loss": 0.0114, "step": 20710 }, { "grad_norm": 0.1293000429868698, "learning_rate": 8.852697955753433e-05, "loss": 0.0121, "step": 20720 }, { "grad_norm": 0.14058393239974976, "learning_rate": 8.851380240801399e-05, "loss": 0.014, "step": 20730 }, { "grad_norm": 0.18946631252765656, "learning_rate": 8.850061867753818e-05, "loss": 0.0138, "step": 20740 }, { "grad_norm": 0.17703579366207123, "learning_rate": 8.848742836835963e-05, "loss": 0.0119, "step": 20750 }, { "grad_norm": 0.15583887696266174, "learning_rate": 8.847423148273221e-05, "loss": 0.0134, "step": 20760 }, { "grad_norm": 0.15635308623313904, "learning_rate": 8.846102802291092e-05, "loss": 0.0138, "step": 20770 }, { "grad_norm": 0.1508684605360031, "learning_rate": 8.844781799115183e-05, "loss": 0.01, "step": 20780 }, { "grad_norm": 0.17322184145450592, "learning_rate": 8.84346013897122e-05, "loss": 0.0116, "step": 20790 }, { "grad_norm": 0.1700904220342636, "learning_rate": 8.842137822085038e-05, "loss": 0.0104, "step": 20800 }, { "grad_norm": 0.16595683991909027, "learning_rate": 8.840814848682585e-05, "loss": 0.0095, "step": 20810 }, { "grad_norm": 0.16415858268737793, "learning_rate": 8.83949121898992e-05, "loss": 0.011, "step": 20820 }, { "grad_norm": 0.18262618780136108, "learning_rate": 8.838166933233217e-05, "loss": 0.0125, "step": 20830 }, { "grad_norm": 0.1705017238855362, "learning_rate": 8.83684199163876e-05, "loss": 0.0103, "step": 20840 }, { "grad_norm": 0.17584963142871857, "learning_rate": 8.835516394432943e-05, "loss": 0.0114, "step": 20850 }, { "grad_norm": 0.20132732391357422, "learning_rate": 8.834190141842276e-05, "loss": 0.0121, "step": 20860 }, { "grad_norm": 0.15260112285614014, "learning_rate": 8.83286323409338e-05, "loss": 0.0107, "step": 20870 }, { "grad_norm": 0.14367951452732086, "learning_rate": 8.831535671412986e-05, "loss": 0.01, "step": 20880 }, { "grad_norm": 0.18161381781101227, "learning_rate": 8.830207454027938e-05, "loss": 0.0096, "step": 20890 }, { "grad_norm": 0.13274091482162476, "learning_rate": 8.828878582165192e-05, "loss": 0.0091, "step": 20900 }, { "grad_norm": 0.15800048410892487, "learning_rate": 8.827549056051818e-05, "loss": 0.0116, "step": 20910 }, { "grad_norm": 0.1520797610282898, "learning_rate": 8.826218875914993e-05, "loss": 0.0105, "step": 20920 }, { "grad_norm": 0.11368506401777267, "learning_rate": 8.82488804198201e-05, "loss": 0.0094, "step": 20930 }, { "grad_norm": 0.13301809132099152, "learning_rate": 8.82355655448027e-05, "loss": 0.0138, "step": 20940 }, { "grad_norm": 0.19692616164684296, "learning_rate": 8.822224413637293e-05, "loss": 0.01, "step": 20950 }, { "grad_norm": 0.2031371295452118, "learning_rate": 8.820891619680697e-05, "loss": 0.0112, "step": 20960 }, { "grad_norm": 0.11873187869787216, "learning_rate": 8.819558172838227e-05, "loss": 0.0108, "step": 20970 }, { "grad_norm": 0.09289693832397461, "learning_rate": 8.818224073337731e-05, "loss": 0.0116, "step": 20980 }, { "grad_norm": 0.17854800820350647, "learning_rate": 8.816889321407169e-05, "loss": 0.009, "step": 20990 }, { "grad_norm": 0.14146655797958374, "learning_rate": 8.815553917274615e-05, "loss": 0.0135, "step": 21000 }, { "grad_norm": 0.12884464859962463, "learning_rate": 8.81421786116825e-05, "loss": 0.0103, "step": 21010 }, { "grad_norm": 0.15909318625926971, "learning_rate": 8.812881153316373e-05, "loss": 0.0127, "step": 21020 }, { "grad_norm": 0.12469946593046188, "learning_rate": 8.81154379394739e-05, "loss": 0.0108, "step": 21030 }, { "grad_norm": 0.13694672286510468, "learning_rate": 8.810205783289818e-05, "loss": 0.0111, "step": 21040 }, { "grad_norm": 0.1862248182296753, "learning_rate": 8.808867121572286e-05, "loss": 0.0132, "step": 21050 }, { "grad_norm": 0.14403550326824188, "learning_rate": 8.807527809023537e-05, "loss": 0.0107, "step": 21060 }, { "grad_norm": 0.14099009335041046, "learning_rate": 8.80618784587242e-05, "loss": 0.0134, "step": 21070 }, { "grad_norm": 0.15888145565986633, "learning_rate": 8.804847232347902e-05, "loss": 0.0113, "step": 21080 }, { "grad_norm": 0.18112853169441223, "learning_rate": 8.803505968679054e-05, "loss": 0.0113, "step": 21090 }, { "grad_norm": 0.1398758888244629, "learning_rate": 8.802164055095061e-05, "loss": 0.0127, "step": 21100 }, { "grad_norm": 0.15626251697540283, "learning_rate": 8.80082149182522e-05, "loss": 0.0134, "step": 21110 }, { "grad_norm": 0.13858360052108765, "learning_rate": 8.79947827909894e-05, "loss": 0.0103, "step": 21120 }, { "grad_norm": 0.15208424627780914, "learning_rate": 8.798134417145738e-05, "loss": 0.0124, "step": 21130 }, { "grad_norm": 0.1549694836139679, "learning_rate": 8.796789906195243e-05, "loss": 0.0123, "step": 21140 }, { "grad_norm": 0.1533842831850052, "learning_rate": 8.795444746477195e-05, "loss": 0.0132, "step": 21150 }, { "grad_norm": 0.13995563983917236, "learning_rate": 8.794098938221446e-05, "loss": 0.0118, "step": 21160 }, { "grad_norm": 0.14509163796901703, "learning_rate": 8.792752481657957e-05, "loss": 0.0118, "step": 21170 }, { "grad_norm": 0.1473347246646881, "learning_rate": 8.791405377016802e-05, "loss": 0.012, "step": 21180 }, { "grad_norm": 0.18034154176712036, "learning_rate": 8.790057624528163e-05, "loss": 0.0139, "step": 21190 }, { "grad_norm": 0.13520485162734985, "learning_rate": 8.788709224422333e-05, "loss": 0.0116, "step": 21200 }, { "grad_norm": 0.1911444514989853, "learning_rate": 8.787360176929717e-05, "loss": 0.0097, "step": 21210 }, { "grad_norm": 0.16137322783470154, "learning_rate": 8.786010482280834e-05, "loss": 0.0114, "step": 21220 }, { "grad_norm": 0.1576577126979828, "learning_rate": 8.784660140706306e-05, "loss": 0.0116, "step": 21230 }, { "grad_norm": 0.15573424100875854, "learning_rate": 8.783309152436872e-05, "loss": 0.0115, "step": 21240 }, { "grad_norm": 0.1893535852432251, "learning_rate": 8.781957517703375e-05, "loss": 0.0101, "step": 21250 }, { "grad_norm": 0.1719680279493332, "learning_rate": 8.780605236736776e-05, "loss": 0.0108, "step": 21260 }, { "grad_norm": 0.15852464735507965, "learning_rate": 8.779252309768142e-05, "loss": 0.0132, "step": 21270 }, { "grad_norm": 0.13767223060131073, "learning_rate": 8.777898737028652e-05, "loss": 0.0106, "step": 21280 }, { "grad_norm": 0.1504620611667633, "learning_rate": 8.776544518749591e-05, "loss": 0.0122, "step": 21290 }, { "grad_norm": 0.1758297234773636, "learning_rate": 8.775189655162364e-05, "loss": 0.0098, "step": 21300 }, { "grad_norm": 0.1975189447402954, "learning_rate": 8.773834146498474e-05, "loss": 0.0143, "step": 21310 }, { "grad_norm": 0.12653344869613647, "learning_rate": 8.772477992989545e-05, "loss": 0.0123, "step": 21320 }, { "grad_norm": 0.14428628981113434, "learning_rate": 8.771121194867304e-05, "loss": 0.0078, "step": 21330 }, { "grad_norm": 0.15094320476055145, "learning_rate": 8.769763752363589e-05, "loss": 0.0111, "step": 21340 }, { "grad_norm": 0.18421176075935364, "learning_rate": 8.768405665710352e-05, "loss": 0.0126, "step": 21350 }, { "grad_norm": 0.1905149668455124, "learning_rate": 8.767046935139655e-05, "loss": 0.0113, "step": 21360 }, { "grad_norm": 0.1003955751657486, "learning_rate": 8.765687560883666e-05, "loss": 0.0118, "step": 21370 }, { "grad_norm": 0.14089147746562958, "learning_rate": 8.764327543174664e-05, "loss": 0.0087, "step": 21380 }, { "grad_norm": 0.14988797903060913, "learning_rate": 8.762966882245038e-05, "loss": 0.0121, "step": 21390 }, { "grad_norm": 0.19502435624599457, "learning_rate": 8.761605578327291e-05, "loss": 0.0107, "step": 21400 }, { "grad_norm": 0.13687998056411743, "learning_rate": 8.76024363165403e-05, "loss": 0.0107, "step": 21410 }, { "grad_norm": 0.12377427518367767, "learning_rate": 8.758881042457976e-05, "loss": 0.0105, "step": 21420 }, { "grad_norm": 0.17894545197486877, "learning_rate": 8.757517810971957e-05, "loss": 0.0099, "step": 21430 }, { "grad_norm": 0.17260730266571045, "learning_rate": 8.756153937428913e-05, "loss": 0.0118, "step": 21440 }, { "grad_norm": 0.15683017671108246, "learning_rate": 8.754789422061889e-05, "loss": 0.0143, "step": 21450 }, { "grad_norm": 0.16130034625530243, "learning_rate": 8.753424265104052e-05, "loss": 0.0094, "step": 21460 }, { "grad_norm": 0.16738088428974152, "learning_rate": 8.752058466788659e-05, "loss": 0.0085, "step": 21470 }, { "grad_norm": 0.1482723355293274, "learning_rate": 8.750692027349097e-05, "loss": 0.0095, "step": 21480 }, { "grad_norm": 0.11811571568250656, "learning_rate": 8.749324947018847e-05, "loss": 0.0112, "step": 21490 }, { "grad_norm": 0.11167469620704651, "learning_rate": 8.747957226031507e-05, "loss": 0.012, "step": 21500 }, { "grad_norm": 0.12873610854148865, "learning_rate": 8.746588864620787e-05, "loss": 0.0094, "step": 21510 }, { "grad_norm": 0.15159282088279724, "learning_rate": 8.745219863020498e-05, "loss": 0.0105, "step": 21520 }, { "grad_norm": 0.1277497112751007, "learning_rate": 8.743850221464564e-05, "loss": 0.013, "step": 21530 }, { "grad_norm": 0.13907276093959808, "learning_rate": 8.742479940187026e-05, "loss": 0.013, "step": 21540 }, { "grad_norm": 0.20528477430343628, "learning_rate": 8.74110901942202e-05, "loss": 0.011, "step": 21550 }, { "grad_norm": 0.13686944544315338, "learning_rate": 8.739737459403803e-05, "loss": 0.0114, "step": 21560 }, { "grad_norm": 0.1446331888437271, "learning_rate": 8.738365260366737e-05, "loss": 0.0104, "step": 21570 }, { "grad_norm": 0.14814874529838562, "learning_rate": 8.736992422545292e-05, "loss": 0.0149, "step": 21580 }, { "grad_norm": 0.18331196904182434, "learning_rate": 8.73561894617405e-05, "loss": 0.0125, "step": 21590 }, { "grad_norm": 0.16057859361171722, "learning_rate": 8.734244831487697e-05, "loss": 0.0137, "step": 21600 }, { "grad_norm": 0.1539321392774582, "learning_rate": 8.732870078721035e-05, "loss": 0.0109, "step": 21610 }, { "grad_norm": 0.14056232571601868, "learning_rate": 8.731494688108972e-05, "loss": 0.0115, "step": 21620 }, { "grad_norm": 0.17786549031734467, "learning_rate": 8.730118659886523e-05, "loss": 0.0122, "step": 21630 }, { "grad_norm": 0.1869589388370514, "learning_rate": 8.728741994288814e-05, "loss": 0.0116, "step": 21640 }, { "grad_norm": 0.1583506166934967, "learning_rate": 8.727364691551079e-05, "loss": 0.0106, "step": 21650 }, { "grad_norm": 0.19903351366519928, "learning_rate": 8.725986751908661e-05, "loss": 0.0126, "step": 21660 }, { "grad_norm": 0.13894589245319366, "learning_rate": 8.724608175597016e-05, "loss": 0.0124, "step": 21670 }, { "grad_norm": 0.19592325389385223, "learning_rate": 8.723228962851699e-05, "loss": 0.0122, "step": 21680 }, { "grad_norm": 0.1402994841337204, "learning_rate": 8.721849113908385e-05, "loss": 0.0129, "step": 21690 }, { "grad_norm": 0.1535908281803131, "learning_rate": 8.720468629002848e-05, "loss": 0.0123, "step": 21700 }, { "grad_norm": 0.1481708586215973, "learning_rate": 8.719087508370978e-05, "loss": 0.0124, "step": 21710 }, { "grad_norm": 0.1512443870306015, "learning_rate": 8.717705752248772e-05, "loss": 0.0162, "step": 21720 }, { "grad_norm": 0.1750049740076065, "learning_rate": 8.71632336087233e-05, "loss": 0.0148, "step": 21730 }, { "grad_norm": 0.12742535769939423, "learning_rate": 8.71494033447787e-05, "loss": 0.0105, "step": 21740 }, { "grad_norm": 0.1604977250099182, "learning_rate": 8.713556673301708e-05, "loss": 0.0118, "step": 21750 }, { "grad_norm": 0.14391577243804932, "learning_rate": 8.712172377580278e-05, "loss": 0.0128, "step": 21760 }, { "grad_norm": 0.16379842162132263, "learning_rate": 8.710787447550114e-05, "loss": 0.0147, "step": 21770 }, { "grad_norm": 0.19204729795455933, "learning_rate": 8.70940188344787e-05, "loss": 0.0118, "step": 21780 }, { "grad_norm": 0.19148312509059906, "learning_rate": 8.708015685510293e-05, "loss": 0.0106, "step": 21790 }, { "grad_norm": 0.16093993186950684, "learning_rate": 8.706628853974252e-05, "loss": 0.0111, "step": 21800 }, { "grad_norm": 0.13103614747524261, "learning_rate": 8.705241389076715e-05, "loss": 0.0122, "step": 21810 }, { "grad_norm": 0.12429405748844147, "learning_rate": 8.703853291054764e-05, "loss": 0.0091, "step": 21820 }, { "grad_norm": 0.17267780005931854, "learning_rate": 8.702464560145587e-05, "loss": 0.0112, "step": 21830 }, { "grad_norm": 0.14139743149280548, "learning_rate": 8.701075196586476e-05, "loss": 0.0105, "step": 21840 }, { "grad_norm": 0.2548775374889374, "learning_rate": 8.699685200614842e-05, "loss": 0.0145, "step": 21850 }, { "grad_norm": 0.1486174762248993, "learning_rate": 8.698294572468193e-05, "loss": 0.0114, "step": 21860 }, { "grad_norm": 0.1586291790008545, "learning_rate": 8.696903312384148e-05, "loss": 0.0136, "step": 21870 }, { "grad_norm": 0.11847754567861557, "learning_rate": 8.695511420600439e-05, "loss": 0.0095, "step": 21880 }, { "grad_norm": 0.13856787979602814, "learning_rate": 8.694118897354901e-05, "loss": 0.0104, "step": 21890 }, { "grad_norm": 0.1883256882429123, "learning_rate": 8.692725742885478e-05, "loss": 0.0113, "step": 21900 }, { "grad_norm": 0.15409007668495178, "learning_rate": 8.691331957430221e-05, "loss": 0.0106, "step": 21910 }, { "grad_norm": 0.1455124318599701, "learning_rate": 8.68993754122729e-05, "loss": 0.0091, "step": 21920 }, { "grad_norm": 0.1478487104177475, "learning_rate": 8.688542494514955e-05, "loss": 0.0101, "step": 21930 }, { "grad_norm": 0.15200573205947876, "learning_rate": 8.68714681753159e-05, "loss": 0.0107, "step": 21940 }, { "grad_norm": 0.1668579876422882, "learning_rate": 8.685750510515676e-05, "loss": 0.0098, "step": 21950 }, { "grad_norm": 0.12547387182712555, "learning_rate": 8.684353573705805e-05, "loss": 0.0102, "step": 21960 }, { "grad_norm": 0.1120753139257431, "learning_rate": 8.682956007340677e-05, "loss": 0.012, "step": 21970 }, { "grad_norm": 0.16847814619541168, "learning_rate": 8.681557811659095e-05, "loss": 0.0129, "step": 21980 }, { "grad_norm": 0.15053650736808777, "learning_rate": 8.680158986899974e-05, "loss": 0.0098, "step": 21990 }, { "grad_norm": 0.14676055312156677, "learning_rate": 8.678759533302335e-05, "loss": 0.0082, "step": 22000 }, { "grad_norm": 0.12206882983446121, "learning_rate": 8.677359451105308e-05, "loss": 0.0095, "step": 22010 }, { "grad_norm": 0.1527787446975708, "learning_rate": 8.675958740548123e-05, "loss": 0.0105, "step": 22020 }, { "grad_norm": 0.10568951815366745, "learning_rate": 8.674557401870129e-05, "loss": 0.0099, "step": 22030 }, { "grad_norm": 0.1802118569612503, "learning_rate": 8.673155435310775e-05, "loss": 0.0113, "step": 22040 }, { "grad_norm": 0.13639897108078003, "learning_rate": 8.671752841109617e-05, "loss": 0.0096, "step": 22050 }, { "grad_norm": 0.12000580132007599, "learning_rate": 8.670349619506321e-05, "loss": 0.0081, "step": 22060 }, { "grad_norm": 0.1686873883008957, "learning_rate": 8.66894577074066e-05, "loss": 0.0114, "step": 22070 }, { "grad_norm": 0.15575549006462097, "learning_rate": 8.667541295052513e-05, "loss": 0.0094, "step": 22080 }, { "grad_norm": 0.21171307563781738, "learning_rate": 8.666136192681865e-05, "loss": 0.0101, "step": 22090 }, { "grad_norm": 0.15530270338058472, "learning_rate": 8.664730463868811e-05, "loss": 0.0127, "step": 22100 }, { "grad_norm": 0.1537129282951355, "learning_rate": 8.663324108853552e-05, "loss": 0.0104, "step": 22110 }, { "grad_norm": 0.1555793583393097, "learning_rate": 8.661917127876395e-05, "loss": 0.0101, "step": 22120 }, { "grad_norm": 0.2067607045173645, "learning_rate": 8.660509521177754e-05, "loss": 0.0142, "step": 22130 }, { "grad_norm": 0.10359224677085876, "learning_rate": 8.65910128899815e-05, "loss": 0.0106, "step": 22140 }, { "grad_norm": 0.1302315592765808, "learning_rate": 8.657692431578214e-05, "loss": 0.01, "step": 22150 }, { "grad_norm": 0.17331545054912567, "learning_rate": 8.656282949158679e-05, "loss": 0.009, "step": 22160 }, { "grad_norm": 0.13677215576171875, "learning_rate": 8.654872841980388e-05, "loss": 0.0102, "step": 22170 }, { "grad_norm": 0.20422004163265228, "learning_rate": 8.653462110284289e-05, "loss": 0.011, "step": 22180 }, { "grad_norm": 0.16545212268829346, "learning_rate": 8.652050754311437e-05, "loss": 0.0098, "step": 22190 }, { "grad_norm": 0.16540829837322235, "learning_rate": 8.650638774302995e-05, "loss": 0.0111, "step": 22200 }, { "grad_norm": 0.12651313841342926, "learning_rate": 8.649226170500233e-05, "loss": 0.0107, "step": 22210 }, { "grad_norm": 0.18528695404529572, "learning_rate": 8.647812943144524e-05, "loss": 0.0133, "step": 22220 }, { "grad_norm": 0.17902134358882904, "learning_rate": 8.646399092477351e-05, "loss": 0.0109, "step": 22230 }, { "grad_norm": 0.10338567197322845, "learning_rate": 8.644984618740301e-05, "loss": 0.0119, "step": 22240 }, { "grad_norm": 0.1398153305053711, "learning_rate": 8.643569522175073e-05, "loss": 0.0171, "step": 22250 }, { "grad_norm": 0.19068825244903564, "learning_rate": 8.642153803023463e-05, "loss": 0.0123, "step": 22260 }, { "grad_norm": 0.1108858659863472, "learning_rate": 8.640737461527383e-05, "loss": 0.0107, "step": 22270 }, { "grad_norm": 0.15128277242183685, "learning_rate": 8.639320497928845e-05, "loss": 0.0101, "step": 22280 }, { "grad_norm": 0.14570985734462738, "learning_rate": 8.637902912469969e-05, "loss": 0.0117, "step": 22290 }, { "grad_norm": 0.09554315358400345, "learning_rate": 8.636484705392982e-05, "loss": 0.0089, "step": 22300 }, { "grad_norm": 0.1392698436975479, "learning_rate": 8.635065876940216e-05, "loss": 0.0112, "step": 22310 }, { "grad_norm": 0.1725059151649475, "learning_rate": 8.633646427354112e-05, "loss": 0.0102, "step": 22320 }, { "grad_norm": 0.13008806109428406, "learning_rate": 8.632226356877213e-05, "loss": 0.0127, "step": 22330 }, { "grad_norm": 0.1397358775138855, "learning_rate": 8.630805665752173e-05, "loss": 0.0118, "step": 22340 }, { "grad_norm": 0.1526365876197815, "learning_rate": 8.629384354221748e-05, "loss": 0.0104, "step": 22350 }, { "grad_norm": 0.17609411478042603, "learning_rate": 8.627962422528797e-05, "loss": 0.011, "step": 22360 }, { "grad_norm": 0.1259680539369583, "learning_rate": 8.626539870916296e-05, "loss": 0.0108, "step": 22370 }, { "grad_norm": 0.13587617874145508, "learning_rate": 8.625116699627317e-05, "loss": 0.0089, "step": 22380 }, { "grad_norm": 0.1988106667995453, "learning_rate": 8.623692908905041e-05, "loss": 0.0127, "step": 22390 }, { "grad_norm": 0.13416683673858643, "learning_rate": 8.622268498992755e-05, "loss": 0.0123, "step": 22400 }, { "grad_norm": 0.1632896065711975, "learning_rate": 8.620843470133851e-05, "loss": 0.0099, "step": 22410 }, { "grad_norm": 0.1655997782945633, "learning_rate": 8.619417822571829e-05, "loss": 0.0122, "step": 22420 }, { "grad_norm": 0.14560270309448242, "learning_rate": 8.617991556550292e-05, "loss": 0.0123, "step": 22430 }, { "grad_norm": 0.1638733297586441, "learning_rate": 8.616564672312952e-05, "loss": 0.0128, "step": 22440 }, { "grad_norm": 0.16956216096878052, "learning_rate": 8.61513717010362e-05, "loss": 0.0105, "step": 22450 }, { "grad_norm": 0.21399244666099548, "learning_rate": 8.613709050166221e-05, "loss": 0.0106, "step": 22460 }, { "grad_norm": 0.13189925253391266, "learning_rate": 8.61228031274478e-05, "loss": 0.0111, "step": 22470 }, { "grad_norm": 0.16133207082748413, "learning_rate": 8.610850958083431e-05, "loss": 0.0091, "step": 22480 }, { "grad_norm": 0.14419029653072357, "learning_rate": 8.609420986426409e-05, "loss": 0.0094, "step": 22490 }, { "grad_norm": 0.13913752138614655, "learning_rate": 8.60799039801806e-05, "loss": 0.011, "step": 22500 }, { "grad_norm": 0.17219369113445282, "learning_rate": 8.606559193102828e-05, "loss": 0.0116, "step": 22510 }, { "grad_norm": 0.13520748913288116, "learning_rate": 8.605127371925273e-05, "loss": 0.01, "step": 22520 }, { "grad_norm": 0.15474188327789307, "learning_rate": 8.603694934730047e-05, "loss": 0.0121, "step": 22530 }, { "grad_norm": 0.199337899684906, "learning_rate": 8.602261881761919e-05, "loss": 0.0111, "step": 22540 }, { "grad_norm": 0.15656618773937225, "learning_rate": 8.600828213265759e-05, "loss": 0.01, "step": 22550 }, { "grad_norm": 0.22464142739772797, "learning_rate": 8.599393929486539e-05, "loss": 0.0107, "step": 22560 }, { "grad_norm": 0.11821101605892181, "learning_rate": 8.59795903066934e-05, "loss": 0.0106, "step": 22570 }, { "grad_norm": 0.13508273661136627, "learning_rate": 8.596523517059347e-05, "loss": 0.0086, "step": 22580 }, { "grad_norm": 0.1592911332845688, "learning_rate": 8.59508738890185e-05, "loss": 0.01, "step": 22590 }, { "grad_norm": 0.16534724831581116, "learning_rate": 8.593650646442246e-05, "loss": 0.0107, "step": 22600 }, { "grad_norm": 0.19822704792022705, "learning_rate": 8.59221328992603e-05, "loss": 0.0099, "step": 22610 }, { "grad_norm": 0.12569478154182434, "learning_rate": 8.590775319598813e-05, "loss": 0.0086, "step": 22620 }, { "grad_norm": 0.12880952656269073, "learning_rate": 8.589336735706301e-05, "loss": 0.0096, "step": 22630 }, { "grad_norm": 0.1634906381368637, "learning_rate": 8.587897538494307e-05, "loss": 0.0115, "step": 22640 }, { "grad_norm": 0.1674504280090332, "learning_rate": 8.586457728208756e-05, "loss": 0.014, "step": 22650 }, { "grad_norm": 0.119742751121521, "learning_rate": 8.585017305095667e-05, "loss": 0.01, "step": 22660 }, { "grad_norm": 0.19915956258773804, "learning_rate": 8.583576269401173e-05, "loss": 0.0149, "step": 22670 }, { "grad_norm": 0.1575595587491989, "learning_rate": 8.582134621371504e-05, "loss": 0.0105, "step": 22680 }, { "grad_norm": 0.20081676542758942, "learning_rate": 8.580692361253e-05, "loss": 0.0121, "step": 22690 }, { "grad_norm": 0.19673828780651093, "learning_rate": 8.579249489292104e-05, "loss": 0.0129, "step": 22700 }, { "grad_norm": 0.19934095442295074, "learning_rate": 8.577806005735363e-05, "loss": 0.0134, "step": 22710 }, { "grad_norm": 0.211992084980011, "learning_rate": 8.576361910829429e-05, "loss": 0.0117, "step": 22720 }, { "grad_norm": 0.1777563989162445, "learning_rate": 8.574917204821057e-05, "loss": 0.0108, "step": 22730 }, { "grad_norm": 0.1642960160970688, "learning_rate": 8.57347188795711e-05, "loss": 0.0122, "step": 22740 }, { "grad_norm": 0.1404837965965271, "learning_rate": 8.572025960484551e-05, "loss": 0.0112, "step": 22750 }, { "grad_norm": 0.1289154291152954, "learning_rate": 8.57057942265045e-05, "loss": 0.0103, "step": 22760 }, { "grad_norm": 0.1503516435623169, "learning_rate": 8.569132274701984e-05, "loss": 0.0123, "step": 22770 }, { "grad_norm": 0.13648921251296997, "learning_rate": 8.567684516886427e-05, "loss": 0.0119, "step": 22780 }, { "grad_norm": 0.15696284174919128, "learning_rate": 8.56623614945116e-05, "loss": 0.0101, "step": 22790 }, { "grad_norm": 0.14632125198841095, "learning_rate": 8.564787172643675e-05, "loss": 0.0097, "step": 22800 }, { "grad_norm": 0.18837596476078033, "learning_rate": 8.563337586711559e-05, "loss": 0.0135, "step": 22810 }, { "grad_norm": 0.1519288867712021, "learning_rate": 8.561887391902506e-05, "loss": 0.0088, "step": 22820 }, { "grad_norm": 0.1925792247056961, "learning_rate": 8.560436588464316e-05, "loss": 0.012, "step": 22830 }, { "grad_norm": 0.1603798121213913, "learning_rate": 8.55898517664489e-05, "loss": 0.0106, "step": 22840 }, { "grad_norm": 0.2120799571275711, "learning_rate": 8.557533156692236e-05, "loss": 0.0119, "step": 22850 }, { "grad_norm": 0.15714965760707855, "learning_rate": 8.556080528854467e-05, "loss": 0.0125, "step": 22860 }, { "grad_norm": 0.1739831268787384, "learning_rate": 8.554627293379791e-05, "loss": 0.0122, "step": 22870 }, { "grad_norm": 0.24144300818443298, "learning_rate": 8.553173450516531e-05, "loss": 0.0112, "step": 22880 }, { "grad_norm": 0.16092686355113983, "learning_rate": 8.551719000513108e-05, "loss": 0.0114, "step": 22890 }, { "grad_norm": 0.16443054378032684, "learning_rate": 8.550263943618049e-05, "loss": 0.0126, "step": 22900 }, { "grad_norm": 0.1066005676984787, "learning_rate": 8.54880828007998e-05, "loss": 0.0117, "step": 22910 }, { "grad_norm": 0.10730506479740143, "learning_rate": 8.547352010147637e-05, "loss": 0.0101, "step": 22920 }, { "grad_norm": 0.16590550541877747, "learning_rate": 8.545895134069855e-05, "loss": 0.0097, "step": 22930 }, { "grad_norm": 0.15284915268421173, "learning_rate": 8.544437652095576e-05, "loss": 0.0116, "step": 22940 }, { "grad_norm": 0.17046532034873962, "learning_rate": 8.542979564473843e-05, "loss": 0.0122, "step": 22950 }, { "grad_norm": 0.20464114844799042, "learning_rate": 8.541520871453802e-05, "loss": 0.0129, "step": 22960 }, { "grad_norm": 0.18676282465457916, "learning_rate": 8.540061573284705e-05, "loss": 0.0129, "step": 22970 }, { "grad_norm": 0.1216818243265152, "learning_rate": 8.538601670215906e-05, "loss": 0.0098, "step": 22980 }, { "grad_norm": 0.1228928342461586, "learning_rate": 8.537141162496864e-05, "loss": 0.009, "step": 22990 }, { "grad_norm": 0.14442762732505798, "learning_rate": 8.535680050377137e-05, "loss": 0.0122, "step": 23000 }, { "grad_norm": 0.17124326527118683, "learning_rate": 8.534218334106391e-05, "loss": 0.011, "step": 23010 }, { "grad_norm": 0.17750465869903564, "learning_rate": 8.532756013934393e-05, "loss": 0.0102, "step": 23020 }, { "grad_norm": 0.12735874950885773, "learning_rate": 8.531293090111012e-05, "loss": 0.0083, "step": 23030 }, { "grad_norm": 0.2815723717212677, "learning_rate": 8.529829562886225e-05, "loss": 0.0123, "step": 23040 }, { "grad_norm": 0.13518014550209045, "learning_rate": 8.528365432510105e-05, "loss": 0.0108, "step": 23050 }, { "grad_norm": 0.1703868955373764, "learning_rate": 8.526900699232833e-05, "loss": 0.0119, "step": 23060 }, { "grad_norm": 0.15397138893604279, "learning_rate": 8.525435363304695e-05, "loss": 0.0107, "step": 23070 }, { "grad_norm": 0.12034644186496735, "learning_rate": 8.523969424976072e-05, "loss": 0.0124, "step": 23080 }, { "grad_norm": 0.17190666496753693, "learning_rate": 8.522502884497457e-05, "loss": 0.0112, "step": 23090 }, { "grad_norm": 0.16687756776809692, "learning_rate": 8.521035742119437e-05, "loss": 0.0101, "step": 23100 }, { "grad_norm": 0.11818434298038483, "learning_rate": 8.519567998092712e-05, "loss": 0.0094, "step": 23110 }, { "grad_norm": 0.15862292051315308, "learning_rate": 8.518099652668075e-05, "loss": 0.0088, "step": 23120 }, { "grad_norm": 0.18161486089229584, "learning_rate": 8.516630706096429e-05, "loss": 0.0097, "step": 23130 }, { "grad_norm": 0.15389806032180786, "learning_rate": 8.515161158628773e-05, "loss": 0.0095, "step": 23140 }, { "grad_norm": 0.10199613124132156, "learning_rate": 8.513691010516216e-05, "loss": 0.0115, "step": 23150 }, { "grad_norm": 0.13009192049503326, "learning_rate": 8.512220262009966e-05, "loss": 0.0106, "step": 23160 }, { "grad_norm": 0.14528606832027435, "learning_rate": 8.510748913361332e-05, "loss": 0.0097, "step": 23170 }, { "grad_norm": 0.15579549968242645, "learning_rate": 8.509276964821726e-05, "loss": 0.0092, "step": 23180 }, { "grad_norm": 0.18560446798801422, "learning_rate": 8.507804416642669e-05, "loss": 0.0122, "step": 23190 }, { "grad_norm": 0.2599985599517822, "learning_rate": 8.506331269075774e-05, "loss": 0.0154, "step": 23200 }, { "grad_norm": 0.19050562381744385, "learning_rate": 8.504857522372765e-05, "loss": 0.0098, "step": 23210 }, { "grad_norm": 0.1625426709651947, "learning_rate": 8.503383176785461e-05, "loss": 0.0094, "step": 23220 }, { "grad_norm": 0.1414598822593689, "learning_rate": 8.501908232565792e-05, "loss": 0.0107, "step": 23230 }, { "grad_norm": 0.1723068654537201, "learning_rate": 8.50043268996578e-05, "loss": 0.0106, "step": 23240 }, { "grad_norm": 0.14029112458229065, "learning_rate": 8.498956549237562e-05, "loss": 0.0078, "step": 23250 }, { "grad_norm": 0.13488425314426422, "learning_rate": 8.497479810633366e-05, "loss": 0.009, "step": 23260 }, { "grad_norm": 0.14028264582157135, "learning_rate": 8.496002474405525e-05, "loss": 0.0098, "step": 23270 }, { "grad_norm": 0.14283417165279388, "learning_rate": 8.494524540806478e-05, "loss": 0.0107, "step": 23280 }, { "grad_norm": 0.16780933737754822, "learning_rate": 8.493046010088761e-05, "loss": 0.01, "step": 23290 }, { "grad_norm": 0.12671217322349548, "learning_rate": 8.491566882505018e-05, "loss": 0.0095, "step": 23300 }, { "grad_norm": 0.14702461659908295, "learning_rate": 8.490087158307988e-05, "loss": 0.0106, "step": 23310 }, { "grad_norm": 0.15553128719329834, "learning_rate": 8.488606837750518e-05, "loss": 0.0124, "step": 23320 }, { "grad_norm": 0.13927322626113892, "learning_rate": 8.487125921085552e-05, "loss": 0.0093, "step": 23330 }, { "grad_norm": 0.13910599052906036, "learning_rate": 8.485644408566141e-05, "loss": 0.0078, "step": 23340 }, { "grad_norm": 0.17507101595401764, "learning_rate": 8.484162300445431e-05, "loss": 0.0097, "step": 23350 }, { "grad_norm": 0.16180962324142456, "learning_rate": 8.482679596976676e-05, "loss": 0.0103, "step": 23360 }, { "grad_norm": 0.10008885711431503, "learning_rate": 8.48119629841323e-05, "loss": 0.0111, "step": 23370 }, { "grad_norm": 0.13130956888198853, "learning_rate": 8.479712405008547e-05, "loss": 0.013, "step": 23380 }, { "grad_norm": 0.137446790933609, "learning_rate": 8.478227917016184e-05, "loss": 0.0103, "step": 23390 }, { "grad_norm": 0.14197660982608795, "learning_rate": 8.476742834689801e-05, "loss": 0.01, "step": 23400 }, { "grad_norm": 0.10334105789661407, "learning_rate": 8.475257158283157e-05, "loss": 0.0104, "step": 23410 }, { "grad_norm": 0.16221357882022858, "learning_rate": 8.473770888050112e-05, "loss": 0.0134, "step": 23420 }, { "grad_norm": 0.16224753856658936, "learning_rate": 8.47228402424463e-05, "loss": 0.0135, "step": 23430 }, { "grad_norm": 0.18125098943710327, "learning_rate": 8.470796567120775e-05, "loss": 0.0117, "step": 23440 }, { "grad_norm": 0.13555870950222015, "learning_rate": 8.469308516932714e-05, "loss": 0.0132, "step": 23450 }, { "grad_norm": 0.16663046181201935, "learning_rate": 8.467819873934714e-05, "loss": 0.0113, "step": 23460 }, { "grad_norm": 0.1983271837234497, "learning_rate": 8.466330638381143e-05, "loss": 0.0114, "step": 23470 }, { "grad_norm": 0.15693539381027222, "learning_rate": 8.464840810526469e-05, "loss": 0.0098, "step": 23480 }, { "grad_norm": 0.1427878439426422, "learning_rate": 8.463350390625264e-05, "loss": 0.0082, "step": 23490 }, { "grad_norm": 0.14768068492412567, "learning_rate": 8.4618593789322e-05, "loss": 0.0115, "step": 23500 }, { "grad_norm": 0.15705183148384094, "learning_rate": 8.46036777570205e-05, "loss": 0.0133, "step": 23510 }, { "grad_norm": 0.20822295546531677, "learning_rate": 8.458875581189688e-05, "loss": 0.0119, "step": 23520 }, { "grad_norm": 0.17864446341991425, "learning_rate": 8.457382795650092e-05, "loss": 0.0134, "step": 23530 }, { "grad_norm": 0.16661295294761658, "learning_rate": 8.455889419338335e-05, "loss": 0.0133, "step": 23540 }, { "grad_norm": 0.14221033453941345, "learning_rate": 8.454395452509593e-05, "loss": 0.0102, "step": 23550 }, { "grad_norm": 0.1513720601797104, "learning_rate": 8.452900895419146e-05, "loss": 0.0079, "step": 23560 }, { "grad_norm": 0.1290135681629181, "learning_rate": 8.451405748322376e-05, "loss": 0.0094, "step": 23570 }, { "grad_norm": 0.1594943404197693, "learning_rate": 8.449910011474759e-05, "loss": 0.0134, "step": 23580 }, { "grad_norm": 0.12956669926643372, "learning_rate": 8.448413685131876e-05, "loss": 0.0111, "step": 23590 }, { "grad_norm": 0.1940804123878479, "learning_rate": 8.446916769549407e-05, "loss": 0.0109, "step": 23600 }, { "grad_norm": 0.14993885159492493, "learning_rate": 8.445419264983136e-05, "loss": 0.0094, "step": 23610 }, { "grad_norm": 0.10945229977369308, "learning_rate": 8.443921171688947e-05, "loss": 0.0082, "step": 23620 }, { "grad_norm": 0.12618309259414673, "learning_rate": 8.442422489922819e-05, "loss": 0.0099, "step": 23630 }, { "grad_norm": 0.14215797185897827, "learning_rate": 8.440923219940838e-05, "loss": 0.0099, "step": 23640 }, { "grad_norm": 0.13577990233898163, "learning_rate": 8.439423361999189e-05, "loss": 0.0085, "step": 23650 }, { "grad_norm": 0.15165652334690094, "learning_rate": 8.437922916354155e-05, "loss": 0.0085, "step": 23660 }, { "grad_norm": 0.16374889016151428, "learning_rate": 8.436421883262123e-05, "loss": 0.0107, "step": 23670 }, { "grad_norm": 0.12207591533660889, "learning_rate": 8.434920262979577e-05, "loss": 0.0081, "step": 23680 }, { "grad_norm": 0.15937501192092896, "learning_rate": 8.433418055763104e-05, "loss": 0.0082, "step": 23690 }, { "grad_norm": 0.14048779010772705, "learning_rate": 8.431915261869389e-05, "loss": 0.0101, "step": 23700 }, { "grad_norm": 0.13290441036224365, "learning_rate": 8.43041188155522e-05, "loss": 0.0087, "step": 23710 }, { "grad_norm": 0.17112773656845093, "learning_rate": 8.428907915077481e-05, "loss": 0.0104, "step": 23720 }, { "grad_norm": 0.20426052808761597, "learning_rate": 8.42740336269316e-05, "loss": 0.0104, "step": 23730 }, { "grad_norm": 0.14968416094779968, "learning_rate": 8.425898224659345e-05, "loss": 0.0123, "step": 23740 }, { "grad_norm": 0.1558118760585785, "learning_rate": 8.42439250123322e-05, "loss": 0.0105, "step": 23750 }, { "grad_norm": 0.1705397069454193, "learning_rate": 8.422886192672076e-05, "loss": 0.012, "step": 23760 }, { "grad_norm": 0.1557684987783432, "learning_rate": 8.421379299233297e-05, "loss": 0.011, "step": 23770 }, { "grad_norm": 0.147865429520607, "learning_rate": 8.419871821174371e-05, "loss": 0.0108, "step": 23780 }, { "grad_norm": 0.16095735132694244, "learning_rate": 8.418363758752884e-05, "loss": 0.0104, "step": 23790 }, { "grad_norm": 0.14705340564250946, "learning_rate": 8.416855112226523e-05, "loss": 0.0111, "step": 23800 }, { "grad_norm": 0.1921127736568451, "learning_rate": 8.415345881853075e-05, "loss": 0.0102, "step": 23810 }, { "grad_norm": 0.15902011096477509, "learning_rate": 8.413836067890426e-05, "loss": 0.0115, "step": 23820 }, { "grad_norm": 0.15214800834655762, "learning_rate": 8.41232567059656e-05, "loss": 0.0087, "step": 23830 }, { "grad_norm": 0.14337395131587982, "learning_rate": 8.410814690229565e-05, "loss": 0.0108, "step": 23840 }, { "grad_norm": 0.1786877065896988, "learning_rate": 8.409303127047626e-05, "loss": 0.0107, "step": 23850 }, { "grad_norm": 0.19587260484695435, "learning_rate": 8.407790981309028e-05, "loss": 0.0093, "step": 23860 }, { "grad_norm": 0.16189736127853394, "learning_rate": 8.406278253272153e-05, "loss": 0.0098, "step": 23870 }, { "grad_norm": 0.15423178672790527, "learning_rate": 8.404764943195487e-05, "loss": 0.0098, "step": 23880 }, { "grad_norm": 0.14720547199249268, "learning_rate": 8.403251051337613e-05, "loss": 0.0113, "step": 23890 }, { "grad_norm": 0.1617274135351181, "learning_rate": 8.401736577957214e-05, "loss": 0.0112, "step": 23900 }, { "grad_norm": 0.16908034682273865, "learning_rate": 8.40022152331307e-05, "loss": 0.0096, "step": 23910 }, { "grad_norm": 0.1547011286020279, "learning_rate": 8.398705887664064e-05, "loss": 0.0095, "step": 23920 }, { "grad_norm": 0.15325768291950226, "learning_rate": 8.397189671269177e-05, "loss": 0.0088, "step": 23930 }, { "grad_norm": 0.16092990338802338, "learning_rate": 8.395672874387488e-05, "loss": 0.0109, "step": 23940 }, { "grad_norm": 0.18239367008209229, "learning_rate": 8.394155497278177e-05, "loss": 0.0112, "step": 23950 }, { "grad_norm": 0.12603570520877838, "learning_rate": 8.392637540200523e-05, "loss": 0.012, "step": 23960 }, { "grad_norm": 0.11097846180200577, "learning_rate": 8.391119003413902e-05, "loss": 0.0077, "step": 23970 }, { "grad_norm": 0.18259654939174652, "learning_rate": 8.38959988717779e-05, "loss": 0.0081, "step": 23980 }, { "grad_norm": 0.15965016186237335, "learning_rate": 8.388080191751764e-05, "loss": 0.01, "step": 23990 }, { "grad_norm": 0.14323502779006958, "learning_rate": 8.386559917395496e-05, "loss": 0.0101, "step": 24000 }, { "grad_norm": 0.10483338683843613, "learning_rate": 8.385039064368761e-05, "loss": 0.0094, "step": 24010 }, { "grad_norm": 0.13263168931007385, "learning_rate": 8.383517632931431e-05, "loss": 0.0101, "step": 24020 }, { "grad_norm": 0.13581937551498413, "learning_rate": 8.381995623343477e-05, "loss": 0.0087, "step": 24030 }, { "grad_norm": 0.18687419593334198, "learning_rate": 8.380473035864968e-05, "loss": 0.0093, "step": 24040 }, { "grad_norm": 0.1380118429660797, "learning_rate": 8.378949870756076e-05, "loss": 0.0108, "step": 24050 }, { "grad_norm": 0.1416977494955063, "learning_rate": 8.377426128277063e-05, "loss": 0.0117, "step": 24060 }, { "grad_norm": 0.1517494022846222, "learning_rate": 8.375901808688298e-05, "loss": 0.0093, "step": 24070 }, { "grad_norm": 0.17509661614894867, "learning_rate": 8.374376912250246e-05, "loss": 0.0093, "step": 24080 }, { "grad_norm": 0.22355270385742188, "learning_rate": 8.372851439223468e-05, "loss": 0.0123, "step": 24090 }, { "grad_norm": 0.11785156279802322, "learning_rate": 8.371325389868627e-05, "loss": 0.0102, "step": 24100 }, { "grad_norm": 0.1434745192527771, "learning_rate": 8.369798764446482e-05, "loss": 0.011, "step": 24110 }, { "grad_norm": 0.12353044003248215, "learning_rate": 8.368271563217893e-05, "loss": 0.012, "step": 24120 }, { "grad_norm": 0.12611830234527588, "learning_rate": 8.366743786443817e-05, "loss": 0.01, "step": 24130 }, { "grad_norm": 0.1470280885696411, "learning_rate": 8.365215434385309e-05, "loss": 0.0099, "step": 24140 }, { "grad_norm": 0.09148278087377548, "learning_rate": 8.36368650730352e-05, "loss": 0.009, "step": 24150 }, { "grad_norm": 0.11353746056556702, "learning_rate": 8.362157005459705e-05, "loss": 0.0104, "step": 24160 }, { "grad_norm": 0.1737150400876999, "learning_rate": 8.360626929115213e-05, "loss": 0.0091, "step": 24170 }, { "grad_norm": 0.17463766038417816, "learning_rate": 8.359096278531492e-05, "loss": 0.0139, "step": 24180 }, { "grad_norm": 0.14651630818843842, "learning_rate": 8.357565053970088e-05, "loss": 0.0144, "step": 24190 }, { "grad_norm": 0.15399345755577087, "learning_rate": 8.356033255692647e-05, "loss": 0.0098, "step": 24200 }, { "grad_norm": 0.19062983989715576, "learning_rate": 8.354500883960911e-05, "loss": 0.0099, "step": 24210 }, { "grad_norm": 0.17598429322242737, "learning_rate": 8.352967939036717e-05, "loss": 0.0148, "step": 24220 }, { "grad_norm": 0.1774178296327591, "learning_rate": 8.35143442118201e-05, "loss": 0.0126, "step": 24230 }, { "grad_norm": 0.1517697423696518, "learning_rate": 8.349900330658819e-05, "loss": 0.0117, "step": 24240 }, { "grad_norm": 0.17545674741268158, "learning_rate": 8.348365667729284e-05, "loss": 0.0115, "step": 24250 }, { "grad_norm": 0.16048812866210938, "learning_rate": 8.346830432655633e-05, "loss": 0.0105, "step": 24260 }, { "grad_norm": 0.17809148132801056, "learning_rate": 8.345294625700195e-05, "loss": 0.0095, "step": 24270 }, { "grad_norm": 0.14733995497226715, "learning_rate": 8.343758247125402e-05, "loss": 0.0101, "step": 24280 }, { "grad_norm": 0.16804048418998718, "learning_rate": 8.342221297193776e-05, "loss": 0.0138, "step": 24290 }, { "grad_norm": 0.10113783180713654, "learning_rate": 8.34068377616794e-05, "loss": 0.0113, "step": 24300 }, { "grad_norm": 0.1599409282207489, "learning_rate": 8.339145684310615e-05, "loss": 0.0116, "step": 24310 }, { "grad_norm": 0.12537330389022827, "learning_rate": 8.337607021884618e-05, "loss": 0.0105, "step": 24320 }, { "grad_norm": 0.13252876698970795, "learning_rate": 8.336067789152867e-05, "loss": 0.0107, "step": 24330 }, { "grad_norm": 0.12772402167320251, "learning_rate": 8.334527986378369e-05, "loss": 0.0104, "step": 24340 }, { "grad_norm": 0.15124216675758362, "learning_rate": 8.332987613824239e-05, "loss": 0.0101, "step": 24350 }, { "grad_norm": 0.1698225885629654, "learning_rate": 8.331446671753685e-05, "loss": 0.0119, "step": 24360 }, { "grad_norm": 0.1727313995361328, "learning_rate": 8.329905160430007e-05, "loss": 0.0118, "step": 24370 }, { "grad_norm": 0.14761611819267273, "learning_rate": 8.328363080116611e-05, "loss": 0.0108, "step": 24380 }, { "grad_norm": 0.1839747577905655, "learning_rate": 8.326820431076997e-05, "loss": 0.0116, "step": 24390 }, { "grad_norm": 0.13836245238780975, "learning_rate": 8.325277213574759e-05, "loss": 0.0105, "step": 24400 }, { "grad_norm": 0.15717944502830505, "learning_rate": 8.32373342787359e-05, "loss": 0.0113, "step": 24410 }, { "grad_norm": 0.1496301293373108, "learning_rate": 8.322189074237285e-05, "loss": 0.0099, "step": 24420 }, { "grad_norm": 0.14464478194713593, "learning_rate": 8.32064415292973e-05, "loss": 0.0117, "step": 24430 }, { "grad_norm": 0.16556864976882935, "learning_rate": 8.319098664214907e-05, "loss": 0.013, "step": 24440 }, { "grad_norm": 0.15709631145000458, "learning_rate": 8.3175526083569e-05, "loss": 0.0091, "step": 24450 }, { "grad_norm": 0.1407906860113144, "learning_rate": 8.316005985619889e-05, "loss": 0.0096, "step": 24460 }, { "grad_norm": 0.17710749804973602, "learning_rate": 8.314458796268147e-05, "loss": 0.0103, "step": 24470 }, { "grad_norm": 0.11971883475780487, "learning_rate": 8.312911040566047e-05, "loss": 0.0104, "step": 24480 }, { "grad_norm": 0.21361811459064484, "learning_rate": 8.31136271877806e-05, "loss": 0.0125, "step": 24490 }, { "grad_norm": 0.20627593994140625, "learning_rate": 8.309813831168748e-05, "loss": 0.013, "step": 24500 }, { "grad_norm": 0.13928496837615967, "learning_rate": 8.308264378002777e-05, "loss": 0.0112, "step": 24510 }, { "grad_norm": 0.17538483440876007, "learning_rate": 8.306714359544906e-05, "loss": 0.0121, "step": 24520 }, { "grad_norm": 0.13537782430648804, "learning_rate": 8.30516377605999e-05, "loss": 0.0103, "step": 24530 }, { "grad_norm": 0.17773425579071045, "learning_rate": 8.30361262781298e-05, "loss": 0.0148, "step": 24540 }, { "grad_norm": 0.13497553765773773, "learning_rate": 8.302060915068924e-05, "loss": 0.0111, "step": 24550 }, { "grad_norm": 0.17739364504814148, "learning_rate": 8.300508638092972e-05, "loss": 0.0113, "step": 24560 }, { "grad_norm": 0.11115768551826477, "learning_rate": 8.298955797150361e-05, "loss": 0.0124, "step": 24570 }, { "grad_norm": 0.13792870938777924, "learning_rate": 8.297402392506433e-05, "loss": 0.0091, "step": 24580 }, { "grad_norm": 0.12319064885377884, "learning_rate": 8.295848424426617e-05, "loss": 0.0084, "step": 24590 }, { "grad_norm": 0.13299769163131714, "learning_rate": 8.29429389317645e-05, "loss": 0.0093, "step": 24600 }, { "grad_norm": 0.18574301898479462, "learning_rate": 8.292738799021556e-05, "loss": 0.0095, "step": 24610 }, { "grad_norm": 0.14203014969825745, "learning_rate": 8.291183142227656e-05, "loss": 0.0104, "step": 24620 }, { "grad_norm": 0.19154758751392365, "learning_rate": 8.289626923060572e-05, "loss": 0.009, "step": 24630 }, { "grad_norm": 0.11926411837339401, "learning_rate": 8.288070141786218e-05, "loss": 0.0081, "step": 24640 }, { "grad_norm": 0.13008737564086914, "learning_rate": 8.286512798670605e-05, "loss": 0.0092, "step": 24650 }, { "grad_norm": 0.15168249607086182, "learning_rate": 8.284954893979842e-05, "loss": 0.0093, "step": 24660 }, { "grad_norm": 0.16561923921108246, "learning_rate": 8.283396427980131e-05, "loss": 0.01, "step": 24670 }, { "grad_norm": 0.15106594562530518, "learning_rate": 8.281837400937771e-05, "loss": 0.0111, "step": 24680 }, { "grad_norm": 0.15620893239974976, "learning_rate": 8.28027781311916e-05, "loss": 0.0127, "step": 24690 }, { "grad_norm": 0.12938189506530762, "learning_rate": 8.278717664790785e-05, "loss": 0.0098, "step": 24700 }, { "grad_norm": 0.16284866631031036, "learning_rate": 8.277156956219234e-05, "loss": 0.009, "step": 24710 }, { "grad_norm": 0.18671059608459473, "learning_rate": 8.275595687671189e-05, "loss": 0.0128, "step": 24720 }, { "grad_norm": 0.15967310965061188, "learning_rate": 8.27403385941343e-05, "loss": 0.0102, "step": 24730 }, { "grad_norm": 0.15908044576644897, "learning_rate": 8.272471471712828e-05, "loss": 0.0096, "step": 24740 }, { "grad_norm": 0.12673422694206238, "learning_rate": 8.270908524836355e-05, "loss": 0.0103, "step": 24750 }, { "grad_norm": 0.21680207550525665, "learning_rate": 8.269345019051074e-05, "loss": 0.0132, "step": 24760 }, { "grad_norm": 0.16773949563503265, "learning_rate": 8.267780954624147e-05, "loss": 0.0093, "step": 24770 }, { "grad_norm": 0.13706983625888824, "learning_rate": 8.266216331822827e-05, "loss": 0.0097, "step": 24780 }, { "grad_norm": 0.12358648329973221, "learning_rate": 8.264651150914469e-05, "loss": 0.011, "step": 24790 }, { "grad_norm": 0.1535429209470749, "learning_rate": 8.263085412166517e-05, "loss": 0.0102, "step": 24800 }, { "grad_norm": 0.21285320818424225, "learning_rate": 8.261519115846514e-05, "loss": 0.0126, "step": 24810 }, { "grad_norm": 0.1915537714958191, "learning_rate": 8.259952262222096e-05, "loss": 0.0087, "step": 24820 }, { "grad_norm": 0.17236806452274323, "learning_rate": 8.258384851560997e-05, "loss": 0.0086, "step": 24830 }, { "grad_norm": 0.157316192984581, "learning_rate": 8.256816884131044e-05, "loss": 0.0093, "step": 24840 }, { "grad_norm": 0.13758531212806702, "learning_rate": 8.255248360200159e-05, "loss": 0.0106, "step": 24850 }, { "grad_norm": 0.10450322180986404, "learning_rate": 8.253679280036359e-05, "loss": 0.0078, "step": 24860 }, { "grad_norm": 0.1412089616060257, "learning_rate": 8.252109643907762e-05, "loss": 0.0089, "step": 24870 }, { "grad_norm": 0.13349968194961548, "learning_rate": 8.250539452082569e-05, "loss": 0.0095, "step": 24880 }, { "grad_norm": 0.12169268727302551, "learning_rate": 8.248968704829087e-05, "loss": 0.0073, "step": 24890 }, { "grad_norm": 0.12597057223320007, "learning_rate": 8.247397402415714e-05, "loss": 0.0089, "step": 24900 }, { "grad_norm": 0.15225152671337128, "learning_rate": 8.24582554511094e-05, "loss": 0.0089, "step": 24910 }, { "grad_norm": 0.11615768820047379, "learning_rate": 8.244253133183355e-05, "loss": 0.0116, "step": 24920 }, { "grad_norm": 0.19797921180725098, "learning_rate": 8.24268016690164e-05, "loss": 0.009, "step": 24930 }, { "grad_norm": 0.15516403317451477, "learning_rate": 8.241106646534571e-05, "loss": 0.0097, "step": 24940 }, { "grad_norm": 0.14782269299030304, "learning_rate": 8.23953257235102e-05, "loss": 0.0099, "step": 24950 }, { "grad_norm": 0.13438542187213898, "learning_rate": 8.237957944619956e-05, "loss": 0.0099, "step": 24960 }, { "grad_norm": 0.17542721331119537, "learning_rate": 8.236382763610437e-05, "loss": 0.0087, "step": 24970 }, { "grad_norm": 0.1666145771741867, "learning_rate": 8.234807029591619e-05, "loss": 0.0115, "step": 24980 }, { "grad_norm": 0.14120206236839294, "learning_rate": 8.233230742832752e-05, "loss": 0.0115, "step": 24990 }, { "grad_norm": 0.13503733277320862, "learning_rate": 8.231653903603178e-05, "loss": 0.0102, "step": 25000 }, { "grad_norm": 0.13983723521232605, "learning_rate": 8.23007651217234e-05, "loss": 0.0095, "step": 25010 }, { "grad_norm": 0.16855131089687347, "learning_rate": 8.228498568809769e-05, "loss": 0.0097, "step": 25020 }, { "grad_norm": 0.16464866697788239, "learning_rate": 8.22692007378509e-05, "loss": 0.0107, "step": 25030 }, { "grad_norm": 0.14742670953273773, "learning_rate": 8.225341027368028e-05, "loss": 0.0097, "step": 25040 }, { "grad_norm": 0.17585910856723785, "learning_rate": 8.223761429828399e-05, "loss": 0.0107, "step": 25050 }, { "grad_norm": 0.14321257174015045, "learning_rate": 8.22218128143611e-05, "loss": 0.0109, "step": 25060 }, { "grad_norm": 0.195227712392807, "learning_rate": 8.220600582461166e-05, "loss": 0.0123, "step": 25070 }, { "grad_norm": 0.16069604456424713, "learning_rate": 8.219019333173668e-05, "loss": 0.0105, "step": 25080 }, { "grad_norm": 0.15125444531440735, "learning_rate": 8.217437533843805e-05, "loss": 0.0097, "step": 25090 }, { "grad_norm": 0.1466570496559143, "learning_rate": 8.215855184741867e-05, "loss": 0.0112, "step": 25100 }, { "grad_norm": 0.12974727153778076, "learning_rate": 8.21427228613823e-05, "loss": 0.0089, "step": 25110 }, { "grad_norm": 0.13349570333957672, "learning_rate": 8.21268883830337e-05, "loss": 0.0116, "step": 25120 }, { "grad_norm": 0.17342422902584076, "learning_rate": 8.211104841507855e-05, "loss": 0.0103, "step": 25130 }, { "grad_norm": 0.2547677159309387, "learning_rate": 8.209520296022346e-05, "loss": 0.0116, "step": 25140 }, { "grad_norm": 0.16513535380363464, "learning_rate": 8.207935202117599e-05, "loss": 0.0109, "step": 25150 }, { "grad_norm": 0.13122646510601044, "learning_rate": 8.206349560064463e-05, "loss": 0.0092, "step": 25160 }, { "grad_norm": 0.15502069890499115, "learning_rate": 8.204763370133881e-05, "loss": 0.009, "step": 25170 }, { "grad_norm": 0.1517941951751709, "learning_rate": 8.203176632596892e-05, "loss": 0.0102, "step": 25180 }, { "grad_norm": 0.14845994114875793, "learning_rate": 8.20158934772462e-05, "loss": 0.0099, "step": 25190 }, { "grad_norm": 0.14316321909427643, "learning_rate": 8.200001515788294e-05, "loss": 0.0079, "step": 25200 }, { "grad_norm": 0.1590443253517151, "learning_rate": 8.198413137059228e-05, "loss": 0.0085, "step": 25210 }, { "grad_norm": 0.1784525066614151, "learning_rate": 8.196824211808835e-05, "loss": 0.0121, "step": 25220 }, { "grad_norm": 0.15754899382591248, "learning_rate": 8.195234740308617e-05, "loss": 0.0108, "step": 25230 }, { "grad_norm": 0.1929130107164383, "learning_rate": 8.193644722830171e-05, "loss": 0.0101, "step": 25240 }, { "grad_norm": 0.13501562178134918, "learning_rate": 8.19205415964519e-05, "loss": 0.0077, "step": 25250 }, { "grad_norm": 0.12885905802249908, "learning_rate": 8.190463051025456e-05, "loss": 0.0083, "step": 25260 }, { "grad_norm": 0.12255407869815826, "learning_rate": 8.188871397242843e-05, "loss": 0.0114, "step": 25270 }, { "grad_norm": 0.12494876235723495, "learning_rate": 8.187279198569326e-05, "loss": 0.0096, "step": 25280 }, { "grad_norm": 0.14013954997062683, "learning_rate": 8.185686455276966e-05, "loss": 0.012, "step": 25290 }, { "grad_norm": 0.1307227462530136, "learning_rate": 8.184093167637921e-05, "loss": 0.0103, "step": 25300 }, { "grad_norm": 0.126358762383461, "learning_rate": 8.182499335924437e-05, "loss": 0.0126, "step": 25310 }, { "grad_norm": 0.1527029424905777, "learning_rate": 8.18090496040886e-05, "loss": 0.0108, "step": 25320 }, { "grad_norm": 0.15351319313049316, "learning_rate": 8.179310041363621e-05, "loss": 0.0087, "step": 25330 }, { "grad_norm": 0.10927922278642654, "learning_rate": 8.17771457906125e-05, "loss": 0.0086, "step": 25340 }, { "grad_norm": 0.11321414262056351, "learning_rate": 8.176118573774371e-05, "loss": 0.0082, "step": 25350 }, { "grad_norm": 0.22858428955078125, "learning_rate": 8.174522025775692e-05, "loss": 0.0094, "step": 25360 }, { "grad_norm": 0.18876735866069794, "learning_rate": 8.172924935338022e-05, "loss": 0.0119, "step": 25370 }, { "grad_norm": 0.18043631315231323, "learning_rate": 8.171327302734262e-05, "loss": 0.0112, "step": 25380 }, { "grad_norm": 0.13098543882369995, "learning_rate": 8.169729128237401e-05, "loss": 0.0112, "step": 25390 }, { "grad_norm": 0.14522893726825714, "learning_rate": 8.168130412120525e-05, "loss": 0.0112, "step": 25400 }, { "grad_norm": 0.1412489414215088, "learning_rate": 8.16653115465681e-05, "loss": 0.0136, "step": 25410 }, { "grad_norm": 0.11026723682880402, "learning_rate": 8.164931356119526e-05, "loss": 0.0123, "step": 25420 }, { "grad_norm": 0.1112285777926445, "learning_rate": 8.163331016782032e-05, "loss": 0.0092, "step": 25430 }, { "grad_norm": 0.12769725918769836, "learning_rate": 8.161730136917785e-05, "loss": 0.0127, "step": 25440 }, { "grad_norm": 0.13059696555137634, "learning_rate": 8.160128716800333e-05, "loss": 0.0076, "step": 25450 }, { "grad_norm": 0.12935422360897064, "learning_rate": 8.158526756703313e-05, "loss": 0.0119, "step": 25460 }, { "grad_norm": 0.1317603588104248, "learning_rate": 8.156924256900455e-05, "loss": 0.0133, "step": 25470 }, { "grad_norm": 0.1555503010749817, "learning_rate": 8.155321217665584e-05, "loss": 0.0123, "step": 25480 }, { "grad_norm": 0.18233250081539154, "learning_rate": 8.153717639272614e-05, "loss": 0.0122, "step": 25490 }, { "grad_norm": 0.20601046085357666, "learning_rate": 8.152113521995555e-05, "loss": 0.0098, "step": 25500 }, { "grad_norm": 0.14304335415363312, "learning_rate": 8.150508866108505e-05, "loss": 0.011, "step": 25510 }, { "grad_norm": 0.14632956683635712, "learning_rate": 8.148903671885657e-05, "loss": 0.0125, "step": 25520 }, { "grad_norm": 0.1713321954011917, "learning_rate": 8.147297939601292e-05, "loss": 0.012, "step": 25530 }, { "grad_norm": 0.1893218457698822, "learning_rate": 8.145691669529792e-05, "loss": 0.0136, "step": 25540 }, { "grad_norm": 0.1504809409379959, "learning_rate": 8.144084861945618e-05, "loss": 0.0085, "step": 25550 }, { "grad_norm": 0.14891861379146576, "learning_rate": 8.142477517123333e-05, "loss": 0.0112, "step": 25560 }, { "grad_norm": 0.12118101119995117, "learning_rate": 8.140869635337586e-05, "loss": 0.0113, "step": 25570 }, { "grad_norm": 0.1244325265288353, "learning_rate": 8.139261216863123e-05, "loss": 0.0114, "step": 25580 }, { "grad_norm": 0.1621164083480835, "learning_rate": 8.137652261974776e-05, "loss": 0.012, "step": 25590 }, { "grad_norm": 0.1626526117324829, "learning_rate": 8.136042770947472e-05, "loss": 0.0128, "step": 25600 }, { "grad_norm": 0.16186940670013428, "learning_rate": 8.134432744056228e-05, "loss": 0.0095, "step": 25610 }, { "grad_norm": 0.15709134936332703, "learning_rate": 8.132822181576158e-05, "loss": 0.0113, "step": 25620 }, { "grad_norm": 0.16744635999202728, "learning_rate": 8.131211083782459e-05, "loss": 0.012, "step": 25630 }, { "grad_norm": 0.14942659437656403, "learning_rate": 8.129599450950424e-05, "loss": 0.0098, "step": 25640 }, { "grad_norm": 0.12680236995220184, "learning_rate": 8.127987283355438e-05, "loss": 0.0095, "step": 25650 }, { "grad_norm": 0.13645413517951965, "learning_rate": 8.126374581272976e-05, "loss": 0.0104, "step": 25660 }, { "grad_norm": 0.10659750550985336, "learning_rate": 8.124761344978605e-05, "loss": 0.0134, "step": 25670 }, { "grad_norm": 0.13136345148086548, "learning_rate": 8.12314757474798e-05, "loss": 0.011, "step": 25680 }, { "grad_norm": 0.1231384128332138, "learning_rate": 8.121533270856856e-05, "loss": 0.008, "step": 25690 }, { "grad_norm": 0.1708422154188156, "learning_rate": 8.119918433581069e-05, "loss": 0.0109, "step": 25700 }, { "grad_norm": 0.14400498569011688, "learning_rate": 8.118303063196551e-05, "loss": 0.0111, "step": 25710 }, { "grad_norm": 0.10860549658536911, "learning_rate": 8.116687159979326e-05, "loss": 0.0086, "step": 25720 }, { "grad_norm": 0.16673347353935242, "learning_rate": 8.115070724205508e-05, "loss": 0.0116, "step": 25730 }, { "grad_norm": 0.21856272220611572, "learning_rate": 8.113453756151296e-05, "loss": 0.0122, "step": 25740 }, { "grad_norm": 0.18213866651058197, "learning_rate": 8.111836256092995e-05, "loss": 0.0121, "step": 25750 }, { "grad_norm": 0.2073948234319687, "learning_rate": 8.110218224306985e-05, "loss": 0.0112, "step": 25760 }, { "grad_norm": 0.12119423598051071, "learning_rate": 8.108599661069745e-05, "loss": 0.0095, "step": 25770 }, { "grad_norm": 0.1532314121723175, "learning_rate": 8.106980566657845e-05, "loss": 0.0106, "step": 25780 }, { "grad_norm": 0.16381168365478516, "learning_rate": 8.10536094134794e-05, "loss": 0.0101, "step": 25790 }, { "grad_norm": 0.09381364285945892, "learning_rate": 8.103740785416783e-05, "loss": 0.0089, "step": 25800 }, { "grad_norm": 0.13957759737968445, "learning_rate": 8.102120099141212e-05, "loss": 0.0093, "step": 25810 }, { "grad_norm": 0.13430450856685638, "learning_rate": 8.100498882798163e-05, "loss": 0.0102, "step": 25820 }, { "grad_norm": 0.10939735174179077, "learning_rate": 8.09887713666465e-05, "loss": 0.012, "step": 25830 }, { "grad_norm": 0.11790559440851212, "learning_rate": 8.09725486101779e-05, "loss": 0.0094, "step": 25840 }, { "grad_norm": 0.11198148876428604, "learning_rate": 8.095632056134784e-05, "loss": 0.0112, "step": 25850 }, { "grad_norm": 0.15457703173160553, "learning_rate": 8.094008722292925e-05, "loss": 0.0144, "step": 25860 }, { "grad_norm": 0.13146713376045227, "learning_rate": 8.092384859769598e-05, "loss": 0.0091, "step": 25870 }, { "grad_norm": 0.14049267768859863, "learning_rate": 8.090760468842275e-05, "loss": 0.0124, "step": 25880 }, { "grad_norm": 0.16035234928131104, "learning_rate": 8.089135549788521e-05, "loss": 0.01, "step": 25890 }, { "grad_norm": 0.16312086582183838, "learning_rate": 8.087510102885987e-05, "loss": 0.0099, "step": 25900 }, { "grad_norm": 0.1834365874528885, "learning_rate": 8.085884128412422e-05, "loss": 0.012, "step": 25910 }, { "grad_norm": 0.15273599326610565, "learning_rate": 8.084257626645659e-05, "loss": 0.0113, "step": 25920 }, { "grad_norm": 0.1274670511484146, "learning_rate": 8.08263059786362e-05, "loss": 0.0082, "step": 25930 }, { "grad_norm": 0.10886474698781967, "learning_rate": 8.081003042344325e-05, "loss": 0.0082, "step": 25940 }, { "grad_norm": 0.16828545928001404, "learning_rate": 8.079374960365872e-05, "loss": 0.0099, "step": 25950 }, { "grad_norm": 0.1294051706790924, "learning_rate": 8.077746352206463e-05, "loss": 0.0084, "step": 25960 }, { "grad_norm": 0.1222711130976677, "learning_rate": 8.076117218144377e-05, "loss": 0.0079, "step": 25970 }, { "grad_norm": 0.12425137311220169, "learning_rate": 8.074487558457991e-05, "loss": 0.0098, "step": 25980 }, { "grad_norm": 0.15863816440105438, "learning_rate": 8.072857373425768e-05, "loss": 0.0091, "step": 25990 }, { "grad_norm": 0.13462965190410614, "learning_rate": 8.071226663326264e-05, "loss": 0.0089, "step": 26000 }, { "grad_norm": 0.1774282157421112, "learning_rate": 8.069595428438121e-05, "loss": 0.0143, "step": 26010 }, { "grad_norm": 0.13477583229541779, "learning_rate": 8.067963669040072e-05, "loss": 0.0087, "step": 26020 }, { "grad_norm": 0.1475704461336136, "learning_rate": 8.066331385410942e-05, "loss": 0.0112, "step": 26030 }, { "grad_norm": 0.14023436605930328, "learning_rate": 8.064698577829641e-05, "loss": 0.0097, "step": 26040 }, { "grad_norm": 0.17832162976264954, "learning_rate": 8.063065246575175e-05, "loss": 0.0083, "step": 26050 }, { "grad_norm": 0.1345026195049286, "learning_rate": 8.061431391926631e-05, "loss": 0.0104, "step": 26060 }, { "grad_norm": 0.15427522361278534, "learning_rate": 8.059797014163195e-05, "loss": 0.0114, "step": 26070 }, { "grad_norm": 0.12571701407432556, "learning_rate": 8.058162113564133e-05, "loss": 0.0118, "step": 26080 }, { "grad_norm": 0.17707139253616333, "learning_rate": 8.056526690408806e-05, "loss": 0.0114, "step": 26090 }, { "grad_norm": 0.17567896842956543, "learning_rate": 8.054890744976666e-05, "loss": 0.0097, "step": 26100 }, { "grad_norm": 0.17541758716106415, "learning_rate": 8.053254277547248e-05, "loss": 0.0094, "step": 26110 }, { "grad_norm": 0.13958095014095306, "learning_rate": 8.051617288400182e-05, "loss": 0.0084, "step": 26120 }, { "grad_norm": 0.1548875868320465, "learning_rate": 8.049979777815182e-05, "loss": 0.0101, "step": 26130 }, { "grad_norm": 0.1420048028230667, "learning_rate": 8.048341746072054e-05, "loss": 0.0104, "step": 26140 }, { "grad_norm": 0.1212569922208786, "learning_rate": 8.046703193450696e-05, "loss": 0.0122, "step": 26150 }, { "grad_norm": 0.12857390940189362, "learning_rate": 8.04506412023109e-05, "loss": 0.0081, "step": 26160 }, { "grad_norm": 0.13382816314697266, "learning_rate": 8.043424526693306e-05, "loss": 0.0114, "step": 26170 }, { "grad_norm": 0.13129359483718872, "learning_rate": 8.04178441311751e-05, "loss": 0.0081, "step": 26180 }, { "grad_norm": 0.13565169274806976, "learning_rate": 8.04014377978395e-05, "loss": 0.0081, "step": 26190 }, { "grad_norm": 0.1493297964334488, "learning_rate": 8.038502626972967e-05, "loss": 0.0123, "step": 26200 }, { "grad_norm": 0.19704362750053406, "learning_rate": 8.036860954964989e-05, "loss": 0.0118, "step": 26210 }, { "grad_norm": 0.10672622919082642, "learning_rate": 8.035218764040531e-05, "loss": 0.01, "step": 26220 }, { "grad_norm": 0.18836288154125214, "learning_rate": 8.033576054480199e-05, "loss": 0.014, "step": 26230 }, { "grad_norm": 0.13820123672485352, "learning_rate": 8.031932826564688e-05, "loss": 0.0093, "step": 26240 }, { "grad_norm": 0.14444415271282196, "learning_rate": 8.030289080574782e-05, "loss": 0.0081, "step": 26250 }, { "grad_norm": 0.12577934563159943, "learning_rate": 8.028644816791349e-05, "loss": 0.0096, "step": 26260 }, { "grad_norm": 0.11194011569023132, "learning_rate": 8.027000035495351e-05, "loss": 0.0087, "step": 26270 }, { "grad_norm": 0.12227760255336761, "learning_rate": 8.025354736967836e-05, "loss": 0.0082, "step": 26280 }, { "grad_norm": 0.14408771693706512, "learning_rate": 8.023708921489941e-05, "loss": 0.0086, "step": 26290 }, { "grad_norm": 0.12226834148168564, "learning_rate": 8.022062589342887e-05, "loss": 0.0115, "step": 26300 }, { "grad_norm": 0.16838142275810242, "learning_rate": 8.020415740807993e-05, "loss": 0.0087, "step": 26310 }, { "grad_norm": 0.119357168674469, "learning_rate": 8.018768376166656e-05, "loss": 0.0116, "step": 26320 }, { "grad_norm": 0.15305253863334656, "learning_rate": 8.017120495700368e-05, "loss": 0.0118, "step": 26330 }, { "grad_norm": 0.12433955818414688, "learning_rate": 8.015472099690704e-05, "loss": 0.0106, "step": 26340 }, { "grad_norm": 0.10989074409008026, "learning_rate": 8.013823188419332e-05, "loss": 0.0119, "step": 26350 }, { "grad_norm": 0.1775762438774109, "learning_rate": 8.012173762168006e-05, "loss": 0.0098, "step": 26360 }, { "grad_norm": 0.1457039713859558, "learning_rate": 8.010523821218567e-05, "loss": 0.0089, "step": 26370 }, { "grad_norm": 0.13503436744213104, "learning_rate": 8.008873365852945e-05, "loss": 0.0116, "step": 26380 }, { "grad_norm": 0.14314144849777222, "learning_rate": 8.007222396353157e-05, "loss": 0.0114, "step": 26390 }, { "grad_norm": 0.10307356715202332, "learning_rate": 8.00557091300131e-05, "loss": 0.0092, "step": 26400 }, { "grad_norm": 0.09913527965545654, "learning_rate": 8.003918916079597e-05, "loss": 0.0111, "step": 26410 }, { "grad_norm": 0.16825753450393677, "learning_rate": 8.002266405870298e-05, "loss": 0.0105, "step": 26420 }, { "grad_norm": 0.11362767219543457, "learning_rate": 8.000613382655782e-05, "loss": 0.0095, "step": 26430 }, { "grad_norm": 0.13079236447811127, "learning_rate": 7.998959846718505e-05, "loss": 0.0089, "step": 26440 }, { "grad_norm": 0.12136693298816681, "learning_rate": 7.997305798341012e-05, "loss": 0.0086, "step": 26450 }, { "grad_norm": 0.16225886344909668, "learning_rate": 7.995651237805937e-05, "loss": 0.0087, "step": 26460 }, { "grad_norm": 0.1518617570400238, "learning_rate": 7.993996165395996e-05, "loss": 0.0092, "step": 26470 }, { "grad_norm": 0.13423503935337067, "learning_rate": 7.992340581393996e-05, "loss": 0.0083, "step": 26480 }, { "grad_norm": 0.20525513589382172, "learning_rate": 7.990684486082831e-05, "loss": 0.0099, "step": 26490 }, { "grad_norm": 0.11508144438266754, "learning_rate": 7.989027879745482e-05, "loss": 0.0086, "step": 26500 }, { "grad_norm": 0.1602199524641037, "learning_rate": 7.98737076266502e-05, "loss": 0.0106, "step": 26510 }, { "grad_norm": 0.12633277475833893, "learning_rate": 7.985713135124598e-05, "loss": 0.0091, "step": 26520 }, { "grad_norm": 0.1003943458199501, "learning_rate": 7.98405499740746e-05, "loss": 0.0091, "step": 26530 }, { "grad_norm": 0.15843644738197327, "learning_rate": 7.98239634979694e-05, "loss": 0.0111, "step": 26540 }, { "grad_norm": 0.16846533119678497, "learning_rate": 7.98073719257645e-05, "loss": 0.0107, "step": 26550 }, { "grad_norm": 0.20612668991088867, "learning_rate": 7.979077526029499e-05, "loss": 0.009, "step": 26560 }, { "grad_norm": 0.11702809482812881, "learning_rate": 7.977417350439675e-05, "loss": 0.012, "step": 26570 }, { "grad_norm": 0.11797116696834564, "learning_rate": 7.97575666609066e-05, "loss": 0.0128, "step": 26580 }, { "grad_norm": 0.11525493860244751, "learning_rate": 7.974095473266216e-05, "loss": 0.0079, "step": 26590 }, { "grad_norm": 0.11963772028684616, "learning_rate": 7.972433772250198e-05, "loss": 0.0079, "step": 26600 }, { "grad_norm": 0.14755432307720184, "learning_rate": 7.970771563326544e-05, "loss": 0.0126, "step": 26610 }, { "grad_norm": 0.11242948472499847, "learning_rate": 7.96910884677928e-05, "loss": 0.0108, "step": 26620 }, { "grad_norm": 0.15060283243656158, "learning_rate": 7.967445622892523e-05, "loss": 0.0089, "step": 26630 }, { "grad_norm": 0.11142991483211517, "learning_rate": 7.965781891950465e-05, "loss": 0.009, "step": 26640 }, { "grad_norm": 0.09790969640016556, "learning_rate": 7.964117654237397e-05, "loss": 0.0079, "step": 26650 }, { "grad_norm": 0.13905228674411774, "learning_rate": 7.962452910037692e-05, "loss": 0.0099, "step": 26660 }, { "grad_norm": 0.13262039422988892, "learning_rate": 7.96078765963581e-05, "loss": 0.0094, "step": 26670 }, { "grad_norm": 0.11216285079717636, "learning_rate": 7.95912190331629e-05, "loss": 0.0093, "step": 26680 }, { "grad_norm": 0.14625269174575806, "learning_rate": 7.957455641363772e-05, "loss": 0.0091, "step": 26690 }, { "grad_norm": 0.13930821418762207, "learning_rate": 7.955788874062968e-05, "loss": 0.0079, "step": 26700 }, { "grad_norm": 0.10999716818332672, "learning_rate": 7.95412160169869e-05, "loss": 0.0108, "step": 26710 }, { "grad_norm": 0.1306498646736145, "learning_rate": 7.952453824555824e-05, "loss": 0.0094, "step": 26720 }, { "grad_norm": 0.13561482727527618, "learning_rate": 7.95078554291935e-05, "loss": 0.0083, "step": 26730 }, { "grad_norm": 0.1588091254234314, "learning_rate": 7.94911675707433e-05, "loss": 0.0076, "step": 26740 }, { "grad_norm": 0.10296177119016647, "learning_rate": 7.947447467305915e-05, "loss": 0.0081, "step": 26750 }, { "grad_norm": 0.12449721246957779, "learning_rate": 7.94577767389934e-05, "loss": 0.0091, "step": 26760 }, { "grad_norm": 0.15686523914337158, "learning_rate": 7.944107377139928e-05, "loss": 0.0088, "step": 26770 }, { "grad_norm": 0.17092178761959076, "learning_rate": 7.942436577313088e-05, "loss": 0.012, "step": 26780 }, { "grad_norm": 0.10965313017368317, "learning_rate": 7.940765274704312e-05, "loss": 0.0071, "step": 26790 }, { "grad_norm": 0.14292272925376892, "learning_rate": 7.939093469599181e-05, "loss": 0.0096, "step": 26800 }, { "grad_norm": 0.14713260531425476, "learning_rate": 7.93742116228336e-05, "loss": 0.0088, "step": 26810 }, { "grad_norm": 0.12986908853054047, "learning_rate": 7.935748353042602e-05, "loss": 0.01, "step": 26820 }, { "grad_norm": 0.18124032020568848, "learning_rate": 7.934075042162744e-05, "loss": 0.011, "step": 26830 }, { "grad_norm": 0.1331399381160736, "learning_rate": 7.932401229929705e-05, "loss": 0.01, "step": 26840 }, { "grad_norm": 0.15302950143814087, "learning_rate": 7.9307269166295e-05, "loss": 0.0097, "step": 26850 }, { "grad_norm": 0.16787189245224, "learning_rate": 7.92905210254822e-05, "loss": 0.0088, "step": 26860 }, { "grad_norm": 0.2010776549577713, "learning_rate": 7.927376787972045e-05, "loss": 0.0113, "step": 26870 }, { "grad_norm": 0.21395210921764374, "learning_rate": 7.92570097318724e-05, "loss": 0.0104, "step": 26880 }, { "grad_norm": 0.1490587294101715, "learning_rate": 7.924024658480158e-05, "loss": 0.0106, "step": 26890 }, { "grad_norm": 0.11633452028036118, "learning_rate": 7.922347844137233e-05, "loss": 0.0098, "step": 26900 }, { "grad_norm": 0.15302571654319763, "learning_rate": 7.92067053044499e-05, "loss": 0.0092, "step": 26910 }, { "grad_norm": 0.08944582939147949, "learning_rate": 7.918992717690031e-05, "loss": 0.0089, "step": 26920 }, { "grad_norm": 0.19499847292900085, "learning_rate": 7.917314406159053e-05, "loss": 0.0099, "step": 26930 }, { "grad_norm": 0.07227826863527298, "learning_rate": 7.915635596138832e-05, "loss": 0.0088, "step": 26940 }, { "grad_norm": 0.15887568891048431, "learning_rate": 7.913956287916228e-05, "loss": 0.0121, "step": 26950 }, { "grad_norm": 0.14649413526058197, "learning_rate": 7.912276481778193e-05, "loss": 0.0101, "step": 26960 }, { "grad_norm": 0.1366005539894104, "learning_rate": 7.910596178011759e-05, "loss": 0.0094, "step": 26970 }, { "grad_norm": 0.18610593676567078, "learning_rate": 7.908915376904043e-05, "loss": 0.0092, "step": 26980 }, { "grad_norm": 0.1642826795578003, "learning_rate": 7.907234078742247e-05, "loss": 0.0104, "step": 26990 }, { "grad_norm": 0.17891936004161835, "learning_rate": 7.90555228381366e-05, "loss": 0.0088, "step": 27000 }, { "grad_norm": 0.18246829509735107, "learning_rate": 7.903869992405656e-05, "loss": 0.0107, "step": 27010 }, { "grad_norm": 0.18249131739139557, "learning_rate": 7.902187204805691e-05, "loss": 0.0101, "step": 27020 }, { "grad_norm": 0.14358572661876678, "learning_rate": 7.900503921301308e-05, "loss": 0.0098, "step": 27030 }, { "grad_norm": 0.12724553048610687, "learning_rate": 7.898820142180133e-05, "loss": 0.0093, "step": 27040 }, { "grad_norm": 0.119776651263237, "learning_rate": 7.897135867729879e-05, "loss": 0.0082, "step": 27050 }, { "grad_norm": 0.12214304506778717, "learning_rate": 7.89545109823834e-05, "loss": 0.0109, "step": 27060 }, { "grad_norm": 0.12368496507406235, "learning_rate": 7.8937658339934e-05, "loss": 0.0103, "step": 27070 }, { "grad_norm": 0.12078572064638138, "learning_rate": 7.892080075283026e-05, "loss": 0.0107, "step": 27080 }, { "grad_norm": 0.12254457920789719, "learning_rate": 7.890393822395263e-05, "loss": 0.0079, "step": 27090 }, { "grad_norm": 0.09987819939851761, "learning_rate": 7.88870707561825e-05, "loss": 0.009, "step": 27100 }, { "grad_norm": 0.15142112970352173, "learning_rate": 7.887019835240203e-05, "loss": 0.0082, "step": 27110 }, { "grad_norm": 0.08070217072963715, "learning_rate": 7.885332101549427e-05, "loss": 0.0089, "step": 27120 }, { "grad_norm": 0.17384009063243866, "learning_rate": 7.883643874834308e-05, "loss": 0.0105, "step": 27130 }, { "grad_norm": 0.12886549532413483, "learning_rate": 7.881955155383321e-05, "loss": 0.0088, "step": 27140 }, { "grad_norm": 0.15038776397705078, "learning_rate": 7.880265943485017e-05, "loss": 0.0111, "step": 27150 }, { "grad_norm": 0.12483564019203186, "learning_rate": 7.878576239428038e-05, "loss": 0.0112, "step": 27160 }, { "grad_norm": 0.17036888003349304, "learning_rate": 7.87688604350111e-05, "loss": 0.0127, "step": 27170 }, { "grad_norm": 0.12043438851833344, "learning_rate": 7.875195355993042e-05, "loss": 0.01, "step": 27180 }, { "grad_norm": 0.1414370834827423, "learning_rate": 7.873504177192724e-05, "loss": 0.0126, "step": 27190 }, { "grad_norm": 0.1499384492635727, "learning_rate": 7.87181250738913e-05, "loss": 0.0099, "step": 27200 }, { "grad_norm": 0.126982644200325, "learning_rate": 7.870120346871324e-05, "loss": 0.0103, "step": 27210 }, { "grad_norm": 0.13079877197742462, "learning_rate": 7.86842769592845e-05, "loss": 0.008, "step": 27220 }, { "grad_norm": 0.13904911279678345, "learning_rate": 7.866734554849732e-05, "loss": 0.0095, "step": 27230 }, { "grad_norm": 0.15982533991336823, "learning_rate": 7.865040923924486e-05, "loss": 0.0087, "step": 27240 }, { "grad_norm": 0.13628527522087097, "learning_rate": 7.863346803442104e-05, "loss": 0.0091, "step": 27250 }, { "grad_norm": 0.1534327268600464, "learning_rate": 7.861652193692067e-05, "loss": 0.0086, "step": 27260 }, { "grad_norm": 0.1487753540277481, "learning_rate": 7.859957094963937e-05, "loss": 0.0101, "step": 27270 }, { "grad_norm": 0.1033586636185646, "learning_rate": 7.858261507547357e-05, "loss": 0.0082, "step": 27280 }, { "grad_norm": 0.09866758435964584, "learning_rate": 7.856565431732061e-05, "loss": 0.0085, "step": 27290 }, { "grad_norm": 0.13837893307209015, "learning_rate": 7.854868867807859e-05, "loss": 0.0117, "step": 27300 }, { "grad_norm": 0.11306546628475189, "learning_rate": 7.85317181606465e-05, "loss": 0.0104, "step": 27310 }, { "grad_norm": 0.16399508714675903, "learning_rate": 7.85147427679241e-05, "loss": 0.0105, "step": 27320 }, { "grad_norm": 0.14491117000579834, "learning_rate": 7.849776250281205e-05, "loss": 0.0075, "step": 27330 }, { "grad_norm": 0.0937919095158577, "learning_rate": 7.84807773682118e-05, "loss": 0.0082, "step": 27340 }, { "grad_norm": 0.15149140357971191, "learning_rate": 7.846378736702565e-05, "loss": 0.0085, "step": 27350 }, { "grad_norm": 0.11302176117897034, "learning_rate": 7.844679250215671e-05, "loss": 0.0084, "step": 27360 }, { "grad_norm": 0.10683540999889374, "learning_rate": 7.842979277650898e-05, "loss": 0.0116, "step": 27370 }, { "grad_norm": 0.14495816826820374, "learning_rate": 7.84127881929872e-05, "loss": 0.0098, "step": 27380 }, { "grad_norm": 0.11163219809532166, "learning_rate": 7.839577875449704e-05, "loss": 0.0087, "step": 27390 }, { "grad_norm": 0.12157165259122849, "learning_rate": 7.837876446394489e-05, "loss": 0.009, "step": 27400 }, { "grad_norm": 0.09608137607574463, "learning_rate": 7.836174532423805e-05, "loss": 0.0091, "step": 27410 }, { "grad_norm": 0.09544240683317184, "learning_rate": 7.834472133828466e-05, "loss": 0.0084, "step": 27420 }, { "grad_norm": 0.1392141878604889, "learning_rate": 7.832769250899359e-05, "loss": 0.0115, "step": 27430 }, { "grad_norm": 0.14924122393131256, "learning_rate": 7.831065883927464e-05, "loss": 0.0089, "step": 27440 }, { "grad_norm": 0.1532513052225113, "learning_rate": 7.829362033203841e-05, "loss": 0.0118, "step": 27450 }, { "grad_norm": 0.12955720722675323, "learning_rate": 7.827657699019628e-05, "loss": 0.0089, "step": 27460 }, { "grad_norm": 0.11598782986402512, "learning_rate": 7.825952881666052e-05, "loss": 0.0107, "step": 27470 }, { "grad_norm": 0.12870889902114868, "learning_rate": 7.824247581434418e-05, "loss": 0.0102, "step": 27480 }, { "grad_norm": 0.13751541078090668, "learning_rate": 7.822541798616116e-05, "loss": 0.0082, "step": 27490 }, { "grad_norm": 0.14925840497016907, "learning_rate": 7.820835533502617e-05, "loss": 0.0097, "step": 27500 }, { "grad_norm": 0.17838826775550842, "learning_rate": 7.819128786385475e-05, "loss": 0.0089, "step": 27510 }, { "grad_norm": 0.1467950940132141, "learning_rate": 7.817421557556329e-05, "loss": 0.0086, "step": 27520 }, { "grad_norm": 0.14594526588916779, "learning_rate": 7.815713847306893e-05, "loss": 0.0071, "step": 27530 }, { "grad_norm": 0.12619534134864807, "learning_rate": 7.81400565592897e-05, "loss": 0.0089, "step": 27540 }, { "grad_norm": 0.14293676614761353, "learning_rate": 7.812296983714444e-05, "loss": 0.0096, "step": 27550 }, { "grad_norm": 0.11189907789230347, "learning_rate": 7.810587830955281e-05, "loss": 0.0076, "step": 27560 }, { "grad_norm": 0.1380767524242401, "learning_rate": 7.808878197943528e-05, "loss": 0.0103, "step": 27570 }, { "grad_norm": 0.12001606076955795, "learning_rate": 7.807168084971312e-05, "loss": 0.0098, "step": 27580 }, { "grad_norm": 0.12803491950035095, "learning_rate": 7.805457492330849e-05, "loss": 0.0117, "step": 27590 }, { "grad_norm": 0.15792310237884521, "learning_rate": 7.803746420314428e-05, "loss": 0.0093, "step": 27600 }, { "grad_norm": 0.08943185210227966, "learning_rate": 7.802034869214428e-05, "loss": 0.0102, "step": 27610 }, { "grad_norm": 0.13115429878234863, "learning_rate": 7.800322839323303e-05, "loss": 0.013, "step": 27620 }, { "grad_norm": 0.15553584694862366, "learning_rate": 7.798610330933593e-05, "loss": 0.0111, "step": 27630 }, { "grad_norm": 0.13461361825466156, "learning_rate": 7.796897344337922e-05, "loss": 0.009, "step": 27640 }, { "grad_norm": 0.12450680136680603, "learning_rate": 7.795183879828989e-05, "loss": 0.0094, "step": 27650 }, { "grad_norm": 0.1323726773262024, "learning_rate": 7.793469937699579e-05, "loss": 0.008, "step": 27660 }, { "grad_norm": 0.14928577840328217, "learning_rate": 7.791755518242558e-05, "loss": 0.008, "step": 27670 }, { "grad_norm": 0.12418066710233688, "learning_rate": 7.790040621750876e-05, "loss": 0.0074, "step": 27680 }, { "grad_norm": 0.12461895495653152, "learning_rate": 7.788325248517558e-05, "loss": 0.0097, "step": 27690 }, { "grad_norm": 0.16742439568042755, "learning_rate": 7.786609398835715e-05, "loss": 0.011, "step": 27700 }, { "grad_norm": 0.1619286686182022, "learning_rate": 7.784893072998541e-05, "loss": 0.0089, "step": 27710 }, { "grad_norm": 0.09970352053642273, "learning_rate": 7.783176271299306e-05, "loss": 0.0084, "step": 27720 }, { "grad_norm": 0.11236795783042908, "learning_rate": 7.781458994031368e-05, "loss": 0.008, "step": 27730 }, { "grad_norm": 0.16285111010074615, "learning_rate": 7.779741241488161e-05, "loss": 0.0091, "step": 27740 }, { "grad_norm": 0.11833478510379791, "learning_rate": 7.7780230139632e-05, "loss": 0.0096, "step": 27750 }, { "grad_norm": 0.09740398079156876, "learning_rate": 7.776304311750087e-05, "loss": 0.0081, "step": 27760 }, { "grad_norm": 0.11014789342880249, "learning_rate": 7.7745851351425e-05, "loss": 0.0092, "step": 27770 }, { "grad_norm": 0.1349346786737442, "learning_rate": 7.772865484434197e-05, "loss": 0.0068, "step": 27780 }, { "grad_norm": 0.14740486443042755, "learning_rate": 7.77114535991902e-05, "loss": 0.0088, "step": 27790 }, { "grad_norm": 0.12358909100294113, "learning_rate": 7.769424761890893e-05, "loss": 0.0098, "step": 27800 }, { "grad_norm": 0.15830907225608826, "learning_rate": 7.767703690643817e-05, "loss": 0.0087, "step": 27810 }, { "grad_norm": 0.18462195992469788, "learning_rate": 7.76598214647188e-05, "loss": 0.0103, "step": 27820 }, { "grad_norm": 0.1284036636352539, "learning_rate": 7.764260129669241e-05, "loss": 0.0078, "step": 27830 }, { "grad_norm": 0.1187584176659584, "learning_rate": 7.76253764053015e-05, "loss": 0.0068, "step": 27840 }, { "grad_norm": 0.1543491780757904, "learning_rate": 7.760814679348932e-05, "loss": 0.0082, "step": 27850 }, { "grad_norm": 0.09788200259208679, "learning_rate": 7.759091246419992e-05, "loss": 0.0086, "step": 27860 }, { "grad_norm": 0.15388935804367065, "learning_rate": 7.75736734203782e-05, "loss": 0.0102, "step": 27870 }, { "grad_norm": 0.1517609804868698, "learning_rate": 7.755642966496985e-05, "loss": 0.0097, "step": 27880 }, { "grad_norm": 0.1555309295654297, "learning_rate": 7.753918120092132e-05, "loss": 0.0114, "step": 27890 }, { "grad_norm": 0.14701835811138153, "learning_rate": 7.752192803117993e-05, "loss": 0.0082, "step": 27900 }, { "grad_norm": 0.16290652751922607, "learning_rate": 7.750467015869377e-05, "loss": 0.0116, "step": 27910 }, { "grad_norm": 0.15310049057006836, "learning_rate": 7.748740758641174e-05, "loss": 0.0077, "step": 27920 }, { "grad_norm": 0.14260736107826233, "learning_rate": 7.74701403172835e-05, "loss": 0.0114, "step": 27930 }, { "grad_norm": 0.14974309504032135, "learning_rate": 7.745286835425962e-05, "loss": 0.008, "step": 27940 }, { "grad_norm": 0.12839917838573456, "learning_rate": 7.743559170029138e-05, "loss": 0.0119, "step": 27950 }, { "grad_norm": 0.1369580179452896, "learning_rate": 7.741831035833087e-05, "loss": 0.0087, "step": 27960 }, { "grad_norm": 0.11223772168159485, "learning_rate": 7.740102433133102e-05, "loss": 0.0102, "step": 27970 }, { "grad_norm": 0.1624779850244522, "learning_rate": 7.738373362224553e-05, "loss": 0.0089, "step": 27980 }, { "grad_norm": 0.14139117300510406, "learning_rate": 7.73664382340289e-05, "loss": 0.0117, "step": 27990 }, { "grad_norm": 0.12747283279895782, "learning_rate": 7.734913816963647e-05, "loss": 0.0086, "step": 28000 }, { "grad_norm": 0.15713642537593842, "learning_rate": 7.73318334320243e-05, "loss": 0.0105, "step": 28010 }, { "grad_norm": 0.13179638981819153, "learning_rate": 7.731452402414934e-05, "loss": 0.0091, "step": 28020 }, { "grad_norm": 0.1794220209121704, "learning_rate": 7.729720994896928e-05, "loss": 0.0099, "step": 28030 }, { "grad_norm": 0.10494772344827652, "learning_rate": 7.727989120944262e-05, "loss": 0.0092, "step": 28040 }, { "grad_norm": 0.14797763526439667, "learning_rate": 7.726256780852865e-05, "loss": 0.0099, "step": 28050 }, { "grad_norm": 0.18173538148403168, "learning_rate": 7.724523974918749e-05, "loss": 0.0085, "step": 28060 }, { "grad_norm": 0.1810363084077835, "learning_rate": 7.722790703438002e-05, "loss": 0.0086, "step": 28070 }, { "grad_norm": 0.14862662553787231, "learning_rate": 7.72105696670679e-05, "loss": 0.0087, "step": 28080 }, { "grad_norm": 0.14137423038482666, "learning_rate": 7.719322765021364e-05, "loss": 0.0072, "step": 28090 }, { "grad_norm": 0.19487830996513367, "learning_rate": 7.717588098678051e-05, "loss": 0.0108, "step": 28100 }, { "grad_norm": 0.14628851413726807, "learning_rate": 7.715852967973258e-05, "loss": 0.0083, "step": 28110 }, { "grad_norm": 0.13379545509815216, "learning_rate": 7.714117373203474e-05, "loss": 0.0093, "step": 28120 }, { "grad_norm": 0.1192362830042839, "learning_rate": 7.712381314665259e-05, "loss": 0.0092, "step": 28130 }, { "grad_norm": 0.15532413125038147, "learning_rate": 7.710644792655261e-05, "loss": 0.0098, "step": 28140 }, { "grad_norm": 0.11512377858161926, "learning_rate": 7.708907807470207e-05, "loss": 0.0084, "step": 28150 }, { "grad_norm": 0.09723109006881714, "learning_rate": 7.707170359406896e-05, "loss": 0.0091, "step": 28160 }, { "grad_norm": 0.12499096989631653, "learning_rate": 7.705432448762213e-05, "loss": 0.0102, "step": 28170 }, { "grad_norm": 0.10317961126565933, "learning_rate": 7.703694075833117e-05, "loss": 0.0084, "step": 28180 }, { "grad_norm": 0.11882846802473068, "learning_rate": 7.70195524091665e-05, "loss": 0.0101, "step": 28190 }, { "grad_norm": 0.14410185813903809, "learning_rate": 7.70021594430993e-05, "loss": 0.0094, "step": 28200 }, { "grad_norm": 0.10513525456190109, "learning_rate": 7.698476186310157e-05, "loss": 0.0092, "step": 28210 }, { "grad_norm": 0.15520143508911133, "learning_rate": 7.696735967214608e-05, "loss": 0.011, "step": 28220 }, { "grad_norm": 0.1322605013847351, "learning_rate": 7.694995287320636e-05, "loss": 0.0074, "step": 28230 }, { "grad_norm": 0.10036326944828033, "learning_rate": 7.693254146925679e-05, "loss": 0.0097, "step": 28240 }, { "grad_norm": 0.11491569876670837, "learning_rate": 7.691512546327251e-05, "loss": 0.0099, "step": 28250 }, { "grad_norm": 0.10727521777153015, "learning_rate": 7.689770485822939e-05, "loss": 0.0099, "step": 28260 }, { "grad_norm": 0.13736574351787567, "learning_rate": 7.688027965710416e-05, "loss": 0.0081, "step": 28270 }, { "grad_norm": 0.10442491620779037, "learning_rate": 7.686284986287433e-05, "loss": 0.0092, "step": 28280 }, { "grad_norm": 0.12355868518352509, "learning_rate": 7.684541547851817e-05, "loss": 0.0082, "step": 28290 }, { "grad_norm": 0.11899201571941376, "learning_rate": 7.68279765070147e-05, "loss": 0.0086, "step": 28300 }, { "grad_norm": 0.10815779119729996, "learning_rate": 7.68105329513438e-05, "loss": 0.0093, "step": 28310 }, { "grad_norm": 0.12400398403406143, "learning_rate": 7.67930848144861e-05, "loss": 0.0097, "step": 28320 }, { "grad_norm": 0.0996285229921341, "learning_rate": 7.6775632099423e-05, "loss": 0.0078, "step": 28330 }, { "grad_norm": 0.09592723846435547, "learning_rate": 7.675817480913667e-05, "loss": 0.0072, "step": 28340 }, { "grad_norm": 0.13431988656520844, "learning_rate": 7.674071294661011e-05, "loss": 0.0073, "step": 28350 }, { "grad_norm": 0.1168004721403122, "learning_rate": 7.672324651482707e-05, "loss": 0.0071, "step": 28360 }, { "grad_norm": 0.1469714492559433, "learning_rate": 7.670577551677209e-05, "loss": 0.0079, "step": 28370 }, { "grad_norm": 0.13020603358745575, "learning_rate": 7.668829995543047e-05, "loss": 0.0097, "step": 28380 }, { "grad_norm": 0.16329838335514069, "learning_rate": 7.667081983378832e-05, "loss": 0.0097, "step": 28390 }, { "grad_norm": 0.14773839712142944, "learning_rate": 7.66533351548325e-05, "loss": 0.0075, "step": 28400 }, { "grad_norm": 0.10970510542392731, "learning_rate": 7.663584592155069e-05, "loss": 0.0132, "step": 28410 }, { "grad_norm": 0.10972238332033157, "learning_rate": 7.661835213693129e-05, "loss": 0.0086, "step": 28420 }, { "grad_norm": 0.11632126569747925, "learning_rate": 7.660085380396353e-05, "loss": 0.0094, "step": 28430 }, { "grad_norm": 0.15206724405288696, "learning_rate": 7.658335092563738e-05, "loss": 0.0088, "step": 28440 }, { "grad_norm": 0.15429401397705078, "learning_rate": 7.656584350494362e-05, "loss": 0.0098, "step": 28450 }, { "grad_norm": 0.16988033056259155, "learning_rate": 7.654833154487378e-05, "loss": 0.0089, "step": 28460 }, { "grad_norm": 0.19420146942138672, "learning_rate": 7.653081504842017e-05, "loss": 0.0106, "step": 28470 }, { "grad_norm": 0.15317587554454803, "learning_rate": 7.65132940185759e-05, "loss": 0.0098, "step": 28480 }, { "grad_norm": 0.12216871231794357, "learning_rate": 7.649576845833481e-05, "loss": 0.0082, "step": 28490 }, { "grad_norm": 0.10461904108524323, "learning_rate": 7.647823837069156e-05, "loss": 0.0083, "step": 28500 }, { "grad_norm": 0.0956147164106369, "learning_rate": 7.646070375864156e-05, "loss": 0.007, "step": 28510 }, { "grad_norm": 0.11778079718351364, "learning_rate": 7.644316462518099e-05, "loss": 0.0083, "step": 28520 }, { "grad_norm": 0.10657733678817749, "learning_rate": 7.642562097330679e-05, "loss": 0.009, "step": 28530 }, { "grad_norm": 0.14046545326709747, "learning_rate": 7.640807280601671e-05, "loss": 0.0086, "step": 28540 }, { "grad_norm": 0.13234873116016388, "learning_rate": 7.639052012630927e-05, "loss": 0.0091, "step": 28550 }, { "grad_norm": 0.11687830835580826, "learning_rate": 7.63729629371837e-05, "loss": 0.01, "step": 28560 }, { "grad_norm": 0.15119720995426178, "learning_rate": 7.635540124164009e-05, "loss": 0.0103, "step": 28570 }, { "grad_norm": 0.1162041649222374, "learning_rate": 7.633783504267922e-05, "loss": 0.0092, "step": 28580 }, { "grad_norm": 0.14727018773555756, "learning_rate": 7.632026434330269e-05, "loss": 0.0092, "step": 28590 }, { "grad_norm": 0.12125667184591293, "learning_rate": 7.630268914651282e-05, "loss": 0.0071, "step": 28600 }, { "grad_norm": 0.168350949883461, "learning_rate": 7.628510945531278e-05, "loss": 0.0092, "step": 28610 }, { "grad_norm": 0.1255166381597519, "learning_rate": 7.626752527270641e-05, "loss": 0.0087, "step": 28620 }, { "grad_norm": 0.14199180901050568, "learning_rate": 7.62499366016984e-05, "loss": 0.0112, "step": 28630 }, { "grad_norm": 0.11065270006656647, "learning_rate": 7.623234344529416e-05, "loss": 0.0086, "step": 28640 }, { "grad_norm": 0.15566134452819824, "learning_rate": 7.62147458064999e-05, "loss": 0.0092, "step": 28650 }, { "grad_norm": 0.11308500170707703, "learning_rate": 7.619714368832254e-05, "loss": 0.0068, "step": 28660 }, { "grad_norm": 0.09052170068025589, "learning_rate": 7.61795370937698e-05, "loss": 0.0078, "step": 28670 }, { "grad_norm": 0.12625765800476074, "learning_rate": 7.61619260258502e-05, "loss": 0.0094, "step": 28680 }, { "grad_norm": 0.1620987504720688, "learning_rate": 7.614431048757298e-05, "loss": 0.0096, "step": 28690 }, { "grad_norm": 0.13086193799972534, "learning_rate": 7.612669048194814e-05, "loss": 0.0098, "step": 28700 }, { "grad_norm": 0.14908285439014435, "learning_rate": 7.610906601198646e-05, "loss": 0.0096, "step": 28710 }, { "grad_norm": 0.14968959987163544, "learning_rate": 7.60914370806995e-05, "loss": 0.0077, "step": 28720 }, { "grad_norm": 0.16772885620594025, "learning_rate": 7.607380369109953e-05, "loss": 0.0095, "step": 28730 }, { "grad_norm": 0.14985691010951996, "learning_rate": 7.605616584619961e-05, "loss": 0.0083, "step": 28740 }, { "grad_norm": 0.1293623447418213, "learning_rate": 7.603852354901362e-05, "loss": 0.0108, "step": 28750 }, { "grad_norm": 0.13767728209495544, "learning_rate": 7.602087680255609e-05, "loss": 0.0096, "step": 28760 }, { "grad_norm": 0.1413411945104599, "learning_rate": 7.600322560984238e-05, "loss": 0.0081, "step": 28770 }, { "grad_norm": 0.12410459667444229, "learning_rate": 7.598556997388863e-05, "loss": 0.0105, "step": 28780 }, { "grad_norm": 0.12424052506685257, "learning_rate": 7.596790989771166e-05, "loss": 0.008, "step": 28790 }, { "grad_norm": 0.08453797549009323, "learning_rate": 7.595024538432914e-05, "loss": 0.0078, "step": 28800 }, { "grad_norm": 0.10085631906986237, "learning_rate": 7.59325764367594e-05, "loss": 0.0079, "step": 28810 }, { "grad_norm": 0.1247897669672966, "learning_rate": 7.59149030580216e-05, "loss": 0.0109, "step": 28820 }, { "grad_norm": 0.11824203282594681, "learning_rate": 7.589722525113562e-05, "loss": 0.0074, "step": 28830 }, { "grad_norm": 0.12138554453849792, "learning_rate": 7.587954301912216e-05, "loss": 0.0086, "step": 28840 }, { "grad_norm": 0.11229289323091507, "learning_rate": 7.586185636500263e-05, "loss": 0.0089, "step": 28850 }, { "grad_norm": 0.1349518895149231, "learning_rate": 7.584416529179914e-05, "loss": 0.0079, "step": 28860 }, { "grad_norm": 0.11735469102859497, "learning_rate": 7.582646980253465e-05, "loss": 0.0082, "step": 28870 }, { "grad_norm": 0.13586582243442535, "learning_rate": 7.580876990023282e-05, "loss": 0.0086, "step": 28880 }, { "grad_norm": 0.104008249938488, "learning_rate": 7.579106558791809e-05, "loss": 0.0075, "step": 28890 }, { "grad_norm": 0.17221559584140778, "learning_rate": 7.577335686861565e-05, "loss": 0.0095, "step": 28900 }, { "grad_norm": 0.16582165658473969, "learning_rate": 7.575564374535141e-05, "loss": 0.008, "step": 28910 }, { "grad_norm": 0.11760534346103668, "learning_rate": 7.573792622115207e-05, "loss": 0.0075, "step": 28920 }, { "grad_norm": 0.11631323397159576, "learning_rate": 7.572020429904507e-05, "loss": 0.0102, "step": 28930 }, { "grad_norm": 0.10792902857065201, "learning_rate": 7.570247798205861e-05, "loss": 0.0065, "step": 28940 }, { "grad_norm": 0.11862902343273163, "learning_rate": 7.568474727322164e-05, "loss": 0.0083, "step": 28950 }, { "grad_norm": 0.14772187173366547, "learning_rate": 7.566701217556384e-05, "loss": 0.0111, "step": 28960 }, { "grad_norm": 0.08613169193267822, "learning_rate": 7.564927269211564e-05, "loss": 0.0087, "step": 28970 }, { "grad_norm": 0.10678637027740479, "learning_rate": 7.563152882590824e-05, "loss": 0.0082, "step": 28980 }, { "grad_norm": 0.16240966320037842, "learning_rate": 7.56137805799736e-05, "loss": 0.0093, "step": 28990 }, { "grad_norm": 0.1237974539399147, "learning_rate": 7.559602795734439e-05, "loss": 0.009, "step": 29000 }, { "grad_norm": 0.12644222378730774, "learning_rate": 7.557827096105408e-05, "loss": 0.0081, "step": 29010 }, { "grad_norm": 0.1100897490978241, "learning_rate": 7.55605095941368e-05, "loss": 0.0091, "step": 29020 }, { "grad_norm": 0.19067028164863586, "learning_rate": 7.554274385962752e-05, "loss": 0.0094, "step": 29030 }, { "grad_norm": 0.12158691883087158, "learning_rate": 7.552497376056191e-05, "loss": 0.012, "step": 29040 }, { "grad_norm": 0.11800594627857208, "learning_rate": 7.550719929997639e-05, "loss": 0.0065, "step": 29050 }, { "grad_norm": 0.08027803152799606, "learning_rate": 7.548942048090813e-05, "loss": 0.0078, "step": 29060 }, { "grad_norm": 0.14235220849514008, "learning_rate": 7.547163730639506e-05, "loss": 0.011, "step": 29070 }, { "grad_norm": 0.13086539506912231, "learning_rate": 7.545384977947583e-05, "loss": 0.0078, "step": 29080 }, { "grad_norm": 0.13184307515621185, "learning_rate": 7.543605790318981e-05, "loss": 0.0095, "step": 29090 }, { "grad_norm": 0.18569959700107574, "learning_rate": 7.54182616805772e-05, "loss": 0.0098, "step": 29100 }, { "grad_norm": 0.1444113850593567, "learning_rate": 7.540046111467885e-05, "loss": 0.0079, "step": 29110 }, { "grad_norm": 0.1840764880180359, "learning_rate": 7.53826562085364e-05, "loss": 0.0095, "step": 29120 }, { "grad_norm": 0.12754161655902863, "learning_rate": 7.536484696519221e-05, "loss": 0.0099, "step": 29130 }, { "grad_norm": 0.1447249948978424, "learning_rate": 7.534703338768942e-05, "loss": 0.0113, "step": 29140 }, { "grad_norm": 0.17827436327934265, "learning_rate": 7.532921547907185e-05, "loss": 0.0098, "step": 29150 }, { "grad_norm": 0.14724613726139069, "learning_rate": 7.531139324238412e-05, "loss": 0.0086, "step": 29160 }, { "grad_norm": 0.09886711090803146, "learning_rate": 7.529356668067157e-05, "loss": 0.0092, "step": 29170 }, { "grad_norm": 0.12461794167757034, "learning_rate": 7.527573579698023e-05, "loss": 0.0085, "step": 29180 }, { "grad_norm": 0.14292699098587036, "learning_rate": 7.525790059435693e-05, "loss": 0.0082, "step": 29190 }, { "grad_norm": 0.09177558124065399, "learning_rate": 7.524006107584926e-05, "loss": 0.007, "step": 29200 }, { "grad_norm": 0.11170072853565216, "learning_rate": 7.522221724450544e-05, "loss": 0.0074, "step": 29210 }, { "grad_norm": 0.09247367829084396, "learning_rate": 7.520436910337451e-05, "loss": 0.009, "step": 29220 }, { "grad_norm": 0.19031409919261932, "learning_rate": 7.518651665550627e-05, "loss": 0.0089, "step": 29230 }, { "grad_norm": 0.13419204950332642, "learning_rate": 7.516865990395117e-05, "loss": 0.0124, "step": 29240 }, { "grad_norm": 0.12711018323898315, "learning_rate": 7.515079885176047e-05, "loss": 0.0091, "step": 29250 }, { "grad_norm": 0.11175156384706497, "learning_rate": 7.513293350198612e-05, "loss": 0.0077, "step": 29260 }, { "grad_norm": 0.11903607100248337, "learning_rate": 7.511506385768081e-05, "loss": 0.0078, "step": 29270 }, { "grad_norm": 0.1432015299797058, "learning_rate": 7.509718992189801e-05, "loss": 0.0079, "step": 29280 }, { "grad_norm": 0.12325069308280945, "learning_rate": 7.507931169769182e-05, "loss": 0.0093, "step": 29290 }, { "grad_norm": 0.10275040566921234, "learning_rate": 7.506142918811722e-05, "loss": 0.0102, "step": 29300 }, { "grad_norm": 0.15118730068206787, "learning_rate": 7.504354239622978e-05, "loss": 0.0087, "step": 29310 }, { "grad_norm": 0.14371444284915924, "learning_rate": 7.50256513250859e-05, "loss": 0.0081, "step": 29320 }, { "grad_norm": 0.10947857052087784, "learning_rate": 7.500775597774265e-05, "loss": 0.009, "step": 29330 }, { "grad_norm": 0.10372000932693481, "learning_rate": 7.498985635725788e-05, "loss": 0.008, "step": 29340 }, { "grad_norm": 0.14823445677757263, "learning_rate": 7.497195246669012e-05, "loss": 0.0092, "step": 29350 }, { "grad_norm": 0.1469978392124176, "learning_rate": 7.495404430909868e-05, "loss": 0.0089, "step": 29360 }, { "grad_norm": 0.16545085608959198, "learning_rate": 7.493613188754356e-05, "loss": 0.0086, "step": 29370 }, { "grad_norm": 0.15893298387527466, "learning_rate": 7.49182152050855e-05, "loss": 0.0078, "step": 29380 }, { "grad_norm": 0.19522720575332642, "learning_rate": 7.490029426478598e-05, "loss": 0.0085, "step": 29390 }, { "grad_norm": 0.1567622721195221, "learning_rate": 7.488236906970719e-05, "loss": 0.0086, "step": 29400 }, { "grad_norm": 0.17104659974575043, "learning_rate": 7.486443962291207e-05, "loss": 0.0068, "step": 29410 }, { "grad_norm": 0.10825195163488388, "learning_rate": 7.484650592746424e-05, "loss": 0.0069, "step": 29420 }, { "grad_norm": 0.1076744943857193, "learning_rate": 7.482856798642811e-05, "loss": 0.0094, "step": 29430 }, { "grad_norm": 0.11809355020523071, "learning_rate": 7.481062580286878e-05, "loss": 0.0081, "step": 29440 }, { "grad_norm": 0.13997700810432434, "learning_rate": 7.479267937985208e-05, "loss": 0.0072, "step": 29450 }, { "grad_norm": 0.13737529516220093, "learning_rate": 7.477472872044456e-05, "loss": 0.009, "step": 29460 }, { "grad_norm": 0.17871473729610443, "learning_rate": 7.475677382771347e-05, "loss": 0.007, "step": 29470 }, { "grad_norm": 0.1433309018611908, "learning_rate": 7.473881470472683e-05, "loss": 0.0094, "step": 29480 }, { "grad_norm": 0.11444041132926941, "learning_rate": 7.472085135455337e-05, "loss": 0.0072, "step": 29490 }, { "grad_norm": 0.14387646317481995, "learning_rate": 7.470288378026256e-05, "loss": 0.0085, "step": 29500 }, { "grad_norm": 0.10678965598344803, "learning_rate": 7.468491198492451e-05, "loss": 0.0092, "step": 29510 }, { "grad_norm": 0.14922021329402924, "learning_rate": 7.466693597161013e-05, "loss": 0.0087, "step": 29520 }, { "grad_norm": 0.11131960898637772, "learning_rate": 7.464895574339104e-05, "loss": 0.0073, "step": 29530 }, { "grad_norm": 0.1306486278772354, "learning_rate": 7.463097130333958e-05, "loss": 0.0117, "step": 29540 }, { "grad_norm": 0.11212944239377975, "learning_rate": 7.461298265452876e-05, "loss": 0.0078, "step": 29550 }, { "grad_norm": 0.17676633596420288, "learning_rate": 7.459498980003239e-05, "loss": 0.0091, "step": 29560 }, { "grad_norm": 0.10291837155818939, "learning_rate": 7.457699274292493e-05, "loss": 0.0075, "step": 29570 }, { "grad_norm": 0.1000533401966095, "learning_rate": 7.455899148628159e-05, "loss": 0.0081, "step": 29580 }, { "grad_norm": 0.10186289250850677, "learning_rate": 7.45409860331783e-05, "loss": 0.0082, "step": 29590 }, { "grad_norm": 0.11534623056650162, "learning_rate": 7.452297638669169e-05, "loss": 0.0104, "step": 29600 }, { "grad_norm": 0.17258623242378235, "learning_rate": 7.450496254989911e-05, "loss": 0.0109, "step": 29610 }, { "grad_norm": 0.16348297894001007, "learning_rate": 7.448694452587866e-05, "loss": 0.0098, "step": 29620 }, { "grad_norm": 0.13269919157028198, "learning_rate": 7.44689223177091e-05, "loss": 0.0071, "step": 29630 }, { "grad_norm": 0.11285321414470673, "learning_rate": 7.445089592846994e-05, "loss": 0.0073, "step": 29640 }, { "grad_norm": 0.08672850579023361, "learning_rate": 7.443286536124141e-05, "loss": 0.008, "step": 29650 }, { "grad_norm": 0.13784395158290863, "learning_rate": 7.441483061910443e-05, "loss": 0.0079, "step": 29660 }, { "grad_norm": 0.13524094223976135, "learning_rate": 7.439679170514064e-05, "loss": 0.0076, "step": 29670 }, { "grad_norm": 0.14562425017356873, "learning_rate": 7.43787486224324e-05, "loss": 0.0081, "step": 29680 }, { "grad_norm": 0.10585056245326996, "learning_rate": 7.436070137406276e-05, "loss": 0.0085, "step": 29690 }, { "grad_norm": 0.1636708378791809, "learning_rate": 7.434264996311556e-05, "loss": 0.0093, "step": 29700 }, { "grad_norm": 0.17124010622501373, "learning_rate": 7.432459439267525e-05, "loss": 0.0093, "step": 29710 }, { "grad_norm": 0.14431822299957275, "learning_rate": 7.430653466582701e-05, "loss": 0.0081, "step": 29720 }, { "grad_norm": 0.12670251727104187, "learning_rate": 7.42884707856568e-05, "loss": 0.0122, "step": 29730 }, { "grad_norm": 0.14324291050434113, "learning_rate": 7.427040275525122e-05, "loss": 0.0073, "step": 29740 }, { "grad_norm": 0.10959525406360626, "learning_rate": 7.42523305776976e-05, "loss": 0.0102, "step": 29750 }, { "grad_norm": 0.12449794262647629, "learning_rate": 7.4234254256084e-05, "loss": 0.0083, "step": 29760 }, { "grad_norm": 0.10098195821046829, "learning_rate": 7.421617379349915e-05, "loss": 0.0101, "step": 29770 }, { "grad_norm": 0.15895088016986847, "learning_rate": 7.41980891930325e-05, "loss": 0.0081, "step": 29780 }, { "grad_norm": 0.12396474182605743, "learning_rate": 7.418000045777425e-05, "loss": 0.0084, "step": 29790 }, { "grad_norm": 0.15762832760810852, "learning_rate": 7.416190759081523e-05, "loss": 0.011, "step": 29800 }, { "grad_norm": 0.13658079504966736, "learning_rate": 7.414381059524704e-05, "loss": 0.0069, "step": 29810 }, { "grad_norm": 0.11726830154657364, "learning_rate": 7.412570947416195e-05, "loss": 0.0086, "step": 29820 }, { "grad_norm": 0.11295782029628754, "learning_rate": 7.410760423065295e-05, "loss": 0.0093, "step": 29830 }, { "grad_norm": 0.13836808502674103, "learning_rate": 7.408949486781372e-05, "loss": 0.0077, "step": 29840 }, { "grad_norm": 0.18437470495700836, "learning_rate": 7.407138138873868e-05, "loss": 0.0101, "step": 29850 }, { "grad_norm": 0.1354842334985733, "learning_rate": 7.405326379652292e-05, "loss": 0.0097, "step": 29860 }, { "grad_norm": 0.14957086741924286, "learning_rate": 7.403514209426222e-05, "loss": 0.0094, "step": 29870 }, { "grad_norm": 0.16696304082870483, "learning_rate": 7.40170162850531e-05, "loss": 0.0112, "step": 29880 }, { "grad_norm": 0.10851151496171951, "learning_rate": 7.399888637199278e-05, "loss": 0.0108, "step": 29890 }, { "grad_norm": 0.14266180992126465, "learning_rate": 7.398075235817914e-05, "loss": 0.0087, "step": 29900 }, { "grad_norm": 0.1600572168827057, "learning_rate": 7.39626142467108e-05, "loss": 0.0102, "step": 29910 }, { "grad_norm": 0.12735240161418915, "learning_rate": 7.394447204068706e-05, "loss": 0.0089, "step": 29920 }, { "grad_norm": 0.10062438994646072, "learning_rate": 7.392632574320793e-05, "loss": 0.0098, "step": 29930 }, { "grad_norm": 0.0839054211974144, "learning_rate": 7.390817535737411e-05, "loss": 0.0098, "step": 29940 }, { "grad_norm": 0.1359008550643921, "learning_rate": 7.389002088628703e-05, "loss": 0.0093, "step": 29950 }, { "grad_norm": 0.0795733854174614, "learning_rate": 7.387186233304877e-05, "loss": 0.0065, "step": 29960 }, { "grad_norm": 0.0902395099401474, "learning_rate": 7.385369970076212e-05, "loss": 0.0068, "step": 29970 }, { "grad_norm": 0.1245250552892685, "learning_rate": 7.38355329925306e-05, "loss": 0.0062, "step": 29980 }, { "grad_norm": 0.10051299631595612, "learning_rate": 7.381736221145838e-05, "loss": 0.0069, "step": 29990 }, { "grad_norm": 0.08459696173667908, "learning_rate": 7.37991873606504e-05, "loss": 0.0084, "step": 30000 }, { "grad_norm": 0.10548318922519684, "learning_rate": 7.378100844321218e-05, "loss": 0.008, "step": 30010 }, { "grad_norm": 0.13181905448436737, "learning_rate": 7.376282546225004e-05, "loss": 0.0123, "step": 30020 }, { "grad_norm": 0.1106996089220047, "learning_rate": 7.374463842087094e-05, "loss": 0.0088, "step": 30030 }, { "grad_norm": 0.09624435007572174, "learning_rate": 7.372644732218254e-05, "loss": 0.0071, "step": 30040 }, { "grad_norm": 0.09967365860939026, "learning_rate": 7.370825216929322e-05, "loss": 0.0084, "step": 30050 }, { "grad_norm": 0.14190077781677246, "learning_rate": 7.369005296531205e-05, "loss": 0.0107, "step": 30060 }, { "grad_norm": 0.11561845242977142, "learning_rate": 7.367184971334873e-05, "loss": 0.0081, "step": 30070 }, { "grad_norm": 0.11688084155321121, "learning_rate": 7.365364241651371e-05, "loss": 0.0081, "step": 30080 }, { "grad_norm": 0.07943795621395111, "learning_rate": 7.363543107791815e-05, "loss": 0.0099, "step": 30090 }, { "grad_norm": 0.18178299069404602, "learning_rate": 7.361721570067384e-05, "loss": 0.01, "step": 30100 }, { "grad_norm": 0.14292119443416595, "learning_rate": 7.359899628789331e-05, "loss": 0.0085, "step": 30110 }, { "grad_norm": 0.1307406723499298, "learning_rate": 7.358077284268974e-05, "loss": 0.0087, "step": 30120 }, { "grad_norm": 0.16615885496139526, "learning_rate": 7.356254536817702e-05, "loss": 0.0079, "step": 30130 }, { "grad_norm": 0.15408240258693695, "learning_rate": 7.354431386746973e-05, "loss": 0.0119, "step": 30140 }, { "grad_norm": 0.10339371114969254, "learning_rate": 7.352607834368316e-05, "loss": 0.0098, "step": 30150 }, { "grad_norm": 0.1746719628572464, "learning_rate": 7.350783879993324e-05, "loss": 0.0098, "step": 30160 }, { "grad_norm": 0.17286154627799988, "learning_rate": 7.348959523933658e-05, "loss": 0.0118, "step": 30170 }, { "grad_norm": 0.14545838534832, "learning_rate": 7.347134766501057e-05, "loss": 0.009, "step": 30180 }, { "grad_norm": 0.14466208219528198, "learning_rate": 7.345309608007315e-05, "loss": 0.0084, "step": 30190 }, { "grad_norm": 0.1321561187505722, "learning_rate": 7.343484048764308e-05, "loss": 0.0101, "step": 30200 }, { "grad_norm": 0.10703225433826447, "learning_rate": 7.341658089083972e-05, "loss": 0.0106, "step": 30210 }, { "grad_norm": 0.1506739854812622, "learning_rate": 7.339831729278313e-05, "loss": 0.0112, "step": 30220 }, { "grad_norm": 0.07665491104125977, "learning_rate": 7.338004969659404e-05, "loss": 0.0075, "step": 30230 }, { "grad_norm": 0.15751506388187408, "learning_rate": 7.336177810539391e-05, "loss": 0.0083, "step": 30240 }, { "grad_norm": 0.1540769338607788, "learning_rate": 7.334350252230485e-05, "loss": 0.0117, "step": 30250 }, { "grad_norm": 0.11448697745800018, "learning_rate": 7.332522295044965e-05, "loss": 0.0075, "step": 30260 }, { "grad_norm": 0.1262480467557907, "learning_rate": 7.33069393929518e-05, "loss": 0.0094, "step": 30270 }, { "grad_norm": 0.15235957503318787, "learning_rate": 7.328865185293545e-05, "loss": 0.0116, "step": 30280 }, { "grad_norm": 0.1556386500597, "learning_rate": 7.327036033352546e-05, "loss": 0.0097, "step": 30290 }, { "grad_norm": 0.14718599617481232, "learning_rate": 7.325206483784733e-05, "loss": 0.0088, "step": 30300 }, { "grad_norm": 0.1302109658718109, "learning_rate": 7.323376536902724e-05, "loss": 0.0071, "step": 30310 }, { "grad_norm": 0.10566077381372452, "learning_rate": 7.321546193019213e-05, "loss": 0.0102, "step": 30320 }, { "grad_norm": 0.12025662511587143, "learning_rate": 7.31971545244695e-05, "loss": 0.0094, "step": 30330 }, { "grad_norm": 0.1594366729259491, "learning_rate": 7.31788431549876e-05, "loss": 0.0108, "step": 30340 }, { "grad_norm": 0.14107376337051392, "learning_rate": 7.316052782487534e-05, "loss": 0.0085, "step": 30350 }, { "grad_norm": 0.12545789778232574, "learning_rate": 7.314220853726234e-05, "loss": 0.0086, "step": 30360 }, { "grad_norm": 0.1516544371843338, "learning_rate": 7.312388529527884e-05, "loss": 0.0098, "step": 30370 }, { "grad_norm": 0.1484028846025467, "learning_rate": 7.310555810205577e-05, "loss": 0.009, "step": 30380 }, { "grad_norm": 0.13010302186012268, "learning_rate": 7.308722696072476e-05, "loss": 0.0089, "step": 30390 }, { "grad_norm": 0.09658820927143097, "learning_rate": 7.306889187441811e-05, "loss": 0.0072, "step": 30400 }, { "grad_norm": 0.11862804740667343, "learning_rate": 7.305055284626876e-05, "loss": 0.0095, "step": 30410 }, { "grad_norm": 0.11775655299425125, "learning_rate": 7.303220987941037e-05, "loss": 0.0096, "step": 30420 }, { "grad_norm": 0.11409790068864822, "learning_rate": 7.301386297697726e-05, "loss": 0.0056, "step": 30430 }, { "grad_norm": 0.1602911800146103, "learning_rate": 7.299551214210438e-05, "loss": 0.0076, "step": 30440 }, { "grad_norm": 0.14827461540699005, "learning_rate": 7.297715737792738e-05, "loss": 0.0085, "step": 30450 }, { "grad_norm": 0.10371328890323639, "learning_rate": 7.295879868758265e-05, "loss": 0.0077, "step": 30460 }, { "grad_norm": 0.1353980153799057, "learning_rate": 7.294043607420713e-05, "loss": 0.0083, "step": 30470 }, { "grad_norm": 0.1194119080901146, "learning_rate": 7.292206954093852e-05, "loss": 0.0076, "step": 30480 }, { "grad_norm": 0.16145777702331543, "learning_rate": 7.290369909091515e-05, "loss": 0.0083, "step": 30490 }, { "grad_norm": 0.1309974044561386, "learning_rate": 7.2885324727276e-05, "loss": 0.0096, "step": 30500 }, { "grad_norm": 0.13514916598796844, "learning_rate": 7.286694645316076e-05, "loss": 0.0084, "step": 30510 }, { "grad_norm": 0.14443419873714447, "learning_rate": 7.284856427170982e-05, "loss": 0.0089, "step": 30520 }, { "grad_norm": 0.10492212325334549, "learning_rate": 7.283017818606414e-05, "loss": 0.0101, "step": 30530 }, { "grad_norm": 0.11226128041744232, "learning_rate": 7.28117881993654e-05, "loss": 0.0076, "step": 30540 }, { "grad_norm": 0.13660743832588196, "learning_rate": 7.279339431475598e-05, "loss": 0.0074, "step": 30550 }, { "grad_norm": 0.1373579204082489, "learning_rate": 7.277499653537887e-05, "loss": 0.0091, "step": 30560 }, { "grad_norm": 0.11874216794967651, "learning_rate": 7.275659486437776e-05, "loss": 0.0114, "step": 30570 }, { "grad_norm": 0.11103326082229614, "learning_rate": 7.273818930489695e-05, "loss": 0.01, "step": 30580 }, { "grad_norm": 0.1272185891866684, "learning_rate": 7.271977986008151e-05, "loss": 0.007, "step": 30590 }, { "grad_norm": 0.12886479496955872, "learning_rate": 7.270136653307705e-05, "loss": 0.0072, "step": 30600 }, { "grad_norm": 0.11391771584749222, "learning_rate": 7.268294932702994e-05, "loss": 0.0094, "step": 30610 }, { "grad_norm": 0.112686388194561, "learning_rate": 7.266452824508719e-05, "loss": 0.006, "step": 30620 }, { "grad_norm": 0.11358291655778885, "learning_rate": 7.264610329039643e-05, "loss": 0.0089, "step": 30630 }, { "grad_norm": 0.14393602311611176, "learning_rate": 7.262767446610599e-05, "loss": 0.0088, "step": 30640 }, { "grad_norm": 0.11761122941970825, "learning_rate": 7.260924177536485e-05, "loss": 0.0077, "step": 30650 }, { "grad_norm": 0.15418672561645508, "learning_rate": 7.259080522132265e-05, "loss": 0.0098, "step": 30660 }, { "grad_norm": 0.11305716633796692, "learning_rate": 7.257236480712972e-05, "loss": 0.0067, "step": 30670 }, { "grad_norm": 0.12810397148132324, "learning_rate": 7.255392053593697e-05, "loss": 0.0084, "step": 30680 }, { "grad_norm": 0.12254936993122101, "learning_rate": 7.253547241089607e-05, "loss": 0.0077, "step": 30690 }, { "grad_norm": 0.15163840353488922, "learning_rate": 7.251702043515927e-05, "loss": 0.0083, "step": 30700 }, { "grad_norm": 0.12341313064098358, "learning_rate": 7.249856461187952e-05, "loss": 0.0083, "step": 30710 }, { "grad_norm": 0.10940379649400711, "learning_rate": 7.248010494421042e-05, "loss": 0.0092, "step": 30720 }, { "grad_norm": 0.14762359857559204, "learning_rate": 7.246164143530622e-05, "loss": 0.0088, "step": 30730 }, { "grad_norm": 0.12474416941404343, "learning_rate": 7.244317408832181e-05, "loss": 0.009, "step": 30740 }, { "grad_norm": 0.10385578870773315, "learning_rate": 7.242470290641279e-05, "loss": 0.007, "step": 30750 }, { "grad_norm": 0.1102539449930191, "learning_rate": 7.240622789273536e-05, "loss": 0.0087, "step": 30760 }, { "grad_norm": 0.1380193829536438, "learning_rate": 7.238774905044638e-05, "loss": 0.0085, "step": 30770 }, { "grad_norm": 0.08872491121292114, "learning_rate": 7.236926638270341e-05, "loss": 0.0099, "step": 30780 }, { "grad_norm": 0.15545307099819183, "learning_rate": 7.23507798926646e-05, "loss": 0.0067, "step": 30790 }, { "grad_norm": 0.13696196675300598, "learning_rate": 7.23322895834888e-05, "loss": 0.0093, "step": 30800 }, { "grad_norm": 0.16027085483074188, "learning_rate": 7.231379545833552e-05, "loss": 0.0089, "step": 30810 }, { "grad_norm": 0.14814095199108124, "learning_rate": 7.229529752036487e-05, "loss": 0.0094, "step": 30820 }, { "grad_norm": 0.09717617928981781, "learning_rate": 7.227679577273765e-05, "loss": 0.0104, "step": 30830 }, { "grad_norm": 0.1278049200773239, "learning_rate": 7.225829021861529e-05, "loss": 0.0071, "step": 30840 }, { "grad_norm": 0.09856053441762924, "learning_rate": 7.223978086115992e-05, "loss": 0.0071, "step": 30850 }, { "grad_norm": 0.12474533170461655, "learning_rate": 7.222126770353425e-05, "loss": 0.0094, "step": 30860 }, { "grad_norm": 0.13494102656841278, "learning_rate": 7.22027507489017e-05, "loss": 0.0116, "step": 30870 }, { "grad_norm": 0.1726561188697815, "learning_rate": 7.218423000042627e-05, "loss": 0.0091, "step": 30880 }, { "grad_norm": 0.12840166687965393, "learning_rate": 7.216570546127268e-05, "loss": 0.0076, "step": 30890 }, { "grad_norm": 0.10936382412910461, "learning_rate": 7.214717713460626e-05, "loss": 0.0082, "step": 30900 }, { "grad_norm": 0.13133220374584198, "learning_rate": 7.2128645023593e-05, "loss": 0.0082, "step": 30910 }, { "grad_norm": 0.13613508641719818, "learning_rate": 7.211010913139951e-05, "loss": 0.0095, "step": 30920 }, { "grad_norm": 0.09273558109998703, "learning_rate": 7.209156946119308e-05, "loss": 0.0084, "step": 30930 }, { "grad_norm": 0.10412030667066574, "learning_rate": 7.207302601614166e-05, "loss": 0.0075, "step": 30940 }, { "grad_norm": 0.11350452154874802, "learning_rate": 7.205447879941378e-05, "loss": 0.0077, "step": 30950 }, { "grad_norm": 0.19706369936466217, "learning_rate": 7.203592781417866e-05, "loss": 0.0089, "step": 30960 }, { "grad_norm": 0.21676324307918549, "learning_rate": 7.201737306360617e-05, "loss": 0.0074, "step": 30970 }, { "grad_norm": 0.14314623177051544, "learning_rate": 7.19988145508668e-05, "loss": 0.0078, "step": 30980 }, { "grad_norm": 0.1315770447254181, "learning_rate": 7.198025227913168e-05, "loss": 0.0096, "step": 30990 }, { "grad_norm": 0.13724054396152496, "learning_rate": 7.196168625157261e-05, "loss": 0.0083, "step": 31000 }, { "grad_norm": 0.14057928323745728, "learning_rate": 7.194311647136201e-05, "loss": 0.0095, "step": 31010 }, { "grad_norm": 0.1238984763622284, "learning_rate": 7.192454294167297e-05, "loss": 0.0088, "step": 31020 }, { "grad_norm": 0.1114596277475357, "learning_rate": 7.190596566567917e-05, "loss": 0.0069, "step": 31030 }, { "grad_norm": 0.12527212500572205, "learning_rate": 7.188738464655496e-05, "loss": 0.0109, "step": 31040 }, { "grad_norm": 0.11351510882377625, "learning_rate": 7.186879988747533e-05, "loss": 0.0079, "step": 31050 }, { "grad_norm": 0.11213116347789764, "learning_rate": 7.185021139161592e-05, "loss": 0.0072, "step": 31060 }, { "grad_norm": 0.11831565201282501, "learning_rate": 7.1831619162153e-05, "loss": 0.0075, "step": 31070 }, { "grad_norm": 0.10593404620885849, "learning_rate": 7.181302320226345e-05, "loss": 0.0097, "step": 31080 }, { "grad_norm": 0.09796573221683502, "learning_rate": 7.179442351512482e-05, "loss": 0.0074, "step": 31090 }, { "grad_norm": 0.12975814938545227, "learning_rate": 7.177582010391528e-05, "loss": 0.0088, "step": 31100 }, { "grad_norm": 0.12770618498325348, "learning_rate": 7.175721297181366e-05, "loss": 0.0073, "step": 31110 }, { "grad_norm": 0.16725298762321472, "learning_rate": 7.173860212199942e-05, "loss": 0.0094, "step": 31120 }, { "grad_norm": 0.13404227793216705, "learning_rate": 7.171998755765263e-05, "loss": 0.0095, "step": 31130 }, { "grad_norm": 0.12654288113117218, "learning_rate": 7.170136928195398e-05, "loss": 0.0105, "step": 31140 }, { "grad_norm": 0.14655724167823792, "learning_rate": 7.168274729808489e-05, "loss": 0.0083, "step": 31150 }, { "grad_norm": 0.15489940345287323, "learning_rate": 7.166412160922728e-05, "loss": 0.0082, "step": 31160 }, { "grad_norm": 0.20745235681533813, "learning_rate": 7.164549221856382e-05, "loss": 0.0085, "step": 31170 }, { "grad_norm": 0.13425858318805695, "learning_rate": 7.162685912927775e-05, "loss": 0.0085, "step": 31180 }, { "grad_norm": 0.13504599034786224, "learning_rate": 7.160822234455294e-05, "loss": 0.0073, "step": 31190 }, { "grad_norm": 0.13511422276496887, "learning_rate": 7.158958186757391e-05, "loss": 0.0079, "step": 31200 }, { "grad_norm": 0.07883056253194809, "learning_rate": 7.157093770152582e-05, "loss": 0.0097, "step": 31210 }, { "grad_norm": 0.12040740251541138, "learning_rate": 7.155228984959446e-05, "loss": 0.01, "step": 31220 }, { "grad_norm": 0.10777921229600906, "learning_rate": 7.153363831496621e-05, "loss": 0.0088, "step": 31230 }, { "grad_norm": 0.08927645534276962, "learning_rate": 7.151498310082811e-05, "loss": 0.0106, "step": 31240 }, { "grad_norm": 0.09300798177719116, "learning_rate": 7.149632421036784e-05, "loss": 0.0076, "step": 31250 }, { "grad_norm": 0.1057017594575882, "learning_rate": 7.147766164677369e-05, "loss": 0.0065, "step": 31260 }, { "grad_norm": 0.12870121002197266, "learning_rate": 7.145899541323459e-05, "loss": 0.0086, "step": 31270 }, { "grad_norm": 0.11146261543035507, "learning_rate": 7.144032551294007e-05, "loss": 0.0103, "step": 31280 }, { "grad_norm": 0.16563484072685242, "learning_rate": 7.14216519490803e-05, "loss": 0.0106, "step": 31290 }, { "grad_norm": 0.20540101826190948, "learning_rate": 7.140297472484609e-05, "loss": 0.0114, "step": 31300 }, { "grad_norm": 0.1538989543914795, "learning_rate": 7.138429384342891e-05, "loss": 0.0072, "step": 31310 }, { "grad_norm": 0.16394886374473572, "learning_rate": 7.136560930802074e-05, "loss": 0.008, "step": 31320 }, { "grad_norm": 0.1178411990404129, "learning_rate": 7.134692112181431e-05, "loss": 0.0085, "step": 31330 }, { "grad_norm": 0.11486823111772537, "learning_rate": 7.13282292880029e-05, "loss": 0.0079, "step": 31340 }, { "grad_norm": 0.10234140604734421, "learning_rate": 7.130953380978043e-05, "loss": 0.0096, "step": 31350 }, { "grad_norm": 0.12815067172050476, "learning_rate": 7.129083469034144e-05, "loss": 0.0091, "step": 31360 }, { "grad_norm": 0.08588073402643204, "learning_rate": 7.127213193288112e-05, "loss": 0.0091, "step": 31370 }, { "grad_norm": 0.0933118611574173, "learning_rate": 7.125342554059522e-05, "loss": 0.0089, "step": 31380 }, { "grad_norm": 0.16061320900917053, "learning_rate": 7.12347155166802e-05, "loss": 0.0123, "step": 31390 }, { "grad_norm": 0.11272738128900528, "learning_rate": 7.121600186433306e-05, "loss": 0.0087, "step": 31400 }, { "grad_norm": 0.15333589911460876, "learning_rate": 7.119728458675148e-05, "loss": 0.0096, "step": 31410 }, { "grad_norm": 0.1377459168434143, "learning_rate": 7.117856368713369e-05, "loss": 0.01, "step": 31420 }, { "grad_norm": 0.12110261619091034, "learning_rate": 7.115983916867861e-05, "loss": 0.0078, "step": 31430 }, { "grad_norm": 0.12084157764911652, "learning_rate": 7.114111103458574e-05, "loss": 0.0083, "step": 31440 }, { "grad_norm": 0.13966113328933716, "learning_rate": 7.11223792880552e-05, "loss": 0.0073, "step": 31450 }, { "grad_norm": 0.14766386151313782, "learning_rate": 7.110364393228773e-05, "loss": 0.0088, "step": 31460 }, { "grad_norm": 0.10906198620796204, "learning_rate": 7.108490497048471e-05, "loss": 0.0074, "step": 31470 }, { "grad_norm": 0.14236806333065033, "learning_rate": 7.10661624058481e-05, "loss": 0.0086, "step": 31480 }, { "grad_norm": 0.19728530943393707, "learning_rate": 7.10474162415805e-05, "loss": 0.0099, "step": 31490 }, { "grad_norm": 0.09366823732852936, "learning_rate": 7.102866648088511e-05, "loss": 0.0095, "step": 31500 }, { "grad_norm": 0.14175264537334442, "learning_rate": 7.100991312696576e-05, "loss": 0.0081, "step": 31510 }, { "grad_norm": 0.1150505542755127, "learning_rate": 7.099115618302686e-05, "loss": 0.0074, "step": 31520 }, { "grad_norm": 0.09443344175815582, "learning_rate": 7.097239565227349e-05, "loss": 0.0072, "step": 31530 }, { "grad_norm": 0.14183013141155243, "learning_rate": 7.09536315379113e-05, "loss": 0.0102, "step": 31540 }, { "grad_norm": 0.14422111213207245, "learning_rate": 7.093486384314656e-05, "loss": 0.0124, "step": 31550 }, { "grad_norm": 0.12224980443716049, "learning_rate": 7.091609257118616e-05, "loss": 0.008, "step": 31560 }, { "grad_norm": 0.11228565871715546, "learning_rate": 7.08973177252376e-05, "loss": 0.0066, "step": 31570 }, { "grad_norm": 0.13615329563617706, "learning_rate": 7.087853930850898e-05, "loss": 0.0071, "step": 31580 }, { "grad_norm": 0.10671135038137436, "learning_rate": 7.085975732420903e-05, "loss": 0.0073, "step": 31590 }, { "grad_norm": 0.1034955382347107, "learning_rate": 7.084097177554706e-05, "loss": 0.0075, "step": 31600 }, { "grad_norm": 0.08367142826318741, "learning_rate": 7.082218266573301e-05, "loss": 0.0068, "step": 31610 }, { "grad_norm": 0.17888666689395905, "learning_rate": 7.080338999797743e-05, "loss": 0.0115, "step": 31620 }, { "grad_norm": 0.13923141360282898, "learning_rate": 7.07845937754915e-05, "loss": 0.0064, "step": 31630 }, { "grad_norm": 0.11838140338659286, "learning_rate": 7.076579400148693e-05, "loss": 0.0087, "step": 31640 }, { "grad_norm": 0.12552855908870697, "learning_rate": 7.074699067917611e-05, "loss": 0.0093, "step": 31650 }, { "grad_norm": 0.12633851170539856, "learning_rate": 7.072818381177201e-05, "loss": 0.0097, "step": 31660 }, { "grad_norm": 0.11597015708684921, "learning_rate": 7.070937340248823e-05, "loss": 0.0103, "step": 31670 }, { "grad_norm": 0.1293923258781433, "learning_rate": 7.069055945453893e-05, "loss": 0.0095, "step": 31680 }, { "grad_norm": 0.17075476050376892, "learning_rate": 7.067174197113892e-05, "loss": 0.0082, "step": 31690 }, { "grad_norm": 0.11965290457010269, "learning_rate": 7.065292095550355e-05, "loss": 0.0087, "step": 31700 }, { "grad_norm": 0.15804094076156616, "learning_rate": 7.063409641084887e-05, "loss": 0.0073, "step": 31710 }, { "grad_norm": 0.14521750807762146, "learning_rate": 7.061526834039145e-05, "loss": 0.0071, "step": 31720 }, { "grad_norm": 0.15390776097774506, "learning_rate": 7.05964367473485e-05, "loss": 0.0077, "step": 31730 }, { "grad_norm": 0.10428856313228607, "learning_rate": 7.057760163493783e-05, "loss": 0.0083, "step": 31740 }, { "grad_norm": 0.11935844272375107, "learning_rate": 7.055876300637783e-05, "loss": 0.0088, "step": 31750 }, { "grad_norm": 0.1034611240029335, "learning_rate": 7.053992086488753e-05, "loss": 0.0079, "step": 31760 }, { "grad_norm": 0.1468246579170227, "learning_rate": 7.052107521368651e-05, "loss": 0.0091, "step": 31770 }, { "grad_norm": 0.09788819402456284, "learning_rate": 7.0502226055995e-05, "loss": 0.008, "step": 31780 }, { "grad_norm": 0.18465556204319, "learning_rate": 7.048337339503379e-05, "loss": 0.0083, "step": 31790 }, { "grad_norm": 0.1672765016555786, "learning_rate": 7.046451723402427e-05, "loss": 0.0081, "step": 31800 }, { "grad_norm": 0.10502176731824875, "learning_rate": 7.044565757618848e-05, "loss": 0.0088, "step": 31810 }, { "grad_norm": 0.10032041370868683, "learning_rate": 7.042679442474899e-05, "loss": 0.0076, "step": 31820 }, { "grad_norm": 0.12605729699134827, "learning_rate": 7.040792778292902e-05, "loss": 0.0088, "step": 31830 }, { "grad_norm": 0.11196228861808777, "learning_rate": 7.038905765395234e-05, "loss": 0.0083, "step": 31840 }, { "grad_norm": 0.1355600655078888, "learning_rate": 7.037018404104334e-05, "loss": 0.0079, "step": 31850 }, { "grad_norm": 0.1345556080341339, "learning_rate": 7.035130694742702e-05, "loss": 0.0074, "step": 31860 }, { "grad_norm": 0.11410818248987198, "learning_rate": 7.033242637632897e-05, "loss": 0.0077, "step": 31870 }, { "grad_norm": 0.1255785971879959, "learning_rate": 7.031354233097534e-05, "loss": 0.0078, "step": 31880 }, { "grad_norm": 0.09731972217559814, "learning_rate": 7.029465481459289e-05, "loss": 0.0063, "step": 31890 }, { "grad_norm": 0.11462642252445221, "learning_rate": 7.027576383040898e-05, "loss": 0.0074, "step": 31900 }, { "grad_norm": 0.12139038741588593, "learning_rate": 7.025686938165159e-05, "loss": 0.0085, "step": 31910 }, { "grad_norm": 0.13810279965400696, "learning_rate": 7.023797147154924e-05, "loss": 0.0076, "step": 31920 }, { "grad_norm": 0.1113911047577858, "learning_rate": 7.021907010333111e-05, "loss": 0.0072, "step": 31930 }, { "grad_norm": 0.1344677358865738, "learning_rate": 7.020016528022685e-05, "loss": 0.0104, "step": 31940 }, { "grad_norm": 0.12786251306533813, "learning_rate": 7.018125700546683e-05, "loss": 0.0076, "step": 31950 }, { "grad_norm": 0.150066539645195, "learning_rate": 7.016234528228196e-05, "loss": 0.0118, "step": 31960 }, { "grad_norm": 0.0908655971288681, "learning_rate": 7.014343011390372e-05, "loss": 0.0071, "step": 31970 }, { "grad_norm": 0.10260460525751114, "learning_rate": 7.01245115035642e-05, "loss": 0.0094, "step": 31980 }, { "grad_norm": 0.10528786480426788, "learning_rate": 7.010558945449606e-05, "loss": 0.0064, "step": 31990 }, { "grad_norm": 0.133759006857872, "learning_rate": 7.008666396993258e-05, "loss": 0.008, "step": 32000 }, { "grad_norm": 0.13021215796470642, "learning_rate": 7.006773505310759e-05, "loss": 0.0098, "step": 32010 }, { "grad_norm": 0.1531195491552353, "learning_rate": 7.004880270725553e-05, "loss": 0.0074, "step": 32020 }, { "grad_norm": 0.1251632571220398, "learning_rate": 7.002986693561144e-05, "loss": 0.0069, "step": 32030 }, { "grad_norm": 0.13182789087295532, "learning_rate": 7.001092774141089e-05, "loss": 0.0061, "step": 32040 }, { "grad_norm": 0.12257375568151474, "learning_rate": 6.999198512789009e-05, "loss": 0.0059, "step": 32050 }, { "grad_norm": 0.10492163151502609, "learning_rate": 6.997303909828584e-05, "loss": 0.0085, "step": 32060 }, { "grad_norm": 0.10401502996683121, "learning_rate": 6.995408965583544e-05, "loss": 0.0064, "step": 32070 }, { "grad_norm": 0.17348258197307587, "learning_rate": 6.993513680377688e-05, "loss": 0.0085, "step": 32080 }, { "grad_norm": 0.11932116001844406, "learning_rate": 6.991618054534868e-05, "loss": 0.0091, "step": 32090 }, { "grad_norm": 0.10192184150218964, "learning_rate": 6.989722088378991e-05, "loss": 0.0063, "step": 32100 }, { "grad_norm": 0.07670719921588898, "learning_rate": 6.987825782234027e-05, "loss": 0.0073, "step": 32110 }, { "grad_norm": 0.14404882490634918, "learning_rate": 6.985929136424006e-05, "loss": 0.0086, "step": 32120 }, { "grad_norm": 0.14012163877487183, "learning_rate": 6.984032151273012e-05, "loss": 0.0077, "step": 32130 }, { "grad_norm": 0.09773598611354828, "learning_rate": 6.982134827105186e-05, "loss": 0.0084, "step": 32140 }, { "grad_norm": 0.10953442007303238, "learning_rate": 6.980237164244729e-05, "loss": 0.0092, "step": 32150 }, { "grad_norm": 0.1367567777633667, "learning_rate": 6.9783391630159e-05, "loss": 0.008, "step": 32160 }, { "grad_norm": 0.12463083118200302, "learning_rate": 6.976440823743015e-05, "loss": 0.0075, "step": 32170 }, { "grad_norm": 0.14088760316371918, "learning_rate": 6.974542146750451e-05, "loss": 0.0078, "step": 32180 }, { "grad_norm": 0.12555067241191864, "learning_rate": 6.972643132362637e-05, "loss": 0.006, "step": 32190 }, { "grad_norm": 0.10597957670688629, "learning_rate": 6.970743780904064e-05, "loss": 0.0066, "step": 32200 }, { "grad_norm": 0.12955085933208466, "learning_rate": 6.968844092699277e-05, "loss": 0.0114, "step": 32210 }, { "grad_norm": 0.17660006880760193, "learning_rate": 6.966944068072883e-05, "loss": 0.0087, "step": 32220 }, { "grad_norm": 0.1410764753818512, "learning_rate": 6.965043707349545e-05, "loss": 0.0087, "step": 32230 }, { "grad_norm": 0.143118754029274, "learning_rate": 6.963143010853982e-05, "loss": 0.009, "step": 32240 }, { "grad_norm": 0.11593569815158844, "learning_rate": 6.961241978910971e-05, "loss": 0.0066, "step": 32250 }, { "grad_norm": 0.11671987920999527, "learning_rate": 6.959340611845344e-05, "loss": 0.0106, "step": 32260 }, { "grad_norm": 0.11109524220228195, "learning_rate": 6.957438909981995e-05, "loss": 0.0074, "step": 32270 }, { "grad_norm": 0.13769946992397308, "learning_rate": 6.955536873645872e-05, "loss": 0.0096, "step": 32280 }, { "grad_norm": 0.16208438575267792, "learning_rate": 6.953634503161982e-05, "loss": 0.0083, "step": 32290 }, { "grad_norm": 0.13924860954284668, "learning_rate": 6.951731798855387e-05, "loss": 0.0089, "step": 32300 }, { "grad_norm": 0.11251307278871536, "learning_rate": 6.949828761051208e-05, "loss": 0.0075, "step": 32310 }, { "grad_norm": 0.11012037843465805, "learning_rate": 6.947925390074622e-05, "loss": 0.0059, "step": 32320 }, { "grad_norm": 0.10861045122146606, "learning_rate": 6.946021686250863e-05, "loss": 0.0081, "step": 32330 }, { "grad_norm": 0.11813953518867493, "learning_rate": 6.94411764990522e-05, "loss": 0.0081, "step": 32340 }, { "grad_norm": 0.09756546467542648, "learning_rate": 6.942213281363044e-05, "loss": 0.0079, "step": 32350 }, { "grad_norm": 0.11752190440893173, "learning_rate": 6.940308580949737e-05, "loss": 0.0082, "step": 32360 }, { "grad_norm": 0.13145744800567627, "learning_rate": 6.93840354899076e-05, "loss": 0.0076, "step": 32370 }, { "grad_norm": 0.1641514152288437, "learning_rate": 6.936498185811633e-05, "loss": 0.0074, "step": 32380 }, { "grad_norm": 0.1359100192785263, "learning_rate": 6.93459249173793e-05, "loss": 0.0107, "step": 32390 }, { "grad_norm": 0.15355880558490753, "learning_rate": 6.932686467095279e-05, "loss": 0.009, "step": 32400 }, { "grad_norm": 0.1603732407093048, "learning_rate": 6.930780112209373e-05, "loss": 0.0114, "step": 32410 }, { "grad_norm": 0.13276910781860352, "learning_rate": 6.92887342740595e-05, "loss": 0.0083, "step": 32420 }, { "grad_norm": 0.12563456594944, "learning_rate": 6.926966413010816e-05, "loss": 0.0076, "step": 32430 }, { "grad_norm": 0.118159219622612, "learning_rate": 6.925059069349824e-05, "loss": 0.0072, "step": 32440 }, { "grad_norm": 0.1486779898405075, "learning_rate": 6.923151396748886e-05, "loss": 0.0085, "step": 32450 }, { "grad_norm": 0.09561353176832199, "learning_rate": 6.921243395533974e-05, "loss": 0.0087, "step": 32460 }, { "grad_norm": 0.12476080656051636, "learning_rate": 6.919335066031109e-05, "loss": 0.0097, "step": 32470 }, { "grad_norm": 0.11715209484100342, "learning_rate": 6.917426408566379e-05, "loss": 0.0087, "step": 32480 }, { "grad_norm": 0.11411207169294357, "learning_rate": 6.915517423465916e-05, "loss": 0.0079, "step": 32490 }, { "grad_norm": 0.13767701387405396, "learning_rate": 6.913608111055914e-05, "loss": 0.0064, "step": 32500 }, { "grad_norm": 0.10353484749794006, "learning_rate": 6.911698471662623e-05, "loss": 0.0065, "step": 32510 }, { "grad_norm": 0.15893518924713135, "learning_rate": 6.90978850561235e-05, "loss": 0.0096, "step": 32520 }, { "grad_norm": 0.13665258884429932, "learning_rate": 6.907878213231454e-05, "loss": 0.0077, "step": 32530 }, { "grad_norm": 0.1469576209783554, "learning_rate": 6.90596759484635e-05, "loss": 0.0139, "step": 32540 }, { "grad_norm": 0.10463418066501617, "learning_rate": 6.904056650783514e-05, "loss": 0.0076, "step": 32550 }, { "grad_norm": 0.11742162704467773, "learning_rate": 6.902145381369471e-05, "loss": 0.0068, "step": 32560 }, { "grad_norm": 0.13953496515750885, "learning_rate": 6.900233786930808e-05, "loss": 0.0088, "step": 32570 }, { "grad_norm": 0.1766936331987381, "learning_rate": 6.898321867794161e-05, "loss": 0.0093, "step": 32580 }, { "grad_norm": 0.11755155026912689, "learning_rate": 6.896409624286226e-05, "loss": 0.0088, "step": 32590 }, { "grad_norm": 0.12124498933553696, "learning_rate": 6.894497056733754e-05, "loss": 0.0111, "step": 32600 }, { "grad_norm": 0.09616102278232574, "learning_rate": 6.89258416546355e-05, "loss": 0.006, "step": 32610 }, { "grad_norm": 0.09380650520324707, "learning_rate": 6.890670950802474e-05, "loss": 0.0055, "step": 32620 }, { "grad_norm": 0.10562245547771454, "learning_rate": 6.88875741307744e-05, "loss": 0.009, "step": 32630 }, { "grad_norm": 0.12110965698957443, "learning_rate": 6.886843552615425e-05, "loss": 0.0074, "step": 32640 }, { "grad_norm": 0.1294708102941513, "learning_rate": 6.884929369743451e-05, "loss": 0.0056, "step": 32650 }, { "grad_norm": 0.11663272231817245, "learning_rate": 6.8830148647886e-05, "loss": 0.009, "step": 32660 }, { "grad_norm": 0.10134205222129822, "learning_rate": 6.88110003807801e-05, "loss": 0.0079, "step": 32670 }, { "grad_norm": 0.11146332323551178, "learning_rate": 6.87918488993887e-05, "loss": 0.0105, "step": 32680 }, { "grad_norm": 0.11925246566534042, "learning_rate": 6.877269420698431e-05, "loss": 0.0068, "step": 32690 }, { "grad_norm": 0.10670619457960129, "learning_rate": 6.875353630683989e-05, "loss": 0.0076, "step": 32700 }, { "grad_norm": 0.10470431298017502, "learning_rate": 6.873437520222905e-05, "loss": 0.0062, "step": 32710 }, { "grad_norm": 0.12681587040424347, "learning_rate": 6.871521089642585e-05, "loss": 0.0067, "step": 32720 }, { "grad_norm": 0.1371477246284485, "learning_rate": 6.869604339270498e-05, "loss": 0.0087, "step": 32730 }, { "grad_norm": 0.12835656106472015, "learning_rate": 6.867687269434164e-05, "loss": 0.0054, "step": 32740 }, { "grad_norm": 0.13399440050125122, "learning_rate": 6.865769880461156e-05, "loss": 0.0089, "step": 32750 }, { "grad_norm": 0.12671524286270142, "learning_rate": 6.863852172679104e-05, "loss": 0.0075, "step": 32760 }, { "grad_norm": 0.1252758502960205, "learning_rate": 6.861934146415693e-05, "loss": 0.0107, "step": 32770 }, { "grad_norm": 0.12865997850894928, "learning_rate": 6.86001580199866e-05, "loss": 0.0082, "step": 32780 }, { "grad_norm": 0.09094788879156113, "learning_rate": 6.858097139755798e-05, "loss": 0.0067, "step": 32790 }, { "grad_norm": 0.1292952299118042, "learning_rate": 6.856178160014955e-05, "loss": 0.0072, "step": 32800 }, { "grad_norm": 0.1179303526878357, "learning_rate": 6.85425886310403e-05, "loss": 0.0061, "step": 32810 }, { "grad_norm": 0.0976872518658638, "learning_rate": 6.852339249350979e-05, "loss": 0.006, "step": 32820 }, { "grad_norm": 0.15998299419879913, "learning_rate": 6.850419319083812e-05, "loss": 0.0101, "step": 32830 }, { "grad_norm": 0.14306792616844177, "learning_rate": 6.848499072630592e-05, "loss": 0.0069, "step": 32840 }, { "grad_norm": 0.10155331343412399, "learning_rate": 6.846578510319439e-05, "loss": 0.0069, "step": 32850 }, { "grad_norm": 0.11583258956670761, "learning_rate": 6.844657632478519e-05, "loss": 0.0068, "step": 32860 }, { "grad_norm": 0.09734643250703812, "learning_rate": 6.842736439436063e-05, "loss": 0.0064, "step": 32870 }, { "grad_norm": 0.12223807722330093, "learning_rate": 6.84081493152035e-05, "loss": 0.0083, "step": 32880 }, { "grad_norm": 0.10819839686155319, "learning_rate": 6.83889310905971e-05, "loss": 0.0061, "step": 32890 }, { "grad_norm": 0.13122183084487915, "learning_rate": 6.836970972382533e-05, "loss": 0.0076, "step": 32900 }, { "grad_norm": 0.11632873117923737, "learning_rate": 6.835048521817257e-05, "loss": 0.0061, "step": 32910 }, { "grad_norm": 0.12766127288341522, "learning_rate": 6.833125757692379e-05, "loss": 0.0081, "step": 32920 }, { "grad_norm": 0.10179566591978073, "learning_rate": 6.831202680336441e-05, "loss": 0.0075, "step": 32930 }, { "grad_norm": 0.1386221945285797, "learning_rate": 6.829279290078052e-05, "loss": 0.01, "step": 32940 }, { "grad_norm": 0.10755296796560287, "learning_rate": 6.827355587245863e-05, "loss": 0.008, "step": 32950 }, { "grad_norm": 0.10594051331281662, "learning_rate": 6.82543157216858e-05, "loss": 0.0062, "step": 32960 }, { "grad_norm": 0.13045142590999603, "learning_rate": 6.823507245174969e-05, "loss": 0.0068, "step": 32970 }, { "grad_norm": 0.11398009210824966, "learning_rate": 6.821582606593841e-05, "loss": 0.0074, "step": 32980 }, { "grad_norm": 0.14329637587070465, "learning_rate": 6.81965765675407e-05, "loss": 0.0076, "step": 32990 }, { "grad_norm": 0.10823756456375122, "learning_rate": 6.81773239598457e-05, "loss": 0.0083, "step": 33000 }, { "grad_norm": 0.11288958042860031, "learning_rate": 6.815806824614319e-05, "loss": 0.0071, "step": 33010 }, { "grad_norm": 0.13986243307590485, "learning_rate": 6.813880942972343e-05, "loss": 0.0073, "step": 33020 }, { "grad_norm": 0.1230233684182167, "learning_rate": 6.811954751387726e-05, "loss": 0.0097, "step": 33030 }, { "grad_norm": 0.1071830540895462, "learning_rate": 6.810028250189598e-05, "loss": 0.0081, "step": 33040 }, { "grad_norm": 0.13422007858753204, "learning_rate": 6.808101439707147e-05, "loss": 0.0101, "step": 33050 }, { "grad_norm": 0.13683460652828217, "learning_rate": 6.806174320269609e-05, "loss": 0.0091, "step": 33060 }, { "grad_norm": 0.13275498151779175, "learning_rate": 6.804246892206281e-05, "loss": 0.0078, "step": 33070 }, { "grad_norm": 0.09771660715341568, "learning_rate": 6.802319155846506e-05, "loss": 0.0089, "step": 33080 }, { "grad_norm": 0.11535326391458511, "learning_rate": 6.800391111519679e-05, "loss": 0.0096, "step": 33090 }, { "grad_norm": 0.1236812099814415, "learning_rate": 6.798462759555253e-05, "loss": 0.0077, "step": 33100 }, { "grad_norm": 0.12007763236761093, "learning_rate": 6.79653410028273e-05, "loss": 0.0073, "step": 33110 }, { "grad_norm": 0.12076909095048904, "learning_rate": 6.794605134031663e-05, "loss": 0.0081, "step": 33120 }, { "grad_norm": 0.1151728704571724, "learning_rate": 6.792675861131661e-05, "loss": 0.007, "step": 33130 }, { "grad_norm": 0.1681024730205536, "learning_rate": 6.790746281912386e-05, "loss": 0.0123, "step": 33140 }, { "grad_norm": 0.11967961490154266, "learning_rate": 6.788816396703546e-05, "loss": 0.0093, "step": 33150 }, { "grad_norm": 0.110141821205616, "learning_rate": 6.78688620583491e-05, "loss": 0.008, "step": 33160 }, { "grad_norm": 0.114516980946064, "learning_rate": 6.784955709636292e-05, "loss": 0.0097, "step": 33170 }, { "grad_norm": 0.11588897556066513, "learning_rate": 6.783024908437564e-05, "loss": 0.0093, "step": 33180 }, { "grad_norm": 0.15377923846244812, "learning_rate": 6.781093802568641e-05, "loss": 0.0073, "step": 33190 }, { "grad_norm": 0.1083783432841301, "learning_rate": 6.779162392359504e-05, "loss": 0.0076, "step": 33200 }, { "grad_norm": 0.14927923679351807, "learning_rate": 6.777230678140172e-05, "loss": 0.0101, "step": 33210 }, { "grad_norm": 0.10443057119846344, "learning_rate": 6.775298660240726e-05, "loss": 0.008, "step": 33220 }, { "grad_norm": 0.12522941827774048, "learning_rate": 6.773366338991292e-05, "loss": 0.0075, "step": 33230 }, { "grad_norm": 0.14250321686267853, "learning_rate": 6.771433714722052e-05, "loss": 0.0081, "step": 33240 }, { "grad_norm": 0.09163759648799896, "learning_rate": 6.769500787763239e-05, "loss": 0.0066, "step": 33250 }, { "grad_norm": 0.1338879019021988, "learning_rate": 6.76756755844514e-05, "loss": 0.0119, "step": 33260 }, { "grad_norm": 0.1045001968741417, "learning_rate": 6.765634027098087e-05, "loss": 0.0084, "step": 33270 }, { "grad_norm": 0.13053716719150543, "learning_rate": 6.763700194052468e-05, "loss": 0.0086, "step": 33280 }, { "grad_norm": 0.11188817024230957, "learning_rate": 6.761766059638723e-05, "loss": 0.008, "step": 33290 }, { "grad_norm": 0.11248800158500671, "learning_rate": 6.759831624187345e-05, "loss": 0.0072, "step": 33300 }, { "grad_norm": 0.10748646408319473, "learning_rate": 6.757896888028871e-05, "loss": 0.0076, "step": 33310 }, { "grad_norm": 0.11478061228990555, "learning_rate": 6.7559618514939e-05, "loss": 0.0061, "step": 33320 }, { "grad_norm": 0.16713163256645203, "learning_rate": 6.754026514913073e-05, "loss": 0.0107, "step": 33330 }, { "grad_norm": 0.14097096025943756, "learning_rate": 6.752090878617087e-05, "loss": 0.0082, "step": 33340 }, { "grad_norm": 0.10595100373029709, "learning_rate": 6.75015494293669e-05, "loss": 0.0067, "step": 33350 }, { "grad_norm": 0.12858590483665466, "learning_rate": 6.74821870820268e-05, "loss": 0.0083, "step": 33360 }, { "grad_norm": 0.11439955979585648, "learning_rate": 6.746282174745907e-05, "loss": 0.0077, "step": 33370 }, { "grad_norm": 0.1015995666384697, "learning_rate": 6.744345342897271e-05, "loss": 0.0089, "step": 33380 }, { "grad_norm": 0.12857979536056519, "learning_rate": 6.742408212987724e-05, "loss": 0.008, "step": 33390 }, { "grad_norm": 0.07937169075012207, "learning_rate": 6.740470785348269e-05, "loss": 0.0071, "step": 33400 }, { "grad_norm": 0.11479898542165756, "learning_rate": 6.738533060309958e-05, "loss": 0.0081, "step": 33410 }, { "grad_norm": 0.157680481672287, "learning_rate": 6.736595038203894e-05, "loss": 0.0068, "step": 33420 }, { "grad_norm": 0.11381212621927261, "learning_rate": 6.734656719361236e-05, "loss": 0.0067, "step": 33430 }, { "grad_norm": 0.15425029397010803, "learning_rate": 6.732718104113189e-05, "loss": 0.0114, "step": 33440 }, { "grad_norm": 0.12141603976488113, "learning_rate": 6.730779192791006e-05, "loss": 0.0073, "step": 33450 }, { "grad_norm": 0.10912593454122543, "learning_rate": 6.728839985725997e-05, "loss": 0.006, "step": 33460 }, { "grad_norm": 0.09496293216943741, "learning_rate": 6.726900483249517e-05, "loss": 0.0088, "step": 33470 }, { "grad_norm": 0.11368045955896378, "learning_rate": 6.724960685692976e-05, "loss": 0.0099, "step": 33480 }, { "grad_norm": 0.11005327850580215, "learning_rate": 6.723020593387833e-05, "loss": 0.0078, "step": 33490 }, { "grad_norm": 0.10942862927913666, "learning_rate": 6.721080206665593e-05, "loss": 0.0071, "step": 33500 }, { "grad_norm": 0.14784295856952667, "learning_rate": 6.719139525857819e-05, "loss": 0.0107, "step": 33510 }, { "grad_norm": 0.09814063459634781, "learning_rate": 6.717198551296117e-05, "loss": 0.006, "step": 33520 }, { "grad_norm": 0.13753949105739594, "learning_rate": 6.715257283312148e-05, "loss": 0.0092, "step": 33530 }, { "grad_norm": 0.08814740180969238, "learning_rate": 6.713315722237623e-05, "loss": 0.0058, "step": 33540 }, { "grad_norm": 0.09909692406654358, "learning_rate": 6.7113738684043e-05, "loss": 0.0076, "step": 33550 }, { "grad_norm": 0.10967062413692474, "learning_rate": 6.709431722143989e-05, "loss": 0.0082, "step": 33560 }, { "grad_norm": 0.11248348653316498, "learning_rate": 6.70748928378855e-05, "loss": 0.0064, "step": 33570 }, { "grad_norm": 0.12993629276752472, "learning_rate": 6.705546553669891e-05, "loss": 0.009, "step": 33580 }, { "grad_norm": 0.11481542885303497, "learning_rate": 6.703603532119974e-05, "loss": 0.0071, "step": 33590 }, { "grad_norm": 0.138380765914917, "learning_rate": 6.701660219470808e-05, "loss": 0.0059, "step": 33600 }, { "grad_norm": 0.10400243103504181, "learning_rate": 6.69971661605445e-05, "loss": 0.0083, "step": 33610 }, { "grad_norm": 0.1040729209780693, "learning_rate": 6.697772722203008e-05, "loss": 0.0068, "step": 33620 }, { "grad_norm": 0.13184131681919098, "learning_rate": 6.695828538248643e-05, "loss": 0.0084, "step": 33630 }, { "grad_norm": 0.10154879838228226, "learning_rate": 6.693884064523563e-05, "loss": 0.008, "step": 33640 }, { "grad_norm": 0.14778585731983185, "learning_rate": 6.691939301360023e-05, "loss": 0.0097, "step": 33650 }, { "grad_norm": 0.13417242467403412, "learning_rate": 6.689994249090333e-05, "loss": 0.0075, "step": 33660 }, { "grad_norm": 0.14753451943397522, "learning_rate": 6.688048908046845e-05, "loss": 0.0083, "step": 33670 }, { "grad_norm": 0.0976901575922966, "learning_rate": 6.686103278561969e-05, "loss": 0.0076, "step": 33680 }, { "grad_norm": 0.154434934258461, "learning_rate": 6.684157360968156e-05, "loss": 0.0107, "step": 33690 }, { "grad_norm": 0.1354566365480423, "learning_rate": 6.682211155597911e-05, "loss": 0.0073, "step": 33700 }, { "grad_norm": 0.144205704331398, "learning_rate": 6.680264662783789e-05, "loss": 0.0094, "step": 33710 }, { "grad_norm": 0.09313508123159409, "learning_rate": 6.678317882858391e-05, "loss": 0.009, "step": 33720 }, { "grad_norm": 0.1292140781879425, "learning_rate": 6.67637081615437e-05, "loss": 0.0073, "step": 33730 }, { "grad_norm": 0.129378080368042, "learning_rate": 6.674423463004427e-05, "loss": 0.0067, "step": 33740 }, { "grad_norm": 0.10051275789737701, "learning_rate": 6.672475823741308e-05, "loss": 0.0074, "step": 33750 }, { "grad_norm": 0.11528625339269638, "learning_rate": 6.670527898697811e-05, "loss": 0.0074, "step": 33760 }, { "grad_norm": 0.14633670449256897, "learning_rate": 6.668579688206788e-05, "loss": 0.0082, "step": 33770 }, { "grad_norm": 0.11907761543989182, "learning_rate": 6.666631192601131e-05, "loss": 0.0082, "step": 33780 }, { "grad_norm": 0.1378534585237503, "learning_rate": 6.664682412213785e-05, "loss": 0.0085, "step": 33790 }, { "grad_norm": 0.13332675397396088, "learning_rate": 6.662733347377745e-05, "loss": 0.01, "step": 33800 }, { "grad_norm": 0.1308380365371704, "learning_rate": 6.660783998426051e-05, "loss": 0.0082, "step": 33810 }, { "grad_norm": 0.09940933436155319, "learning_rate": 6.658834365691794e-05, "loss": 0.0055, "step": 33820 }, { "grad_norm": 0.13104526698589325, "learning_rate": 6.656884449508115e-05, "loss": 0.0082, "step": 33830 }, { "grad_norm": 0.08201302587985992, "learning_rate": 6.654934250208198e-05, "loss": 0.0063, "step": 33840 }, { "grad_norm": 0.11658404767513275, "learning_rate": 6.65298376812528e-05, "loss": 0.0084, "step": 33850 }, { "grad_norm": 0.08632419258356094, "learning_rate": 6.651033003592646e-05, "loss": 0.0061, "step": 33860 }, { "grad_norm": 0.10451420396566391, "learning_rate": 6.649081956943626e-05, "loss": 0.0063, "step": 33870 }, { "grad_norm": 0.15538765490055084, "learning_rate": 6.647130628511604e-05, "loss": 0.009, "step": 33880 }, { "grad_norm": 0.14200425148010254, "learning_rate": 6.645179018630005e-05, "loss": 0.0085, "step": 33890 }, { "grad_norm": 0.13589243590831757, "learning_rate": 6.643227127632309e-05, "loss": 0.0098, "step": 33900 }, { "grad_norm": 0.06651029735803604, "learning_rate": 6.641274955852038e-05, "loss": 0.0062, "step": 33910 }, { "grad_norm": 0.09316147863864899, "learning_rate": 6.639322503622768e-05, "loss": 0.0075, "step": 33920 }, { "grad_norm": 0.12505091726779938, "learning_rate": 6.637369771278116e-05, "loss": 0.011, "step": 33930 }, { "grad_norm": 0.09707702696323395, "learning_rate": 6.635416759151751e-05, "loss": 0.0073, "step": 33940 }, { "grad_norm": 0.10944080352783203, "learning_rate": 6.633463467577394e-05, "loss": 0.0066, "step": 33950 }, { "grad_norm": 0.0775148794054985, "learning_rate": 6.631509896888803e-05, "loss": 0.0082, "step": 33960 }, { "grad_norm": 0.11909259110689163, "learning_rate": 6.629556047419794e-05, "loss": 0.0082, "step": 33970 }, { "grad_norm": 0.12358745187520981, "learning_rate": 6.627601919504223e-05, "loss": 0.01, "step": 33980 }, { "grad_norm": 0.16984190046787262, "learning_rate": 6.625647513476001e-05, "loss": 0.0084, "step": 33990 }, { "grad_norm": 0.11303991824388504, "learning_rate": 6.62369282966908e-05, "loss": 0.0099, "step": 34000 }, { "grad_norm": 0.0797605961561203, "learning_rate": 6.621737868417464e-05, "loss": 0.0057, "step": 34010 }, { "grad_norm": 0.10298983752727509, "learning_rate": 6.619782630055198e-05, "loss": 0.0084, "step": 34020 }, { "grad_norm": 0.1065872460603714, "learning_rate": 6.617827114916382e-05, "loss": 0.0097, "step": 34030 }, { "grad_norm": 0.09800373017787933, "learning_rate": 6.615871323335161e-05, "loss": 0.0072, "step": 34040 }, { "grad_norm": 0.12415821850299835, "learning_rate": 6.613915255645725e-05, "loss": 0.0072, "step": 34050 }, { "grad_norm": 0.12326277792453766, "learning_rate": 6.611958912182312e-05, "loss": 0.0114, "step": 34060 }, { "grad_norm": 0.1129799410700798, "learning_rate": 6.610002293279207e-05, "loss": 0.0073, "step": 34070 }, { "grad_norm": 0.12397532165050507, "learning_rate": 6.608045399270746e-05, "loss": 0.0078, "step": 34080 }, { "grad_norm": 0.10250557214021683, "learning_rate": 6.606088230491304e-05, "loss": 0.0096, "step": 34090 }, { "grad_norm": 0.1439984291791916, "learning_rate": 6.604130787275312e-05, "loss": 0.0066, "step": 34100 }, { "grad_norm": 0.1100539043545723, "learning_rate": 6.602173069957242e-05, "loss": 0.0105, "step": 34110 }, { "grad_norm": 0.13430236279964447, "learning_rate": 6.600215078871612e-05, "loss": 0.0095, "step": 34120 }, { "grad_norm": 0.14714524149894714, "learning_rate": 6.598256814352992e-05, "loss": 0.0092, "step": 34130 }, { "grad_norm": 0.09795965999364853, "learning_rate": 6.596298276735995e-05, "loss": 0.0069, "step": 34140 }, { "grad_norm": 0.14283178746700287, "learning_rate": 6.594339466355282e-05, "loss": 0.0084, "step": 34150 }, { "grad_norm": 0.14275994896888733, "learning_rate": 6.592380383545558e-05, "loss": 0.0072, "step": 34160 }, { "grad_norm": 0.12232343852519989, "learning_rate": 6.590421028641577e-05, "loss": 0.0081, "step": 34170 }, { "grad_norm": 0.11385267227888107, "learning_rate": 6.588461401978143e-05, "loss": 0.0076, "step": 34180 }, { "grad_norm": 0.11562991887331009, "learning_rate": 6.586501503890099e-05, "loss": 0.0073, "step": 34190 }, { "grad_norm": 0.12217962741851807, "learning_rate": 6.584541334712338e-05, "loss": 0.007, "step": 34200 }, { "grad_norm": 0.11748335510492325, "learning_rate": 6.582580894779802e-05, "loss": 0.0064, "step": 34210 }, { "grad_norm": 0.1419123113155365, "learning_rate": 6.580620184427473e-05, "loss": 0.0072, "step": 34220 }, { "grad_norm": 0.18191097676753998, "learning_rate": 6.578659203990385e-05, "loss": 0.0086, "step": 34230 }, { "grad_norm": 0.14283369481563568, "learning_rate": 6.576697953803615e-05, "loss": 0.0083, "step": 34240 }, { "grad_norm": 0.1074056625366211, "learning_rate": 6.57473643420229e-05, "loss": 0.0079, "step": 34250 }, { "grad_norm": 0.13469430804252625, "learning_rate": 6.572774645521574e-05, "loss": 0.0092, "step": 34260 }, { "grad_norm": 0.13682571053504944, "learning_rate": 6.570812588096688e-05, "loss": 0.0076, "step": 34270 }, { "grad_norm": 0.10313938558101654, "learning_rate": 6.568850262262893e-05, "loss": 0.0092, "step": 34280 }, { "grad_norm": 0.13034586608409882, "learning_rate": 6.566887668355497e-05, "loss": 0.0062, "step": 34290 }, { "grad_norm": 0.10593893378973007, "learning_rate": 6.564924806709851e-05, "loss": 0.0064, "step": 34300 }, { "grad_norm": 0.08907656371593475, "learning_rate": 6.562961677661359e-05, "loss": 0.005, "step": 34310 }, { "grad_norm": 0.10897619277238846, "learning_rate": 6.56099828154546e-05, "loss": 0.0086, "step": 34320 }, { "grad_norm": 0.08231784403324127, "learning_rate": 6.55903461869765e-05, "loss": 0.0063, "step": 34330 }, { "grad_norm": 0.10589294135570526, "learning_rate": 6.557070689453465e-05, "loss": 0.0061, "step": 34340 }, { "grad_norm": 0.10274112224578857, "learning_rate": 6.555106494148482e-05, "loss": 0.0081, "step": 34350 }, { "grad_norm": 0.1745225042104721, "learning_rate": 6.553142033118333e-05, "loss": 0.0084, "step": 34360 }, { "grad_norm": 0.13871000707149506, "learning_rate": 6.551177306698688e-05, "loss": 0.0068, "step": 34370 }, { "grad_norm": 0.11985702812671661, "learning_rate": 6.549212315225267e-05, "loss": 0.0076, "step": 34380 }, { "grad_norm": 0.18689094483852386, "learning_rate": 6.547247059033833e-05, "loss": 0.0112, "step": 34390 }, { "grad_norm": 0.11640423536300659, "learning_rate": 6.545281538460193e-05, "loss": 0.0063, "step": 34400 }, { "grad_norm": 0.08441706001758575, "learning_rate": 6.543315753840202e-05, "loss": 0.006, "step": 34410 }, { "grad_norm": 0.10198865830898285, "learning_rate": 6.541349705509758e-05, "loss": 0.0065, "step": 34420 }, { "grad_norm": 0.1031864732503891, "learning_rate": 6.539383393804805e-05, "loss": 0.0071, "step": 34430 }, { "grad_norm": 0.08627837896347046, "learning_rate": 6.537416819061333e-05, "loss": 0.0067, "step": 34440 }, { "grad_norm": 0.11452341079711914, "learning_rate": 6.535449981615375e-05, "loss": 0.0082, "step": 34450 }, { "grad_norm": 0.10774845629930496, "learning_rate": 6.53348288180301e-05, "loss": 0.0071, "step": 34460 }, { "grad_norm": 0.11249487102031708, "learning_rate": 6.531515519960361e-05, "loss": 0.0071, "step": 34470 }, { "grad_norm": 0.12086918205022812, "learning_rate": 6.529547896423597e-05, "loss": 0.0073, "step": 34480 }, { "grad_norm": 0.109283946454525, "learning_rate": 6.52758001152893e-05, "loss": 0.0054, "step": 34490 }, { "grad_norm": 0.11914874613285065, "learning_rate": 6.525611865612618e-05, "loss": 0.0072, "step": 34500 }, { "grad_norm": 0.13579434156417847, "learning_rate": 6.523643459010966e-05, "loss": 0.0085, "step": 34510 }, { "grad_norm": 0.13454443216323853, "learning_rate": 6.521674792060317e-05, "loss": 0.0076, "step": 34520 }, { "grad_norm": 0.11352312564849854, "learning_rate": 6.519705865097063e-05, "loss": 0.0083, "step": 34530 }, { "grad_norm": 0.12151233851909637, "learning_rate": 6.517736678457641e-05, "loss": 0.007, "step": 34540 }, { "grad_norm": 0.5975819230079651, "learning_rate": 6.515767232478534e-05, "loss": 0.0126, "step": 34550 }, { "grad_norm": 0.12627200782299042, "learning_rate": 6.51379752749626e-05, "loss": 0.0096, "step": 34560 }, { "grad_norm": 0.15724478662014008, "learning_rate": 6.511827563847393e-05, "loss": 0.0109, "step": 34570 }, { "grad_norm": 0.09845959395170212, "learning_rate": 6.509857341868542e-05, "loss": 0.0138, "step": 34580 }, { "grad_norm": 0.14334502816200256, "learning_rate": 6.507886861896367e-05, "loss": 0.011, "step": 34590 }, { "grad_norm": 0.20882852375507355, "learning_rate": 6.505916124267567e-05, "loss": 0.009, "step": 34600 }, { "grad_norm": 0.10958492010831833, "learning_rate": 6.503945129318891e-05, "loss": 0.0088, "step": 34610 }, { "grad_norm": 0.1428297907114029, "learning_rate": 6.501973877387122e-05, "loss": 0.0096, "step": 34620 }, { "grad_norm": 0.1009015440940857, "learning_rate": 6.500002368809098e-05, "loss": 0.0099, "step": 34630 }, { "grad_norm": 0.10269288718700409, "learning_rate": 6.498030603921694e-05, "loss": 0.0098, "step": 34640 }, { "grad_norm": 0.13161621987819672, "learning_rate": 6.496058583061832e-05, "loss": 0.0119, "step": 34650 }, { "grad_norm": 0.0918576568365097, "learning_rate": 6.494086306566475e-05, "loss": 0.0067, "step": 34660 }, { "grad_norm": 0.13972876965999603, "learning_rate": 6.492113774772632e-05, "loss": 0.0066, "step": 34670 }, { "grad_norm": 0.10113205015659332, "learning_rate": 6.490140988017354e-05, "loss": 0.0088, "step": 34680 }, { "grad_norm": 0.13483655452728271, "learning_rate": 6.488167946637736e-05, "loss": 0.0071, "step": 34690 }, { "grad_norm": 0.1374330222606659, "learning_rate": 6.486194650970915e-05, "loss": 0.0085, "step": 34700 }, { "grad_norm": 0.09841830283403397, "learning_rate": 6.48422110135408e-05, "loss": 0.0081, "step": 34710 }, { "grad_norm": 0.12447965145111084, "learning_rate": 6.482247298124451e-05, "loss": 0.0081, "step": 34720 }, { "grad_norm": 0.1422521471977234, "learning_rate": 6.480273241619297e-05, "loss": 0.007, "step": 34730 }, { "grad_norm": 0.13232676684856415, "learning_rate": 6.478298932175933e-05, "loss": 0.008, "step": 34740 }, { "grad_norm": 0.12707914412021637, "learning_rate": 6.476324370131712e-05, "loss": 0.0086, "step": 34750 }, { "grad_norm": 0.11713019013404846, "learning_rate": 6.474349555824036e-05, "loss": 0.0069, "step": 34760 }, { "grad_norm": 0.15160520374774933, "learning_rate": 6.472374489590342e-05, "loss": 0.0076, "step": 34770 }, { "grad_norm": 0.08561159670352936, "learning_rate": 6.470399171768118e-05, "loss": 0.0059, "step": 34780 }, { "grad_norm": 0.09839658439159393, "learning_rate": 6.468423602694891e-05, "loss": 0.0069, "step": 34790 }, { "grad_norm": 0.11445393413305283, "learning_rate": 6.466447782708232e-05, "loss": 0.0065, "step": 34800 }, { "grad_norm": 0.12281063199043274, "learning_rate": 6.464471712145754e-05, "loss": 0.0083, "step": 34810 }, { "grad_norm": 0.11125575006008148, "learning_rate": 6.462495391345114e-05, "loss": 0.0067, "step": 34820 }, { "grad_norm": 0.08351079374551773, "learning_rate": 6.46051882064401e-05, "loss": 0.0079, "step": 34830 }, { "grad_norm": 0.1054607555270195, "learning_rate": 6.458542000380186e-05, "loss": 0.0063, "step": 34840 }, { "grad_norm": 0.09005733579397202, "learning_rate": 6.456564930891424e-05, "loss": 0.0066, "step": 34850 }, { "grad_norm": 0.10705135762691498, "learning_rate": 6.454587612515555e-05, "loss": 0.0055, "step": 34860 }, { "grad_norm": 0.11737534403800964, "learning_rate": 6.452610045590444e-05, "loss": 0.0091, "step": 34870 }, { "grad_norm": 0.089551642537117, "learning_rate": 6.450632230454005e-05, "loss": 0.007, "step": 34880 }, { "grad_norm": 0.09751211851835251, "learning_rate": 6.448654167444195e-05, "loss": 0.0078, "step": 34890 }, { "grad_norm": 0.10635056346654892, "learning_rate": 6.446675856899005e-05, "loss": 0.0062, "step": 34900 }, { "grad_norm": 0.1369198113679886, "learning_rate": 6.444697299156481e-05, "loss": 0.0065, "step": 34910 }, { "grad_norm": 0.0923485979437828, "learning_rate": 6.442718494554701e-05, "loss": 0.006, "step": 34920 }, { "grad_norm": 0.1281081587076187, "learning_rate": 6.440739443431787e-05, "loss": 0.0073, "step": 34930 }, { "grad_norm": 0.10571734607219696, "learning_rate": 6.438760146125906e-05, "loss": 0.0084, "step": 34940 }, { "grad_norm": 0.13459455966949463, "learning_rate": 6.436780602975267e-05, "loss": 0.0067, "step": 34950 }, { "grad_norm": 0.13689514994621277, "learning_rate": 6.43480081431812e-05, "loss": 0.007, "step": 34960 }, { "grad_norm": 0.09126511216163635, "learning_rate": 6.432820780492756e-05, "loss": 0.0067, "step": 34970 }, { "grad_norm": 0.10499921441078186, "learning_rate": 6.430840501837506e-05, "loss": 0.0101, "step": 34980 }, { "grad_norm": 0.1420610398054123, "learning_rate": 6.428859978690748e-05, "loss": 0.0071, "step": 34990 }, { "grad_norm": 0.08829671144485474, "learning_rate": 6.426879211390901e-05, "loss": 0.0067, "step": 35000 }, { "grad_norm": 0.13213016092777252, "learning_rate": 6.424898200276422e-05, "loss": 0.0125, "step": 35010 }, { "grad_norm": 0.11910593509674072, "learning_rate": 6.42291694568581e-05, "loss": 0.0063, "step": 35020 }, { "grad_norm": 0.13793246448040009, "learning_rate": 6.42093544795761e-05, "loss": 0.0084, "step": 35030 }, { "grad_norm": 0.10908116400241852, "learning_rate": 6.418953707430403e-05, "loss": 0.007, "step": 35040 }, { "grad_norm": 0.11027493327856064, "learning_rate": 6.416971724442819e-05, "loss": 0.0069, "step": 35050 }, { "grad_norm": 0.09511706978082657, "learning_rate": 6.414989499333519e-05, "loss": 0.0059, "step": 35060 }, { "grad_norm": 0.1365094780921936, "learning_rate": 6.413007032441214e-05, "loss": 0.0068, "step": 35070 }, { "grad_norm": 0.13597385585308075, "learning_rate": 6.411024324104653e-05, "loss": 0.0114, "step": 35080 }, { "grad_norm": 0.14264634251594543, "learning_rate": 6.409041374662628e-05, "loss": 0.0083, "step": 35090 }, { "grad_norm": 0.127044215798378, "learning_rate": 6.407058184453967e-05, "loss": 0.0082, "step": 35100 }, { "grad_norm": 0.1339179128408432, "learning_rate": 6.405074753817548e-05, "loss": 0.0099, "step": 35110 }, { "grad_norm": 0.11250170320272446, "learning_rate": 6.40309108309228e-05, "loss": 0.0078, "step": 35120 }, { "grad_norm": 0.10299038141965866, "learning_rate": 6.401107172617122e-05, "loss": 0.0071, "step": 35130 }, { "grad_norm": 0.11947577446699142, "learning_rate": 6.399123022731068e-05, "loss": 0.0063, "step": 35140 }, { "grad_norm": 0.10694131255149841, "learning_rate": 6.397138633773157e-05, "loss": 0.0063, "step": 35150 }, { "grad_norm": 0.09635505080223083, "learning_rate": 6.395154006082463e-05, "loss": 0.0065, "step": 35160 }, { "grad_norm": 0.1123913899064064, "learning_rate": 6.393169139998109e-05, "loss": 0.006, "step": 35170 }, { "grad_norm": 0.11682234704494476, "learning_rate": 6.39118403585925e-05, "loss": 0.0099, "step": 35180 }, { "grad_norm": 0.13757838308811188, "learning_rate": 6.38919869400509e-05, "loss": 0.0097, "step": 35190 }, { "grad_norm": 0.11985457688570023, "learning_rate": 6.387213114774865e-05, "loss": 0.0097, "step": 35200 }, { "grad_norm": 0.09225067496299744, "learning_rate": 6.385227298507863e-05, "loss": 0.0064, "step": 35210 }, { "grad_norm": 0.14687545597553253, "learning_rate": 6.3832412455434e-05, "loss": 0.0067, "step": 35220 }, { "grad_norm": 0.12889568507671356, "learning_rate": 6.381254956220841e-05, "loss": 0.0078, "step": 35230 }, { "grad_norm": 0.13202936947345734, "learning_rate": 6.379268430879586e-05, "loss": 0.0081, "step": 35240 }, { "grad_norm": 0.12556768953800201, "learning_rate": 6.37728166985908e-05, "loss": 0.0074, "step": 35250 }, { "grad_norm": 0.1411169171333313, "learning_rate": 6.375294673498804e-05, "loss": 0.0078, "step": 35260 }, { "grad_norm": 0.13777053356170654, "learning_rate": 6.373307442138284e-05, "loss": 0.0069, "step": 35270 }, { "grad_norm": 0.10212031751871109, "learning_rate": 6.371319976117081e-05, "loss": 0.005, "step": 35280 }, { "grad_norm": 0.08732541650533676, "learning_rate": 6.3693322757748e-05, "loss": 0.0077, "step": 35290 }, { "grad_norm": 0.09404940903186798, "learning_rate": 6.367344341451086e-05, "loss": 0.0072, "step": 35300 }, { "grad_norm": 0.15339398384094238, "learning_rate": 6.36535617348562e-05, "loss": 0.0073, "step": 35310 }, { "grad_norm": 0.16407355666160583, "learning_rate": 6.363367772218128e-05, "loss": 0.0086, "step": 35320 }, { "grad_norm": 0.12656988203525543, "learning_rate": 6.36137913798837e-05, "loss": 0.0083, "step": 35330 }, { "grad_norm": 0.10384569317102432, "learning_rate": 6.359390271136151e-05, "loss": 0.0074, "step": 35340 }, { "grad_norm": 0.13296785950660706, "learning_rate": 6.357401172001314e-05, "loss": 0.0061, "step": 35350 }, { "grad_norm": 0.11847536265850067, "learning_rate": 6.355411840923742e-05, "loss": 0.0068, "step": 35360 }, { "grad_norm": 0.11714331805706024, "learning_rate": 6.353422278243358e-05, "loss": 0.0098, "step": 35370 }, { "grad_norm": 0.11279837787151337, "learning_rate": 6.351432484300121e-05, "loss": 0.0057, "step": 35380 }, { "grad_norm": 0.11492615193128586, "learning_rate": 6.349442459434036e-05, "loss": 0.0085, "step": 35390 }, { "grad_norm": 0.15614105761051178, "learning_rate": 6.34745220398514e-05, "loss": 0.0072, "step": 35400 }, { "grad_norm": 0.10065457224845886, "learning_rate": 6.345461718293518e-05, "loss": 0.0061, "step": 35410 }, { "grad_norm": 0.13789132237434387, "learning_rate": 6.343471002699286e-05, "loss": 0.0097, "step": 35420 }, { "grad_norm": 0.12613628804683685, "learning_rate": 6.341480057542602e-05, "loss": 0.0089, "step": 35430 }, { "grad_norm": 0.13187992572784424, "learning_rate": 6.339488883163667e-05, "loss": 0.011, "step": 35440 }, { "grad_norm": 0.1221231147646904, "learning_rate": 6.337497479902716e-05, "loss": 0.0078, "step": 35450 }, { "grad_norm": 0.09255003184080124, "learning_rate": 6.335505848100027e-05, "loss": 0.0084, "step": 35460 }, { "grad_norm": 0.14506825804710388, "learning_rate": 6.333513988095915e-05, "loss": 0.0078, "step": 35470 }, { "grad_norm": 0.10580259561538696, "learning_rate": 6.331521900230735e-05, "loss": 0.0079, "step": 35480 }, { "grad_norm": 0.0883297473192215, "learning_rate": 6.329529584844878e-05, "loss": 0.0041, "step": 35490 }, { "grad_norm": 0.08703397959470749, "learning_rate": 6.327537042278777e-05, "loss": 0.0072, "step": 35500 }, { "grad_norm": 0.13177257776260376, "learning_rate": 6.325544272872905e-05, "loss": 0.0105, "step": 35510 }, { "grad_norm": 0.10971875488758087, "learning_rate": 6.323551276967771e-05, "loss": 0.0085, "step": 35520 }, { "grad_norm": 0.11538208276033401, "learning_rate": 6.321558054903922e-05, "loss": 0.0081, "step": 35530 }, { "grad_norm": 0.15148618817329407, "learning_rate": 6.319564607021947e-05, "loss": 0.0083, "step": 35540 }, { "grad_norm": 0.1439383625984192, "learning_rate": 6.31757093366247e-05, "loss": 0.0065, "step": 35550 }, { "grad_norm": 0.09042482823133469, "learning_rate": 6.315577035166154e-05, "loss": 0.0079, "step": 35560 }, { "grad_norm": 0.11551506817340851, "learning_rate": 6.313582911873708e-05, "loss": 0.0082, "step": 35570 }, { "grad_norm": 0.13748030364513397, "learning_rate": 6.311588564125865e-05, "loss": 0.0069, "step": 35580 }, { "grad_norm": 0.14104263484477997, "learning_rate": 6.30959399226341e-05, "loss": 0.0085, "step": 35590 }, { "grad_norm": 0.11551681905984879, "learning_rate": 6.30759919662716e-05, "loss": 0.0071, "step": 35600 }, { "grad_norm": 0.11157647520303726, "learning_rate": 6.30560417755797e-05, "loss": 0.007, "step": 35610 }, { "grad_norm": 0.18346025049686432, "learning_rate": 6.303608935396735e-05, "loss": 0.0095, "step": 35620 }, { "grad_norm": 0.1164461001753807, "learning_rate": 6.301613470484386e-05, "loss": 0.0091, "step": 35630 }, { "grad_norm": 0.15242020785808563, "learning_rate": 6.299617783161893e-05, "loss": 0.0094, "step": 35640 }, { "grad_norm": 0.11499565839767456, "learning_rate": 6.297621873770266e-05, "loss": 0.0071, "step": 35650 }, { "grad_norm": 0.07291830331087112, "learning_rate": 6.29562574265055e-05, "loss": 0.0068, "step": 35660 }, { "grad_norm": 0.09214100986719131, "learning_rate": 6.293629390143834e-05, "loss": 0.0068, "step": 35670 }, { "grad_norm": 0.11036615073680878, "learning_rate": 6.291632816591232e-05, "loss": 0.0074, "step": 35680 }, { "grad_norm": 0.11725582927465439, "learning_rate": 6.28963602233391e-05, "loss": 0.0082, "step": 35690 }, { "grad_norm": 0.14808513224124908, "learning_rate": 6.287639007713062e-05, "loss": 0.0076, "step": 35700 }, { "grad_norm": 0.09286350011825562, "learning_rate": 6.285641773069926e-05, "loss": 0.0069, "step": 35710 }, { "grad_norm": 0.11982521414756775, "learning_rate": 6.283644318745773e-05, "loss": 0.0101, "step": 35720 }, { "grad_norm": 0.0994466245174408, "learning_rate": 6.281646645081912e-05, "loss": 0.0077, "step": 35730 }, { "grad_norm": 0.1519741714000702, "learning_rate": 6.279648752419693e-05, "loss": 0.0065, "step": 35740 }, { "grad_norm": 0.12067423015832901, "learning_rate": 6.2776506411005e-05, "loss": 0.0065, "step": 35750 }, { "grad_norm": 0.11081679910421371, "learning_rate": 6.275652311465758e-05, "loss": 0.0068, "step": 35760 }, { "grad_norm": 0.08629626780748367, "learning_rate": 6.273653763856926e-05, "loss": 0.0088, "step": 35770 }, { "grad_norm": 0.0874653086066246, "learning_rate": 6.271654998615501e-05, "loss": 0.0051, "step": 35780 }, { "grad_norm": 0.12070801109075546, "learning_rate": 6.269656016083013e-05, "loss": 0.0054, "step": 35790 }, { "grad_norm": 0.11581446975469589, "learning_rate": 6.267656816601038e-05, "loss": 0.0073, "step": 35800 }, { "grad_norm": 0.13313165307044983, "learning_rate": 6.265657400511185e-05, "loss": 0.0092, "step": 35810 }, { "grad_norm": 0.13683435320854187, "learning_rate": 6.263657768155098e-05, "loss": 0.0058, "step": 35820 }, { "grad_norm": 0.10509780049324036, "learning_rate": 6.261657919874457e-05, "loss": 0.0058, "step": 35830 }, { "grad_norm": 0.13104048371315002, "learning_rate": 6.259657856010986e-05, "loss": 0.0069, "step": 35840 }, { "grad_norm": 0.08970591425895691, "learning_rate": 6.257657576906439e-05, "loss": 0.0058, "step": 35850 }, { "grad_norm": 0.10101807117462158, "learning_rate": 6.255657082902609e-05, "loss": 0.0053, "step": 35860 }, { "grad_norm": 0.10585292428731918, "learning_rate": 6.253656374341325e-05, "loss": 0.009, "step": 35870 }, { "grad_norm": 0.1325748711824417, "learning_rate": 6.251655451564457e-05, "loss": 0.0055, "step": 35880 }, { "grad_norm": 0.10323511064052582, "learning_rate": 6.249654314913902e-05, "loss": 0.0064, "step": 35890 }, { "grad_norm": 0.08991889655590057, "learning_rate": 6.247652964731604e-05, "loss": 0.0057, "step": 35900 }, { "grad_norm": 0.11257924884557724, "learning_rate": 6.245651401359537e-05, "loss": 0.0092, "step": 35910 }, { "grad_norm": 0.115573450922966, "learning_rate": 6.243649625139715e-05, "loss": 0.0055, "step": 35920 }, { "grad_norm": 0.07988741993904114, "learning_rate": 6.241647636414185e-05, "loss": 0.0058, "step": 35930 }, { "grad_norm": 0.07959599047899246, "learning_rate": 6.239645435525034e-05, "loss": 0.0063, "step": 35940 }, { "grad_norm": 0.07056096196174622, "learning_rate": 6.237643022814381e-05, "loss": 0.0059, "step": 35950 }, { "grad_norm": 0.07658150792121887, "learning_rate": 6.235640398624386e-05, "loss": 0.0054, "step": 35960 }, { "grad_norm": 0.14637941122055054, "learning_rate": 6.233637563297243e-05, "loss": 0.0076, "step": 35970 }, { "grad_norm": 0.10847117751836777, "learning_rate": 6.23163451717518e-05, "loss": 0.0065, "step": 35980 }, { "grad_norm": 0.1482595056295395, "learning_rate": 6.229631260600463e-05, "loss": 0.008, "step": 35990 }, { "grad_norm": 0.13039261102676392, "learning_rate": 6.227627793915392e-05, "loss": 0.0072, "step": 36000 }, { "grad_norm": 0.0916108638048172, "learning_rate": 6.225624117462309e-05, "loss": 0.0074, "step": 36010 }, { "grad_norm": 0.13109399378299713, "learning_rate": 6.223620231583586e-05, "loss": 0.005, "step": 36020 }, { "grad_norm": 0.11787163466215134, "learning_rate": 6.221616136621629e-05, "loss": 0.0089, "step": 36030 }, { "grad_norm": 0.0846395418047905, "learning_rate": 6.219611832918887e-05, "loss": 0.0078, "step": 36040 }, { "grad_norm": 0.11355404555797577, "learning_rate": 6.217607320817838e-05, "loss": 0.0058, "step": 36050 }, { "grad_norm": 0.09932220727205276, "learning_rate": 6.215602600661001e-05, "loss": 0.0076, "step": 36060 }, { "grad_norm": 0.16636031866073608, "learning_rate": 6.213597672790925e-05, "loss": 0.0069, "step": 36070 }, { "grad_norm": 0.1403188407421112, "learning_rate": 6.2115925375502e-05, "loss": 0.0089, "step": 36080 }, { "grad_norm": 0.16284233331680298, "learning_rate": 6.209587195281447e-05, "loss": 0.0087, "step": 36090 }, { "grad_norm": 0.09937015920877457, "learning_rate": 6.207581646327324e-05, "loss": 0.0085, "step": 36100 }, { "grad_norm": 0.1189359650015831, "learning_rate": 6.205575891030526e-05, "loss": 0.0078, "step": 36110 }, { "grad_norm": 0.09160592406988144, "learning_rate": 6.203569929733781e-05, "loss": 0.0074, "step": 36120 }, { "grad_norm": 0.1076212227344513, "learning_rate": 6.201563762779852e-05, "loss": 0.0065, "step": 36130 }, { "grad_norm": 0.07713580131530762, "learning_rate": 6.199557390511538e-05, "loss": 0.0059, "step": 36140 }, { "grad_norm": 0.11259203404188156, "learning_rate": 6.197550813271675e-05, "loss": 0.0068, "step": 36150 }, { "grad_norm": 0.11429649591445923, "learning_rate": 6.195544031403131e-05, "loss": 0.0064, "step": 36160 }, { "grad_norm": 0.10450203716754913, "learning_rate": 6.19353704524881e-05, "loss": 0.0092, "step": 36170 }, { "grad_norm": 0.08982928097248077, "learning_rate": 6.191529855151652e-05, "loss": 0.006, "step": 36180 }, { "grad_norm": 0.11228685826063156, "learning_rate": 6.189522461454629e-05, "loss": 0.0065, "step": 36190 }, { "grad_norm": 0.163313090801239, "learning_rate": 6.187514864500752e-05, "loss": 0.0064, "step": 36200 }, { "grad_norm": 0.08046547323465347, "learning_rate": 6.185507064633062e-05, "loss": 0.0065, "step": 36210 }, { "grad_norm": 0.08880160003900528, "learning_rate": 6.18349906219464e-05, "loss": 0.0055, "step": 36220 }, { "grad_norm": 0.10454404354095459, "learning_rate": 6.181490857528596e-05, "loss": 0.0091, "step": 36230 }, { "grad_norm": 0.14903570711612701, "learning_rate": 6.179482450978077e-05, "loss": 0.0079, "step": 36240 }, { "grad_norm": 0.10927315056324005, "learning_rate": 6.177473842886269e-05, "loss": 0.0075, "step": 36250 }, { "grad_norm": 0.12373427301645279, "learning_rate": 6.175465033596382e-05, "loss": 0.0069, "step": 36260 }, { "grad_norm": 0.08791036158800125, "learning_rate": 6.173456023451671e-05, "loss": 0.007, "step": 36270 }, { "grad_norm": 0.10833849757909775, "learning_rate": 6.171446812795422e-05, "loss": 0.0075, "step": 36280 }, { "grad_norm": 0.13342231512069702, "learning_rate": 6.169437401970949e-05, "loss": 0.007, "step": 36290 }, { "grad_norm": 0.12638749182224274, "learning_rate": 6.16742779132161e-05, "loss": 0.0061, "step": 36300 }, { "grad_norm": 0.13307157158851624, "learning_rate": 6.165417981190789e-05, "loss": 0.0065, "step": 36310 }, { "grad_norm": 0.08480724692344666, "learning_rate": 6.16340797192191e-05, "loss": 0.0064, "step": 36320 }, { "grad_norm": 0.1490246206521988, "learning_rate": 6.161397763858427e-05, "loss": 0.0082, "step": 36330 }, { "grad_norm": 0.1425299048423767, "learning_rate": 6.159387357343834e-05, "loss": 0.0091, "step": 36340 }, { "grad_norm": 0.10169410705566406, "learning_rate": 6.157376752721648e-05, "loss": 0.0075, "step": 36350 }, { "grad_norm": 0.08815211802721024, "learning_rate": 6.155365950335428e-05, "loss": 0.0059, "step": 36360 }, { "grad_norm": 0.11304861307144165, "learning_rate": 6.153354950528768e-05, "loss": 0.0068, "step": 36370 }, { "grad_norm": 0.15750521421432495, "learning_rate": 6.151343753645293e-05, "loss": 0.0079, "step": 36380 }, { "grad_norm": 0.11328013986349106, "learning_rate": 6.149332360028657e-05, "loss": 0.009, "step": 36390 }, { "grad_norm": 0.10050570219755173, "learning_rate": 6.147320770022555e-05, "loss": 0.0065, "step": 36400 }, { "grad_norm": 0.1536591649055481, "learning_rate": 6.145308983970715e-05, "loss": 0.0067, "step": 36410 }, { "grad_norm": 0.11280779540538788, "learning_rate": 6.143297002216892e-05, "loss": 0.0068, "step": 36420 }, { "grad_norm": 0.09019284695386887, "learning_rate": 6.141284825104882e-05, "loss": 0.0078, "step": 36430 }, { "grad_norm": 0.08608116209506989, "learning_rate": 6.13927245297851e-05, "loss": 0.0066, "step": 36440 }, { "grad_norm": 0.14580315351486206, "learning_rate": 6.137259886181633e-05, "loss": 0.0068, "step": 36450 }, { "grad_norm": 0.09133361279964447, "learning_rate": 6.135247125058145e-05, "loss": 0.0053, "step": 36460 }, { "grad_norm": 0.12129972130060196, "learning_rate": 6.133234169951974e-05, "loss": 0.0059, "step": 36470 }, { "grad_norm": 0.09126780927181244, "learning_rate": 6.131221021207078e-05, "loss": 0.0056, "step": 36480 }, { "grad_norm": 0.13291697204113007, "learning_rate": 6.129207679167448e-05, "loss": 0.0061, "step": 36490 }, { "grad_norm": 0.14834794402122498, "learning_rate": 6.127194144177109e-05, "loss": 0.0059, "step": 36500 }, { "grad_norm": 0.12644745409488678, "learning_rate": 6.125180416580118e-05, "loss": 0.0051, "step": 36510 }, { "grad_norm": 0.09175247699022293, "learning_rate": 6.123166496720571e-05, "loss": 0.005, "step": 36520 }, { "grad_norm": 0.06844156980514526, "learning_rate": 6.121152384942588e-05, "loss": 0.0047, "step": 36530 }, { "grad_norm": 0.13642515242099762, "learning_rate": 6.119138081590324e-05, "loss": 0.0075, "step": 36540 }, { "grad_norm": 0.09307030588388443, "learning_rate": 6.117123587007971e-05, "loss": 0.0058, "step": 36550 }, { "grad_norm": 0.09732282906770706, "learning_rate": 6.11510890153975e-05, "loss": 0.0054, "step": 36560 }, { "grad_norm": 0.0933971107006073, "learning_rate": 6.113094025529916e-05, "loss": 0.006, "step": 36570 }, { "grad_norm": 0.12676957249641418, "learning_rate": 6.111078959322757e-05, "loss": 0.0102, "step": 36580 }, { "grad_norm": 0.1358274221420288, "learning_rate": 6.109063703262592e-05, "loss": 0.0087, "step": 36590 }, { "grad_norm": 0.12883898615837097, "learning_rate": 6.107048257693772e-05, "loss": 0.008, "step": 36600 }, { "grad_norm": 0.0915105938911438, "learning_rate": 6.105032622960683e-05, "loss": 0.0082, "step": 36610 }, { "grad_norm": 0.08748017996549606, "learning_rate": 6.103016799407743e-05, "loss": 0.0078, "step": 36620 }, { "grad_norm": 0.13090750575065613, "learning_rate": 6.1010007873793984e-05, "loss": 0.006, "step": 36630 }, { "grad_norm": 0.13796916604042053, "learning_rate": 6.098984587220131e-05, "loss": 0.0058, "step": 36640 }, { "grad_norm": 0.0989459678530693, "learning_rate": 6.096968199274456e-05, "loss": 0.0078, "step": 36650 }, { "grad_norm": 0.12049460411071777, "learning_rate": 6.0949516238869166e-05, "loss": 0.0058, "step": 36660 }, { "grad_norm": 0.11188188195228577, "learning_rate": 6.092934861402092e-05, "loss": 0.0066, "step": 36670 }, { "grad_norm": 0.11265577375888824, "learning_rate": 6.0909179121645924e-05, "loss": 0.0063, "step": 36680 }, { "grad_norm": 0.08879607170820236, "learning_rate": 6.0889007765190576e-05, "loss": 0.0078, "step": 36690 }, { "grad_norm": 0.10993573814630508, "learning_rate": 6.086883454810162e-05, "loss": 0.0065, "step": 36700 }, { "grad_norm": 0.12582935392856598, "learning_rate": 6.0848659473826084e-05, "loss": 0.0068, "step": 36710 }, { "grad_norm": 0.10457294434309006, "learning_rate": 6.082848254581138e-05, "loss": 0.0073, "step": 36720 }, { "grad_norm": 0.10356298089027405, "learning_rate": 6.080830376750517e-05, "loss": 0.0063, "step": 36730 }, { "grad_norm": 0.11885710060596466, "learning_rate": 6.0788123142355445e-05, "loss": 0.0069, "step": 36740 }, { "grad_norm": 0.12487290054559708, "learning_rate": 6.076794067381052e-05, "loss": 0.01, "step": 36750 }, { "grad_norm": 0.09915650635957718, "learning_rate": 6.074775636531905e-05, "loss": 0.0091, "step": 36760 }, { "grad_norm": 0.10889173299074173, "learning_rate": 6.072757022032997e-05, "loss": 0.005, "step": 36770 }, { "grad_norm": 0.10094565153121948, "learning_rate": 6.070738224229253e-05, "loss": 0.0063, "step": 36780 }, { "grad_norm": 0.1186944991350174, "learning_rate": 6.0687192434656314e-05, "loss": 0.0066, "step": 36790 }, { "grad_norm": 0.12333716452121735, "learning_rate": 6.066700080087121e-05, "loss": 0.0083, "step": 36800 }, { "grad_norm": 0.101898692548275, "learning_rate": 6.0646807344387424e-05, "loss": 0.006, "step": 36810 }, { "grad_norm": 0.12781886756420135, "learning_rate": 6.062661206865543e-05, "loss": 0.0081, "step": 36820 }, { "grad_norm": 0.14406166970729828, "learning_rate": 6.06064149771261e-05, "loss": 0.0086, "step": 36830 }, { "grad_norm": 0.11372888088226318, "learning_rate": 6.058621607325051e-05, "loss": 0.0063, "step": 36840 }, { "grad_norm": 0.12362180650234222, "learning_rate": 6.056601536048014e-05, "loss": 0.0071, "step": 36850 }, { "grad_norm": 0.12684109807014465, "learning_rate": 6.0545812842266725e-05, "loss": 0.0076, "step": 36860 }, { "grad_norm": 0.14454318583011627, "learning_rate": 6.052560852206232e-05, "loss": 0.0061, "step": 36870 }, { "grad_norm": 0.09869930148124695, "learning_rate": 6.05054024033193e-05, "loss": 0.0077, "step": 36880 }, { "grad_norm": 0.12655144929885864, "learning_rate": 6.048519448949032e-05, "loss": 0.0059, "step": 36890 }, { "grad_norm": 0.1450597494840622, "learning_rate": 6.046498478402839e-05, "loss": 0.0064, "step": 36900 }, { "grad_norm": 0.13118453323841095, "learning_rate": 6.044477329038677e-05, "loss": 0.0102, "step": 36910 }, { "grad_norm": 0.12635934352874756, "learning_rate": 6.042456001201906e-05, "loss": 0.0066, "step": 36920 }, { "grad_norm": 0.0894128829240799, "learning_rate": 6.040434495237917e-05, "loss": 0.0057, "step": 36930 }, { "grad_norm": 0.15677829086780548, "learning_rate": 6.0384128114921256e-05, "loss": 0.0062, "step": 36940 }, { "grad_norm": 0.14695291221141815, "learning_rate": 6.036390950309987e-05, "loss": 0.01, "step": 36950 }, { "grad_norm": 0.1085531935095787, "learning_rate": 6.0343689120369805e-05, "loss": 0.0096, "step": 36960 }, { "grad_norm": 0.10319355130195618, "learning_rate": 6.032346697018616e-05, "loss": 0.0078, "step": 36970 }, { "grad_norm": 0.10747858136892319, "learning_rate": 6.0303243056004375e-05, "loss": 0.0062, "step": 36980 }, { "grad_norm": 0.14897912740707397, "learning_rate": 6.0283017381280136e-05, "loss": 0.0059, "step": 36990 }, { "grad_norm": 0.11753976345062256, "learning_rate": 6.026278994946948e-05, "loss": 0.0081, "step": 37000 }, { "grad_norm": 0.16086238622665405, "learning_rate": 6.02425607640287e-05, "loss": 0.0074, "step": 37010 }, { "grad_norm": 0.10444991290569305, "learning_rate": 6.022232982841441e-05, "loss": 0.006, "step": 37020 }, { "grad_norm": 0.10741551965475082, "learning_rate": 6.020209714608355e-05, "loss": 0.0057, "step": 37030 }, { "grad_norm": 0.09144550561904907, "learning_rate": 6.018186272049331e-05, "loss": 0.009, "step": 37040 }, { "grad_norm": 0.10455091297626495, "learning_rate": 6.01616265551012e-05, "loss": 0.0077, "step": 37050 }, { "grad_norm": 0.09327354282140732, "learning_rate": 6.014138865336503e-05, "loss": 0.0082, "step": 37060 }, { "grad_norm": 0.10875718295574188, "learning_rate": 6.0121149018742905e-05, "loss": 0.007, "step": 37070 }, { "grad_norm": 0.12952324748039246, "learning_rate": 6.010090765469325e-05, "loss": 0.0068, "step": 37080 }, { "grad_norm": 0.15278002619743347, "learning_rate": 6.008066456467473e-05, "loss": 0.0087, "step": 37090 }, { "grad_norm": 0.1287579983472824, "learning_rate": 6.0060419752146335e-05, "loss": 0.0071, "step": 37100 }, { "grad_norm": 0.13015678524971008, "learning_rate": 6.0040173220567353e-05, "loss": 0.0069, "step": 37110 }, { "grad_norm": 0.16127392649650574, "learning_rate": 6.001992497339737e-05, "loss": 0.0057, "step": 37120 }, { "grad_norm": 0.09271649271249771, "learning_rate": 5.999967501409626e-05, "loss": 0.0063, "step": 37130 }, { "grad_norm": 0.07371029257774353, "learning_rate": 5.997942334612418e-05, "loss": 0.0049, "step": 37140 }, { "grad_norm": 0.11969602853059769, "learning_rate": 5.995916997294158e-05, "loss": 0.0069, "step": 37150 }, { "grad_norm": 0.1500859558582306, "learning_rate": 5.9938914898009214e-05, "loss": 0.0063, "step": 37160 }, { "grad_norm": 0.09219396114349365, "learning_rate": 5.991865812478813e-05, "loss": 0.0081, "step": 37170 }, { "grad_norm": 0.09651771932840347, "learning_rate": 5.989839965673964e-05, "loss": 0.0072, "step": 37180 }, { "grad_norm": 0.13139817118644714, "learning_rate": 5.987813949732539e-05, "loss": 0.011, "step": 37190 }, { "grad_norm": 0.10100707411766052, "learning_rate": 5.9857877650007255e-05, "loss": 0.0066, "step": 37200 }, { "grad_norm": 0.13830086588859558, "learning_rate": 5.983761411824744e-05, "loss": 0.0098, "step": 37210 }, { "grad_norm": 0.11105433106422424, "learning_rate": 5.981734890550844e-05, "loss": 0.0081, "step": 37220 }, { "grad_norm": 0.14346855878829956, "learning_rate": 5.979708201525301e-05, "loss": 0.0102, "step": 37230 }, { "grad_norm": 0.1461424082517624, "learning_rate": 5.977681345094422e-05, "loss": 0.0064, "step": 37240 }, { "grad_norm": 0.10246875882148743, "learning_rate": 5.97565432160454e-05, "loss": 0.0065, "step": 37250 }, { "grad_norm": 0.1527624875307083, "learning_rate": 5.9736271314020186e-05, "loss": 0.0063, "step": 37260 }, { "grad_norm": 0.18998843431472778, "learning_rate": 5.971599774833251e-05, "loss": 0.0068, "step": 37270 }, { "grad_norm": 0.15960228443145752, "learning_rate": 5.9695722522446525e-05, "loss": 0.0083, "step": 37280 }, { "grad_norm": 0.10025704652070999, "learning_rate": 5.9675445639826765e-05, "loss": 0.0055, "step": 37290 }, { "grad_norm": 0.12474166601896286, "learning_rate": 5.965516710393796e-05, "loss": 0.0078, "step": 37300 }, { "grad_norm": 0.09244231134653091, "learning_rate": 5.963488691824516e-05, "loss": 0.0057, "step": 37310 }, { "grad_norm": 0.12694567441940308, "learning_rate": 5.96146050862137e-05, "loss": 0.01, "step": 37320 }, { "grad_norm": 0.14257769286632538, "learning_rate": 5.959432161130919e-05, "loss": 0.009, "step": 37330 }, { "grad_norm": 0.10446295887231827, "learning_rate": 5.9574036496997545e-05, "loss": 0.0077, "step": 37340 }, { "grad_norm": 0.08413319289684296, "learning_rate": 5.955374974674488e-05, "loss": 0.0082, "step": 37350 }, { "grad_norm": 0.097695492208004, "learning_rate": 5.9533461364017696e-05, "loss": 0.0085, "step": 37360 }, { "grad_norm": 0.11254110187292099, "learning_rate": 5.9513171352282716e-05, "loss": 0.0056, "step": 37370 }, { "grad_norm": 0.11639968305826187, "learning_rate": 5.949287971500692e-05, "loss": 0.0071, "step": 37380 }, { "grad_norm": 0.10642924904823303, "learning_rate": 5.947258645565762e-05, "loss": 0.0082, "step": 37390 }, { "grad_norm": 0.15121425688266754, "learning_rate": 5.945229157770237e-05, "loss": 0.0075, "step": 37400 }, { "grad_norm": 0.09543044120073318, "learning_rate": 5.9431995084609006e-05, "loss": 0.0095, "step": 37410 }, { "grad_norm": 0.11038319021463394, "learning_rate": 5.941169697984564e-05, "loss": 0.0058, "step": 37420 }, { "grad_norm": 0.10190442204475403, "learning_rate": 5.9391397266880675e-05, "loss": 0.0056, "step": 37430 }, { "grad_norm": 0.10374820232391357, "learning_rate": 5.937109594918279e-05, "loss": 0.0093, "step": 37440 }, { "grad_norm": 0.1041339710354805, "learning_rate": 5.9350793030220884e-05, "loss": 0.0071, "step": 37450 }, { "grad_norm": 0.12758684158325195, "learning_rate": 5.933048851346421e-05, "loss": 0.008, "step": 37460 }, { "grad_norm": 0.11459345370531082, "learning_rate": 5.931018240238224e-05, "loss": 0.0069, "step": 37470 }, { "grad_norm": 0.11963999271392822, "learning_rate": 5.928987470044471e-05, "loss": 0.007, "step": 37480 }, { "grad_norm": 0.1088697761297226, "learning_rate": 5.9269565411121695e-05, "loss": 0.0078, "step": 37490 }, { "grad_norm": 0.1388007253408432, "learning_rate": 5.924925453788347e-05, "loss": 0.0068, "step": 37500 }, { "grad_norm": 0.1043984442949295, "learning_rate": 5.92289420842006e-05, "loss": 0.0062, "step": 37510 }, { "grad_norm": 0.09875649213790894, "learning_rate": 5.9208628053543945e-05, "loss": 0.0076, "step": 37520 }, { "grad_norm": 0.16160908341407776, "learning_rate": 5.918831244938462e-05, "loss": 0.0069, "step": 37530 }, { "grad_norm": 0.09165581315755844, "learning_rate": 5.9167995275194e-05, "loss": 0.006, "step": 37540 }, { "grad_norm": 0.08913175761699677, "learning_rate": 5.914767653444373e-05, "loss": 0.0069, "step": 37550 }, { "grad_norm": 0.10114456713199615, "learning_rate": 5.912735623060572e-05, "loss": 0.0069, "step": 37560 }, { "grad_norm": 0.1309092491865158, "learning_rate": 5.910703436715217e-05, "loss": 0.0084, "step": 37570 }, { "grad_norm": 0.09261654317378998, "learning_rate": 5.908671094755552e-05, "loss": 0.0063, "step": 37580 }, { "grad_norm": 0.1344870626926422, "learning_rate": 5.906638597528851e-05, "loss": 0.0065, "step": 37590 }, { "grad_norm": 0.10647115111351013, "learning_rate": 5.9046059453824076e-05, "loss": 0.0039, "step": 37600 }, { "grad_norm": 0.10821785777807236, "learning_rate": 5.9025731386635505e-05, "loss": 0.0062, "step": 37610 }, { "grad_norm": 0.0979500412940979, "learning_rate": 5.900540177719629e-05, "loss": 0.0051, "step": 37620 }, { "grad_norm": 0.10999786853790283, "learning_rate": 5.898507062898021e-05, "loss": 0.0053, "step": 37630 }, { "grad_norm": 0.13136456906795502, "learning_rate": 5.8964737945461316e-05, "loss": 0.0055, "step": 37640 }, { "grad_norm": 0.11140567809343338, "learning_rate": 5.8944403730113885e-05, "loss": 0.007, "step": 37650 }, { "grad_norm": 0.12562470138072968, "learning_rate": 5.892406798641248e-05, "loss": 0.0072, "step": 37660 }, { "grad_norm": 0.10622037202119827, "learning_rate": 5.890373071783193e-05, "loss": 0.0092, "step": 37670 }, { "grad_norm": 0.15069590508937836, "learning_rate": 5.888339192784732e-05, "loss": 0.0113, "step": 37680 }, { "grad_norm": 0.07748518139123917, "learning_rate": 5.8863051619934003e-05, "loss": 0.0082, "step": 37690 }, { "grad_norm": 0.0817558765411377, "learning_rate": 5.8842709797567554e-05, "loss": 0.0055, "step": 37700 }, { "grad_norm": 0.11686762422323227, "learning_rate": 5.8822366464223855e-05, "loss": 0.0068, "step": 37710 }, { "grad_norm": 0.09113839268684387, "learning_rate": 5.880202162337901e-05, "loss": 0.0064, "step": 37720 }, { "grad_norm": 0.0813322514295578, "learning_rate": 5.8781675278509405e-05, "loss": 0.006, "step": 37730 }, { "grad_norm": 0.17827843129634857, "learning_rate": 5.8761327433091696e-05, "loss": 0.0075, "step": 37740 }, { "grad_norm": 0.12114131450653076, "learning_rate": 5.874097809060275e-05, "loss": 0.0078, "step": 37750 }, { "grad_norm": 0.10837281495332718, "learning_rate": 5.87206272545197e-05, "loss": 0.0071, "step": 37760 }, { "grad_norm": 0.12312263250350952, "learning_rate": 5.8700274928319955e-05, "loss": 0.0063, "step": 37770 }, { "grad_norm": 0.11895299702882767, "learning_rate": 5.867992111548118e-05, "loss": 0.0084, "step": 37780 }, { "grad_norm": 0.14843788743019104, "learning_rate": 5.865956581948131e-05, "loss": 0.0054, "step": 37790 }, { "grad_norm": 0.10754388570785522, "learning_rate": 5.863920904379845e-05, "loss": 0.0063, "step": 37800 }, { "grad_norm": 0.08253086358308792, "learning_rate": 5.8618850791911064e-05, "loss": 0.0066, "step": 37810 }, { "grad_norm": 0.1180901750922203, "learning_rate": 5.859849106729779e-05, "loss": 0.0069, "step": 37820 }, { "grad_norm": 0.13613222539424896, "learning_rate": 5.857812987343758e-05, "loss": 0.0052, "step": 37830 }, { "grad_norm": 0.11802192777395248, "learning_rate": 5.855776721380957e-05, "loss": 0.0062, "step": 37840 }, { "grad_norm": 0.11991949379444122, "learning_rate": 5.8537403091893217e-05, "loss": 0.0079, "step": 37850 }, { "grad_norm": 0.09225823730230331, "learning_rate": 5.851703751116816e-05, "loss": 0.006, "step": 37860 }, { "grad_norm": 0.08668193966150284, "learning_rate": 5.8496670475114336e-05, "loss": 0.0054, "step": 37870 }, { "grad_norm": 0.07736897468566895, "learning_rate": 5.84763019872119e-05, "loss": 0.006, "step": 37880 }, { "grad_norm": 0.09136413782835007, "learning_rate": 5.845593205094131e-05, "loss": 0.006, "step": 37890 }, { "grad_norm": 0.0896737352013588, "learning_rate": 5.843556066978318e-05, "loss": 0.0055, "step": 37900 }, { "grad_norm": 0.11637437343597412, "learning_rate": 5.8415187847218455e-05, "loss": 0.0055, "step": 37910 }, { "grad_norm": 0.14819595217704773, "learning_rate": 5.839481358672827e-05, "loss": 0.0055, "step": 37920 }, { "grad_norm": 0.09902819246053696, "learning_rate": 5.837443789179407e-05, "loss": 0.0065, "step": 37930 }, { "grad_norm": 0.09327009320259094, "learning_rate": 5.8354060765897445e-05, "loss": 0.0058, "step": 37940 }, { "grad_norm": 0.08238674700260162, "learning_rate": 5.8333682212520334e-05, "loss": 0.0062, "step": 37950 }, { "grad_norm": 0.07113194465637207, "learning_rate": 5.831330223514486e-05, "loss": 0.0072, "step": 37960 }, { "grad_norm": 0.1078394278883934, "learning_rate": 5.8292920837253396e-05, "loss": 0.0085, "step": 37970 }, { "grad_norm": 0.1425457000732422, "learning_rate": 5.827253802232857e-05, "loss": 0.0069, "step": 37980 }, { "grad_norm": 0.07425778359174728, "learning_rate": 5.825215379385327e-05, "loss": 0.0054, "step": 37990 }, { "grad_norm": 0.12527549266815186, "learning_rate": 5.823176815531057e-05, "loss": 0.0087, "step": 38000 }, { "grad_norm": 0.09941098093986511, "learning_rate": 5.8211381110183826e-05, "loss": 0.0073, "step": 38010 }, { "grad_norm": 0.10446766018867493, "learning_rate": 5.8190992661956645e-05, "loss": 0.0062, "step": 38020 }, { "grad_norm": 0.12472948431968689, "learning_rate": 5.817060281411284e-05, "loss": 0.0053, "step": 38030 }, { "grad_norm": 0.09799707680940628, "learning_rate": 5.815021157013647e-05, "loss": 0.0056, "step": 38040 }, { "grad_norm": 0.1108945980668068, "learning_rate": 5.8129818933511856e-05, "loss": 0.0073, "step": 38050 }, { "grad_norm": 0.09127496927976608, "learning_rate": 5.8109424907723544e-05, "loss": 0.0062, "step": 38060 }, { "grad_norm": 0.09062156826257706, "learning_rate": 5.80890294962563e-05, "loss": 0.0051, "step": 38070 }, { "grad_norm": 0.11022497713565826, "learning_rate": 5.806863270259515e-05, "loss": 0.0071, "step": 38080 }, { "grad_norm": 0.10629478842020035, "learning_rate": 5.804823453022536e-05, "loss": 0.0071, "step": 38090 }, { "grad_norm": 0.11261138319969177, "learning_rate": 5.80278349826324e-05, "loss": 0.0087, "step": 38100 }, { "grad_norm": 0.14454250037670135, "learning_rate": 5.8007434063302014e-05, "loss": 0.007, "step": 38110 }, { "grad_norm": 0.12033852189779282, "learning_rate": 5.7987031775720136e-05, "loss": 0.0076, "step": 38120 }, { "grad_norm": 0.10917284339666367, "learning_rate": 5.7966628123372976e-05, "loss": 0.0088, "step": 38130 }, { "grad_norm": 0.11482317000627518, "learning_rate": 5.7946223109746956e-05, "loss": 0.006, "step": 38140 }, { "grad_norm": 0.11391275376081467, "learning_rate": 5.7925816738328754e-05, "loss": 0.0056, "step": 38150 }, { "grad_norm": 0.10274679958820343, "learning_rate": 5.790540901260521e-05, "loss": 0.0074, "step": 38160 }, { "grad_norm": 0.1500677764415741, "learning_rate": 5.788499993606351e-05, "loss": 0.0089, "step": 38170 }, { "grad_norm": 0.08443508297204971, "learning_rate": 5.786458951219096e-05, "loss": 0.0049, "step": 38180 }, { "grad_norm": 0.11733312159776688, "learning_rate": 5.784417774447517e-05, "loss": 0.0067, "step": 38190 }, { "grad_norm": 0.09496518224477768, "learning_rate": 5.782376463640393e-05, "loss": 0.0056, "step": 38200 }, { "grad_norm": 0.10390178114175797, "learning_rate": 5.780335019146531e-05, "loss": 0.0068, "step": 38210 }, { "grad_norm": 0.11491063237190247, "learning_rate": 5.778293441314755e-05, "loss": 0.0059, "step": 38220 }, { "grad_norm": 0.09995701164007187, "learning_rate": 5.776251730493917e-05, "loss": 0.0075, "step": 38230 }, { "grad_norm": 0.07488124817609787, "learning_rate": 5.774209887032887e-05, "loss": 0.0058, "step": 38240 }, { "grad_norm": 0.11265438050031662, "learning_rate": 5.772167911280565e-05, "loss": 0.0075, "step": 38250 }, { "grad_norm": 0.09977946430444717, "learning_rate": 5.770125803585864e-05, "loss": 0.0069, "step": 38260 }, { "grad_norm": 0.12129625678062439, "learning_rate": 5.768083564297726e-05, "loss": 0.0062, "step": 38270 }, { "grad_norm": 0.1300533264875412, "learning_rate": 5.766041193765114e-05, "loss": 0.0075, "step": 38280 }, { "grad_norm": 0.1161583736538887, "learning_rate": 5.763998692337015e-05, "loss": 0.0069, "step": 38290 }, { "grad_norm": 0.09294114261865616, "learning_rate": 5.761956060362433e-05, "loss": 0.0062, "step": 38300 }, { "grad_norm": 0.1369217038154602, "learning_rate": 5.7599132981904e-05, "loss": 0.0072, "step": 38310 }, { "grad_norm": 0.0872621014714241, "learning_rate": 5.75787040616997e-05, "loss": 0.0063, "step": 38320 }, { "grad_norm": 0.1409919261932373, "learning_rate": 5.755827384650212e-05, "loss": 0.0088, "step": 38330 }, { "grad_norm": 0.08437249064445496, "learning_rate": 5.753784233980228e-05, "loss": 0.011, "step": 38340 }, { "grad_norm": 0.13139981031417847, "learning_rate": 5.751740954509135e-05, "loss": 0.0071, "step": 38350 }, { "grad_norm": 0.10344234853982925, "learning_rate": 5.7496975465860715e-05, "loss": 0.0072, "step": 38360 }, { "grad_norm": 0.1654195785522461, "learning_rate": 5.747654010560202e-05, "loss": 0.0079, "step": 38370 }, { "grad_norm": 0.1396128535270691, "learning_rate": 5.7456103467807097e-05, "loss": 0.0077, "step": 38380 }, { "grad_norm": 0.1274474710226059, "learning_rate": 5.7435665555968046e-05, "loss": 0.0064, "step": 38390 }, { "grad_norm": 0.11234787106513977, "learning_rate": 5.74152263735771e-05, "loss": 0.0066, "step": 38400 }, { "grad_norm": 0.13813917338848114, "learning_rate": 5.739478592412677e-05, "loss": 0.0092, "step": 38410 }, { "grad_norm": 0.11939436197280884, "learning_rate": 5.7374344211109766e-05, "loss": 0.0082, "step": 38420 }, { "grad_norm": 0.11448163539171219, "learning_rate": 5.735390123801904e-05, "loss": 0.0089, "step": 38430 }, { "grad_norm": 0.1457098126411438, "learning_rate": 5.7333457008347704e-05, "loss": 0.0056, "step": 38440 }, { "grad_norm": 0.11231798678636551, "learning_rate": 5.7313011525589156e-05, "loss": 0.0065, "step": 38450 }, { "grad_norm": 0.09432128816843033, "learning_rate": 5.729256479323694e-05, "loss": 0.0088, "step": 38460 }, { "grad_norm": 0.09881498664617538, "learning_rate": 5.727211681478485e-05, "loss": 0.008, "step": 38470 }, { "grad_norm": 0.07582104951143265, "learning_rate": 5.7251667593726886e-05, "loss": 0.0074, "step": 38480 }, { "grad_norm": 0.11095913499593735, "learning_rate": 5.723121713355728e-05, "loss": 0.0067, "step": 38490 }, { "grad_norm": 0.09608910232782364, "learning_rate": 5.721076543777044e-05, "loss": 0.0086, "step": 38500 }, { "grad_norm": 0.12355665862560272, "learning_rate": 5.7190312509860986e-05, "loss": 0.0083, "step": 38510 }, { "grad_norm": 0.13749611377716064, "learning_rate": 5.716985835332379e-05, "loss": 0.0058, "step": 38520 }, { "grad_norm": 0.11351760476827621, "learning_rate": 5.714940297165389e-05, "loss": 0.0066, "step": 38530 }, { "grad_norm": 0.1104939803481102, "learning_rate": 5.712894636834656e-05, "loss": 0.0066, "step": 38540 }, { "grad_norm": 0.1041731983423233, "learning_rate": 5.7108488546897276e-05, "loss": 0.0095, "step": 38550 }, { "grad_norm": 0.10754118114709854, "learning_rate": 5.708802951080172e-05, "loss": 0.0056, "step": 38560 }, { "grad_norm": 0.10221482813358307, "learning_rate": 5.706756926355576e-05, "loss": 0.0069, "step": 38570 }, { "grad_norm": 0.10459697991609573, "learning_rate": 5.704710780865554e-05, "loss": 0.007, "step": 38580 }, { "grad_norm": 0.12362749129533768, "learning_rate": 5.7026645149597325e-05, "loss": 0.0065, "step": 38590 }, { "grad_norm": 0.10394193232059479, "learning_rate": 5.700618128987764e-05, "loss": 0.0076, "step": 38600 }, { "grad_norm": 0.09468533098697662, "learning_rate": 5.698571623299317e-05, "loss": 0.0065, "step": 38610 }, { "grad_norm": 0.10546896606683731, "learning_rate": 5.696524998244086e-05, "loss": 0.0091, "step": 38620 }, { "grad_norm": 0.08174819499254227, "learning_rate": 5.6944782541717836e-05, "loss": 0.0058, "step": 38630 }, { "grad_norm": 0.10327578336000443, "learning_rate": 5.69243139143214e-05, "loss": 0.007, "step": 38640 }, { "grad_norm": 0.11562992632389069, "learning_rate": 5.6903844103749125e-05, "loss": 0.0063, "step": 38650 }, { "grad_norm": 0.12161682546138763, "learning_rate": 5.688337311349869e-05, "loss": 0.0071, "step": 38660 }, { "grad_norm": 0.13675463199615479, "learning_rate": 5.6862900947068074e-05, "loss": 0.0059, "step": 38670 }, { "grad_norm": 0.10623250901699066, "learning_rate": 5.6842427607955374e-05, "loss": 0.007, "step": 38680 }, { "grad_norm": 0.11531991511583328, "learning_rate": 5.682195309965893e-05, "loss": 0.007, "step": 38690 }, { "grad_norm": 0.09999381750822067, "learning_rate": 5.6801477425677294e-05, "loss": 0.0061, "step": 38700 }, { "grad_norm": 0.09957104921340942, "learning_rate": 5.678100058950917e-05, "loss": 0.0071, "step": 38710 }, { "grad_norm": 0.1588425487279892, "learning_rate": 5.676052259465352e-05, "loss": 0.007, "step": 38720 }, { "grad_norm": 0.15166552364826202, "learning_rate": 5.674004344460945e-05, "loss": 0.0086, "step": 38730 }, { "grad_norm": 0.10567908734083176, "learning_rate": 5.6719563142876295e-05, "loss": 0.0076, "step": 38740 }, { "grad_norm": 0.1018625795841217, "learning_rate": 5.669908169295359e-05, "loss": 0.0085, "step": 38750 }, { "grad_norm": 0.10325255990028381, "learning_rate": 5.667859909834105e-05, "loss": 0.006, "step": 38760 }, { "grad_norm": 0.12785302102565765, "learning_rate": 5.6658115362538565e-05, "loss": 0.0083, "step": 38770 }, { "grad_norm": 0.0968879982829094, "learning_rate": 5.663763048904628e-05, "loss": 0.0059, "step": 38780 }, { "grad_norm": 0.0970780998468399, "learning_rate": 5.661714448136447e-05, "loss": 0.0066, "step": 38790 }, { "grad_norm": 0.10291755944490433, "learning_rate": 5.659665734299366e-05, "loss": 0.0074, "step": 38800 }, { "grad_norm": 0.11172687262296677, "learning_rate": 5.6576169077434516e-05, "loss": 0.008, "step": 38810 }, { "grad_norm": 0.11918318271636963, "learning_rate": 5.6555679688187944e-05, "loss": 0.0075, "step": 38820 }, { "grad_norm": 0.1302880346775055, "learning_rate": 5.6535189178755e-05, "loss": 0.0074, "step": 38830 }, { "grad_norm": 0.1124061718583107, "learning_rate": 5.651469755263698e-05, "loss": 0.0073, "step": 38840 }, { "grad_norm": 0.1346454620361328, "learning_rate": 5.6494204813335316e-05, "loss": 0.0063, "step": 38850 }, { "grad_norm": 0.10062511265277863, "learning_rate": 5.647371096435168e-05, "loss": 0.0097, "step": 38860 }, { "grad_norm": 0.16661478579044342, "learning_rate": 5.645321600918788e-05, "loss": 0.0072, "step": 38870 }, { "grad_norm": 0.12427490949630737, "learning_rate": 5.643271995134597e-05, "loss": 0.0068, "step": 38880 }, { "grad_norm": 0.10905709117650986, "learning_rate": 5.641222279432814e-05, "loss": 0.0065, "step": 38890 }, { "grad_norm": 0.11621692776679993, "learning_rate": 5.6391724541636834e-05, "loss": 0.0079, "step": 38900 }, { "grad_norm": 0.085780069231987, "learning_rate": 5.6371225196774605e-05, "loss": 0.007, "step": 38910 }, { "grad_norm": 0.11921681463718414, "learning_rate": 5.635072476324423e-05, "loss": 0.006, "step": 38920 }, { "grad_norm": 0.13987277448177338, "learning_rate": 5.63302232445487e-05, "loss": 0.0046, "step": 38930 }, { "grad_norm": 0.07860442996025085, "learning_rate": 5.6309720644191144e-05, "loss": 0.0048, "step": 38940 }, { "grad_norm": 0.13310277462005615, "learning_rate": 5.628921696567491e-05, "loss": 0.0066, "step": 38950 }, { "grad_norm": 0.11199060082435608, "learning_rate": 5.62687122125035e-05, "loss": 0.0056, "step": 38960 }, { "grad_norm": 0.10109202563762665, "learning_rate": 5.624820638818062e-05, "loss": 0.0061, "step": 38970 }, { "grad_norm": 0.10781466960906982, "learning_rate": 5.6227699496210164e-05, "loss": 0.0064, "step": 38980 }, { "grad_norm": 0.11517178267240524, "learning_rate": 5.6207191540096195e-05, "loss": 0.008, "step": 38990 }, { "grad_norm": 0.11531689763069153, "learning_rate": 5.618668252334296e-05, "loss": 0.0063, "step": 39000 }, { "grad_norm": 0.15210650861263275, "learning_rate": 5.616617244945488e-05, "loss": 0.0074, "step": 39010 }, { "grad_norm": 0.1296418458223343, "learning_rate": 5.614566132193656e-05, "loss": 0.0058, "step": 39020 }, { "grad_norm": 0.0718693658709526, "learning_rate": 5.612514914429282e-05, "loss": 0.007, "step": 39030 }, { "grad_norm": 0.13471238315105438, "learning_rate": 5.610463592002863e-05, "loss": 0.008, "step": 39040 }, { "grad_norm": 0.09149951487779617, "learning_rate": 5.608412165264909e-05, "loss": 0.0051, "step": 39050 }, { "grad_norm": 0.11131764203310013, "learning_rate": 5.606360634565959e-05, "loss": 0.0076, "step": 39060 }, { "grad_norm": 0.1115449070930481, "learning_rate": 5.604309000256559e-05, "loss": 0.008, "step": 39070 }, { "grad_norm": 0.1254732757806778, "learning_rate": 5.6022572626872785e-05, "loss": 0.0051, "step": 39080 }, { "grad_norm": 0.11182893812656403, "learning_rate": 5.600205422208704e-05, "loss": 0.0074, "step": 39090 }, { "grad_norm": 0.08035213500261307, "learning_rate": 5.59815347917144e-05, "loss": 0.0081, "step": 39100 }, { "grad_norm": 0.11053906381130219, "learning_rate": 5.596101433926103e-05, "loss": 0.0059, "step": 39110 }, { "grad_norm": 0.15005740523338318, "learning_rate": 5.5940492868233364e-05, "loss": 0.0101, "step": 39120 }, { "grad_norm": 0.11994971334934235, "learning_rate": 5.591997038213793e-05, "loss": 0.01, "step": 39130 }, { "grad_norm": 0.13621996343135834, "learning_rate": 5.5899446884481475e-05, "loss": 0.0072, "step": 39140 }, { "grad_norm": 0.10185687243938446, "learning_rate": 5.5878922378770906e-05, "loss": 0.0058, "step": 39150 }, { "grad_norm": 0.14019569754600525, "learning_rate": 5.5858396868513285e-05, "loss": 0.0062, "step": 39160 }, { "grad_norm": 0.21256406605243683, "learning_rate": 5.583787035721586e-05, "loss": 0.006, "step": 39170 }, { "grad_norm": 0.08261097967624664, "learning_rate": 5.581734284838606e-05, "loss": 0.0057, "step": 39180 }, { "grad_norm": 0.08723831921815872, "learning_rate": 5.579681434553147e-05, "loss": 0.0071, "step": 39190 }, { "grad_norm": 0.12573936581611633, "learning_rate": 5.5776284852159854e-05, "loss": 0.0063, "step": 39200 }, { "grad_norm": 0.13174165785312653, "learning_rate": 5.575575437177913e-05, "loss": 0.0074, "step": 39210 }, { "grad_norm": 0.10099100321531296, "learning_rate": 5.573522290789742e-05, "loss": 0.0073, "step": 39220 }, { "grad_norm": 0.13425971567630768, "learning_rate": 5.571469046402298e-05, "loss": 0.0069, "step": 39230 }, { "grad_norm": 0.18908163905143738, "learning_rate": 5.5694157043664205e-05, "loss": 0.0096, "step": 39240 }, { "grad_norm": 0.09743241220712662, "learning_rate": 5.567362265032975e-05, "loss": 0.0043, "step": 39250 }, { "grad_norm": 0.11856742203235626, "learning_rate": 5.565308728752836e-05, "loss": 0.0051, "step": 39260 }, { "grad_norm": 0.11005856841802597, "learning_rate": 5.5632550958768956e-05, "loss": 0.0054, "step": 39270 }, { "grad_norm": 0.11362922191619873, "learning_rate": 5.5612013667560636e-05, "loss": 0.0084, "step": 39280 }, { "grad_norm": 0.09934241324663162, "learning_rate": 5.5591475417412676e-05, "loss": 0.0087, "step": 39290 }, { "grad_norm": 0.09928170591592789, "learning_rate": 5.557093621183451e-05, "loss": 0.0062, "step": 39300 }, { "grad_norm": 0.10472323000431061, "learning_rate": 5.55503960543357e-05, "loss": 0.0066, "step": 39310 }, { "grad_norm": 0.13911741971969604, "learning_rate": 5.552985494842601e-05, "loss": 0.0079, "step": 39320 }, { "grad_norm": 0.09792725741863251, "learning_rate": 5.550931289761534e-05, "loss": 0.0071, "step": 39330 }, { "grad_norm": 0.0846845954656601, "learning_rate": 5.548876990541378e-05, "loss": 0.0049, "step": 39340 }, { "grad_norm": 0.11377035081386566, "learning_rate": 5.5468225975331564e-05, "loss": 0.0085, "step": 39350 }, { "grad_norm": 0.0912972018122673, "learning_rate": 5.544768111087909e-05, "loss": 0.0062, "step": 39360 }, { "grad_norm": 0.12636852264404297, "learning_rate": 5.54271353155669e-05, "loss": 0.0057, "step": 39370 }, { "grad_norm": 0.08076784759759903, "learning_rate": 5.5406588592905715e-05, "loss": 0.006, "step": 39380 }, { "grad_norm": 0.09776671230792999, "learning_rate": 5.5386040946406416e-05, "loss": 0.0062, "step": 39390 }, { "grad_norm": 0.12827849388122559, "learning_rate": 5.536549237958004e-05, "loss": 0.005, "step": 39400 }, { "grad_norm": 0.124713234603405, "learning_rate": 5.5344942895937744e-05, "loss": 0.0069, "step": 39410 }, { "grad_norm": 0.08769248425960541, "learning_rate": 5.5324392498990904e-05, "loss": 0.0064, "step": 39420 }, { "grad_norm": 0.1305142492055893, "learning_rate": 5.5303841192251e-05, "loss": 0.0078, "step": 39430 }, { "grad_norm": 0.11295658349990845, "learning_rate": 5.52832889792297e-05, "loss": 0.0096, "step": 39440 }, { "grad_norm": 0.08736222982406616, "learning_rate": 5.526273586343881e-05, "loss": 0.007, "step": 39450 }, { "grad_norm": 0.1153746247291565, "learning_rate": 5.5242181848390306e-05, "loss": 0.0066, "step": 39460 }, { "grad_norm": 0.1200363039970398, "learning_rate": 5.5221626937596285e-05, "loss": 0.0053, "step": 39470 }, { "grad_norm": 0.10611401498317719, "learning_rate": 5.520107113456903e-05, "loss": 0.0055, "step": 39480 }, { "grad_norm": 0.11514423042535782, "learning_rate": 5.5180514442820974e-05, "loss": 0.0066, "step": 39490 }, { "grad_norm": 0.08762456476688385, "learning_rate": 5.515995686586469e-05, "loss": 0.0051, "step": 39500 }, { "grad_norm": 0.13562345504760742, "learning_rate": 5.5139398407212916e-05, "loss": 0.0104, "step": 39510 }, { "grad_norm": 0.10530165582895279, "learning_rate": 5.511883907037849e-05, "loss": 0.0082, "step": 39520 }, { "grad_norm": 0.08218970149755478, "learning_rate": 5.509827885887449e-05, "loss": 0.0043, "step": 39530 }, { "grad_norm": 0.13021816313266754, "learning_rate": 5.507771777621406e-05, "loss": 0.0065, "step": 39540 }, { "grad_norm": 0.09458205103874207, "learning_rate": 5.505715582591052e-05, "loss": 0.0054, "step": 39550 }, { "grad_norm": 0.07359318435192108, "learning_rate": 5.50365930114774e-05, "loss": 0.0054, "step": 39560 }, { "grad_norm": 0.10433749109506607, "learning_rate": 5.5016029336428255e-05, "loss": 0.0055, "step": 39570 }, { "grad_norm": 0.10525969415903091, "learning_rate": 5.49954648042769e-05, "loss": 0.0065, "step": 39580 }, { "grad_norm": 0.14329375326633453, "learning_rate": 5.4974899418537226e-05, "loss": 0.0071, "step": 39590 }, { "grad_norm": 0.13891790807247162, "learning_rate": 5.4954333182723316e-05, "loss": 0.0059, "step": 39600 }, { "grad_norm": 0.10001451522111893, "learning_rate": 5.493376610034937e-05, "loss": 0.0071, "step": 39610 }, { "grad_norm": 0.12434898316860199, "learning_rate": 5.4913198174929735e-05, "loss": 0.0047, "step": 39620 }, { "grad_norm": 0.0996004194021225, "learning_rate": 5.489262940997889e-05, "loss": 0.0056, "step": 39630 }, { "grad_norm": 0.09201151132583618, "learning_rate": 5.487205980901151e-05, "loss": 0.005, "step": 39640 }, { "grad_norm": 0.07651291787624359, "learning_rate": 5.485148937554234e-05, "loss": 0.0043, "step": 39650 }, { "grad_norm": 0.11739496141672134, "learning_rate": 5.483091811308635e-05, "loss": 0.0061, "step": 39660 }, { "grad_norm": 0.09924117475748062, "learning_rate": 5.4810346025158564e-05, "loss": 0.0063, "step": 39670 }, { "grad_norm": 0.10515230894088745, "learning_rate": 5.478977311527421e-05, "loss": 0.0058, "step": 39680 }, { "grad_norm": 0.08716732263565063, "learning_rate": 5.476919938694863e-05, "loss": 0.0058, "step": 39690 }, { "grad_norm": 0.12795117497444153, "learning_rate": 5.474862484369733e-05, "loss": 0.0072, "step": 39700 }, { "grad_norm": 0.11730898171663284, "learning_rate": 5.472804948903589e-05, "loss": 0.0073, "step": 39710 }, { "grad_norm": 0.1079150140285492, "learning_rate": 5.470747332648013e-05, "loss": 0.0058, "step": 39720 }, { "grad_norm": 0.11557819694280624, "learning_rate": 5.468689635954591e-05, "loss": 0.0073, "step": 39730 }, { "grad_norm": 0.09767027944326401, "learning_rate": 5.46663185917493e-05, "loss": 0.0094, "step": 39740 }, { "grad_norm": 0.12910759449005127, "learning_rate": 5.464574002660645e-05, "loss": 0.0051, "step": 39750 }, { "grad_norm": 0.12392237782478333, "learning_rate": 5.4625160667633724e-05, "loss": 0.0066, "step": 39760 }, { "grad_norm": 0.13745050132274628, "learning_rate": 5.4604580518347505e-05, "loss": 0.0054, "step": 39770 }, { "grad_norm": 0.10494199395179749, "learning_rate": 5.458399958226442e-05, "loss": 0.0058, "step": 39780 }, { "grad_norm": 0.07834099978208542, "learning_rate": 5.456341786290119e-05, "loss": 0.0075, "step": 39790 }, { "grad_norm": 0.0931984931230545, "learning_rate": 5.454283536377465e-05, "loss": 0.0067, "step": 39800 }, { "grad_norm": 0.08583234250545502, "learning_rate": 5.452225208840179e-05, "loss": 0.0066, "step": 39810 }, { "grad_norm": 0.09628414362668991, "learning_rate": 5.450166804029975e-05, "loss": 0.0078, "step": 39820 }, { "grad_norm": 0.09191329777240753, "learning_rate": 5.448108322298574e-05, "loss": 0.0063, "step": 39830 }, { "grad_norm": 0.1586533933877945, "learning_rate": 5.446049763997717e-05, "loss": 0.0073, "step": 39840 }, { "grad_norm": 0.16377747058868408, "learning_rate": 5.4439911294791546e-05, "loss": 0.0081, "step": 39850 }, { "grad_norm": 0.10038342326879501, "learning_rate": 5.441932419094652e-05, "loss": 0.0067, "step": 39860 }, { "grad_norm": 0.1005031168460846, "learning_rate": 5.439873633195985e-05, "loss": 0.0055, "step": 39870 }, { "grad_norm": 0.07757645100355148, "learning_rate": 5.437814772134947e-05, "loss": 0.0055, "step": 39880 }, { "grad_norm": 0.15098202228546143, "learning_rate": 5.4357558362633366e-05, "loss": 0.0088, "step": 39890 }, { "grad_norm": 0.115908682346344, "learning_rate": 5.4336968259329726e-05, "loss": 0.0079, "step": 39900 }, { "grad_norm": 0.12295825779438019, "learning_rate": 5.431637741495681e-05, "loss": 0.0064, "step": 39910 }, { "grad_norm": 0.09471888840198517, "learning_rate": 5.429578583303307e-05, "loss": 0.0072, "step": 39920 }, { "grad_norm": 0.10239949822425842, "learning_rate": 5.427519351707701e-05, "loss": 0.0082, "step": 39930 }, { "grad_norm": 0.08159995824098587, "learning_rate": 5.42546004706073e-05, "loss": 0.0053, "step": 39940 }, { "grad_norm": 0.06713524460792542, "learning_rate": 5.4234006697142735e-05, "loss": 0.0049, "step": 39950 }, { "grad_norm": 0.14097189903259277, "learning_rate": 5.421341220020224e-05, "loss": 0.0074, "step": 39960 }, { "grad_norm": 0.09919918328523636, "learning_rate": 5.419281698330482e-05, "loss": 0.0056, "step": 39970 }, { "grad_norm": 0.0815005674958229, "learning_rate": 5.4172221049969665e-05, "loss": 0.007, "step": 39980 }, { "grad_norm": 0.10560032725334167, "learning_rate": 5.415162440371604e-05, "loss": 0.0056, "step": 39990 }, { "grad_norm": 0.10871515423059464, "learning_rate": 5.413102704806334e-05, "loss": 0.0051, "step": 40000 } ], "logging_steps": 10, "max_steps": 80000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 10000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 48, "trial_name": null, "trial_params": null }