{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 100000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.4879143238067627, "learning_rate": 1.8e-07, "loss": 1.5418, "step": 10 }, { "grad_norm": 1.2125145196914673, "learning_rate": 3.8e-07, "loss": 1.5586, "step": 20 }, { "grad_norm": 1.0093275308609009, "learning_rate": 5.8e-07, "loss": 1.5233, "step": 30 }, { "grad_norm": 1.284960150718689, "learning_rate": 7.8e-07, "loss": 1.5426, "step": 40 }, { "grad_norm": 1.1905227899551392, "learning_rate": 9.8e-07, "loss": 1.5158, "step": 50 }, { "grad_norm": 1.1193840503692627, "learning_rate": 1.18e-06, "loss": 1.4948, "step": 60 }, { "grad_norm": 1.447500467300415, "learning_rate": 1.3800000000000001e-06, "loss": 1.4409, "step": 70 }, { "grad_norm": 2.457939863204956, "learning_rate": 1.5800000000000003e-06, "loss": 1.3633, "step": 80 }, { "grad_norm": 1.7681373357772827, "learning_rate": 1.7800000000000001e-06, "loss": 1.2764, "step": 90 }, { "grad_norm": 2.5412757396698, "learning_rate": 1.98e-06, "loss": 1.2315, "step": 100 }, { "grad_norm": 1.7988629341125488, "learning_rate": 2.1800000000000003e-06, "loss": 1.1955, "step": 110 }, { "grad_norm": 2.1861133575439453, "learning_rate": 2.38e-06, "loss": 1.2332, "step": 120 }, { "grad_norm": 2.0704996585845947, "learning_rate": 2.5800000000000003e-06, "loss": 1.2089, "step": 130 }, { "grad_norm": 3.294076681137085, "learning_rate": 2.78e-06, "loss": 1.224, "step": 140 }, { "grad_norm": 2.22719669342041, "learning_rate": 2.9800000000000003e-06, "loss": 1.2241, "step": 150 }, { "grad_norm": 1.902626872062683, "learning_rate": 3.1800000000000005e-06, "loss": 1.204, "step": 160 }, { "grad_norm": 2.0503785610198975, "learning_rate": 3.38e-06, "loss": 1.1658, "step": 170 }, { "grad_norm": 2.0173897743225098, "learning_rate": 3.58e-06, "loss": 1.1573, "step": 180 }, { "grad_norm": 2.4189536571502686, "learning_rate": 3.7800000000000002e-06, "loss": 1.1805, "step": 190 }, { "grad_norm": 1.678001046180725, "learning_rate": 3.98e-06, "loss": 1.1674, "step": 200 }, { "grad_norm": 2.0618093013763428, "learning_rate": 4.18e-06, "loss": 1.1681, "step": 210 }, { "grad_norm": 2.0092992782592773, "learning_rate": 4.38e-06, "loss": 1.1373, "step": 220 }, { "grad_norm": 1.525163173675537, "learning_rate": 4.58e-06, "loss": 1.1356, "step": 230 }, { "grad_norm": 1.9010322093963623, "learning_rate": 4.780000000000001e-06, "loss": 1.1214, "step": 240 }, { "grad_norm": 1.8651554584503174, "learning_rate": 4.98e-06, "loss": 1.124, "step": 250 }, { "grad_norm": 2.126697540283203, "learning_rate": 5.18e-06, "loss": 1.1167, "step": 260 }, { "grad_norm": 2.140212297439575, "learning_rate": 5.38e-06, "loss": 1.1172, "step": 270 }, { "grad_norm": 1.9060050249099731, "learning_rate": 5.580000000000001e-06, "loss": 1.1028, "step": 280 }, { "grad_norm": 2.1105449199676514, "learning_rate": 5.78e-06, "loss": 1.1007, "step": 290 }, { "grad_norm": 2.0385308265686035, "learning_rate": 5.98e-06, "loss": 1.0874, "step": 300 }, { "grad_norm": 1.8350483179092407, "learning_rate": 6.18e-06, "loss": 1.1036, "step": 310 }, { "grad_norm": 1.5672540664672852, "learning_rate": 6.38e-06, "loss": 1.0881, "step": 320 }, { "grad_norm": 1.9767330884933472, "learning_rate": 6.58e-06, "loss": 1.0679, "step": 330 }, { "grad_norm": 1.5258127450942993, "learning_rate": 6.78e-06, "loss": 1.0961, "step": 340 }, { "grad_norm": 2.3253910541534424, "learning_rate": 6.98e-06, "loss": 1.0895, "step": 350 }, { "grad_norm": 1.78799569606781, "learning_rate": 7.180000000000001e-06, "loss": 1.0761, "step": 360 }, { "grad_norm": 1.7636773586273193, "learning_rate": 7.3800000000000005e-06, "loss": 1.0558, "step": 370 }, { "grad_norm": 1.6119842529296875, "learning_rate": 7.580000000000001e-06, "loss": 1.1031, "step": 380 }, { "grad_norm": 1.8201050758361816, "learning_rate": 7.78e-06, "loss": 1.0671, "step": 390 }, { "grad_norm": 1.9472323656082153, "learning_rate": 7.98e-06, "loss": 1.0674, "step": 400 }, { "grad_norm": 1.5813192129135132, "learning_rate": 8.18e-06, "loss": 1.0969, "step": 410 }, { "grad_norm": 1.9025641679763794, "learning_rate": 8.380000000000001e-06, "loss": 1.0796, "step": 420 }, { "grad_norm": 1.8873175382614136, "learning_rate": 8.580000000000001e-06, "loss": 1.04, "step": 430 }, { "grad_norm": 2.2516438961029053, "learning_rate": 8.78e-06, "loss": 1.0718, "step": 440 }, { "grad_norm": 1.86641263961792, "learning_rate": 8.98e-06, "loss": 1.0889, "step": 450 }, { "grad_norm": 1.5858174562454224, "learning_rate": 9.180000000000002e-06, "loss": 1.0743, "step": 460 }, { "grad_norm": 2.042987108230591, "learning_rate": 9.38e-06, "loss": 1.0621, "step": 470 }, { "grad_norm": 1.7679263353347778, "learning_rate": 9.58e-06, "loss": 1.0593, "step": 480 }, { "grad_norm": 1.852079153060913, "learning_rate": 9.78e-06, "loss": 1.0866, "step": 490 }, { "grad_norm": 2.238293170928955, "learning_rate": 9.980000000000001e-06, "loss": 1.0253, "step": 500 }, { "grad_norm": 2.0882930755615234, "learning_rate": 1.018e-05, "loss": 1.0709, "step": 510 }, { "grad_norm": 1.7011305093765259, "learning_rate": 1.038e-05, "loss": 1.0685, "step": 520 }, { "grad_norm": 2.039494752883911, "learning_rate": 1.058e-05, "loss": 1.0497, "step": 530 }, { "grad_norm": 2.545200824737549, "learning_rate": 1.0780000000000002e-05, "loss": 1.0386, "step": 540 }, { "grad_norm": 2.0924041271209717, "learning_rate": 1.098e-05, "loss": 1.0131, "step": 550 }, { "grad_norm": 2.545370101928711, "learning_rate": 1.118e-05, "loss": 1.0005, "step": 560 }, { "grad_norm": 4.314510822296143, "learning_rate": 1.1380000000000001e-05, "loss": 0.9734, "step": 570 }, { "grad_norm": 2.3251984119415283, "learning_rate": 1.1580000000000001e-05, "loss": 0.9857, "step": 580 }, { "grad_norm": 3.376749277114868, "learning_rate": 1.178e-05, "loss": 0.9793, "step": 590 }, { "grad_norm": 2.4637064933776855, "learning_rate": 1.198e-05, "loss": 0.9514, "step": 600 }, { "grad_norm": 2.182204484939575, "learning_rate": 1.2180000000000002e-05, "loss": 0.9548, "step": 610 }, { "grad_norm": 2.292041540145874, "learning_rate": 1.238e-05, "loss": 0.904, "step": 620 }, { "grad_norm": 2.302976369857788, "learning_rate": 1.258e-05, "loss": 0.8932, "step": 630 }, { "grad_norm": 2.511971950531006, "learning_rate": 1.278e-05, "loss": 0.8611, "step": 640 }, { "grad_norm": 2.3950374126434326, "learning_rate": 1.2980000000000001e-05, "loss": 0.8397, "step": 650 }, { "grad_norm": 3.4905598163604736, "learning_rate": 1.3180000000000001e-05, "loss": 0.8491, "step": 660 }, { "grad_norm": 2.8224117755889893, "learning_rate": 1.338e-05, "loss": 0.8094, "step": 670 }, { "grad_norm": 2.863330602645874, "learning_rate": 1.358e-05, "loss": 0.8017, "step": 680 }, { "grad_norm": 2.76591420173645, "learning_rate": 1.3780000000000002e-05, "loss": 0.776, "step": 690 }, { "grad_norm": 2.8230764865875244, "learning_rate": 1.3980000000000002e-05, "loss": 0.7843, "step": 700 }, { "grad_norm": 2.787898540496826, "learning_rate": 1.4180000000000001e-05, "loss": 0.764, "step": 710 }, { "grad_norm": 3.0977225303649902, "learning_rate": 1.4380000000000001e-05, "loss": 0.7088, "step": 720 }, { "grad_norm": 2.7768406867980957, "learning_rate": 1.4580000000000003e-05, "loss": 0.7199, "step": 730 }, { "grad_norm": 2.5255544185638428, "learning_rate": 1.4779999999999999e-05, "loss": 0.6972, "step": 740 }, { "grad_norm": 2.2280383110046387, "learning_rate": 1.4979999999999999e-05, "loss": 0.6733, "step": 750 }, { "grad_norm": 3.1569182872772217, "learning_rate": 1.518e-05, "loss": 0.6574, "step": 760 }, { "grad_norm": 3.159153938293457, "learning_rate": 1.538e-05, "loss": 0.6707, "step": 770 }, { "grad_norm": 3.125946521759033, "learning_rate": 1.558e-05, "loss": 0.6425, "step": 780 }, { "grad_norm": 3.243069887161255, "learning_rate": 1.578e-05, "loss": 0.5893, "step": 790 }, { "grad_norm": 3.5320208072662354, "learning_rate": 1.598e-05, "loss": 0.6631, "step": 800 }, { "grad_norm": 2.9484801292419434, "learning_rate": 1.618e-05, "loss": 0.5875, "step": 810 }, { "grad_norm": 3.4420275688171387, "learning_rate": 1.6380000000000002e-05, "loss": 0.5752, "step": 820 }, { "grad_norm": 4.242731094360352, "learning_rate": 1.658e-05, "loss": 0.5452, "step": 830 }, { "grad_norm": 3.2395384311676025, "learning_rate": 1.6780000000000002e-05, "loss": 0.5666, "step": 840 }, { "grad_norm": 3.4473936557769775, "learning_rate": 1.698e-05, "loss": 0.5904, "step": 850 }, { "grad_norm": 3.7186014652252197, "learning_rate": 1.718e-05, "loss": 0.5516, "step": 860 }, { "grad_norm": 3.049288272857666, "learning_rate": 1.7380000000000003e-05, "loss": 0.5455, "step": 870 }, { "grad_norm": 3.2210564613342285, "learning_rate": 1.758e-05, "loss": 0.5182, "step": 880 }, { "grad_norm": 2.797494888305664, "learning_rate": 1.7780000000000003e-05, "loss": 0.488, "step": 890 }, { "grad_norm": 3.2141640186309814, "learning_rate": 1.798e-05, "loss": 0.4596, "step": 900 }, { "grad_norm": 2.865952253341675, "learning_rate": 1.818e-05, "loss": 0.4993, "step": 910 }, { "grad_norm": 2.9059853553771973, "learning_rate": 1.838e-05, "loss": 0.4888, "step": 920 }, { "grad_norm": 2.5972139835357666, "learning_rate": 1.858e-05, "loss": 0.4684, "step": 930 }, { "grad_norm": 4.6604204177856445, "learning_rate": 1.878e-05, "loss": 0.484, "step": 940 }, { "grad_norm": 3.620666027069092, "learning_rate": 1.898e-05, "loss": 0.4797, "step": 950 }, { "grad_norm": 2.745401620864868, "learning_rate": 1.918e-05, "loss": 0.4403, "step": 960 }, { "grad_norm": 3.019735813140869, "learning_rate": 1.938e-05, "loss": 0.4831, "step": 970 }, { "grad_norm": 2.894753932952881, "learning_rate": 1.9580000000000002e-05, "loss": 0.4338, "step": 980 }, { "grad_norm": 3.038259267807007, "learning_rate": 1.978e-05, "loss": 0.4226, "step": 990 }, { "grad_norm": 4.793507099151611, "learning_rate": 1.9980000000000002e-05, "loss": 0.43, "step": 1000 }, { "grad_norm": 3.269639492034912, "learning_rate": 2.0180000000000003e-05, "loss": 0.44, "step": 1010 }, { "grad_norm": 3.100569725036621, "learning_rate": 2.038e-05, "loss": 0.4509, "step": 1020 }, { "grad_norm": 3.619241714477539, "learning_rate": 2.0580000000000003e-05, "loss": 0.413, "step": 1030 }, { "grad_norm": 2.439338207244873, "learning_rate": 2.078e-05, "loss": 0.4081, "step": 1040 }, { "grad_norm": 3.4699034690856934, "learning_rate": 2.098e-05, "loss": 0.4379, "step": 1050 }, { "grad_norm": 3.6267480850219727, "learning_rate": 2.118e-05, "loss": 0.3894, "step": 1060 }, { "grad_norm": 3.3698618412017822, "learning_rate": 2.138e-05, "loss": 0.4087, "step": 1070 }, { "grad_norm": 2.1718332767486572, "learning_rate": 2.158e-05, "loss": 0.3094, "step": 1080 }, { "grad_norm": 3.428133964538574, "learning_rate": 2.178e-05, "loss": 0.4299, "step": 1090 }, { "grad_norm": 2.4986114501953125, "learning_rate": 2.198e-05, "loss": 0.3664, "step": 1100 }, { "grad_norm": 2.257115364074707, "learning_rate": 2.218e-05, "loss": 0.3622, "step": 1110 }, { "grad_norm": 3.091705560684204, "learning_rate": 2.2380000000000003e-05, "loss": 0.3447, "step": 1120 }, { "grad_norm": 2.618730306625366, "learning_rate": 2.258e-05, "loss": 0.3314, "step": 1130 }, { "grad_norm": 2.8509411811828613, "learning_rate": 2.2780000000000002e-05, "loss": 0.3068, "step": 1140 }, { "grad_norm": 2.72190260887146, "learning_rate": 2.298e-05, "loss": 0.3055, "step": 1150 }, { "grad_norm": 3.3422398567199707, "learning_rate": 2.318e-05, "loss": 0.3305, "step": 1160 }, { "grad_norm": 3.426090717315674, "learning_rate": 2.3380000000000003e-05, "loss": 0.3, "step": 1170 }, { "grad_norm": 3.3464083671569824, "learning_rate": 2.358e-05, "loss": 0.327, "step": 1180 }, { "grad_norm": 2.6660149097442627, "learning_rate": 2.3780000000000003e-05, "loss": 0.3054, "step": 1190 }, { "grad_norm": 2.2019383907318115, "learning_rate": 2.398e-05, "loss": 0.3212, "step": 1200 }, { "grad_norm": 2.24572491645813, "learning_rate": 2.418e-05, "loss": 0.3223, "step": 1210 }, { "grad_norm": 2.8494985103607178, "learning_rate": 2.438e-05, "loss": 0.2969, "step": 1220 }, { "grad_norm": 3.10492205619812, "learning_rate": 2.4580000000000002e-05, "loss": 0.3112, "step": 1230 }, { "grad_norm": 2.9594335556030273, "learning_rate": 2.478e-05, "loss": 0.2894, "step": 1240 }, { "grad_norm": 2.758659839630127, "learning_rate": 2.498e-05, "loss": 0.2981, "step": 1250 }, { "grad_norm": 2.8611068725585938, "learning_rate": 2.5180000000000003e-05, "loss": 0.2907, "step": 1260 }, { "grad_norm": 2.831101894378662, "learning_rate": 2.5380000000000004e-05, "loss": 0.2725, "step": 1270 }, { "grad_norm": 2.917311191558838, "learning_rate": 2.5580000000000002e-05, "loss": 0.2795, "step": 1280 }, { "grad_norm": 2.9019153118133545, "learning_rate": 2.5779999999999997e-05, "loss": 0.2597, "step": 1290 }, { "grad_norm": 3.2190239429473877, "learning_rate": 2.598e-05, "loss": 0.2758, "step": 1300 }, { "grad_norm": 2.373692512512207, "learning_rate": 2.618e-05, "loss": 0.2503, "step": 1310 }, { "grad_norm": 2.822981595993042, "learning_rate": 2.6379999999999998e-05, "loss": 0.3085, "step": 1320 }, { "grad_norm": 3.258157730102539, "learning_rate": 2.658e-05, "loss": 0.2196, "step": 1330 }, { "grad_norm": 2.3210678100585938, "learning_rate": 2.678e-05, "loss": 0.2427, "step": 1340 }, { "grad_norm": 2.5877180099487305, "learning_rate": 2.698e-05, "loss": 0.2278, "step": 1350 }, { "grad_norm": 3.0419223308563232, "learning_rate": 2.718e-05, "loss": 0.2843, "step": 1360 }, { "grad_norm": 2.2018883228302, "learning_rate": 2.738e-05, "loss": 0.2362, "step": 1370 }, { "grad_norm": 2.918020486831665, "learning_rate": 2.758e-05, "loss": 0.2535, "step": 1380 }, { "grad_norm": 3.6035544872283936, "learning_rate": 2.778e-05, "loss": 0.2443, "step": 1390 }, { "grad_norm": 2.918701410293579, "learning_rate": 2.798e-05, "loss": 0.2858, "step": 1400 }, { "grad_norm": 2.841735601425171, "learning_rate": 2.818e-05, "loss": 0.2235, "step": 1410 }, { "grad_norm": 3.296907901763916, "learning_rate": 2.8380000000000003e-05, "loss": 0.255, "step": 1420 }, { "grad_norm": 3.115443706512451, "learning_rate": 2.858e-05, "loss": 0.2418, "step": 1430 }, { "grad_norm": 2.193618059158325, "learning_rate": 2.8780000000000002e-05, "loss": 0.2235, "step": 1440 }, { "grad_norm": 1.9395097494125366, "learning_rate": 2.898e-05, "loss": 0.2023, "step": 1450 }, { "grad_norm": 3.3547439575195312, "learning_rate": 2.9180000000000002e-05, "loss": 0.2255, "step": 1460 }, { "grad_norm": 2.250713348388672, "learning_rate": 2.9380000000000003e-05, "loss": 0.1979, "step": 1470 }, { "grad_norm": 2.0857036113739014, "learning_rate": 2.958e-05, "loss": 0.1687, "step": 1480 }, { "grad_norm": 2.652470827102661, "learning_rate": 2.9780000000000003e-05, "loss": 0.2127, "step": 1490 }, { "grad_norm": 1.8348627090454102, "learning_rate": 2.998e-05, "loss": 0.1849, "step": 1500 }, { "grad_norm": 2.3616249561309814, "learning_rate": 3.0180000000000002e-05, "loss": 0.2233, "step": 1510 }, { "grad_norm": 2.330385684967041, "learning_rate": 3.0380000000000004e-05, "loss": 0.1937, "step": 1520 }, { "grad_norm": 2.4650936126708984, "learning_rate": 3.058e-05, "loss": 0.1977, "step": 1530 }, { "grad_norm": 2.000901460647583, "learning_rate": 3.078e-05, "loss": 0.2274, "step": 1540 }, { "grad_norm": 2.460939884185791, "learning_rate": 3.0980000000000005e-05, "loss": 0.2836, "step": 1550 }, { "grad_norm": 2.932310104370117, "learning_rate": 3.118e-05, "loss": 0.2458, "step": 1560 }, { "grad_norm": 2.2964234352111816, "learning_rate": 3.138e-05, "loss": 0.1787, "step": 1570 }, { "grad_norm": 2.4127633571624756, "learning_rate": 3.1580000000000006e-05, "loss": 0.2268, "step": 1580 }, { "grad_norm": 2.416541576385498, "learning_rate": 3.1780000000000004e-05, "loss": 0.2105, "step": 1590 }, { "grad_norm": 2.032120943069458, "learning_rate": 3.198e-05, "loss": 0.1904, "step": 1600 }, { "grad_norm": 2.402925729751587, "learning_rate": 3.218e-05, "loss": 0.1951, "step": 1610 }, { "grad_norm": 2.5955424308776855, "learning_rate": 3.238e-05, "loss": 0.2, "step": 1620 }, { "grad_norm": 2.167024850845337, "learning_rate": 3.2579999999999996e-05, "loss": 0.1968, "step": 1630 }, { "grad_norm": 2.8697853088378906, "learning_rate": 3.278e-05, "loss": 0.2106, "step": 1640 }, { "grad_norm": 2.681098222732544, "learning_rate": 3.298e-05, "loss": 0.2353, "step": 1650 }, { "grad_norm": 2.243069648742676, "learning_rate": 3.318e-05, "loss": 0.1911, "step": 1660 }, { "grad_norm": 2.334779977798462, "learning_rate": 3.338e-05, "loss": 0.2191, "step": 1670 }, { "grad_norm": 2.121624231338501, "learning_rate": 3.358e-05, "loss": 0.2174, "step": 1680 }, { "grad_norm": 2.2416296005249023, "learning_rate": 3.378e-05, "loss": 0.1998, "step": 1690 }, { "grad_norm": 2.1574432849884033, "learning_rate": 3.398e-05, "loss": 0.2171, "step": 1700 }, { "grad_norm": 1.9824497699737549, "learning_rate": 3.418e-05, "loss": 0.1762, "step": 1710 }, { "grad_norm": 2.2487916946411133, "learning_rate": 3.438e-05, "loss": 0.1968, "step": 1720 }, { "grad_norm": 2.351738691329956, "learning_rate": 3.4580000000000004e-05, "loss": 0.2282, "step": 1730 }, { "grad_norm": 2.435720205307007, "learning_rate": 3.478e-05, "loss": 0.206, "step": 1740 }, { "grad_norm": 2.393458366394043, "learning_rate": 3.498e-05, "loss": 0.1812, "step": 1750 }, { "grad_norm": 2.3337159156799316, "learning_rate": 3.518e-05, "loss": 0.1859, "step": 1760 }, { "grad_norm": 2.562011957168579, "learning_rate": 3.5380000000000003e-05, "loss": 0.2121, "step": 1770 }, { "grad_norm": 2.3314943313598633, "learning_rate": 3.558e-05, "loss": 0.1962, "step": 1780 }, { "grad_norm": 1.7478551864624023, "learning_rate": 3.578e-05, "loss": 0.1946, "step": 1790 }, { "grad_norm": 2.381892681121826, "learning_rate": 3.5980000000000004e-05, "loss": 0.2092, "step": 1800 }, { "grad_norm": 1.749424934387207, "learning_rate": 3.618e-05, "loss": 0.1816, "step": 1810 }, { "grad_norm": 1.78411865234375, "learning_rate": 3.638e-05, "loss": 0.2104, "step": 1820 }, { "grad_norm": 2.5870847702026367, "learning_rate": 3.6580000000000006e-05, "loss": 0.169, "step": 1830 }, { "grad_norm": 2.105250120162964, "learning_rate": 3.6780000000000004e-05, "loss": 0.188, "step": 1840 }, { "grad_norm": 2.274521589279175, "learning_rate": 3.698e-05, "loss": 0.1993, "step": 1850 }, { "grad_norm": 1.70295250415802, "learning_rate": 3.7180000000000007e-05, "loss": 0.2231, "step": 1860 }, { "grad_norm": 2.2106411457061768, "learning_rate": 3.7380000000000005e-05, "loss": 0.2057, "step": 1870 }, { "grad_norm": 2.062937021255493, "learning_rate": 3.758e-05, "loss": 0.1915, "step": 1880 }, { "grad_norm": 2.237396717071533, "learning_rate": 3.778000000000001e-05, "loss": 0.1983, "step": 1890 }, { "grad_norm": 1.8403494358062744, "learning_rate": 3.7980000000000006e-05, "loss": 0.2014, "step": 1900 }, { "grad_norm": 1.4870305061340332, "learning_rate": 3.818e-05, "loss": 0.208, "step": 1910 }, { "grad_norm": 2.1955175399780273, "learning_rate": 3.838e-05, "loss": 0.1808, "step": 1920 }, { "grad_norm": 2.0040667057037354, "learning_rate": 3.858e-05, "loss": 0.2387, "step": 1930 }, { "grad_norm": 1.8216995000839233, "learning_rate": 3.878e-05, "loss": 0.204, "step": 1940 }, { "grad_norm": 1.4441386461257935, "learning_rate": 3.898e-05, "loss": 0.1659, "step": 1950 }, { "grad_norm": 1.9493268728256226, "learning_rate": 3.918e-05, "loss": 0.17, "step": 1960 }, { "grad_norm": 2.5457942485809326, "learning_rate": 3.938e-05, "loss": 0.1849, "step": 1970 }, { "grad_norm": 2.0604145526885986, "learning_rate": 3.958e-05, "loss": 0.1826, "step": 1980 }, { "grad_norm": 1.7133742570877075, "learning_rate": 3.978e-05, "loss": 0.1834, "step": 1990 }, { "grad_norm": 2.343719244003296, "learning_rate": 3.998e-05, "loss": 0.1792, "step": 2000 }, { "grad_norm": 1.6772836446762085, "learning_rate": 4.018e-05, "loss": 0.1686, "step": 2010 }, { "grad_norm": 1.9481697082519531, "learning_rate": 4.038e-05, "loss": 0.1695, "step": 2020 }, { "grad_norm": 2.1521847248077393, "learning_rate": 4.058e-05, "loss": 0.2116, "step": 2030 }, { "grad_norm": 2.19183349609375, "learning_rate": 4.078e-05, "loss": 0.199, "step": 2040 }, { "grad_norm": 1.605644941329956, "learning_rate": 4.0980000000000004e-05, "loss": 0.1566, "step": 2050 }, { "grad_norm": 1.791190505027771, "learning_rate": 4.118e-05, "loss": 0.1831, "step": 2060 }, { "grad_norm": 1.5308798551559448, "learning_rate": 4.138e-05, "loss": 0.1799, "step": 2070 }, { "grad_norm": 1.624755620956421, "learning_rate": 4.1580000000000005e-05, "loss": 0.1595, "step": 2080 }, { "grad_norm": 1.5052307844161987, "learning_rate": 4.178e-05, "loss": 0.1962, "step": 2090 }, { "grad_norm": 1.4584252834320068, "learning_rate": 4.198e-05, "loss": 0.182, "step": 2100 }, { "grad_norm": 2.2701358795166016, "learning_rate": 4.2180000000000006e-05, "loss": 0.1963, "step": 2110 }, { "grad_norm": 1.846603512763977, "learning_rate": 4.2380000000000004e-05, "loss": 0.2201, "step": 2120 }, { "grad_norm": 1.9308857917785645, "learning_rate": 4.258e-05, "loss": 0.2115, "step": 2130 }, { "grad_norm": 1.8193215131759644, "learning_rate": 4.278e-05, "loss": 0.2023, "step": 2140 }, { "grad_norm": 1.7198048830032349, "learning_rate": 4.2980000000000005e-05, "loss": 0.2178, "step": 2150 }, { "grad_norm": 1.804165005683899, "learning_rate": 4.318e-05, "loss": 0.2035, "step": 2160 }, { "grad_norm": 1.9347316026687622, "learning_rate": 4.338e-05, "loss": 0.1822, "step": 2170 }, { "grad_norm": 1.4371405839920044, "learning_rate": 4.3580000000000006e-05, "loss": 0.1887, "step": 2180 }, { "grad_norm": 2.0090012550354004, "learning_rate": 4.3780000000000004e-05, "loss": 0.1557, "step": 2190 }, { "grad_norm": 1.693588137626648, "learning_rate": 4.398e-05, "loss": 0.1413, "step": 2200 }, { "grad_norm": 1.805899977684021, "learning_rate": 4.418000000000001e-05, "loss": 0.1943, "step": 2210 }, { "grad_norm": 2.3039276599884033, "learning_rate": 4.438e-05, "loss": 0.2317, "step": 2220 }, { "grad_norm": 1.7563267946243286, "learning_rate": 4.458e-05, "loss": 0.1752, "step": 2230 }, { "grad_norm": 1.5342140197753906, "learning_rate": 4.478e-05, "loss": 0.1349, "step": 2240 }, { "grad_norm": 1.99171781539917, "learning_rate": 4.498e-05, "loss": 0.17, "step": 2250 }, { "grad_norm": 1.5614020824432373, "learning_rate": 4.518e-05, "loss": 0.1994, "step": 2260 }, { "grad_norm": 1.6265157461166382, "learning_rate": 4.538e-05, "loss": 0.2114, "step": 2270 }, { "grad_norm": 1.9307626485824585, "learning_rate": 4.558e-05, "loss": 0.1955, "step": 2280 }, { "grad_norm": 2.0640709400177, "learning_rate": 4.578e-05, "loss": 0.188, "step": 2290 }, { "grad_norm": 1.481408953666687, "learning_rate": 4.5980000000000004e-05, "loss": 0.2179, "step": 2300 }, { "grad_norm": 1.6337895393371582, "learning_rate": 4.618e-05, "loss": 0.1709, "step": 2310 }, { "grad_norm": 1.1006097793579102, "learning_rate": 4.638e-05, "loss": 0.1884, "step": 2320 }, { "grad_norm": 1.837408423423767, "learning_rate": 4.6580000000000005e-05, "loss": 0.1654, "step": 2330 }, { "grad_norm": 1.5918394327163696, "learning_rate": 4.678e-05, "loss": 0.1915, "step": 2340 }, { "grad_norm": 1.5889348983764648, "learning_rate": 4.698e-05, "loss": 0.1602, "step": 2350 }, { "grad_norm": 2.3733112812042236, "learning_rate": 4.718e-05, "loss": 0.1539, "step": 2360 }, { "grad_norm": 2.216536521911621, "learning_rate": 4.7380000000000004e-05, "loss": 0.2146, "step": 2370 }, { "grad_norm": 1.9283525943756104, "learning_rate": 4.758e-05, "loss": 0.1714, "step": 2380 }, { "grad_norm": 1.6451241970062256, "learning_rate": 4.778e-05, "loss": 0.1743, "step": 2390 }, { "grad_norm": 2.0320777893066406, "learning_rate": 4.7980000000000005e-05, "loss": 0.2038, "step": 2400 }, { "grad_norm": 1.9839903116226196, "learning_rate": 4.818e-05, "loss": 0.242, "step": 2410 }, { "grad_norm": 3.7000248432159424, "learning_rate": 4.838e-05, "loss": 0.1811, "step": 2420 }, { "grad_norm": 1.928950548171997, "learning_rate": 4.8580000000000006e-05, "loss": 0.1671, "step": 2430 }, { "grad_norm": 2.003436803817749, "learning_rate": 4.8780000000000004e-05, "loss": 0.1813, "step": 2440 }, { "grad_norm": 1.6874433755874634, "learning_rate": 4.898e-05, "loss": 0.1961, "step": 2450 }, { "grad_norm": 1.5577555894851685, "learning_rate": 4.918000000000001e-05, "loss": 0.147, "step": 2460 }, { "grad_norm": 1.915242075920105, "learning_rate": 4.9380000000000005e-05, "loss": 0.2144, "step": 2470 }, { "grad_norm": 1.822774052619934, "learning_rate": 4.958e-05, "loss": 0.163, "step": 2480 }, { "grad_norm": 1.4075082540512085, "learning_rate": 4.978e-05, "loss": 0.1671, "step": 2490 }, { "grad_norm": 1.4822769165039062, "learning_rate": 4.9980000000000006e-05, "loss": 0.1622, "step": 2500 }, { "grad_norm": 1.3628157377243042, "learning_rate": 5.0180000000000004e-05, "loss": 0.1902, "step": 2510 }, { "grad_norm": 1.171311855316162, "learning_rate": 5.038e-05, "loss": 0.1708, "step": 2520 }, { "grad_norm": 1.6240390539169312, "learning_rate": 5.058000000000001e-05, "loss": 0.154, "step": 2530 }, { "grad_norm": 1.5442895889282227, "learning_rate": 5.0780000000000005e-05, "loss": 0.1607, "step": 2540 }, { "grad_norm": 1.2759464979171753, "learning_rate": 5.098e-05, "loss": 0.1715, "step": 2550 }, { "grad_norm": 1.7439743280410767, "learning_rate": 5.118000000000001e-05, "loss": 0.182, "step": 2560 }, { "grad_norm": 1.5492373704910278, "learning_rate": 5.1380000000000006e-05, "loss": 0.2016, "step": 2570 }, { "grad_norm": 1.482730507850647, "learning_rate": 5.1580000000000004e-05, "loss": 0.1773, "step": 2580 }, { "grad_norm": 1.7895643711090088, "learning_rate": 5.178000000000001e-05, "loss": 0.1832, "step": 2590 }, { "grad_norm": 1.5390466451644897, "learning_rate": 5.198000000000001e-05, "loss": 0.1735, "step": 2600 }, { "grad_norm": 1.60984468460083, "learning_rate": 5.2180000000000005e-05, "loss": 0.1637, "step": 2610 }, { "grad_norm": 1.5727826356887817, "learning_rate": 5.238000000000001e-05, "loss": 0.1776, "step": 2620 }, { "grad_norm": 1.3677459955215454, "learning_rate": 5.258000000000001e-05, "loss": 0.1329, "step": 2630 }, { "grad_norm": 1.4287452697753906, "learning_rate": 5.2780000000000006e-05, "loss": 0.194, "step": 2640 }, { "grad_norm": 1.2490229606628418, "learning_rate": 5.2980000000000004e-05, "loss": 0.1408, "step": 2650 }, { "grad_norm": 1.3467648029327393, "learning_rate": 5.318000000000001e-05, "loss": 0.1502, "step": 2660 }, { "grad_norm": 1.692333459854126, "learning_rate": 5.338000000000001e-05, "loss": 0.1473, "step": 2670 }, { "grad_norm": 1.202362060546875, "learning_rate": 5.3580000000000005e-05, "loss": 0.1545, "step": 2680 }, { "grad_norm": 1.0102238655090332, "learning_rate": 5.378e-05, "loss": 0.1808, "step": 2690 }, { "grad_norm": 1.1401009559631348, "learning_rate": 5.3979999999999995e-05, "loss": 0.2038, "step": 2700 }, { "grad_norm": 1.7584974765777588, "learning_rate": 5.418e-05, "loss": 0.1785, "step": 2710 }, { "grad_norm": 1.6641548871994019, "learning_rate": 5.438e-05, "loss": 0.1719, "step": 2720 }, { "grad_norm": 1.4920963048934937, "learning_rate": 5.4579999999999996e-05, "loss": 0.1412, "step": 2730 }, { "grad_norm": 1.566347360610962, "learning_rate": 5.478e-05, "loss": 0.1729, "step": 2740 }, { "grad_norm": 0.9817872643470764, "learning_rate": 5.498e-05, "loss": 0.1395, "step": 2750 }, { "grad_norm": 0.9796192646026611, "learning_rate": 5.518e-05, "loss": 0.1454, "step": 2760 }, { "grad_norm": 1.2310250997543335, "learning_rate": 5.538e-05, "loss": 0.1656, "step": 2770 }, { "grad_norm": 1.4238178730010986, "learning_rate": 5.558e-05, "loss": 0.1879, "step": 2780 }, { "grad_norm": 1.4195871353149414, "learning_rate": 5.578e-05, "loss": 0.1409, "step": 2790 }, { "grad_norm": 1.4600591659545898, "learning_rate": 5.5979999999999996e-05, "loss": 0.1612, "step": 2800 }, { "grad_norm": 1.4095697402954102, "learning_rate": 5.618e-05, "loss": 0.1821, "step": 2810 }, { "grad_norm": 2.1073272228240967, "learning_rate": 5.638e-05, "loss": 0.1633, "step": 2820 }, { "grad_norm": 1.4055968523025513, "learning_rate": 5.658e-05, "loss": 0.1482, "step": 2830 }, { "grad_norm": 1.6666957139968872, "learning_rate": 5.678e-05, "loss": 0.18, "step": 2840 }, { "grad_norm": 2.052492618560791, "learning_rate": 5.698e-05, "loss": 0.1372, "step": 2850 }, { "grad_norm": 1.2176074981689453, "learning_rate": 5.718e-05, "loss": 0.1422, "step": 2860 }, { "grad_norm": 2.1106131076812744, "learning_rate": 5.738e-05, "loss": 0.1683, "step": 2870 }, { "grad_norm": 1.1379115581512451, "learning_rate": 5.758e-05, "loss": 0.1502, "step": 2880 }, { "grad_norm": 1.1174226999282837, "learning_rate": 5.778e-05, "loss": 0.2045, "step": 2890 }, { "grad_norm": 1.780735731124878, "learning_rate": 5.7980000000000004e-05, "loss": 0.1735, "step": 2900 }, { "grad_norm": 1.0784834623336792, "learning_rate": 5.818e-05, "loss": 0.2117, "step": 2910 }, { "grad_norm": 1.325713872909546, "learning_rate": 5.838e-05, "loss": 0.1487, "step": 2920 }, { "grad_norm": 1.2490450143814087, "learning_rate": 5.858e-05, "loss": 0.173, "step": 2930 }, { "grad_norm": 1.2474552392959595, "learning_rate": 5.878e-05, "loss": 0.1763, "step": 2940 }, { "grad_norm": 1.4659842252731323, "learning_rate": 5.898e-05, "loss": 0.1648, "step": 2950 }, { "grad_norm": 1.6962718963623047, "learning_rate": 5.918e-05, "loss": 0.1572, "step": 2960 }, { "grad_norm": 1.3505069017410278, "learning_rate": 5.9380000000000004e-05, "loss": 0.1468, "step": 2970 }, { "grad_norm": 1.073028326034546, "learning_rate": 5.958e-05, "loss": 0.1813, "step": 2980 }, { "grad_norm": 1.3366897106170654, "learning_rate": 5.978e-05, "loss": 0.1794, "step": 2990 }, { "grad_norm": 1.356711745262146, "learning_rate": 5.9980000000000005e-05, "loss": 0.1673, "step": 3000 }, { "grad_norm": 1.068508505821228, "learning_rate": 6.018e-05, "loss": 0.1567, "step": 3010 }, { "grad_norm": 0.9307708740234375, "learning_rate": 6.038e-05, "loss": 0.1471, "step": 3020 }, { "grad_norm": 0.9680135250091553, "learning_rate": 6.0580000000000006e-05, "loss": 0.1583, "step": 3030 }, { "grad_norm": 1.3700640201568604, "learning_rate": 6.0780000000000004e-05, "loss": 0.1759, "step": 3040 }, { "grad_norm": 1.6502416133880615, "learning_rate": 6.098e-05, "loss": 0.1533, "step": 3050 }, { "grad_norm": 1.051966905593872, "learning_rate": 6.118000000000001e-05, "loss": 0.2045, "step": 3060 }, { "grad_norm": 1.1613327264785767, "learning_rate": 6.138e-05, "loss": 0.1781, "step": 3070 }, { "grad_norm": 1.2332230806350708, "learning_rate": 6.158e-05, "loss": 0.1584, "step": 3080 }, { "grad_norm": 1.3577989339828491, "learning_rate": 6.178000000000001e-05, "loss": 0.1566, "step": 3090 }, { "grad_norm": 1.3144696950912476, "learning_rate": 6.198e-05, "loss": 0.1476, "step": 3100 }, { "grad_norm": 1.7548656463623047, "learning_rate": 6.218e-05, "loss": 0.1979, "step": 3110 }, { "grad_norm": 1.1428053379058838, "learning_rate": 6.238000000000001e-05, "loss": 0.1653, "step": 3120 }, { "grad_norm": 1.03415048122406, "learning_rate": 6.258e-05, "loss": 0.1378, "step": 3130 }, { "grad_norm": 1.2772091627120972, "learning_rate": 6.278e-05, "loss": 0.1661, "step": 3140 }, { "grad_norm": 1.082331657409668, "learning_rate": 6.298000000000001e-05, "loss": 0.1471, "step": 3150 }, { "grad_norm": 1.3580986261367798, "learning_rate": 6.318e-05, "loss": 0.1546, "step": 3160 }, { "grad_norm": 1.8131937980651855, "learning_rate": 6.338e-05, "loss": 0.148, "step": 3170 }, { "grad_norm": 1.5351344347000122, "learning_rate": 6.358000000000001e-05, "loss": 0.183, "step": 3180 }, { "grad_norm": 1.1370158195495605, "learning_rate": 6.378e-05, "loss": 0.1541, "step": 3190 }, { "grad_norm": 1.1405490636825562, "learning_rate": 6.398000000000001e-05, "loss": 0.1467, "step": 3200 }, { "grad_norm": 1.3718599081039429, "learning_rate": 6.418000000000001e-05, "loss": 0.1298, "step": 3210 }, { "grad_norm": 1.1949535608291626, "learning_rate": 6.438e-05, "loss": 0.1379, "step": 3220 }, { "grad_norm": 1.4800058603286743, "learning_rate": 6.458000000000001e-05, "loss": 0.1715, "step": 3230 }, { "grad_norm": 1.2120808362960815, "learning_rate": 6.478000000000001e-05, "loss": 0.1722, "step": 3240 }, { "grad_norm": 1.200016975402832, "learning_rate": 6.498e-05, "loss": 0.1695, "step": 3250 }, { "grad_norm": 1.2679129838943481, "learning_rate": 6.518000000000001e-05, "loss": 0.1999, "step": 3260 }, { "grad_norm": 1.3446252346038818, "learning_rate": 6.538000000000001e-05, "loss": 0.1562, "step": 3270 }, { "grad_norm": 1.2622554302215576, "learning_rate": 6.558e-05, "loss": 0.1769, "step": 3280 }, { "grad_norm": 1.498152494430542, "learning_rate": 6.578000000000001e-05, "loss": 0.1752, "step": 3290 }, { "grad_norm": 1.4777089357376099, "learning_rate": 6.598e-05, "loss": 0.1657, "step": 3300 }, { "grad_norm": 1.1096192598342896, "learning_rate": 6.618e-05, "loss": 0.1556, "step": 3310 }, { "grad_norm": 1.3171495199203491, "learning_rate": 6.638e-05, "loss": 0.1796, "step": 3320 }, { "grad_norm": 1.3058116436004639, "learning_rate": 6.658e-05, "loss": 0.1472, "step": 3330 }, { "grad_norm": 1.4914367198944092, "learning_rate": 6.678e-05, "loss": 0.1536, "step": 3340 }, { "grad_norm": 1.186394214630127, "learning_rate": 6.698e-05, "loss": 0.1544, "step": 3350 }, { "grad_norm": 1.1498695611953735, "learning_rate": 6.718e-05, "loss": 0.1571, "step": 3360 }, { "grad_norm": 1.0018680095672607, "learning_rate": 6.738e-05, "loss": 0.1544, "step": 3370 }, { "grad_norm": 1.265194296836853, "learning_rate": 6.758e-05, "loss": 0.1408, "step": 3380 }, { "grad_norm": 1.0004326105117798, "learning_rate": 6.778e-05, "loss": 0.148, "step": 3390 }, { "grad_norm": 1.0754506587982178, "learning_rate": 6.798e-05, "loss": 0.165, "step": 3400 }, { "grad_norm": 1.184824824333191, "learning_rate": 6.818e-05, "loss": 0.1521, "step": 3410 }, { "grad_norm": 1.8029965162277222, "learning_rate": 6.838e-05, "loss": 0.2094, "step": 3420 }, { "grad_norm": 0.9986934065818787, "learning_rate": 6.858e-05, "loss": 0.1639, "step": 3430 }, { "grad_norm": 1.3199520111083984, "learning_rate": 6.878e-05, "loss": 0.1768, "step": 3440 }, { "grad_norm": 1.2057929039001465, "learning_rate": 6.898e-05, "loss": 0.1616, "step": 3450 }, { "grad_norm": 1.5902833938598633, "learning_rate": 6.918e-05, "loss": 0.1502, "step": 3460 }, { "grad_norm": 1.5680276155471802, "learning_rate": 6.938e-05, "loss": 0.1671, "step": 3470 }, { "grad_norm": 1.3725954294204712, "learning_rate": 6.958e-05, "loss": 0.1275, "step": 3480 }, { "grad_norm": 2.100944995880127, "learning_rate": 6.978e-05, "loss": 0.1585, "step": 3490 }, { "grad_norm": 1.385787010192871, "learning_rate": 6.998e-05, "loss": 0.1774, "step": 3500 }, { "grad_norm": 1.50184965133667, "learning_rate": 7.018e-05, "loss": 0.1677, "step": 3510 }, { "grad_norm": 1.8081368207931519, "learning_rate": 7.038e-05, "loss": 0.1627, "step": 3520 }, { "grad_norm": 1.065475583076477, "learning_rate": 7.058e-05, "loss": 0.1591, "step": 3530 }, { "grad_norm": 1.498306393623352, "learning_rate": 7.078e-05, "loss": 0.1644, "step": 3540 }, { "grad_norm": 1.8333309888839722, "learning_rate": 7.098e-05, "loss": 0.1895, "step": 3550 }, { "grad_norm": 1.376451015472412, "learning_rate": 7.118e-05, "loss": 0.1888, "step": 3560 }, { "grad_norm": 1.2869778871536255, "learning_rate": 7.138e-05, "loss": 0.1483, "step": 3570 }, { "grad_norm": 1.353995680809021, "learning_rate": 7.158e-05, "loss": 0.1639, "step": 3580 }, { "grad_norm": 1.7574104070663452, "learning_rate": 7.178000000000001e-05, "loss": 0.1635, "step": 3590 }, { "grad_norm": 1.128562569618225, "learning_rate": 7.198e-05, "loss": 0.1572, "step": 3600 }, { "grad_norm": 1.5359457731246948, "learning_rate": 7.218e-05, "loss": 0.1637, "step": 3610 }, { "grad_norm": 0.8633236885070801, "learning_rate": 7.238000000000001e-05, "loss": 0.1378, "step": 3620 }, { "grad_norm": 1.296722412109375, "learning_rate": 7.258e-05, "loss": 0.1448, "step": 3630 }, { "grad_norm": 1.542135238647461, "learning_rate": 7.278e-05, "loss": 0.1733, "step": 3640 }, { "grad_norm": 1.1218873262405396, "learning_rate": 7.298000000000001e-05, "loss": 0.136, "step": 3650 }, { "grad_norm": 1.4641057252883911, "learning_rate": 7.318e-05, "loss": 0.1961, "step": 3660 }, { "grad_norm": 1.1983784437179565, "learning_rate": 7.338e-05, "loss": 0.1871, "step": 3670 }, { "grad_norm": 1.1506446599960327, "learning_rate": 7.358000000000001e-05, "loss": 0.1443, "step": 3680 }, { "grad_norm": 1.2497951984405518, "learning_rate": 7.378e-05, "loss": 0.1986, "step": 3690 }, { "grad_norm": 1.1299412250518799, "learning_rate": 7.398e-05, "loss": 0.157, "step": 3700 }, { "grad_norm": 1.9984467029571533, "learning_rate": 7.418000000000001e-05, "loss": 0.1738, "step": 3710 }, { "grad_norm": 1.2445473670959473, "learning_rate": 7.438e-05, "loss": 0.2273, "step": 3720 }, { "grad_norm": 1.2216373682022095, "learning_rate": 7.458000000000001e-05, "loss": 0.1489, "step": 3730 }, { "grad_norm": 2.584022045135498, "learning_rate": 7.478e-05, "loss": 0.161, "step": 3740 }, { "grad_norm": 1.215186357498169, "learning_rate": 7.498e-05, "loss": 0.1457, "step": 3750 }, { "grad_norm": 1.180112600326538, "learning_rate": 7.518000000000001e-05, "loss": 0.1484, "step": 3760 }, { "grad_norm": 1.2346251010894775, "learning_rate": 7.538e-05, "loss": 0.1595, "step": 3770 }, { "grad_norm": 1.3760828971862793, "learning_rate": 7.558e-05, "loss": 0.136, "step": 3780 }, { "grad_norm": 1.1781609058380127, "learning_rate": 7.578000000000001e-05, "loss": 0.1574, "step": 3790 }, { "grad_norm": 0.6908823251724243, "learning_rate": 7.598e-05, "loss": 0.1491, "step": 3800 }, { "grad_norm": 0.9731953740119934, "learning_rate": 7.618e-05, "loss": 0.1578, "step": 3810 }, { "grad_norm": 1.488499402999878, "learning_rate": 7.638000000000001e-05, "loss": 0.1979, "step": 3820 }, { "grad_norm": 1.3937721252441406, "learning_rate": 7.658e-05, "loss": 0.1643, "step": 3830 }, { "grad_norm": 0.9009853005409241, "learning_rate": 7.678000000000001e-05, "loss": 0.1549, "step": 3840 }, { "grad_norm": 1.1514790058135986, "learning_rate": 7.698000000000001e-05, "loss": 0.1507, "step": 3850 }, { "grad_norm": 1.386032223701477, "learning_rate": 7.718e-05, "loss": 0.1383, "step": 3860 }, { "grad_norm": 0.9792682528495789, "learning_rate": 7.738000000000001e-05, "loss": 0.1383, "step": 3870 }, { "grad_norm": 1.9144575595855713, "learning_rate": 7.758000000000001e-05, "loss": 0.1753, "step": 3880 }, { "grad_norm": 0.9101016521453857, "learning_rate": 7.778e-05, "loss": 0.1426, "step": 3890 }, { "grad_norm": 0.8557650446891785, "learning_rate": 7.798000000000001e-05, "loss": 0.1273, "step": 3900 }, { "grad_norm": 0.8576015830039978, "learning_rate": 7.818000000000001e-05, "loss": 0.1641, "step": 3910 }, { "grad_norm": 1.228427767753601, "learning_rate": 7.838e-05, "loss": 0.1836, "step": 3920 }, { "grad_norm": 0.8627099990844727, "learning_rate": 7.858000000000001e-05, "loss": 0.1185, "step": 3930 }, { "grad_norm": 1.2747578620910645, "learning_rate": 7.878e-05, "loss": 0.1442, "step": 3940 }, { "grad_norm": 1.0037366151809692, "learning_rate": 7.897999999999999e-05, "loss": 0.1419, "step": 3950 }, { "grad_norm": 1.1039987802505493, "learning_rate": 7.918e-05, "loss": 0.1914, "step": 3960 }, { "grad_norm": 1.1618479490280151, "learning_rate": 7.938e-05, "loss": 0.1478, "step": 3970 }, { "grad_norm": 1.3023768663406372, "learning_rate": 7.958e-05, "loss": 0.1529, "step": 3980 }, { "grad_norm": 1.4721055030822754, "learning_rate": 7.978e-05, "loss": 0.1965, "step": 3990 }, { "grad_norm": 1.1598759889602661, "learning_rate": 7.998e-05, "loss": 0.1331, "step": 4000 }, { "grad_norm": 1.401652216911316, "learning_rate": 8.018e-05, "loss": 0.1935, "step": 4010 }, { "grad_norm": 1.169985055923462, "learning_rate": 8.038e-05, "loss": 0.2057, "step": 4020 }, { "grad_norm": 1.003322958946228, "learning_rate": 8.058e-05, "loss": 0.1356, "step": 4030 }, { "grad_norm": 0.9441876411437988, "learning_rate": 8.078e-05, "loss": 0.1629, "step": 4040 }, { "grad_norm": 1.1202853918075562, "learning_rate": 8.098e-05, "loss": 0.1154, "step": 4050 }, { "grad_norm": 1.0536205768585205, "learning_rate": 8.118e-05, "loss": 0.1515, "step": 4060 }, { "grad_norm": 1.2352386713027954, "learning_rate": 8.138e-05, "loss": 0.1153, "step": 4070 }, { "grad_norm": 1.368311882019043, "learning_rate": 8.158e-05, "loss": 0.1881, "step": 4080 }, { "grad_norm": 1.2201294898986816, "learning_rate": 8.178e-05, "loss": 0.1427, "step": 4090 }, { "grad_norm": 1.2968238592147827, "learning_rate": 8.198e-05, "loss": 0.1646, "step": 4100 }, { "grad_norm": 1.3340535163879395, "learning_rate": 8.218e-05, "loss": 0.1442, "step": 4110 }, { "grad_norm": 1.2500724792480469, "learning_rate": 8.238000000000001e-05, "loss": 0.1686, "step": 4120 }, { "grad_norm": 1.5601842403411865, "learning_rate": 8.258e-05, "loss": 0.1507, "step": 4130 }, { "grad_norm": 1.2541147470474243, "learning_rate": 8.278e-05, "loss": 0.1563, "step": 4140 }, { "grad_norm": 1.046938419342041, "learning_rate": 8.298000000000001e-05, "loss": 0.1468, "step": 4150 }, { "grad_norm": 0.8945430517196655, "learning_rate": 8.318e-05, "loss": 0.1696, "step": 4160 }, { "grad_norm": 0.9238892793655396, "learning_rate": 8.338e-05, "loss": 0.1298, "step": 4170 }, { "grad_norm": 1.0498936176300049, "learning_rate": 8.358e-05, "loss": 0.1857, "step": 4180 }, { "grad_norm": 1.3224787712097168, "learning_rate": 8.378e-05, "loss": 0.1405, "step": 4190 }, { "grad_norm": 1.1574366092681885, "learning_rate": 8.398e-05, "loss": 0.1599, "step": 4200 }, { "grad_norm": 1.0019066333770752, "learning_rate": 8.418e-05, "loss": 0.1662, "step": 4210 }, { "grad_norm": 1.1783125400543213, "learning_rate": 8.438e-05, "loss": 0.1894, "step": 4220 }, { "grad_norm": 1.924530029296875, "learning_rate": 8.458e-05, "loss": 0.1591, "step": 4230 }, { "grad_norm": 0.9998896718025208, "learning_rate": 8.478e-05, "loss": 0.153, "step": 4240 }, { "grad_norm": 0.8155059814453125, "learning_rate": 8.498e-05, "loss": 0.1599, "step": 4250 }, { "grad_norm": 0.8137487769126892, "learning_rate": 8.518000000000001e-05, "loss": 0.1405, "step": 4260 }, { "grad_norm": 0.9798411130905151, "learning_rate": 8.538e-05, "loss": 0.1267, "step": 4270 }, { "grad_norm": 1.143325686454773, "learning_rate": 8.558e-05, "loss": 0.1313, "step": 4280 }, { "grad_norm": 1.1619101762771606, "learning_rate": 8.578000000000001e-05, "loss": 0.1765, "step": 4290 }, { "grad_norm": 1.3062602281570435, "learning_rate": 8.598e-05, "loss": 0.1675, "step": 4300 }, { "grad_norm": 1.0883409976959229, "learning_rate": 8.618e-05, "loss": 0.1798, "step": 4310 }, { "grad_norm": 1.0177487134933472, "learning_rate": 8.638000000000001e-05, "loss": 0.1523, "step": 4320 }, { "grad_norm": 1.2354040145874023, "learning_rate": 8.658e-05, "loss": 0.1381, "step": 4330 }, { "grad_norm": 1.0772464275360107, "learning_rate": 8.678e-05, "loss": 0.132, "step": 4340 }, { "grad_norm": 1.279889702796936, "learning_rate": 8.698000000000001e-05, "loss": 0.179, "step": 4350 }, { "grad_norm": 1.1399027109146118, "learning_rate": 8.718e-05, "loss": 0.1912, "step": 4360 }, { "grad_norm": 1.1755448579788208, "learning_rate": 8.738000000000001e-05, "loss": 0.1609, "step": 4370 }, { "grad_norm": 1.1986174583435059, "learning_rate": 8.758000000000001e-05, "loss": 0.1503, "step": 4380 }, { "grad_norm": 0.719417154788971, "learning_rate": 8.778e-05, "loss": 0.1463, "step": 4390 }, { "grad_norm": 0.914290726184845, "learning_rate": 8.798000000000001e-05, "loss": 0.138, "step": 4400 }, { "grad_norm": 1.1129611730575562, "learning_rate": 8.818000000000001e-05, "loss": 0.1679, "step": 4410 }, { "grad_norm": 0.8875679969787598, "learning_rate": 8.838e-05, "loss": 0.1369, "step": 4420 }, { "grad_norm": 0.8760764598846436, "learning_rate": 8.858000000000001e-05, "loss": 0.1501, "step": 4430 }, { "grad_norm": 1.385438323020935, "learning_rate": 8.878000000000001e-05, "loss": 0.1509, "step": 4440 }, { "grad_norm": 0.9092206358909607, "learning_rate": 8.898e-05, "loss": 0.1243, "step": 4450 }, { "grad_norm": 1.18179190158844, "learning_rate": 8.918000000000001e-05, "loss": 0.1669, "step": 4460 }, { "grad_norm": 1.2923877239227295, "learning_rate": 8.938e-05, "loss": 0.1495, "step": 4470 }, { "grad_norm": 1.044233798980713, "learning_rate": 8.958e-05, "loss": 0.1265, "step": 4480 }, { "grad_norm": 1.149007797241211, "learning_rate": 8.978000000000001e-05, "loss": 0.143, "step": 4490 }, { "grad_norm": 1.0207208395004272, "learning_rate": 8.998e-05, "loss": 0.16, "step": 4500 }, { "grad_norm": 1.0398691892623901, "learning_rate": 9.018000000000001e-05, "loss": 0.1555, "step": 4510 }, { "grad_norm": 1.0691574811935425, "learning_rate": 9.038000000000001e-05, "loss": 0.1432, "step": 4520 }, { "grad_norm": 1.1882481575012207, "learning_rate": 9.058e-05, "loss": 0.1565, "step": 4530 }, { "grad_norm": 0.9592229723930359, "learning_rate": 9.078000000000001e-05, "loss": 0.1393, "step": 4540 }, { "grad_norm": 0.936359703540802, "learning_rate": 9.098000000000001e-05, "loss": 0.1389, "step": 4550 }, { "grad_norm": 1.1972596645355225, "learning_rate": 9.118e-05, "loss": 0.1512, "step": 4560 }, { "grad_norm": 0.7587670683860779, "learning_rate": 9.138e-05, "loss": 0.1721, "step": 4570 }, { "grad_norm": 1.7536144256591797, "learning_rate": 9.158e-05, "loss": 0.1776, "step": 4580 }, { "grad_norm": 0.737167477607727, "learning_rate": 9.178e-05, "loss": 0.1462, "step": 4590 }, { "grad_norm": 1.0012117624282837, "learning_rate": 9.198e-05, "loss": 0.1497, "step": 4600 }, { "grad_norm": 1.1247097253799438, "learning_rate": 9.218e-05, "loss": 0.1605, "step": 4610 }, { "grad_norm": 0.8472771048545837, "learning_rate": 9.238e-05, "loss": 0.1503, "step": 4620 }, { "grad_norm": 1.2532007694244385, "learning_rate": 9.258e-05, "loss": 0.1234, "step": 4630 }, { "grad_norm": 0.8576799035072327, "learning_rate": 9.278e-05, "loss": 0.1302, "step": 4640 }, { "grad_norm": 1.4112308025360107, "learning_rate": 9.298e-05, "loss": 0.1566, "step": 4650 }, { "grad_norm": 1.0498369932174683, "learning_rate": 9.318e-05, "loss": 0.1524, "step": 4660 }, { "grad_norm": 1.1274765729904175, "learning_rate": 9.338e-05, "loss": 0.1451, "step": 4670 }, { "grad_norm": 0.9639546871185303, "learning_rate": 9.358e-05, "loss": 0.1415, "step": 4680 }, { "grad_norm": 0.6763057708740234, "learning_rate": 9.378e-05, "loss": 0.1775, "step": 4690 }, { "grad_norm": 0.8133760690689087, "learning_rate": 9.398e-05, "loss": 0.1434, "step": 4700 }, { "grad_norm": 1.1117700338363647, "learning_rate": 9.418e-05, "loss": 0.1432, "step": 4710 }, { "grad_norm": 0.9571735858917236, "learning_rate": 9.438e-05, "loss": 0.1703, "step": 4720 }, { "grad_norm": 0.7914835810661316, "learning_rate": 9.458e-05, "loss": 0.1498, "step": 4730 }, { "grad_norm": 0.9418796896934509, "learning_rate": 9.478e-05, "loss": 0.1378, "step": 4740 }, { "grad_norm": 1.267486810684204, "learning_rate": 9.498e-05, "loss": 0.1728, "step": 4750 }, { "grad_norm": 0.8760393857955933, "learning_rate": 9.518000000000001e-05, "loss": 0.1619, "step": 4760 }, { "grad_norm": 1.0746911764144897, "learning_rate": 9.538e-05, "loss": 0.1617, "step": 4770 }, { "grad_norm": 1.0673942565917969, "learning_rate": 9.558e-05, "loss": 0.165, "step": 4780 }, { "grad_norm": 1.2225459814071655, "learning_rate": 9.578000000000001e-05, "loss": 0.1603, "step": 4790 }, { "grad_norm": 1.335009217262268, "learning_rate": 9.598e-05, "loss": 0.1363, "step": 4800 }, { "grad_norm": 0.7508177161216736, "learning_rate": 9.618e-05, "loss": 0.1353, "step": 4810 }, { "grad_norm": 0.998579740524292, "learning_rate": 9.638000000000001e-05, "loss": 0.1473, "step": 4820 }, { "grad_norm": 1.0209482908248901, "learning_rate": 9.658e-05, "loss": 0.217, "step": 4830 }, { "grad_norm": 1.0383579730987549, "learning_rate": 9.678e-05, "loss": 0.1738, "step": 4840 }, { "grad_norm": 1.9238204956054688, "learning_rate": 9.698000000000001e-05, "loss": 0.1615, "step": 4850 }, { "grad_norm": 1.0530461072921753, "learning_rate": 9.718e-05, "loss": 0.1477, "step": 4860 }, { "grad_norm": 1.2157971858978271, "learning_rate": 9.738e-05, "loss": 0.1794, "step": 4870 }, { "grad_norm": 0.9704339504241943, "learning_rate": 9.758000000000001e-05, "loss": 0.1547, "step": 4880 }, { "grad_norm": 0.9925690293312073, "learning_rate": 9.778e-05, "loss": 0.1427, "step": 4890 }, { "grad_norm": 1.2629235982894897, "learning_rate": 9.798000000000001e-05, "loss": 0.1446, "step": 4900 }, { "grad_norm": 0.7826972007751465, "learning_rate": 9.818000000000001e-05, "loss": 0.1465, "step": 4910 }, { "grad_norm": 1.203446626663208, "learning_rate": 9.838e-05, "loss": 0.1464, "step": 4920 }, { "grad_norm": 1.278612494468689, "learning_rate": 9.858000000000001e-05, "loss": 0.1645, "step": 4930 }, { "grad_norm": 0.8602445125579834, "learning_rate": 9.878e-05, "loss": 0.1315, "step": 4940 }, { "grad_norm": 0.9315417408943176, "learning_rate": 9.898e-05, "loss": 0.1604, "step": 4950 }, { "grad_norm": 1.0579164028167725, "learning_rate": 9.918000000000001e-05, "loss": 0.178, "step": 4960 }, { "grad_norm": 1.2438350915908813, "learning_rate": 9.938e-05, "loss": 0.1436, "step": 4970 }, { "grad_norm": 0.9854502081871033, "learning_rate": 9.958e-05, "loss": 0.1372, "step": 4980 }, { "grad_norm": 0.9737119674682617, "learning_rate": 9.978000000000001e-05, "loss": 0.1484, "step": 4990 }, { "grad_norm": 0.9519772529602051, "learning_rate": 9.998e-05, "loss": 0.1256, "step": 5000 }, { "grad_norm": 0.9491996169090271, "learning_rate": 9.999999778549045e-05, "loss": 0.1302, "step": 5010 }, { "grad_norm": 1.2088196277618408, "learning_rate": 9.999999013039593e-05, "loss": 0.1395, "step": 5020 }, { "grad_norm": 0.9389463067054749, "learning_rate": 9.999997700737766e-05, "loss": 0.1516, "step": 5030 }, { "grad_norm": 1.3446253538131714, "learning_rate": 9.999995841643709e-05, "loss": 0.1596, "step": 5040 }, { "grad_norm": 0.9374755620956421, "learning_rate": 9.999993435757623e-05, "loss": 0.1543, "step": 5050 }, { "grad_norm": 0.7997305989265442, "learning_rate": 9.999990483079773e-05, "loss": 0.122, "step": 5060 }, { "grad_norm": 1.3632607460021973, "learning_rate": 9.999986983610481e-05, "loss": 0.1547, "step": 5070 }, { "grad_norm": 1.109691858291626, "learning_rate": 9.99998293735013e-05, "loss": 0.1362, "step": 5080 }, { "grad_norm": 0.8433117866516113, "learning_rate": 9.999978344299161e-05, "loss": 0.1823, "step": 5090 }, { "grad_norm": 0.9824782609939575, "learning_rate": 9.99997320445808e-05, "loss": 0.1271, "step": 5100 }, { "grad_norm": 1.0228302478790283, "learning_rate": 9.999967517827444e-05, "loss": 0.1674, "step": 5110 }, { "grad_norm": 1.3150831460952759, "learning_rate": 9.999961284407879e-05, "loss": 0.1453, "step": 5120 }, { "grad_norm": 1.174692988395691, "learning_rate": 9.999954504200067e-05, "loss": 0.1311, "step": 5130 }, { "grad_norm": 0.9238080382347107, "learning_rate": 9.999947177204744e-05, "loss": 0.1526, "step": 5140 }, { "grad_norm": 0.993269145488739, "learning_rate": 9.999939303422718e-05, "loss": 0.1576, "step": 5150 }, { "grad_norm": 0.9647553563117981, "learning_rate": 9.999930882854847e-05, "loss": 0.1348, "step": 5160 }, { "grad_norm": 1.3037182092666626, "learning_rate": 9.999921915502051e-05, "loss": 0.1582, "step": 5170 }, { "grad_norm": 0.9025166630744934, "learning_rate": 9.99991240136531e-05, "loss": 0.142, "step": 5180 }, { "grad_norm": 1.098415732383728, "learning_rate": 9.999902340445668e-05, "loss": 0.1704, "step": 5190 }, { "grad_norm": 0.8088397979736328, "learning_rate": 9.999891732744224e-05, "loss": 0.1331, "step": 5200 }, { "grad_norm": 1.071036696434021, "learning_rate": 9.999880578262135e-05, "loss": 0.1358, "step": 5210 }, { "grad_norm": 0.8885997533798218, "learning_rate": 9.999868877000624e-05, "loss": 0.1599, "step": 5220 }, { "grad_norm": 1.125024676322937, "learning_rate": 9.99985662896097e-05, "loss": 0.1504, "step": 5230 }, { "grad_norm": 0.8946710824966431, "learning_rate": 9.999843834144513e-05, "loss": 0.1199, "step": 5240 }, { "grad_norm": 0.994486927986145, "learning_rate": 9.99983049255265e-05, "loss": 0.1234, "step": 5250 }, { "grad_norm": 0.8100867867469788, "learning_rate": 9.999816604186843e-05, "loss": 0.1532, "step": 5260 }, { "grad_norm": 1.1229408979415894, "learning_rate": 9.999802169048609e-05, "loss": 0.1571, "step": 5270 }, { "grad_norm": 0.9546085000038147, "learning_rate": 9.999787187139527e-05, "loss": 0.1569, "step": 5280 }, { "grad_norm": 1.2900376319885254, "learning_rate": 9.999771658461234e-05, "loss": 0.1675, "step": 5290 }, { "grad_norm": 1.2018802165985107, "learning_rate": 9.999755583015431e-05, "loss": 0.1439, "step": 5300 }, { "grad_norm": 0.6983299255371094, "learning_rate": 9.999738960803874e-05, "loss": 0.1634, "step": 5310 }, { "grad_norm": 1.2518147230148315, "learning_rate": 9.99972179182838e-05, "loss": 0.1595, "step": 5320 }, { "grad_norm": 1.1781526803970337, "learning_rate": 9.99970407609083e-05, "loss": 0.1659, "step": 5330 }, { "grad_norm": 1.2446311712265015, "learning_rate": 9.999685813593159e-05, "loss": 0.1689, "step": 5340 }, { "grad_norm": 0.9353486895561218, "learning_rate": 9.999667004337362e-05, "loss": 0.1705, "step": 5350 }, { "grad_norm": 1.0606199502944946, "learning_rate": 9.9996476483255e-05, "loss": 0.1767, "step": 5360 }, { "grad_norm": 0.7918563485145569, "learning_rate": 9.999627745559688e-05, "loss": 0.1504, "step": 5370 }, { "grad_norm": 0.9074621200561523, "learning_rate": 9.999607296042101e-05, "loss": 0.1362, "step": 5380 }, { "grad_norm": 1.0680131912231445, "learning_rate": 9.99958629977498e-05, "loss": 0.1465, "step": 5390 }, { "grad_norm": 1.082570195198059, "learning_rate": 9.999564756760615e-05, "loss": 0.1299, "step": 5400 }, { "grad_norm": 1.2158985137939453, "learning_rate": 9.999542667001366e-05, "loss": 0.139, "step": 5410 }, { "grad_norm": 0.9460130333900452, "learning_rate": 9.999520030499647e-05, "loss": 0.1458, "step": 5420 }, { "grad_norm": 0.9973957538604736, "learning_rate": 9.999496847257936e-05, "loss": 0.14, "step": 5430 }, { "grad_norm": 0.8514935970306396, "learning_rate": 9.999473117278764e-05, "loss": 0.1707, "step": 5440 }, { "grad_norm": 0.8374741673469543, "learning_rate": 9.999448840564731e-05, "loss": 0.1508, "step": 5450 }, { "grad_norm": 1.0440653562545776, "learning_rate": 9.999424017118488e-05, "loss": 0.1401, "step": 5460 }, { "grad_norm": 1.0712121725082397, "learning_rate": 9.999398646942751e-05, "loss": 0.121, "step": 5470 }, { "grad_norm": 0.6817668676376343, "learning_rate": 9.999372730040296e-05, "loss": 0.1248, "step": 5480 }, { "grad_norm": 1.0203224420547485, "learning_rate": 9.999346266413953e-05, "loss": 0.1376, "step": 5490 }, { "grad_norm": 1.232972264289856, "learning_rate": 9.99931925606662e-05, "loss": 0.1366, "step": 5500 }, { "grad_norm": 0.8608493804931641, "learning_rate": 9.99929169900125e-05, "loss": 0.1249, "step": 5510 }, { "grad_norm": 1.190839409828186, "learning_rate": 9.999263595220855e-05, "loss": 0.1299, "step": 5520 }, { "grad_norm": 0.8917580842971802, "learning_rate": 9.99923494472851e-05, "loss": 0.1237, "step": 5530 }, { "grad_norm": 1.3571122884750366, "learning_rate": 9.999205747527348e-05, "loss": 0.1574, "step": 5540 }, { "grad_norm": 1.2097225189208984, "learning_rate": 9.999176003620561e-05, "loss": 0.1611, "step": 5550 }, { "grad_norm": 1.1247707605361938, "learning_rate": 9.999145713011405e-05, "loss": 0.1579, "step": 5560 }, { "grad_norm": 0.86859530210495, "learning_rate": 9.999114875703186e-05, "loss": 0.1446, "step": 5570 }, { "grad_norm": 0.9987765550613403, "learning_rate": 9.999083491699281e-05, "loss": 0.1534, "step": 5580 }, { "grad_norm": 0.9756432175636292, "learning_rate": 9.999051561003123e-05, "loss": 0.1248, "step": 5590 }, { "grad_norm": 0.9701782464981079, "learning_rate": 9.999019083618202e-05, "loss": 0.1491, "step": 5600 }, { "grad_norm": 1.0569761991500854, "learning_rate": 9.99898605954807e-05, "loss": 0.1474, "step": 5610 }, { "grad_norm": 1.0472534894943237, "learning_rate": 9.998952488796338e-05, "loss": 0.1946, "step": 5620 }, { "grad_norm": 0.8290643692016602, "learning_rate": 9.998918371366676e-05, "loss": 0.1481, "step": 5630 }, { "grad_norm": 1.2615307569503784, "learning_rate": 9.99888370726282e-05, "loss": 0.1638, "step": 5640 }, { "grad_norm": 1.3338457345962524, "learning_rate": 9.998848496488556e-05, "loss": 0.1455, "step": 5650 }, { "grad_norm": 1.1375221014022827, "learning_rate": 9.998812739047736e-05, "loss": 0.181, "step": 5660 }, { "grad_norm": 1.0272607803344727, "learning_rate": 9.99877643494427e-05, "loss": 0.1692, "step": 5670 }, { "grad_norm": 0.8138233423233032, "learning_rate": 9.998739584182128e-05, "loss": 0.1739, "step": 5680 }, { "grad_norm": 0.9818016290664673, "learning_rate": 9.998702186765342e-05, "loss": 0.1731, "step": 5690 }, { "grad_norm": 1.1786326169967651, "learning_rate": 9.998664242698e-05, "loss": 0.1399, "step": 5700 }, { "grad_norm": 0.9839988350868225, "learning_rate": 9.998625751984251e-05, "loss": 0.1212, "step": 5710 }, { "grad_norm": 1.2975200414657593, "learning_rate": 9.998586714628307e-05, "loss": 0.1458, "step": 5720 }, { "grad_norm": 1.0838433504104614, "learning_rate": 9.998547130634432e-05, "loss": 0.1505, "step": 5730 }, { "grad_norm": 0.8396374583244324, "learning_rate": 9.99850700000696e-05, "loss": 0.1234, "step": 5740 }, { "grad_norm": 1.132948875427246, "learning_rate": 9.998466322750278e-05, "loss": 0.1357, "step": 5750 }, { "grad_norm": 1.0343354940414429, "learning_rate": 9.998425098868834e-05, "loss": 0.1139, "step": 5760 }, { "grad_norm": 0.8200494647026062, "learning_rate": 9.998383328367136e-05, "loss": 0.1508, "step": 5770 }, { "grad_norm": 0.8254609107971191, "learning_rate": 9.99834101124975e-05, "loss": 0.1388, "step": 5780 }, { "grad_norm": 0.750197172164917, "learning_rate": 9.998298147521309e-05, "loss": 0.1227, "step": 5790 }, { "grad_norm": 0.8947212100028992, "learning_rate": 9.998254737186496e-05, "loss": 0.1538, "step": 5800 }, { "grad_norm": 1.2218366861343384, "learning_rate": 9.99821078025006e-05, "loss": 0.1286, "step": 5810 }, { "grad_norm": 1.2924188375473022, "learning_rate": 9.998166276716807e-05, "loss": 0.1431, "step": 5820 }, { "grad_norm": 0.8672069311141968, "learning_rate": 9.998121226591606e-05, "loss": 0.1424, "step": 5830 }, { "grad_norm": 0.9175875782966614, "learning_rate": 9.998075629879382e-05, "loss": 0.1248, "step": 5840 }, { "grad_norm": 0.9629296064376831, "learning_rate": 9.99802948658512e-05, "loss": 0.1197, "step": 5850 }, { "grad_norm": 1.0724444389343262, "learning_rate": 9.99798279671387e-05, "loss": 0.1289, "step": 5860 }, { "grad_norm": 0.7933607697486877, "learning_rate": 9.997935560270734e-05, "loss": 0.127, "step": 5870 }, { "grad_norm": 1.214593768119812, "learning_rate": 9.997887777260879e-05, "loss": 0.1206, "step": 5880 }, { "grad_norm": 1.1244847774505615, "learning_rate": 9.997839447689532e-05, "loss": 0.1221, "step": 5890 }, { "grad_norm": 0.8480269312858582, "learning_rate": 9.997790571561978e-05, "loss": 0.1274, "step": 5900 }, { "grad_norm": 0.9261433482170105, "learning_rate": 9.99774114888356e-05, "loss": 0.1487, "step": 5910 }, { "grad_norm": 0.8206645846366882, "learning_rate": 9.997691179659684e-05, "loss": 0.1146, "step": 5920 }, { "grad_norm": 0.8300005793571472, "learning_rate": 9.997640663895815e-05, "loss": 0.1033, "step": 5930 }, { "grad_norm": 0.9753710031509399, "learning_rate": 9.997589601597477e-05, "loss": 0.1402, "step": 5940 }, { "grad_norm": 0.7924512028694153, "learning_rate": 9.997537992770252e-05, "loss": 0.1314, "step": 5950 }, { "grad_norm": 0.9669636487960815, "learning_rate": 9.997485837419788e-05, "loss": 0.1362, "step": 5960 }, { "grad_norm": 1.1993141174316406, "learning_rate": 9.997433135551786e-05, "loss": 0.1205, "step": 5970 }, { "grad_norm": 1.0269027948379517, "learning_rate": 9.997379887172009e-05, "loss": 0.1639, "step": 5980 }, { "grad_norm": 0.9593906998634338, "learning_rate": 9.997326092286281e-05, "loss": 0.1636, "step": 5990 }, { "grad_norm": 0.8160423040390015, "learning_rate": 9.997271750900486e-05, "loss": 0.1224, "step": 6000 }, { "grad_norm": 0.693603515625, "learning_rate": 9.997216863020565e-05, "loss": 0.1171, "step": 6010 }, { "grad_norm": 0.9251887202262878, "learning_rate": 9.99716142865252e-05, "loss": 0.1294, "step": 6020 }, { "grad_norm": 0.897566556930542, "learning_rate": 9.997105447802415e-05, "loss": 0.1439, "step": 6030 }, { "grad_norm": 0.6686179637908936, "learning_rate": 9.997048920476373e-05, "loss": 0.1192, "step": 6040 }, { "grad_norm": 0.8612222671508789, "learning_rate": 9.996991846680572e-05, "loss": 0.1442, "step": 6050 }, { "grad_norm": 0.7626209259033203, "learning_rate": 9.996934226421257e-05, "loss": 0.1304, "step": 6060 }, { "grad_norm": 0.7152012586593628, "learning_rate": 9.996876059704726e-05, "loss": 0.1187, "step": 6070 }, { "grad_norm": 1.2988406419754028, "learning_rate": 9.996817346537343e-05, "loss": 0.15, "step": 6080 }, { "grad_norm": 0.9677882790565491, "learning_rate": 9.996758086925526e-05, "loss": 0.1235, "step": 6090 }, { "grad_norm": 0.6172149181365967, "learning_rate": 9.996698280875759e-05, "loss": 0.1393, "step": 6100 }, { "grad_norm": 0.9656276702880859, "learning_rate": 9.99663792839458e-05, "loss": 0.1378, "step": 6110 }, { "grad_norm": 0.8294305205345154, "learning_rate": 9.99657702948859e-05, "loss": 0.1378, "step": 6120 }, { "grad_norm": 0.7629187703132629, "learning_rate": 9.996515584164448e-05, "loss": 0.115, "step": 6130 }, { "grad_norm": 1.0521869659423828, "learning_rate": 9.996453592428873e-05, "loss": 0.1277, "step": 6140 }, { "grad_norm": 0.6938046216964722, "learning_rate": 9.996391054288646e-05, "loss": 0.118, "step": 6150 }, { "grad_norm": 0.7403213977813721, "learning_rate": 9.996327969750605e-05, "loss": 0.1283, "step": 6160 }, { "grad_norm": 0.9073890447616577, "learning_rate": 9.996264338821649e-05, "loss": 0.1457, "step": 6170 }, { "grad_norm": 0.8558454513549805, "learning_rate": 9.996200161508735e-05, "loss": 0.1562, "step": 6180 }, { "grad_norm": 0.9779731631278992, "learning_rate": 9.996135437818885e-05, "loss": 0.1336, "step": 6190 }, { "grad_norm": 0.9856162667274475, "learning_rate": 9.996070167759175e-05, "loss": 0.1297, "step": 6200 }, { "grad_norm": 1.137843132019043, "learning_rate": 9.996004351336743e-05, "loss": 0.119, "step": 6210 }, { "grad_norm": 0.9909107685089111, "learning_rate": 9.995937988558785e-05, "loss": 0.1233, "step": 6220 }, { "grad_norm": 0.8272787928581238, "learning_rate": 9.995871079432561e-05, "loss": 0.1143, "step": 6230 }, { "grad_norm": 1.2241337299346924, "learning_rate": 9.995803623965389e-05, "loss": 0.1065, "step": 6240 }, { "grad_norm": 1.1675686836242676, "learning_rate": 9.995735622164641e-05, "loss": 0.1321, "step": 6250 }, { "grad_norm": 1.3412306308746338, "learning_rate": 9.995667074037758e-05, "loss": 0.1531, "step": 6260 }, { "grad_norm": 0.9965149760246277, "learning_rate": 9.995597979592232e-05, "loss": 0.1232, "step": 6270 }, { "grad_norm": 1.0725337266921997, "learning_rate": 9.995528338835625e-05, "loss": 0.1451, "step": 6280 }, { "grad_norm": 1.5188748836517334, "learning_rate": 9.995458151775547e-05, "loss": 0.1248, "step": 6290 }, { "grad_norm": 1.5077462196350098, "learning_rate": 9.995387418419677e-05, "loss": 0.1368, "step": 6300 }, { "grad_norm": 0.9546157717704773, "learning_rate": 9.99531613877575e-05, "loss": 0.1207, "step": 6310 }, { "grad_norm": 1.441286325454712, "learning_rate": 9.995244312851559e-05, "loss": 0.1418, "step": 6320 }, { "grad_norm": 0.7163488268852234, "learning_rate": 9.995171940654961e-05, "loss": 0.1322, "step": 6330 }, { "grad_norm": 0.7958322763442993, "learning_rate": 9.995099022193871e-05, "loss": 0.1235, "step": 6340 }, { "grad_norm": 0.8446505069732666, "learning_rate": 9.995025557476261e-05, "loss": 0.1359, "step": 6350 }, { "grad_norm": 1.105229139328003, "learning_rate": 9.994951546510165e-05, "loss": 0.1531, "step": 6360 }, { "grad_norm": 0.8645409941673279, "learning_rate": 9.994876989303679e-05, "loss": 0.1431, "step": 6370 }, { "grad_norm": 1.0218206644058228, "learning_rate": 9.994801885864955e-05, "loss": 0.1385, "step": 6380 }, { "grad_norm": 0.9250398874282837, "learning_rate": 9.994726236202205e-05, "loss": 0.1294, "step": 6390 }, { "grad_norm": 0.7468629479408264, "learning_rate": 9.994650040323704e-05, "loss": 0.1425, "step": 6400 }, { "grad_norm": 0.9302475452423096, "learning_rate": 9.994573298237784e-05, "loss": 0.1312, "step": 6410 }, { "grad_norm": 0.8840485215187073, "learning_rate": 9.994496009952837e-05, "loss": 0.1211, "step": 6420 }, { "grad_norm": 1.2780442237854004, "learning_rate": 9.994418175477316e-05, "loss": 0.148, "step": 6430 }, { "grad_norm": 0.770529568195343, "learning_rate": 9.994339794819733e-05, "loss": 0.168, "step": 6440 }, { "grad_norm": 0.6401891708374023, "learning_rate": 9.994260867988658e-05, "loss": 0.1476, "step": 6450 }, { "grad_norm": 0.7508208155632019, "learning_rate": 9.994181394992723e-05, "loss": 0.1193, "step": 6460 }, { "grad_norm": 0.9422900676727295, "learning_rate": 9.994101375840618e-05, "loss": 0.1377, "step": 6470 }, { "grad_norm": 0.790124773979187, "learning_rate": 9.994020810541098e-05, "loss": 0.12, "step": 6480 }, { "grad_norm": 0.8415157198905945, "learning_rate": 9.99393969910297e-05, "loss": 0.1054, "step": 6490 }, { "grad_norm": 0.8623924851417542, "learning_rate": 9.993858041535104e-05, "loss": 0.1567, "step": 6500 }, { "grad_norm": 0.7026618123054504, "learning_rate": 9.99377583784643e-05, "loss": 0.1247, "step": 6510 }, { "grad_norm": 0.9278709292411804, "learning_rate": 9.993693088045939e-05, "loss": 0.1248, "step": 6520 }, { "grad_norm": 0.9052345156669617, "learning_rate": 9.99360979214268e-05, "loss": 0.1282, "step": 6530 }, { "grad_norm": 1.1118261814117432, "learning_rate": 9.99352595014576e-05, "loss": 0.1296, "step": 6540 }, { "grad_norm": 1.0549030303955078, "learning_rate": 9.993441562064354e-05, "loss": 0.1271, "step": 6550 }, { "grad_norm": 0.9061239361763, "learning_rate": 9.993356627907685e-05, "loss": 0.125, "step": 6560 }, { "grad_norm": 1.1367186307907104, "learning_rate": 9.99327114768504e-05, "loss": 0.1239, "step": 6570 }, { "grad_norm": 0.9487401843070984, "learning_rate": 9.99318512140577e-05, "loss": 0.1344, "step": 6580 }, { "grad_norm": 0.6874019503593445, "learning_rate": 9.993098549079284e-05, "loss": 0.1247, "step": 6590 }, { "grad_norm": 0.7332987785339355, "learning_rate": 9.993011430715047e-05, "loss": 0.1144, "step": 6600 }, { "grad_norm": 0.8578271269798279, "learning_rate": 9.992923766322586e-05, "loss": 0.1478, "step": 6610 }, { "grad_norm": 0.6942188143730164, "learning_rate": 9.99283555591149e-05, "loss": 0.1152, "step": 6620 }, { "grad_norm": 0.7971992492675781, "learning_rate": 9.992746799491404e-05, "loss": 0.1393, "step": 6630 }, { "grad_norm": 0.8951964378356934, "learning_rate": 9.992657497072033e-05, "loss": 0.1598, "step": 6640 }, { "grad_norm": 0.8991621732711792, "learning_rate": 9.992567648663147e-05, "loss": 0.1225, "step": 6650 }, { "grad_norm": 0.7805278301239014, "learning_rate": 9.992477254274568e-05, "loss": 0.127, "step": 6660 }, { "grad_norm": 0.9330773949623108, "learning_rate": 9.992386313916183e-05, "loss": 0.1217, "step": 6670 }, { "grad_norm": 0.9965944290161133, "learning_rate": 9.992294827597934e-05, "loss": 0.1289, "step": 6680 }, { "grad_norm": 0.7631148099899292, "learning_rate": 9.992202795329831e-05, "loss": 0.1375, "step": 6690 }, { "grad_norm": 0.9698922038078308, "learning_rate": 9.992110217121936e-05, "loss": 0.1262, "step": 6700 }, { "grad_norm": 0.8481687307357788, "learning_rate": 9.992017092984372e-05, "loss": 0.114, "step": 6710 }, { "grad_norm": 0.6220568418502808, "learning_rate": 9.991923422927326e-05, "loss": 0.1203, "step": 6720 }, { "grad_norm": 0.8401553630828857, "learning_rate": 9.991829206961037e-05, "loss": 0.1224, "step": 6730 }, { "grad_norm": 0.9156072735786438, "learning_rate": 9.991734445095813e-05, "loss": 0.1402, "step": 6740 }, { "grad_norm": 1.2140178680419922, "learning_rate": 9.991639137342015e-05, "loss": 0.181, "step": 6750 }, { "grad_norm": 1.1855586767196655, "learning_rate": 9.991543283710064e-05, "loss": 0.1816, "step": 6760 }, { "grad_norm": 1.1642085313796997, "learning_rate": 9.991446884210445e-05, "loss": 0.1371, "step": 6770 }, { "grad_norm": 0.735676646232605, "learning_rate": 9.9913499388537e-05, "loss": 0.113, "step": 6780 }, { "grad_norm": 0.6811376810073853, "learning_rate": 9.99125244765043e-05, "loss": 0.1507, "step": 6790 }, { "grad_norm": 0.8814405798912048, "learning_rate": 9.991154410611296e-05, "loss": 0.1266, "step": 6800 }, { "grad_norm": 1.7524009943008423, "learning_rate": 9.99105582774702e-05, "loss": 0.132, "step": 6810 }, { "grad_norm": 0.8731167316436768, "learning_rate": 9.990956699068384e-05, "loss": 0.1131, "step": 6820 }, { "grad_norm": 0.5997774600982666, "learning_rate": 9.990857024586224e-05, "loss": 0.1179, "step": 6830 }, { "grad_norm": 0.9628852009773254, "learning_rate": 9.990756804311446e-05, "loss": 0.1337, "step": 6840 }, { "grad_norm": 0.814433753490448, "learning_rate": 9.990656038255006e-05, "loss": 0.1139, "step": 6850 }, { "grad_norm": 0.8187528848648071, "learning_rate": 9.990554726427926e-05, "loss": 0.1506, "step": 6860 }, { "grad_norm": 0.6905945539474487, "learning_rate": 9.990452868841284e-05, "loss": 0.1058, "step": 6870 }, { "grad_norm": 0.8690517544746399, "learning_rate": 9.99035046550622e-05, "loss": 0.1684, "step": 6880 }, { "grad_norm": 0.9326394200325012, "learning_rate": 9.99024751643393e-05, "loss": 0.1176, "step": 6890 }, { "grad_norm": 0.8876421451568604, "learning_rate": 9.990144021635677e-05, "loss": 0.127, "step": 6900 }, { "grad_norm": 0.9753937125205994, "learning_rate": 9.990039981122775e-05, "loss": 0.1215, "step": 6910 }, { "grad_norm": 0.787394642829895, "learning_rate": 9.989935394906602e-05, "loss": 0.1345, "step": 6920 }, { "grad_norm": 1.1095926761627197, "learning_rate": 9.989830262998598e-05, "loss": 0.15, "step": 6930 }, { "grad_norm": 0.8732206225395203, "learning_rate": 9.989724585410259e-05, "loss": 0.1344, "step": 6940 }, { "grad_norm": 0.9959803819656372, "learning_rate": 9.989618362153139e-05, "loss": 0.1315, "step": 6950 }, { "grad_norm": 0.9982351660728455, "learning_rate": 9.989511593238859e-05, "loss": 0.1282, "step": 6960 }, { "grad_norm": 1.0656306743621826, "learning_rate": 9.98940427867909e-05, "loss": 0.1354, "step": 6970 }, { "grad_norm": 0.931402325630188, "learning_rate": 9.989296418485573e-05, "loss": 0.129, "step": 6980 }, { "grad_norm": 0.7124209403991699, "learning_rate": 9.989188012670101e-05, "loss": 0.1237, "step": 6990 }, { "grad_norm": 0.6648887395858765, "learning_rate": 9.989079061244528e-05, "loss": 0.0996, "step": 7000 }, { "grad_norm": 0.9193800687789917, "learning_rate": 9.988969564220769e-05, "loss": 0.1164, "step": 7010 }, { "grad_norm": 1.0092887878417969, "learning_rate": 9.988859521610801e-05, "loss": 0.128, "step": 7020 }, { "grad_norm": 0.8886992335319519, "learning_rate": 9.988748933426656e-05, "loss": 0.1107, "step": 7030 }, { "grad_norm": 0.7841851115226746, "learning_rate": 9.988637799680428e-05, "loss": 0.1411, "step": 7040 }, { "grad_norm": 1.0561069250106812, "learning_rate": 9.98852612038427e-05, "loss": 0.1325, "step": 7050 }, { "grad_norm": 0.6732900142669678, "learning_rate": 9.988413895550397e-05, "loss": 0.1169, "step": 7060 }, { "grad_norm": 0.7930694222450256, "learning_rate": 9.98830112519108e-05, "loss": 0.1174, "step": 7070 }, { "grad_norm": 0.8015963435173035, "learning_rate": 9.98818780931865e-05, "loss": 0.1244, "step": 7080 }, { "grad_norm": 0.8899248242378235, "learning_rate": 9.988073947945502e-05, "loss": 0.1252, "step": 7090 }, { "grad_norm": 0.8777888417243958, "learning_rate": 9.987959541084087e-05, "loss": 0.1386, "step": 7100 }, { "grad_norm": 0.8650599718093872, "learning_rate": 9.987844588746915e-05, "loss": 0.1028, "step": 7110 }, { "grad_norm": 0.7319146394729614, "learning_rate": 9.987729090946558e-05, "loss": 0.1037, "step": 7120 }, { "grad_norm": 0.8515529632568359, "learning_rate": 9.987613047695647e-05, "loss": 0.1193, "step": 7130 }, { "grad_norm": 0.7794911861419678, "learning_rate": 9.987496459006871e-05, "loss": 0.131, "step": 7140 }, { "grad_norm": 0.8021753430366516, "learning_rate": 9.987379324892982e-05, "loss": 0.1674, "step": 7150 }, { "grad_norm": 1.024025559425354, "learning_rate": 9.987261645366788e-05, "loss": 0.121, "step": 7160 }, { "grad_norm": 0.7343617081642151, "learning_rate": 9.987143420441158e-05, "loss": 0.1455, "step": 7170 }, { "grad_norm": 1.2292454242706299, "learning_rate": 9.987024650129022e-05, "loss": 0.1343, "step": 7180 }, { "grad_norm": 0.92071533203125, "learning_rate": 9.986905334443368e-05, "loss": 0.1242, "step": 7190 }, { "grad_norm": 0.8757066130638123, "learning_rate": 9.986785473397245e-05, "loss": 0.1356, "step": 7200 }, { "grad_norm": 0.7674590945243835, "learning_rate": 9.98666506700376e-05, "loss": 0.1239, "step": 7210 }, { "grad_norm": 0.8014757633209229, "learning_rate": 9.986544115276081e-05, "loss": 0.1266, "step": 7220 }, { "grad_norm": 1.0212607383728027, "learning_rate": 9.986422618227433e-05, "loss": 0.1104, "step": 7230 }, { "grad_norm": 0.8028773665428162, "learning_rate": 9.986300575871106e-05, "loss": 0.1207, "step": 7240 }, { "grad_norm": 0.6505030393600464, "learning_rate": 9.986177988220444e-05, "loss": 0.0963, "step": 7250 }, { "grad_norm": 0.7709053754806519, "learning_rate": 9.986054855288856e-05, "loss": 0.1015, "step": 7260 }, { "grad_norm": 0.888192355632782, "learning_rate": 9.985931177089802e-05, "loss": 0.1173, "step": 7270 }, { "grad_norm": 0.7717440128326416, "learning_rate": 9.985806953636814e-05, "loss": 0.1029, "step": 7280 }, { "grad_norm": 0.8416531682014465, "learning_rate": 9.985682184943471e-05, "loss": 0.1247, "step": 7290 }, { "grad_norm": 0.7769631743431091, "learning_rate": 9.98555687102342e-05, "loss": 0.1131, "step": 7300 }, { "grad_norm": 0.7966835498809814, "learning_rate": 9.985431011890367e-05, "loss": 0.1367, "step": 7310 }, { "grad_norm": 0.9364163279533386, "learning_rate": 9.985304607558075e-05, "loss": 0.103, "step": 7320 }, { "grad_norm": 0.8431732654571533, "learning_rate": 9.985177658040364e-05, "loss": 0.1224, "step": 7330 }, { "grad_norm": 0.9035516977310181, "learning_rate": 9.985050163351119e-05, "loss": 0.1126, "step": 7340 }, { "grad_norm": 0.6633657217025757, "learning_rate": 9.984922123504286e-05, "loss": 0.1259, "step": 7350 }, { "grad_norm": 0.9980842471122742, "learning_rate": 9.984793538513862e-05, "loss": 0.1179, "step": 7360 }, { "grad_norm": 0.7064674496650696, "learning_rate": 9.984664408393912e-05, "loss": 0.1181, "step": 7370 }, { "grad_norm": 0.6138377785682678, "learning_rate": 9.984534733158556e-05, "loss": 0.1152, "step": 7380 }, { "grad_norm": 0.8513625860214233, "learning_rate": 9.984404512821977e-05, "loss": 0.1247, "step": 7390 }, { "grad_norm": 0.6720279455184937, "learning_rate": 9.984273747398411e-05, "loss": 0.1075, "step": 7400 }, { "grad_norm": 1.0424754619598389, "learning_rate": 9.984142436902165e-05, "loss": 0.1253, "step": 7410 }, { "grad_norm": 1.0313125848770142, "learning_rate": 9.984010581347596e-05, "loss": 0.1131, "step": 7420 }, { "grad_norm": 0.8133149743080139, "learning_rate": 9.983878180749121e-05, "loss": 0.1205, "step": 7430 }, { "grad_norm": 0.8839675188064575, "learning_rate": 9.983745235121222e-05, "loss": 0.1076, "step": 7440 }, { "grad_norm": 0.8758889436721802, "learning_rate": 9.983611744478438e-05, "loss": 0.1114, "step": 7450 }, { "grad_norm": 0.6453133225440979, "learning_rate": 9.983477708835365e-05, "loss": 0.1021, "step": 7460 }, { "grad_norm": 0.7582225799560547, "learning_rate": 9.983343128206664e-05, "loss": 0.1125, "step": 7470 }, { "grad_norm": 0.8486539721488953, "learning_rate": 9.983208002607049e-05, "loss": 0.1035, "step": 7480 }, { "grad_norm": 0.7901632189750671, "learning_rate": 9.9830723320513e-05, "loss": 0.1216, "step": 7490 }, { "grad_norm": 0.8527823090553284, "learning_rate": 9.982936116554254e-05, "loss": 0.1374, "step": 7500 }, { "grad_norm": 0.6914036273956299, "learning_rate": 9.982799356130803e-05, "loss": 0.1053, "step": 7510 }, { "grad_norm": 0.5184653401374817, "learning_rate": 9.982662050795908e-05, "loss": 0.0889, "step": 7520 }, { "grad_norm": 1.0321120023727417, "learning_rate": 9.982524200564583e-05, "loss": 0.1161, "step": 7530 }, { "grad_norm": 1.070760726928711, "learning_rate": 9.982385805451901e-05, "loss": 0.136, "step": 7540 }, { "grad_norm": 0.8314123153686523, "learning_rate": 9.982246865472998e-05, "loss": 0.1248, "step": 7550 }, { "grad_norm": 0.7024849653244019, "learning_rate": 9.982107380643069e-05, "loss": 0.1311, "step": 7560 }, { "grad_norm": 0.775700032711029, "learning_rate": 9.981967350977368e-05, "loss": 0.1217, "step": 7570 }, { "grad_norm": 0.7769746780395508, "learning_rate": 9.981826776491208e-05, "loss": 0.1034, "step": 7580 }, { "grad_norm": 1.0401753187179565, "learning_rate": 9.98168565719996e-05, "loss": 0.1431, "step": 7590 }, { "grad_norm": 0.693110466003418, "learning_rate": 9.98154399311906e-05, "loss": 0.1298, "step": 7600 }, { "grad_norm": 0.814463198184967, "learning_rate": 9.981401784263997e-05, "loss": 0.1095, "step": 7610 }, { "grad_norm": 0.825453519821167, "learning_rate": 9.981259030650326e-05, "loss": 0.1063, "step": 7620 }, { "grad_norm": 0.7643927931785583, "learning_rate": 9.981115732293655e-05, "loss": 0.1037, "step": 7630 }, { "grad_norm": 0.7978413105010986, "learning_rate": 9.980971889209659e-05, "loss": 0.1257, "step": 7640 }, { "grad_norm": 0.8453851938247681, "learning_rate": 9.980827501414064e-05, "loss": 0.1133, "step": 7650 }, { "grad_norm": 0.6481263041496277, "learning_rate": 9.980682568922663e-05, "loss": 0.1055, "step": 7660 }, { "grad_norm": 1.0977550745010376, "learning_rate": 9.980537091751304e-05, "loss": 0.093, "step": 7670 }, { "grad_norm": 0.7531293630599976, "learning_rate": 9.980391069915897e-05, "loss": 0.1038, "step": 7680 }, { "grad_norm": 0.6217013597488403, "learning_rate": 9.98024450343241e-05, "loss": 0.1272, "step": 7690 }, { "grad_norm": 1.176895022392273, "learning_rate": 9.980097392316872e-05, "loss": 0.1355, "step": 7700 }, { "grad_norm": 0.908551037311554, "learning_rate": 9.97994973658537e-05, "loss": 0.1101, "step": 7710 }, { "grad_norm": 0.9558018445968628, "learning_rate": 9.979801536254054e-05, "loss": 0.1268, "step": 7720 }, { "grad_norm": 0.6790947318077087, "learning_rate": 9.979652791339127e-05, "loss": 0.1231, "step": 7730 }, { "grad_norm": 0.8687283396720886, "learning_rate": 9.97950350185686e-05, "loss": 0.1083, "step": 7740 }, { "grad_norm": 0.8819242119789124, "learning_rate": 9.979353667823574e-05, "loss": 0.1027, "step": 7750 }, { "grad_norm": 0.8811890482902527, "learning_rate": 9.979203289255658e-05, "loss": 0.0966, "step": 7760 }, { "grad_norm": 0.9726349711418152, "learning_rate": 9.979052366169557e-05, "loss": 0.1387, "step": 7770 }, { "grad_norm": 0.7419189214706421, "learning_rate": 9.978900898581775e-05, "loss": 0.116, "step": 7780 }, { "grad_norm": 0.930060863494873, "learning_rate": 9.978748886508875e-05, "loss": 0.1266, "step": 7790 }, { "grad_norm": 0.8890635371208191, "learning_rate": 9.978596329967484e-05, "loss": 0.1268, "step": 7800 }, { "grad_norm": 1.0411243438720703, "learning_rate": 9.978443228974284e-05, "loss": 0.1183, "step": 7810 }, { "grad_norm": 0.8783160448074341, "learning_rate": 9.978289583546015e-05, "loss": 0.1081, "step": 7820 }, { "grad_norm": 0.8024159669876099, "learning_rate": 9.978135393699484e-05, "loss": 0.1272, "step": 7830 }, { "grad_norm": 1.0126343965530396, "learning_rate": 9.977980659451548e-05, "loss": 0.1378, "step": 7840 }, { "grad_norm": 0.9186937212944031, "learning_rate": 9.977825380819135e-05, "loss": 0.1319, "step": 7850 }, { "grad_norm": 0.8259017467498779, "learning_rate": 9.97766955781922e-05, "loss": 0.1072, "step": 7860 }, { "grad_norm": 1.1616257429122925, "learning_rate": 9.977513190468848e-05, "loss": 0.1482, "step": 7870 }, { "grad_norm": 0.8007764220237732, "learning_rate": 9.977356278785116e-05, "loss": 0.1488, "step": 7880 }, { "grad_norm": 0.842278242111206, "learning_rate": 9.977198822785184e-05, "loss": 0.1415, "step": 7890 }, { "grad_norm": 1.0089335441589355, "learning_rate": 9.977040822486273e-05, "loss": 0.1105, "step": 7900 }, { "grad_norm": 0.6231544017791748, "learning_rate": 9.97688227790566e-05, "loss": 0.1361, "step": 7910 }, { "grad_norm": 1.0489126443862915, "learning_rate": 9.976723189060684e-05, "loss": 0.1054, "step": 7920 }, { "grad_norm": 0.667710542678833, "learning_rate": 9.976563555968742e-05, "loss": 0.1112, "step": 7930 }, { "grad_norm": 0.821643054485321, "learning_rate": 9.976403378647292e-05, "loss": 0.1252, "step": 7940 }, { "grad_norm": 0.6671634912490845, "learning_rate": 9.97624265711385e-05, "loss": 0.1286, "step": 7950 }, { "grad_norm": 0.6929706335067749, "learning_rate": 9.976081391385993e-05, "loss": 0.0973, "step": 7960 }, { "grad_norm": 0.6921641826629639, "learning_rate": 9.975919581481356e-05, "loss": 0.1446, "step": 7970 }, { "grad_norm": 1.2783989906311035, "learning_rate": 9.975757227417634e-05, "loss": 0.1275, "step": 7980 }, { "grad_norm": 0.876500129699707, "learning_rate": 9.975594329212586e-05, "loss": 0.1302, "step": 7990 }, { "grad_norm": 1.2567235231399536, "learning_rate": 9.97543088688402e-05, "loss": 0.1298, "step": 8000 }, { "grad_norm": 0.7292324304580688, "learning_rate": 9.975266900449814e-05, "loss": 0.1083, "step": 8010 }, { "grad_norm": 0.7812943458557129, "learning_rate": 9.975102369927898e-05, "loss": 0.1511, "step": 8020 }, { "grad_norm": 1.4226701259613037, "learning_rate": 9.974937295336269e-05, "loss": 0.1307, "step": 8030 }, { "grad_norm": 0.8751047849655151, "learning_rate": 9.974771676692975e-05, "loss": 0.1144, "step": 8040 }, { "grad_norm": 0.7914405465126038, "learning_rate": 9.974605514016131e-05, "loss": 0.1061, "step": 8050 }, { "grad_norm": 0.8748986721038818, "learning_rate": 9.974438807323907e-05, "loss": 0.1407, "step": 8060 }, { "grad_norm": 0.8427030444145203, "learning_rate": 9.974271556634535e-05, "loss": 0.1063, "step": 8070 }, { "grad_norm": 0.5528704524040222, "learning_rate": 9.974103761966302e-05, "loss": 0.111, "step": 8080 }, { "grad_norm": 0.9034551382064819, "learning_rate": 9.973935423337563e-05, "loss": 0.1281, "step": 8090 }, { "grad_norm": 0.6743155121803284, "learning_rate": 9.973766540766722e-05, "loss": 0.1123, "step": 8100 }, { "grad_norm": 1.0063539743423462, "learning_rate": 9.97359711427225e-05, "loss": 0.1113, "step": 8110 }, { "grad_norm": 0.8039211630821228, "learning_rate": 9.973427143872677e-05, "loss": 0.115, "step": 8120 }, { "grad_norm": 0.8943960666656494, "learning_rate": 9.973256629586589e-05, "loss": 0.119, "step": 8130 }, { "grad_norm": 0.8311063051223755, "learning_rate": 9.973085571432632e-05, "loss": 0.101, "step": 8140 }, { "grad_norm": 0.9274527430534363, "learning_rate": 9.972913969429513e-05, "loss": 0.1231, "step": 8150 }, { "grad_norm": 0.8187047839164734, "learning_rate": 9.972741823596e-05, "loss": 0.1125, "step": 8160 }, { "grad_norm": 0.7950342893600464, "learning_rate": 9.972569133950917e-05, "loss": 0.1093, "step": 8170 }, { "grad_norm": 0.7081670761108398, "learning_rate": 9.972395900513151e-05, "loss": 0.1066, "step": 8180 }, { "grad_norm": 0.9484191536903381, "learning_rate": 9.972222123301645e-05, "loss": 0.1136, "step": 8190 }, { "grad_norm": 0.8930220603942871, "learning_rate": 9.972047802335403e-05, "loss": 0.1053, "step": 8200 }, { "grad_norm": 0.9597578048706055, "learning_rate": 9.971872937633488e-05, "loss": 0.0923, "step": 8210 }, { "grad_norm": 0.9924896955490112, "learning_rate": 9.971697529215024e-05, "loss": 0.1283, "step": 8220 }, { "grad_norm": 1.1419306993484497, "learning_rate": 9.971521577099192e-05, "loss": 0.1218, "step": 8230 }, { "grad_norm": 1.0880029201507568, "learning_rate": 9.971345081305236e-05, "loss": 0.1105, "step": 8240 }, { "grad_norm": 0.6194501519203186, "learning_rate": 9.971168041852456e-05, "loss": 0.1168, "step": 8250 }, { "grad_norm": 1.053040623664856, "learning_rate": 9.970990458760215e-05, "loss": 0.119, "step": 8260 }, { "grad_norm": 0.9146245718002319, "learning_rate": 9.970812332047929e-05, "loss": 0.1301, "step": 8270 }, { "grad_norm": 1.0751770734786987, "learning_rate": 9.97063366173508e-05, "loss": 0.1129, "step": 8280 }, { "grad_norm": 0.7007060050964355, "learning_rate": 9.970454447841207e-05, "loss": 0.1368, "step": 8290 }, { "grad_norm": 0.9576636552810669, "learning_rate": 9.970274690385909e-05, "loss": 0.1326, "step": 8300 }, { "grad_norm": 1.1980698108673096, "learning_rate": 9.970094389388844e-05, "loss": 0.1205, "step": 8310 }, { "grad_norm": 0.8875174522399902, "learning_rate": 9.969913544869728e-05, "loss": 0.1302, "step": 8320 }, { "grad_norm": 0.6989279985427856, "learning_rate": 9.96973215684834e-05, "loss": 0.0977, "step": 8330 }, { "grad_norm": 0.8175203204154968, "learning_rate": 9.969550225344513e-05, "loss": 0.1113, "step": 8340 }, { "grad_norm": 0.8365489840507507, "learning_rate": 9.969367750378147e-05, "loss": 0.1204, "step": 8350 }, { "grad_norm": 0.6685181856155396, "learning_rate": 9.969184731969194e-05, "loss": 0.1102, "step": 8360 }, { "grad_norm": 0.9253824949264526, "learning_rate": 9.96900117013767e-05, "loss": 0.1266, "step": 8370 }, { "grad_norm": 0.9625654816627502, "learning_rate": 9.96881706490365e-05, "loss": 0.1071, "step": 8380 }, { "grad_norm": 0.9283770322799683, "learning_rate": 9.968632416287265e-05, "loss": 0.1185, "step": 8390 }, { "grad_norm": 0.6507605314254761, "learning_rate": 9.96844722430871e-05, "loss": 0.1098, "step": 8400 }, { "grad_norm": 0.7587424516677856, "learning_rate": 9.968261488988235e-05, "loss": 0.1265, "step": 8410 }, { "grad_norm": 0.6908682584762573, "learning_rate": 9.968075210346155e-05, "loss": 0.1317, "step": 8420 }, { "grad_norm": 0.7822103500366211, "learning_rate": 9.967888388402839e-05, "loss": 0.1403, "step": 8430 }, { "grad_norm": 0.9135851860046387, "learning_rate": 9.967701023178717e-05, "loss": 0.1433, "step": 8440 }, { "grad_norm": 1.0175873041152954, "learning_rate": 9.967513114694282e-05, "loss": 0.1403, "step": 8450 }, { "grad_norm": 1.0256367921829224, "learning_rate": 9.967324662970079e-05, "loss": 0.1564, "step": 8460 }, { "grad_norm": 0.8812273144721985, "learning_rate": 9.96713566802672e-05, "loss": 0.1218, "step": 8470 }, { "grad_norm": 0.9217737913131714, "learning_rate": 9.966946129884873e-05, "loss": 0.1346, "step": 8480 }, { "grad_norm": 0.9942677617073059, "learning_rate": 9.966756048565265e-05, "loss": 0.1207, "step": 8490 }, { "grad_norm": 0.9644351601600647, "learning_rate": 9.966565424088681e-05, "loss": 0.1225, "step": 8500 }, { "grad_norm": 0.9890433549880981, "learning_rate": 9.96637425647597e-05, "loss": 0.1333, "step": 8510 }, { "grad_norm": 0.809492290019989, "learning_rate": 9.966182545748038e-05, "loss": 0.1331, "step": 8520 }, { "grad_norm": 0.9218470454216003, "learning_rate": 9.96599029192585e-05, "loss": 0.1225, "step": 8530 }, { "grad_norm": 0.7333460450172424, "learning_rate": 9.965797495030428e-05, "loss": 0.1143, "step": 8540 }, { "grad_norm": 0.7150700688362122, "learning_rate": 9.96560415508286e-05, "loss": 0.1039, "step": 8550 }, { "grad_norm": 0.8001885414123535, "learning_rate": 9.965410272104286e-05, "loss": 0.1078, "step": 8560 }, { "grad_norm": 0.8262882828712463, "learning_rate": 9.96521584611591e-05, "loss": 0.1242, "step": 8570 }, { "grad_norm": 0.8902608156204224, "learning_rate": 9.965020877138994e-05, "loss": 0.1191, "step": 8580 }, { "grad_norm": 1.035404086112976, "learning_rate": 9.964825365194861e-05, "loss": 0.1091, "step": 8590 }, { "grad_norm": 1.0957155227661133, "learning_rate": 9.96462931030489e-05, "loss": 0.1015, "step": 8600 }, { "grad_norm": 1.6257140636444092, "learning_rate": 9.96443271249052e-05, "loss": 0.1295, "step": 8610 }, { "grad_norm": 0.9859310388565063, "learning_rate": 9.964235571773255e-05, "loss": 0.1315, "step": 8620 }, { "grad_norm": 1.1626089811325073, "learning_rate": 9.96403788817465e-05, "loss": 0.1671, "step": 8630 }, { "grad_norm": 0.9248374104499817, "learning_rate": 9.963839661716325e-05, "loss": 0.1123, "step": 8640 }, { "grad_norm": 0.7718961238861084, "learning_rate": 9.963640892419958e-05, "loss": 0.099, "step": 8650 }, { "grad_norm": 0.7701563239097595, "learning_rate": 9.963441580307286e-05, "loss": 0.1099, "step": 8660 }, { "grad_norm": 1.0225598812103271, "learning_rate": 9.963241725400104e-05, "loss": 0.1022, "step": 8670 }, { "grad_norm": 0.8177255988121033, "learning_rate": 9.963041327720271e-05, "loss": 0.1024, "step": 8680 }, { "grad_norm": 0.600337028503418, "learning_rate": 9.962840387289697e-05, "loss": 0.1019, "step": 8690 }, { "grad_norm": 0.5974397659301758, "learning_rate": 9.962638904130363e-05, "loss": 0.0858, "step": 8700 }, { "grad_norm": 0.8456661701202393, "learning_rate": 9.962436878264298e-05, "loss": 0.0988, "step": 8710 }, { "grad_norm": 0.6406697630882263, "learning_rate": 9.962234309713598e-05, "loss": 0.0902, "step": 8720 }, { "grad_norm": 0.6793524622917175, "learning_rate": 9.962031198500414e-05, "loss": 0.1055, "step": 8730 }, { "grad_norm": 0.7036952376365662, "learning_rate": 9.961827544646958e-05, "loss": 0.1089, "step": 8740 }, { "grad_norm": 0.7797527313232422, "learning_rate": 9.961623348175501e-05, "loss": 0.1006, "step": 8750 }, { "grad_norm": 0.9858496785163879, "learning_rate": 9.961418609108377e-05, "loss": 0.1146, "step": 8760 }, { "grad_norm": 0.826545774936676, "learning_rate": 9.961213327467971e-05, "loss": 0.1141, "step": 8770 }, { "grad_norm": 1.040236234664917, "learning_rate": 9.961007503276736e-05, "loss": 0.1109, "step": 8780 }, { "grad_norm": 0.8258951902389526, "learning_rate": 9.960801136557179e-05, "loss": 0.1408, "step": 8790 }, { "grad_norm": 0.8271641135215759, "learning_rate": 9.960594227331866e-05, "loss": 0.1031, "step": 8800 }, { "grad_norm": 0.8391882181167603, "learning_rate": 9.960386775623429e-05, "loss": 0.1111, "step": 8810 }, { "grad_norm": 0.8786490559577942, "learning_rate": 9.96017878145455e-05, "loss": 0.1145, "step": 8820 }, { "grad_norm": 0.825794517993927, "learning_rate": 9.959970244847977e-05, "loss": 0.1328, "step": 8830 }, { "grad_norm": 0.7474356293678284, "learning_rate": 9.959761165826518e-05, "loss": 0.1327, "step": 8840 }, { "grad_norm": 0.6157569289207458, "learning_rate": 9.959551544413033e-05, "loss": 0.1188, "step": 8850 }, { "grad_norm": 0.8513565063476562, "learning_rate": 9.959341380630448e-05, "loss": 0.0986, "step": 8860 }, { "grad_norm": 0.6832506060600281, "learning_rate": 9.959130674501746e-05, "loss": 0.1158, "step": 8870 }, { "grad_norm": 0.6597596406936646, "learning_rate": 9.958919426049968e-05, "loss": 0.103, "step": 8880 }, { "grad_norm": 0.7097010016441345, "learning_rate": 9.958707635298219e-05, "loss": 0.1514, "step": 8890 }, { "grad_norm": 1.0838520526885986, "learning_rate": 9.958495302269657e-05, "loss": 0.1388, "step": 8900 }, { "grad_norm": 0.9324324131011963, "learning_rate": 9.958282426987503e-05, "loss": 0.1413, "step": 8910 }, { "grad_norm": 0.6953748464584351, "learning_rate": 9.95806900947504e-05, "loss": 0.1035, "step": 8920 }, { "grad_norm": 0.6306325197219849, "learning_rate": 9.957855049755604e-05, "loss": 0.1182, "step": 8930 }, { "grad_norm": 0.9754296541213989, "learning_rate": 9.957640547852593e-05, "loss": 0.1018, "step": 8940 }, { "grad_norm": 0.9606737494468689, "learning_rate": 9.957425503789466e-05, "loss": 0.1024, "step": 8950 }, { "grad_norm": 0.9913280010223389, "learning_rate": 9.957209917589738e-05, "loss": 0.1375, "step": 8960 }, { "grad_norm": 0.9987545013427734, "learning_rate": 9.956993789276987e-05, "loss": 0.1312, "step": 8970 }, { "grad_norm": 0.8497615456581116, "learning_rate": 9.956777118874847e-05, "loss": 0.1328, "step": 8980 }, { "grad_norm": 1.0022529363632202, "learning_rate": 9.956559906407016e-05, "loss": 0.129, "step": 8990 }, { "grad_norm": 0.9611091017723083, "learning_rate": 9.956342151897245e-05, "loss": 0.1104, "step": 9000 }, { "grad_norm": 1.4028019905090332, "learning_rate": 9.956123855369346e-05, "loss": 0.1539, "step": 9010 }, { "grad_norm": 0.6995336413383484, "learning_rate": 9.955905016847196e-05, "loss": 0.1362, "step": 9020 }, { "grad_norm": 0.7966244220733643, "learning_rate": 9.955685636354723e-05, "loss": 0.109, "step": 9030 }, { "grad_norm": 0.9920767545700073, "learning_rate": 9.95546571391592e-05, "loss": 0.1281, "step": 9040 }, { "grad_norm": 0.9242304563522339, "learning_rate": 9.955245249554837e-05, "loss": 0.1201, "step": 9050 }, { "grad_norm": 0.9073213934898376, "learning_rate": 9.955024243295582e-05, "loss": 0.1145, "step": 9060 }, { "grad_norm": 0.8173967003822327, "learning_rate": 9.954802695162328e-05, "loss": 0.121, "step": 9070 }, { "grad_norm": 1.0450756549835205, "learning_rate": 9.954580605179302e-05, "loss": 0.1169, "step": 9080 }, { "grad_norm": 0.6220647096633911, "learning_rate": 9.954357973370788e-05, "loss": 0.138, "step": 9090 }, { "grad_norm": 0.7699528932571411, "learning_rate": 9.954134799761135e-05, "loss": 0.1056, "step": 9100 }, { "grad_norm": 0.5776875019073486, "learning_rate": 9.953911084374748e-05, "loss": 0.0879, "step": 9110 }, { "grad_norm": 0.8340128064155579, "learning_rate": 9.953686827236093e-05, "loss": 0.1251, "step": 9120 }, { "grad_norm": 0.8982893228530884, "learning_rate": 9.953462028369695e-05, "loss": 0.121, "step": 9130 }, { "grad_norm": 1.004469633102417, "learning_rate": 9.953236687800136e-05, "loss": 0.1323, "step": 9140 }, { "grad_norm": 0.717751681804657, "learning_rate": 9.95301080555206e-05, "loss": 0.1222, "step": 9150 }, { "grad_norm": 1.0783109664916992, "learning_rate": 9.952784381650171e-05, "loss": 0.1368, "step": 9160 }, { "grad_norm": 1.1833775043487549, "learning_rate": 9.952557416119226e-05, "loss": 0.1259, "step": 9170 }, { "grad_norm": 0.6902458071708679, "learning_rate": 9.95232990898405e-05, "loss": 0.1091, "step": 9180 }, { "grad_norm": 0.966696560382843, "learning_rate": 9.95210186026952e-05, "loss": 0.1281, "step": 9190 }, { "grad_norm": 0.7917421460151672, "learning_rate": 9.951873270000576e-05, "loss": 0.0986, "step": 9200 }, { "grad_norm": 0.8952944278717041, "learning_rate": 9.951644138202216e-05, "loss": 0.1083, "step": 9210 }, { "grad_norm": 0.8255006670951843, "learning_rate": 9.951414464899498e-05, "loss": 0.1128, "step": 9220 }, { "grad_norm": 0.7115268707275391, "learning_rate": 9.951184250117538e-05, "loss": 0.0847, "step": 9230 }, { "grad_norm": 1.1328471899032593, "learning_rate": 9.950953493881513e-05, "loss": 0.1225, "step": 9240 }, { "grad_norm": 0.935404896736145, "learning_rate": 9.950722196216658e-05, "loss": 0.1067, "step": 9250 }, { "grad_norm": 0.9074404239654541, "learning_rate": 9.950490357148265e-05, "loss": 0.1359, "step": 9260 }, { "grad_norm": 0.7132943272590637, "learning_rate": 9.950257976701692e-05, "loss": 0.1081, "step": 9270 }, { "grad_norm": 1.0092017650604248, "learning_rate": 9.950025054902348e-05, "loss": 0.1201, "step": 9280 }, { "grad_norm": 0.6814562678337097, "learning_rate": 9.949791591775706e-05, "loss": 0.1002, "step": 9290 }, { "grad_norm": 0.8323113918304443, "learning_rate": 9.949557587347298e-05, "loss": 0.1081, "step": 9300 }, { "grad_norm": 1.0548475980758667, "learning_rate": 9.949323041642713e-05, "loss": 0.1367, "step": 9310 }, { "grad_norm": 0.7820980548858643, "learning_rate": 9.949087954687602e-05, "loss": 0.1087, "step": 9320 }, { "grad_norm": 1.0507124662399292, "learning_rate": 9.948852326507672e-05, "loss": 0.116, "step": 9330 }, { "grad_norm": 1.460416316986084, "learning_rate": 9.948616157128694e-05, "loss": 0.1157, "step": 9340 }, { "grad_norm": 0.5411182045936584, "learning_rate": 9.948379446576493e-05, "loss": 0.1072, "step": 9350 }, { "grad_norm": 0.8399465084075928, "learning_rate": 9.948142194876952e-05, "loss": 0.1015, "step": 9360 }, { "grad_norm": 0.913709282875061, "learning_rate": 9.947904402056024e-05, "loss": 0.1169, "step": 9370 }, { "grad_norm": 0.8664935827255249, "learning_rate": 9.947666068139708e-05, "loss": 0.1344, "step": 9380 }, { "grad_norm": 0.6325063705444336, "learning_rate": 9.947427193154071e-05, "loss": 0.099, "step": 9390 }, { "grad_norm": 0.8274492621421814, "learning_rate": 9.947187777125233e-05, "loss": 0.1124, "step": 9400 }, { "grad_norm": 0.7565017342567444, "learning_rate": 9.946947820079377e-05, "loss": 0.1093, "step": 9410 }, { "grad_norm": 0.9171491861343384, "learning_rate": 9.946707322042747e-05, "loss": 0.129, "step": 9420 }, { "grad_norm": 0.6362663507461548, "learning_rate": 9.94646628304164e-05, "loss": 0.1013, "step": 9430 }, { "grad_norm": 0.8524993658065796, "learning_rate": 9.946224703102418e-05, "loss": 0.1106, "step": 9440 }, { "grad_norm": 0.6963754892349243, "learning_rate": 9.945982582251498e-05, "loss": 0.094, "step": 9450 }, { "grad_norm": 0.6026002168655396, "learning_rate": 9.94573992051536e-05, "loss": 0.1391, "step": 9460 }, { "grad_norm": 0.6083542704582214, "learning_rate": 9.94549671792054e-05, "loss": 0.1166, "step": 9470 }, { "grad_norm": 0.6920450329780579, "learning_rate": 9.945252974493635e-05, "loss": 0.0978, "step": 9480 }, { "grad_norm": 0.8270047307014465, "learning_rate": 9.9450086902613e-05, "loss": 0.1068, "step": 9490 }, { "grad_norm": 0.6385084986686707, "learning_rate": 9.944763865250248e-05, "loss": 0.0971, "step": 9500 }, { "grad_norm": 0.6813627481460571, "learning_rate": 9.944518499487254e-05, "loss": 0.1206, "step": 9510 }, { "grad_norm": 0.9943249225616455, "learning_rate": 9.944272592999151e-05, "loss": 0.1168, "step": 9520 }, { "grad_norm": 1.0606979131698608, "learning_rate": 9.94402614581283e-05, "loss": 0.1178, "step": 9530 }, { "grad_norm": 0.9927793741226196, "learning_rate": 9.943779157955244e-05, "loss": 0.1296, "step": 9540 }, { "grad_norm": 0.6809148788452148, "learning_rate": 9.943531629453403e-05, "loss": 0.128, "step": 9550 }, { "grad_norm": 1.1182067394256592, "learning_rate": 9.943283560334375e-05, "loss": 0.1239, "step": 9560 }, { "grad_norm": 0.7893150448799133, "learning_rate": 9.943034950625288e-05, "loss": 0.1019, "step": 9570 }, { "grad_norm": 0.6370822191238403, "learning_rate": 9.942785800353332e-05, "loss": 0.1152, "step": 9580 }, { "grad_norm": 0.6762855052947998, "learning_rate": 9.942536109545751e-05, "loss": 0.1105, "step": 9590 }, { "grad_norm": 0.8808674812316895, "learning_rate": 9.942285878229853e-05, "loss": 0.1221, "step": 9600 }, { "grad_norm": 0.98894864320755, "learning_rate": 9.942035106433001e-05, "loss": 0.092, "step": 9610 }, { "grad_norm": 0.699821412563324, "learning_rate": 9.94178379418262e-05, "loss": 0.1043, "step": 9620 }, { "grad_norm": 0.9244725704193115, "learning_rate": 9.941531941506194e-05, "loss": 0.1034, "step": 9630 }, { "grad_norm": 0.7531698942184448, "learning_rate": 9.941279548431263e-05, "loss": 0.1364, "step": 9640 }, { "grad_norm": 1.0779188871383667, "learning_rate": 9.941026614985431e-05, "loss": 0.1102, "step": 9650 }, { "grad_norm": 0.8780747056007385, "learning_rate": 9.940773141196357e-05, "loss": 0.1032, "step": 9660 }, { "grad_norm": 0.6962528824806213, "learning_rate": 9.94051912709176e-05, "loss": 0.1167, "step": 9670 }, { "grad_norm": 0.7080633044242859, "learning_rate": 9.940264572699421e-05, "loss": 0.1127, "step": 9680 }, { "grad_norm": 0.8929478526115417, "learning_rate": 9.940009478047174e-05, "loss": 0.1125, "step": 9690 }, { "grad_norm": 0.7362813949584961, "learning_rate": 9.939753843162918e-05, "loss": 0.1058, "step": 9700 }, { "grad_norm": 0.677379846572876, "learning_rate": 9.939497668074609e-05, "loss": 0.0959, "step": 9710 }, { "grad_norm": 0.7314031720161438, "learning_rate": 9.93924095281026e-05, "loss": 0.0835, "step": 9720 }, { "grad_norm": 0.769951343536377, "learning_rate": 9.938983697397948e-05, "loss": 0.1113, "step": 9730 }, { "grad_norm": 0.6747838258743286, "learning_rate": 9.938725901865805e-05, "loss": 0.1039, "step": 9740 }, { "grad_norm": 0.8101316094398499, "learning_rate": 9.93846756624202e-05, "loss": 0.1005, "step": 9750 }, { "grad_norm": 0.8609674572944641, "learning_rate": 9.938208690554849e-05, "loss": 0.1133, "step": 9760 }, { "grad_norm": 0.6639533638954163, "learning_rate": 9.9379492748326e-05, "loss": 0.14, "step": 9770 }, { "grad_norm": 0.5951584577560425, "learning_rate": 9.937689319103641e-05, "loss": 0.1284, "step": 9780 }, { "grad_norm": 0.9504291415214539, "learning_rate": 9.937428823396404e-05, "loss": 0.1083, "step": 9790 }, { "grad_norm": 0.8335028290748596, "learning_rate": 9.937167787739372e-05, "loss": 0.091, "step": 9800 }, { "grad_norm": 0.9401015639305115, "learning_rate": 9.936906212161095e-05, "loss": 0.0871, "step": 9810 }, { "grad_norm": 0.6437135934829712, "learning_rate": 9.936644096690176e-05, "loss": 0.1143, "step": 9820 }, { "grad_norm": 1.1333564519882202, "learning_rate": 9.936381441355282e-05, "loss": 0.1298, "step": 9830 }, { "grad_norm": 0.8322449922561646, "learning_rate": 9.936118246185136e-05, "loss": 0.0989, "step": 9840 }, { "grad_norm": 0.6789301633834839, "learning_rate": 9.935854511208518e-05, "loss": 0.1109, "step": 9850 }, { "grad_norm": 0.8170152902603149, "learning_rate": 9.935590236454272e-05, "loss": 0.0988, "step": 9860 }, { "grad_norm": 1.0543527603149414, "learning_rate": 9.935325421951298e-05, "loss": 0.1038, "step": 9870 }, { "grad_norm": 0.9093135595321655, "learning_rate": 9.935060067728557e-05, "loss": 0.1056, "step": 9880 }, { "grad_norm": 0.9441105723381042, "learning_rate": 9.934794173815067e-05, "loss": 0.1193, "step": 9890 }, { "grad_norm": 0.6440089344978333, "learning_rate": 9.934527740239906e-05, "loss": 0.1091, "step": 9900 }, { "grad_norm": 0.7600393295288086, "learning_rate": 9.934260767032209e-05, "loss": 0.114, "step": 9910 }, { "grad_norm": 0.6521167755126953, "learning_rate": 9.933993254221172e-05, "loss": 0.1022, "step": 9920 }, { "grad_norm": 1.0835453271865845, "learning_rate": 9.933725201836053e-05, "loss": 0.1271, "step": 9930 }, { "grad_norm": 0.8052130937576294, "learning_rate": 9.933456609906162e-05, "loss": 0.1179, "step": 9940 }, { "grad_norm": 1.1114721298217773, "learning_rate": 9.933187478460875e-05, "loss": 0.1174, "step": 9950 }, { "grad_norm": 0.9625930190086365, "learning_rate": 9.93291780752962e-05, "loss": 0.125, "step": 9960 }, { "grad_norm": 0.8014451265335083, "learning_rate": 9.932647597141893e-05, "loss": 0.1016, "step": 9970 }, { "grad_norm": 0.6915233731269836, "learning_rate": 9.932376847327239e-05, "loss": 0.1179, "step": 9980 }, { "grad_norm": 0.6553946137428284, "learning_rate": 9.932105558115268e-05, "loss": 0.106, "step": 9990 }, { "grad_norm": 0.8297827243804932, "learning_rate": 9.931833729535651e-05, "loss": 0.1322, "step": 10000 }, { "grad_norm": 0.7766807675361633, "learning_rate": 9.931561361618111e-05, "loss": 0.1139, "step": 10010 }, { "grad_norm": 0.9668408632278442, "learning_rate": 9.931288454392435e-05, "loss": 0.1062, "step": 10020 }, { "grad_norm": 0.6777482032775879, "learning_rate": 9.931015007888467e-05, "loss": 0.1009, "step": 10030 }, { "grad_norm": 0.905738115310669, "learning_rate": 9.930741022136112e-05, "loss": 0.1137, "step": 10040 }, { "grad_norm": 1.0544724464416504, "learning_rate": 9.930466497165333e-05, "loss": 0.1156, "step": 10050 }, { "grad_norm": 0.7115098237991333, "learning_rate": 9.93019143300615e-05, "loss": 0.0869, "step": 10060 }, { "grad_norm": 0.9802873134613037, "learning_rate": 9.929915829688644e-05, "loss": 0.1031, "step": 10070 }, { "grad_norm": 1.0952880382537842, "learning_rate": 9.929639687242955e-05, "loss": 0.1241, "step": 10080 }, { "grad_norm": 1.2039471864700317, "learning_rate": 9.929363005699281e-05, "loss": 0.1379, "step": 10090 }, { "grad_norm": 1.1688437461853027, "learning_rate": 9.92908578508788e-05, "loss": 0.1205, "step": 10100 }, { "grad_norm": 0.6894754767417908, "learning_rate": 9.928808025439069e-05, "loss": 0.1092, "step": 10110 }, { "grad_norm": 0.5535531044006348, "learning_rate": 9.928529726783223e-05, "loss": 0.0939, "step": 10120 }, { "grad_norm": 0.647793173789978, "learning_rate": 9.928250889150774e-05, "loss": 0.1406, "step": 10130 }, { "grad_norm": 0.7100672721862793, "learning_rate": 9.92797151257222e-05, "loss": 0.121, "step": 10140 }, { "grad_norm": 0.8981602787971497, "learning_rate": 9.927691597078108e-05, "loss": 0.1475, "step": 10150 }, { "grad_norm": 1.0397835969924927, "learning_rate": 9.927411142699053e-05, "loss": 0.1146, "step": 10160 }, { "grad_norm": 0.5844125747680664, "learning_rate": 9.927130149465725e-05, "loss": 0.0932, "step": 10170 }, { "grad_norm": 1.0043891668319702, "learning_rate": 9.92684861740885e-05, "loss": 0.097, "step": 10180 }, { "grad_norm": 0.5954045653343201, "learning_rate": 9.926566546559217e-05, "loss": 0.0867, "step": 10190 }, { "grad_norm": 0.9364611506462097, "learning_rate": 9.926283936947673e-05, "loss": 0.1102, "step": 10200 }, { "grad_norm": 0.7732373476028442, "learning_rate": 9.926000788605126e-05, "loss": 0.0984, "step": 10210 }, { "grad_norm": 0.705832302570343, "learning_rate": 9.92571710156254e-05, "loss": 0.1303, "step": 10220 }, { "grad_norm": 0.6213142275810242, "learning_rate": 9.925432875850936e-05, "loss": 0.1226, "step": 10230 }, { "grad_norm": 1.0460277795791626, "learning_rate": 9.925148111501396e-05, "loss": 0.129, "step": 10240 }, { "grad_norm": 0.8085958957672119, "learning_rate": 9.924862808545066e-05, "loss": 0.117, "step": 10250 }, { "grad_norm": 0.7191464304924011, "learning_rate": 9.924576967013141e-05, "loss": 0.1247, "step": 10260 }, { "grad_norm": 0.9280139803886414, "learning_rate": 9.924290586936887e-05, "loss": 0.1271, "step": 10270 }, { "grad_norm": 0.7381138801574707, "learning_rate": 9.924003668347614e-05, "loss": 0.1021, "step": 10280 }, { "grad_norm": 0.7698318958282471, "learning_rate": 9.923716211276704e-05, "loss": 0.1035, "step": 10290 }, { "grad_norm": 1.0531901121139526, "learning_rate": 9.923428215755594e-05, "loss": 0.0985, "step": 10300 }, { "grad_norm": 0.5507748126983643, "learning_rate": 9.923139681815775e-05, "loss": 0.1198, "step": 10310 }, { "grad_norm": 0.715359628200531, "learning_rate": 9.922850609488801e-05, "loss": 0.1026, "step": 10320 }, { "grad_norm": 0.7362277507781982, "learning_rate": 9.922560998806287e-05, "loss": 0.1184, "step": 10330 }, { "grad_norm": 0.9235870242118835, "learning_rate": 9.922270849799905e-05, "loss": 0.1062, "step": 10340 }, { "grad_norm": 0.5344908833503723, "learning_rate": 9.92198016250138e-05, "loss": 0.0918, "step": 10350 }, { "grad_norm": 0.7305009961128235, "learning_rate": 9.921688936942506e-05, "loss": 0.1039, "step": 10360 }, { "grad_norm": 1.090278148651123, "learning_rate": 9.921397173155129e-05, "loss": 0.1285, "step": 10370 }, { "grad_norm": 0.6358738541603088, "learning_rate": 9.921104871171157e-05, "loss": 0.1423, "step": 10380 }, { "grad_norm": 0.655921459197998, "learning_rate": 9.920812031022554e-05, "loss": 0.1002, "step": 10390 }, { "grad_norm": 1.1791774034500122, "learning_rate": 9.920518652741348e-05, "loss": 0.1092, "step": 10400 }, { "grad_norm": 0.7088575959205627, "learning_rate": 9.920224736359618e-05, "loss": 0.108, "step": 10410 }, { "grad_norm": 0.6420704126358032, "learning_rate": 9.91993028190951e-05, "loss": 0.113, "step": 10420 }, { "grad_norm": 0.6575040817260742, "learning_rate": 9.919635289423222e-05, "loss": 0.0928, "step": 10430 }, { "grad_norm": 0.7740283012390137, "learning_rate": 9.919339758933015e-05, "loss": 0.1046, "step": 10440 }, { "grad_norm": 0.643922746181488, "learning_rate": 9.919043690471209e-05, "loss": 0.0834, "step": 10450 }, { "grad_norm": 0.7416548728942871, "learning_rate": 9.91874708407018e-05, "loss": 0.1047, "step": 10460 }, { "grad_norm": 0.9293580651283264, "learning_rate": 9.918449939762367e-05, "loss": 0.0987, "step": 10470 }, { "grad_norm": 1.2742640972137451, "learning_rate": 9.91815225758026e-05, "loss": 0.0991, "step": 10480 }, { "grad_norm": 0.5630201697349548, "learning_rate": 9.917854037556419e-05, "loss": 0.1335, "step": 10490 }, { "grad_norm": 0.7415151596069336, "learning_rate": 9.917555279723454e-05, "loss": 0.1052, "step": 10500 }, { "grad_norm": 0.9488153457641602, "learning_rate": 9.917255984114036e-05, "loss": 0.09, "step": 10510 }, { "grad_norm": 1.2433329820632935, "learning_rate": 9.916956150760896e-05, "loss": 0.1101, "step": 10520 }, { "grad_norm": 0.8105249404907227, "learning_rate": 9.916655779696826e-05, "loss": 0.1206, "step": 10530 }, { "grad_norm": 0.8586105704307556, "learning_rate": 9.916354870954671e-05, "loss": 0.0985, "step": 10540 }, { "grad_norm": 1.163618803024292, "learning_rate": 9.91605342456734e-05, "loss": 0.112, "step": 10550 }, { "grad_norm": 0.6808671951293945, "learning_rate": 9.915751440567795e-05, "loss": 0.1095, "step": 10560 }, { "grad_norm": 0.8006519675254822, "learning_rate": 9.915448918989066e-05, "loss": 0.1471, "step": 10570 }, { "grad_norm": 0.6400066614151001, "learning_rate": 9.915145859864232e-05, "loss": 0.0971, "step": 10580 }, { "grad_norm": 0.6741430163383484, "learning_rate": 9.914842263226437e-05, "loss": 0.1106, "step": 10590 }, { "grad_norm": 0.7583491802215576, "learning_rate": 9.914538129108882e-05, "loss": 0.0946, "step": 10600 }, { "grad_norm": 0.7345970869064331, "learning_rate": 9.914233457544825e-05, "loss": 0.0926, "step": 10610 }, { "grad_norm": 0.6916726231575012, "learning_rate": 9.913928248567586e-05, "loss": 0.0889, "step": 10620 }, { "grad_norm": 0.8399041891098022, "learning_rate": 9.913622502210542e-05, "loss": 0.105, "step": 10630 }, { "grad_norm": 0.6436733603477478, "learning_rate": 9.913316218507128e-05, "loss": 0.0898, "step": 10640 }, { "grad_norm": 0.8945137858390808, "learning_rate": 9.91300939749084e-05, "loss": 0.1339, "step": 10650 }, { "grad_norm": 0.8370975255966187, "learning_rate": 9.91270203919523e-05, "loss": 0.1053, "step": 10660 }, { "grad_norm": 0.7070419192314148, "learning_rate": 9.912394143653912e-05, "loss": 0.102, "step": 10670 }, { "grad_norm": 0.8072584867477417, "learning_rate": 9.912085710900555e-05, "loss": 0.1016, "step": 10680 }, { "grad_norm": 0.7843619585037231, "learning_rate": 9.911776740968892e-05, "loss": 0.0983, "step": 10690 }, { "grad_norm": 0.5779896378517151, "learning_rate": 9.911467233892709e-05, "loss": 0.0914, "step": 10700 }, { "grad_norm": 0.9329432249069214, "learning_rate": 9.911157189705853e-05, "loss": 0.0963, "step": 10710 }, { "grad_norm": 0.7310795187950134, "learning_rate": 9.910846608442229e-05, "loss": 0.1054, "step": 10720 }, { "grad_norm": 0.5494190454483032, "learning_rate": 9.910535490135805e-05, "loss": 0.0893, "step": 10730 }, { "grad_norm": 0.9407819509506226, "learning_rate": 9.910223834820603e-05, "loss": 0.0997, "step": 10740 }, { "grad_norm": 0.8676794171333313, "learning_rate": 9.909911642530703e-05, "loss": 0.0967, "step": 10750 }, { "grad_norm": 0.7446309328079224, "learning_rate": 9.909598913300249e-05, "loss": 0.0904, "step": 10760 }, { "grad_norm": 0.6637923717498779, "learning_rate": 9.909285647163438e-05, "loss": 0.096, "step": 10770 }, { "grad_norm": 0.9395173788070679, "learning_rate": 9.908971844154531e-05, "loss": 0.1025, "step": 10780 }, { "grad_norm": 1.0279070138931274, "learning_rate": 9.908657504307843e-05, "loss": 0.1074, "step": 10790 }, { "grad_norm": 0.6915583610534668, "learning_rate": 9.908342627657751e-05, "loss": 0.0971, "step": 10800 }, { "grad_norm": 0.9946667551994324, "learning_rate": 9.908027214238689e-05, "loss": 0.099, "step": 10810 }, { "grad_norm": 0.6741175055503845, "learning_rate": 9.90771126408515e-05, "loss": 0.0882, "step": 10820 }, { "grad_norm": 0.7071918845176697, "learning_rate": 9.907394777231685e-05, "loss": 0.1084, "step": 10830 }, { "grad_norm": 0.5088467597961426, "learning_rate": 9.907077753712905e-05, "loss": 0.0976, "step": 10840 }, { "grad_norm": 0.6619904637336731, "learning_rate": 9.906760193563482e-05, "loss": 0.1018, "step": 10850 }, { "grad_norm": 0.7793307304382324, "learning_rate": 9.906442096818139e-05, "loss": 0.112, "step": 10860 }, { "grad_norm": 0.7494401931762695, "learning_rate": 9.906123463511665e-05, "loss": 0.1372, "step": 10870 }, { "grad_norm": 0.7428287863731384, "learning_rate": 9.905804293678907e-05, "loss": 0.1175, "step": 10880 }, { "grad_norm": 0.6637319922447205, "learning_rate": 9.905484587354766e-05, "loss": 0.0745, "step": 10890 }, { "grad_norm": 0.7459954619407654, "learning_rate": 9.905164344574205e-05, "loss": 0.1005, "step": 10900 }, { "grad_norm": 0.5535476207733154, "learning_rate": 9.904843565372248e-05, "loss": 0.09, "step": 10910 }, { "grad_norm": 0.6194445490837097, "learning_rate": 9.904522249783972e-05, "loss": 0.0735, "step": 10920 }, { "grad_norm": 0.5185580849647522, "learning_rate": 9.904200397844517e-05, "loss": 0.1114, "step": 10930 }, { "grad_norm": 0.8524879217147827, "learning_rate": 9.903878009589078e-05, "loss": 0.094, "step": 10940 }, { "grad_norm": 0.8749186396598816, "learning_rate": 9.903555085052915e-05, "loss": 0.0916, "step": 10950 }, { "grad_norm": 0.8282402753829956, "learning_rate": 9.903231624271338e-05, "loss": 0.101, "step": 10960 }, { "grad_norm": 0.8759211301803589, "learning_rate": 9.902907627279724e-05, "loss": 0.0988, "step": 10970 }, { "grad_norm": 0.9082358479499817, "learning_rate": 9.902583094113504e-05, "loss": 0.0842, "step": 10980 }, { "grad_norm": 1.1047359704971313, "learning_rate": 9.902258024808168e-05, "loss": 0.0976, "step": 10990 }, { "grad_norm": 0.7921352982521057, "learning_rate": 9.901932419399264e-05, "loss": 0.0857, "step": 11000 }, { "grad_norm": 0.6859539747238159, "learning_rate": 9.9016062779224e-05, "loss": 0.109, "step": 11010 }, { "grad_norm": 0.7887077927589417, "learning_rate": 9.901279600413242e-05, "loss": 0.1087, "step": 11020 }, { "grad_norm": 0.9380043745040894, "learning_rate": 9.900952386907518e-05, "loss": 0.0964, "step": 11030 }, { "grad_norm": 0.6691905856132507, "learning_rate": 9.90062463744101e-05, "loss": 0.0827, "step": 11040 }, { "grad_norm": 0.8367983102798462, "learning_rate": 9.900296352049558e-05, "loss": 0.1111, "step": 11050 }, { "grad_norm": 1.0677865743637085, "learning_rate": 9.899967530769065e-05, "loss": 0.0905, "step": 11060 }, { "grad_norm": 0.9110828638076782, "learning_rate": 9.899638173635489e-05, "loss": 0.0838, "step": 11070 }, { "grad_norm": 0.6959742307662964, "learning_rate": 9.899308280684849e-05, "loss": 0.0983, "step": 11080 }, { "grad_norm": 1.0589993000030518, "learning_rate": 9.898977851953222e-05, "loss": 0.0983, "step": 11090 }, { "grad_norm": 0.7420686483383179, "learning_rate": 9.898646887476741e-05, "loss": 0.0946, "step": 11100 }, { "grad_norm": 0.8830799460411072, "learning_rate": 9.898315387291603e-05, "loss": 0.1018, "step": 11110 }, { "grad_norm": 0.8991459608078003, "learning_rate": 9.89798335143406e-05, "loss": 0.0906, "step": 11120 }, { "grad_norm": 0.7523704767227173, "learning_rate": 9.897650779940419e-05, "loss": 0.1238, "step": 11130 }, { "grad_norm": 0.6186733245849609, "learning_rate": 9.897317672847054e-05, "loss": 0.1001, "step": 11140 }, { "grad_norm": 0.6643883585929871, "learning_rate": 9.89698403019039e-05, "loss": 0.1209, "step": 11150 }, { "grad_norm": 0.9965341687202454, "learning_rate": 9.896649852006917e-05, "loss": 0.1168, "step": 11160 }, { "grad_norm": 0.700030505657196, "learning_rate": 9.896315138333177e-05, "loss": 0.0945, "step": 11170 }, { "grad_norm": 0.5982533097267151, "learning_rate": 9.895979889205774e-05, "loss": 0.099, "step": 11180 }, { "grad_norm": 0.6651837229728699, "learning_rate": 9.895644104661372e-05, "loss": 0.1014, "step": 11190 }, { "grad_norm": 0.5859765410423279, "learning_rate": 9.895307784736691e-05, "loss": 0.112, "step": 11200 }, { "grad_norm": 0.6469642519950867, "learning_rate": 9.894970929468512e-05, "loss": 0.1027, "step": 11210 }, { "grad_norm": 0.8701629042625427, "learning_rate": 9.89463353889367e-05, "loss": 0.0926, "step": 11220 }, { "grad_norm": 0.6070181131362915, "learning_rate": 9.894295613049065e-05, "loss": 0.1024, "step": 11230 }, { "grad_norm": 0.7547380924224854, "learning_rate": 9.893957151971649e-05, "loss": 0.1044, "step": 11240 }, { "grad_norm": 0.7874503135681152, "learning_rate": 9.893618155698436e-05, "loss": 0.0956, "step": 11250 }, { "grad_norm": 0.5118104815483093, "learning_rate": 9.8932786242665e-05, "loss": 0.0773, "step": 11260 }, { "grad_norm": 0.8542701005935669, "learning_rate": 9.89293855771297e-05, "loss": 0.0887, "step": 11270 }, { "grad_norm": 0.620192289352417, "learning_rate": 9.892597956075036e-05, "loss": 0.0891, "step": 11280 }, { "grad_norm": 0.7679996490478516, "learning_rate": 9.892256819389947e-05, "loss": 0.1018, "step": 11290 }, { "grad_norm": 0.5836160182952881, "learning_rate": 9.891915147695006e-05, "loss": 0.1181, "step": 11300 }, { "grad_norm": 0.5364449620246887, "learning_rate": 9.891572941027577e-05, "loss": 0.112, "step": 11310 }, { "grad_norm": 0.8890255689620972, "learning_rate": 9.89123019942509e-05, "loss": 0.1113, "step": 11320 }, { "grad_norm": 0.6142236590385437, "learning_rate": 9.89088692292502e-05, "loss": 0.098, "step": 11330 }, { "grad_norm": 1.0652470588684082, "learning_rate": 9.89054311156491e-05, "loss": 0.1102, "step": 11340 }, { "grad_norm": 0.5502697825431824, "learning_rate": 9.890198765382357e-05, "loss": 0.0894, "step": 11350 }, { "grad_norm": 0.7560873031616211, "learning_rate": 9.889853884415021e-05, "loss": 0.1226, "step": 11360 }, { "grad_norm": 0.7698215246200562, "learning_rate": 9.889508468700614e-05, "loss": 0.0881, "step": 11370 }, { "grad_norm": 0.6082625389099121, "learning_rate": 9.889162518276915e-05, "loss": 0.1175, "step": 11380 }, { "grad_norm": 0.6279037594795227, "learning_rate": 9.888816033181752e-05, "loss": 0.0934, "step": 11390 }, { "grad_norm": 0.9692256450653076, "learning_rate": 9.888469013453018e-05, "loss": 0.1141, "step": 11400 }, { "grad_norm": 0.8189468383789062, "learning_rate": 9.888121459128663e-05, "loss": 0.1016, "step": 11410 }, { "grad_norm": 0.9740139842033386, "learning_rate": 9.887773370246693e-05, "loss": 0.1111, "step": 11420 }, { "grad_norm": 1.1415021419525146, "learning_rate": 9.887424746845177e-05, "loss": 0.0993, "step": 11430 }, { "grad_norm": 0.7763781547546387, "learning_rate": 9.887075588962239e-05, "loss": 0.1047, "step": 11440 }, { "grad_norm": 0.9258647561073303, "learning_rate": 9.88672589663606e-05, "loss": 0.1053, "step": 11450 }, { "grad_norm": 1.0224695205688477, "learning_rate": 9.886375669904886e-05, "loss": 0.0962, "step": 11460 }, { "grad_norm": 0.7597795724868774, "learning_rate": 9.886024908807014e-05, "loss": 0.1247, "step": 11470 }, { "grad_norm": 0.607671320438385, "learning_rate": 9.885673613380806e-05, "loss": 0.1233, "step": 11480 }, { "grad_norm": 0.6625048518180847, "learning_rate": 9.885321783664676e-05, "loss": 0.1073, "step": 11490 }, { "grad_norm": 0.6354547739028931, "learning_rate": 9.884969419697101e-05, "loss": 0.0835, "step": 11500 }, { "grad_norm": 0.768662691116333, "learning_rate": 9.884616521516614e-05, "loss": 0.0918, "step": 11510 }, { "grad_norm": 0.8628070950508118, "learning_rate": 9.88426308916181e-05, "loss": 0.1077, "step": 11520 }, { "grad_norm": 0.8378647565841675, "learning_rate": 9.883909122671335e-05, "loss": 0.087, "step": 11530 }, { "grad_norm": 0.8500966429710388, "learning_rate": 9.883554622083904e-05, "loss": 0.1249, "step": 11540 }, { "grad_norm": 0.8968406915664673, "learning_rate": 9.88319958743828e-05, "loss": 0.1318, "step": 11550 }, { "grad_norm": 0.7039662599563599, "learning_rate": 9.882844018773291e-05, "loss": 0.1053, "step": 11560 }, { "grad_norm": 0.8162010312080383, "learning_rate": 9.882487916127823e-05, "loss": 0.0893, "step": 11570 }, { "grad_norm": 0.9386875629425049, "learning_rate": 9.882131279540815e-05, "loss": 0.0979, "step": 11580 }, { "grad_norm": 0.8303813338279724, "learning_rate": 9.881774109051271e-05, "loss": 0.1242, "step": 11590 }, { "grad_norm": 0.761923611164093, "learning_rate": 9.881416404698252e-05, "loss": 0.0953, "step": 11600 }, { "grad_norm": 0.6349790096282959, "learning_rate": 9.881058166520873e-05, "loss": 0.1145, "step": 11610 }, { "grad_norm": 0.9565579891204834, "learning_rate": 9.880699394558311e-05, "loss": 0.0932, "step": 11620 }, { "grad_norm": 0.8947707414627075, "learning_rate": 9.880340088849801e-05, "loss": 0.1066, "step": 11630 }, { "grad_norm": 0.7884804010391235, "learning_rate": 9.879980249434637e-05, "loss": 0.106, "step": 11640 }, { "grad_norm": 0.8516453504562378, "learning_rate": 9.879619876352168e-05, "loss": 0.1199, "step": 11650 }, { "grad_norm": 0.8626405596733093, "learning_rate": 9.879258969641809e-05, "loss": 0.1156, "step": 11660 }, { "grad_norm": 0.897604763507843, "learning_rate": 9.878897529343023e-05, "loss": 0.109, "step": 11670 }, { "grad_norm": 0.6417223811149597, "learning_rate": 9.878535555495338e-05, "loss": 0.1038, "step": 11680 }, { "grad_norm": 0.8090003132820129, "learning_rate": 9.87817304813834e-05, "loss": 0.1087, "step": 11690 }, { "grad_norm": 0.5688220858573914, "learning_rate": 9.877810007311671e-05, "loss": 0.1009, "step": 11700 }, { "grad_norm": 1.2031370401382446, "learning_rate": 9.877446433055035e-05, "loss": 0.1459, "step": 11710 }, { "grad_norm": 0.7500813007354736, "learning_rate": 9.877082325408191e-05, "loss": 0.103, "step": 11720 }, { "grad_norm": 1.0570425987243652, "learning_rate": 9.876717684410954e-05, "loss": 0.1292, "step": 11730 }, { "grad_norm": 0.716878354549408, "learning_rate": 9.876352510103204e-05, "loss": 0.0985, "step": 11740 }, { "grad_norm": 0.9309397339820862, "learning_rate": 9.875986802524875e-05, "loss": 0.1283, "step": 11750 }, { "grad_norm": 0.7756755948066711, "learning_rate": 9.87562056171596e-05, "loss": 0.1133, "step": 11760 }, { "grad_norm": 0.771512508392334, "learning_rate": 9.875253787716511e-05, "loss": 0.1224, "step": 11770 }, { "grad_norm": 0.6915346384048462, "learning_rate": 9.874886480566637e-05, "loss": 0.0953, "step": 11780 }, { "grad_norm": 0.5761597156524658, "learning_rate": 9.874518640306507e-05, "loss": 0.1033, "step": 11790 }, { "grad_norm": 0.48933690786361694, "learning_rate": 9.874150266976347e-05, "loss": 0.1341, "step": 11800 }, { "grad_norm": 1.0401759147644043, "learning_rate": 9.873781360616443e-05, "loss": 0.1175, "step": 11810 }, { "grad_norm": 1.0268508195877075, "learning_rate": 9.873411921267137e-05, "loss": 0.116, "step": 11820 }, { "grad_norm": 1.4850107431411743, "learning_rate": 9.873041948968829e-05, "loss": 0.0967, "step": 11830 }, { "grad_norm": 0.7734977602958679, "learning_rate": 9.872671443761981e-05, "loss": 0.0941, "step": 11840 }, { "grad_norm": 0.8308189511299133, "learning_rate": 9.872300405687109e-05, "loss": 0.0934, "step": 11850 }, { "grad_norm": 0.707778811454773, "learning_rate": 9.871928834784792e-05, "loss": 0.0951, "step": 11860 }, { "grad_norm": 0.6220777034759521, "learning_rate": 9.871556731095661e-05, "loss": 0.0957, "step": 11870 }, { "grad_norm": 0.6277380585670471, "learning_rate": 9.871184094660411e-05, "loss": 0.1222, "step": 11880 }, { "grad_norm": 1.0463205575942993, "learning_rate": 9.870810925519791e-05, "loss": 0.106, "step": 11890 }, { "grad_norm": 0.7776855230331421, "learning_rate": 9.870437223714612e-05, "loss": 0.1027, "step": 11900 }, { "grad_norm": 0.7695932984352112, "learning_rate": 9.87006298928574e-05, "loss": 0.09, "step": 11910 }, { "grad_norm": 0.5748677253723145, "learning_rate": 9.869688222274103e-05, "loss": 0.0893, "step": 11920 }, { "grad_norm": 0.7910195589065552, "learning_rate": 9.869312922720681e-05, "loss": 0.087, "step": 11930 }, { "grad_norm": 1.3296513557434082, "learning_rate": 9.868937090666521e-05, "loss": 0.107, "step": 11940 }, { "grad_norm": 0.717099130153656, "learning_rate": 9.86856072615272e-05, "loss": 0.0963, "step": 11950 }, { "grad_norm": 0.6635069847106934, "learning_rate": 9.868183829220438e-05, "loss": 0.1079, "step": 11960 }, { "grad_norm": 0.9356710314750671, "learning_rate": 9.867806399910893e-05, "loss": 0.1072, "step": 11970 }, { "grad_norm": 0.7303187847137451, "learning_rate": 9.867428438265356e-05, "loss": 0.1365, "step": 11980 }, { "grad_norm": 0.8523892164230347, "learning_rate": 9.867049944325165e-05, "loss": 0.1099, "step": 11990 }, { "grad_norm": 0.7076554894447327, "learning_rate": 9.86667091813171e-05, "loss": 0.1459, "step": 12000 }, { "grad_norm": 0.896543562412262, "learning_rate": 9.866291359726438e-05, "loss": 0.1056, "step": 12010 }, { "grad_norm": 0.7553380131721497, "learning_rate": 9.865911269150861e-05, "loss": 0.0985, "step": 12020 }, { "grad_norm": 0.5901393294334412, "learning_rate": 9.865530646446544e-05, "loss": 0.0898, "step": 12030 }, { "grad_norm": 0.686752200126648, "learning_rate": 9.86514949165511e-05, "loss": 0.0991, "step": 12040 }, { "grad_norm": 0.8983293175697327, "learning_rate": 9.864767804818243e-05, "loss": 0.1069, "step": 12050 }, { "grad_norm": 0.6381819844245911, "learning_rate": 9.86438558597768e-05, "loss": 0.104, "step": 12060 }, { "grad_norm": 0.9097678661346436, "learning_rate": 9.864002835175225e-05, "loss": 0.115, "step": 12070 }, { "grad_norm": 0.7547919750213623, "learning_rate": 9.863619552452734e-05, "loss": 0.1135, "step": 12080 }, { "grad_norm": 0.9759066104888916, "learning_rate": 9.863235737852119e-05, "loss": 0.1248, "step": 12090 }, { "grad_norm": 0.7237095832824707, "learning_rate": 9.862851391415356e-05, "loss": 0.0981, "step": 12100 }, { "grad_norm": 0.8263908624649048, "learning_rate": 9.862466513184477e-05, "loss": 0.1013, "step": 12110 }, { "grad_norm": 0.8551293611526489, "learning_rate": 9.86208110320157e-05, "loss": 0.0761, "step": 12120 }, { "grad_norm": 0.7624576091766357, "learning_rate": 9.861695161508784e-05, "loss": 0.1062, "step": 12130 }, { "grad_norm": 0.8580217361450195, "learning_rate": 9.861308688148324e-05, "loss": 0.1049, "step": 12140 }, { "grad_norm": 1.452221393585205, "learning_rate": 9.860921683162455e-05, "loss": 0.1184, "step": 12150 }, { "grad_norm": 0.8015516996383667, "learning_rate": 9.860534146593499e-05, "loss": 0.1077, "step": 12160 }, { "grad_norm": 0.6821070909500122, "learning_rate": 9.860146078483836e-05, "loss": 0.1126, "step": 12170 }, { "grad_norm": 0.6696502566337585, "learning_rate": 9.859757478875905e-05, "loss": 0.0995, "step": 12180 }, { "grad_norm": 0.8201799988746643, "learning_rate": 9.859368347812204e-05, "loss": 0.1115, "step": 12190 }, { "grad_norm": 0.5965662002563477, "learning_rate": 9.858978685335285e-05, "loss": 0.1136, "step": 12200 }, { "grad_norm": 0.78339684009552, "learning_rate": 9.858588491487763e-05, "loss": 0.0724, "step": 12210 }, { "grad_norm": 1.0489147901535034, "learning_rate": 9.858197766312308e-05, "loss": 0.0924, "step": 12220 }, { "grad_norm": 0.707145631313324, "learning_rate": 9.857806509851649e-05, "loss": 0.0733, "step": 12230 }, { "grad_norm": 0.6027532815933228, "learning_rate": 9.857414722148574e-05, "loss": 0.102, "step": 12240 }, { "grad_norm": 0.9791891574859619, "learning_rate": 9.857022403245928e-05, "loss": 0.09, "step": 12250 }, { "grad_norm": 0.6536150574684143, "learning_rate": 9.856629553186615e-05, "loss": 0.0833, "step": 12260 }, { "grad_norm": 0.726550281047821, "learning_rate": 9.856236172013595e-05, "loss": 0.1007, "step": 12270 }, { "grad_norm": 0.6691674590110779, "learning_rate": 9.85584225976989e-05, "loss": 0.1002, "step": 12280 }, { "grad_norm": 0.6683257818222046, "learning_rate": 9.855447816498575e-05, "loss": 0.1022, "step": 12290 }, { "grad_norm": 1.0858943462371826, "learning_rate": 9.855052842242787e-05, "loss": 0.1223, "step": 12300 }, { "grad_norm": 0.7920944094657898, "learning_rate": 9.85465733704572e-05, "loss": 0.1142, "step": 12310 }, { "grad_norm": 0.8981068730354309, "learning_rate": 9.854261300950624e-05, "loss": 0.1285, "step": 12320 }, { "grad_norm": 0.9248080253601074, "learning_rate": 9.853864734000813e-05, "loss": 0.1377, "step": 12330 }, { "grad_norm": 0.716264545917511, "learning_rate": 9.85346763623965e-05, "loss": 0.1181, "step": 12340 }, { "grad_norm": 0.7805996537208557, "learning_rate": 9.853070007710564e-05, "loss": 0.0994, "step": 12350 }, { "grad_norm": 0.6968328356742859, "learning_rate": 9.85267184845704e-05, "loss": 0.0968, "step": 12360 }, { "grad_norm": 1.0558916330337524, "learning_rate": 9.852273158522616e-05, "loss": 0.1162, "step": 12370 }, { "grad_norm": 0.7724587917327881, "learning_rate": 9.851873937950896e-05, "loss": 0.1084, "step": 12380 }, { "grad_norm": 0.6956670880317688, "learning_rate": 9.851474186785537e-05, "loss": 0.1065, "step": 12390 }, { "grad_norm": 0.7041724324226379, "learning_rate": 9.851073905070254e-05, "loss": 0.0772, "step": 12400 }, { "grad_norm": 0.6787941455841064, "learning_rate": 9.850673092848824e-05, "loss": 0.1091, "step": 12410 }, { "grad_norm": 0.9220882654190063, "learning_rate": 9.850271750165077e-05, "loss": 0.1099, "step": 12420 }, { "grad_norm": 0.6685525178909302, "learning_rate": 9.849869877062902e-05, "loss": 0.0869, "step": 12430 }, { "grad_norm": 0.9737496376037598, "learning_rate": 9.849467473586252e-05, "loss": 0.0901, "step": 12440 }, { "grad_norm": 0.7535803914070129, "learning_rate": 9.849064539779127e-05, "loss": 0.1001, "step": 12450 }, { "grad_norm": 1.0574328899383545, "learning_rate": 9.848661075685594e-05, "loss": 0.0908, "step": 12460 }, { "grad_norm": 0.7143069505691528, "learning_rate": 9.848257081349778e-05, "loss": 0.1145, "step": 12470 }, { "grad_norm": 0.8009154200553894, "learning_rate": 9.847852556815856e-05, "loss": 0.1048, "step": 12480 }, { "grad_norm": 0.8379195928573608, "learning_rate": 9.847447502128067e-05, "loss": 0.1076, "step": 12490 }, { "grad_norm": 0.6458595991134644, "learning_rate": 9.847041917330708e-05, "loss": 0.1127, "step": 12500 }, { "grad_norm": 0.4746374189853668, "learning_rate": 9.846635802468132e-05, "loss": 0.0784, "step": 12510 }, { "grad_norm": 0.6983257532119751, "learning_rate": 9.84622915758475e-05, "loss": 0.0782, "step": 12520 }, { "grad_norm": 0.8930191397666931, "learning_rate": 9.845821982725034e-05, "loss": 0.0873, "step": 12530 }, { "grad_norm": 0.7405622005462646, "learning_rate": 9.845414277933514e-05, "loss": 0.1015, "step": 12540 }, { "grad_norm": 0.6087504625320435, "learning_rate": 9.845006043254771e-05, "loss": 0.0885, "step": 12550 }, { "grad_norm": 0.7969506978988647, "learning_rate": 9.844597278733451e-05, "loss": 0.0938, "step": 12560 }, { "grad_norm": 0.8438699841499329, "learning_rate": 9.844187984414259e-05, "loss": 0.1024, "step": 12570 }, { "grad_norm": 0.7294280529022217, "learning_rate": 9.84377816034195e-05, "loss": 0.1074, "step": 12580 }, { "grad_norm": 0.7337543368339539, "learning_rate": 9.843367806561345e-05, "loss": 0.1092, "step": 12590 }, { "grad_norm": 0.6927781701087952, "learning_rate": 9.842956923117317e-05, "loss": 0.0853, "step": 12600 }, { "grad_norm": 0.5361444354057312, "learning_rate": 9.842545510054802e-05, "loss": 0.0884, "step": 12610 }, { "grad_norm": 0.5695033669471741, "learning_rate": 9.842133567418792e-05, "loss": 0.0824, "step": 12620 }, { "grad_norm": 0.588502824306488, "learning_rate": 9.841721095254333e-05, "loss": 0.1031, "step": 12630 }, { "grad_norm": 1.1801046133041382, "learning_rate": 9.841308093606537e-05, "loss": 0.0967, "step": 12640 }, { "grad_norm": 1.1532955169677734, "learning_rate": 9.840894562520565e-05, "loss": 0.1117, "step": 12650 }, { "grad_norm": 0.7851071953773499, "learning_rate": 9.840480502041642e-05, "loss": 0.0818, "step": 12660 }, { "grad_norm": 1.1663470268249512, "learning_rate": 9.840065912215049e-05, "loss": 0.1214, "step": 12670 }, { "grad_norm": 1.0086135864257812, "learning_rate": 9.839650793086124e-05, "loss": 0.0913, "step": 12680 }, { "grad_norm": 0.5664210319519043, "learning_rate": 9.839235144700265e-05, "loss": 0.1028, "step": 12690 }, { "grad_norm": 1.0097812414169312, "learning_rate": 9.838818967102926e-05, "loss": 0.0895, "step": 12700 }, { "grad_norm": 0.7684842944145203, "learning_rate": 9.83840226033962e-05, "loss": 0.1024, "step": 12710 }, { "grad_norm": 0.8221323490142822, "learning_rate": 9.837985024455918e-05, "loss": 0.082, "step": 12720 }, { "grad_norm": 0.6864126920700073, "learning_rate": 9.837567259497447e-05, "loss": 0.0832, "step": 12730 }, { "grad_norm": 0.7759485244750977, "learning_rate": 9.837148965509894e-05, "loss": 0.0966, "step": 12740 }, { "grad_norm": 0.8028625845909119, "learning_rate": 9.836730142539001e-05, "loss": 0.0812, "step": 12750 }, { "grad_norm": 0.8802642226219177, "learning_rate": 9.836310790630574e-05, "loss": 0.0854, "step": 12760 }, { "grad_norm": 0.8390129208564758, "learning_rate": 9.83589090983047e-05, "loss": 0.1085, "step": 12770 }, { "grad_norm": 0.8807097673416138, "learning_rate": 9.835470500184605e-05, "loss": 0.1148, "step": 12780 }, { "grad_norm": 0.7123662233352661, "learning_rate": 9.835049561738957e-05, "loss": 0.0833, "step": 12790 }, { "grad_norm": 0.6927765011787415, "learning_rate": 9.834628094539558e-05, "loss": 0.1217, "step": 12800 }, { "grad_norm": 0.6757659912109375, "learning_rate": 9.834206098632499e-05, "loss": 0.0847, "step": 12810 }, { "grad_norm": 0.8390081524848938, "learning_rate": 9.833783574063931e-05, "loss": 0.0934, "step": 12820 }, { "grad_norm": 0.7238256335258484, "learning_rate": 9.833360520880058e-05, "loss": 0.1136, "step": 12830 }, { "grad_norm": 0.6030123233795166, "learning_rate": 9.832936939127144e-05, "loss": 0.0783, "step": 12840 }, { "grad_norm": 0.8074654936790466, "learning_rate": 9.832512828851515e-05, "loss": 0.0797, "step": 12850 }, { "grad_norm": 0.6746381521224976, "learning_rate": 9.832088190099546e-05, "loss": 0.0903, "step": 12860 }, { "grad_norm": 0.8319671750068665, "learning_rate": 9.831663022917679e-05, "loss": 0.0966, "step": 12870 }, { "grad_norm": 0.5225164294242859, "learning_rate": 9.831237327352407e-05, "loss": 0.0878, "step": 12880 }, { "grad_norm": 0.5451517701148987, "learning_rate": 9.830811103450286e-05, "loss": 0.0853, "step": 12890 }, { "grad_norm": 0.8617101907730103, "learning_rate": 9.830384351257924e-05, "loss": 0.1027, "step": 12900 }, { "grad_norm": 0.638081431388855, "learning_rate": 9.829957070821993e-05, "loss": 0.0926, "step": 12910 }, { "grad_norm": 0.49073347449302673, "learning_rate": 9.829529262189218e-05, "loss": 0.0973, "step": 12920 }, { "grad_norm": 0.7758685946464539, "learning_rate": 9.829100925406385e-05, "loss": 0.0949, "step": 12930 }, { "grad_norm": 0.8501649498939514, "learning_rate": 9.828672060520333e-05, "loss": 0.0919, "step": 12940 }, { "grad_norm": 0.4671817719936371, "learning_rate": 9.828242667577966e-05, "loss": 0.0877, "step": 12950 }, { "grad_norm": 0.700238049030304, "learning_rate": 9.82781274662624e-05, "loss": 0.0901, "step": 12960 }, { "grad_norm": 0.7102072834968567, "learning_rate": 9.82738229771217e-05, "loss": 0.0921, "step": 12970 }, { "grad_norm": 0.8152122497558594, "learning_rate": 9.826951320882829e-05, "loss": 0.1167, "step": 12980 }, { "grad_norm": 0.5497161149978638, "learning_rate": 9.826519816185351e-05, "loss": 0.0873, "step": 12990 }, { "grad_norm": 0.4435287117958069, "learning_rate": 9.826087783666921e-05, "loss": 0.0747, "step": 13000 }, { "grad_norm": 0.6674590706825256, "learning_rate": 9.825655223374787e-05, "loss": 0.0785, "step": 13010 }, { "grad_norm": 0.7763223648071289, "learning_rate": 9.825222135356253e-05, "loss": 0.1, "step": 13020 }, { "grad_norm": 0.6874516010284424, "learning_rate": 9.82478851965868e-05, "loss": 0.0997, "step": 13030 }, { "grad_norm": 0.9316011071205139, "learning_rate": 9.82435437632949e-05, "loss": 0.0987, "step": 13040 }, { "grad_norm": 0.8067481517791748, "learning_rate": 9.823919705416158e-05, "loss": 0.0893, "step": 13050 }, { "grad_norm": 0.7190793752670288, "learning_rate": 9.82348450696622e-05, "loss": 0.0839, "step": 13060 }, { "grad_norm": 0.7835870385169983, "learning_rate": 9.823048781027268e-05, "loss": 0.1015, "step": 13070 }, { "grad_norm": 0.6270321011543274, "learning_rate": 9.822612527646953e-05, "loss": 0.0865, "step": 13080 }, { "grad_norm": 0.9686453342437744, "learning_rate": 9.822175746872984e-05, "loss": 0.0882, "step": 13090 }, { "grad_norm": 0.7497661709785461, "learning_rate": 9.821738438753123e-05, "loss": 0.0806, "step": 13100 }, { "grad_norm": 0.9079952836036682, "learning_rate": 9.821300603335196e-05, "loss": 0.1039, "step": 13110 }, { "grad_norm": 0.6707354784011841, "learning_rate": 9.820862240667085e-05, "loss": 0.0949, "step": 13120 }, { "grad_norm": 0.8777199387550354, "learning_rate": 9.820423350796726e-05, "loss": 0.1017, "step": 13130 }, { "grad_norm": 0.6140292882919312, "learning_rate": 9.819983933772118e-05, "loss": 0.1173, "step": 13140 }, { "grad_norm": 0.7949194312095642, "learning_rate": 9.819543989641314e-05, "loss": 0.0848, "step": 13150 }, { "grad_norm": 0.7778784036636353, "learning_rate": 9.819103518452423e-05, "loss": 0.0884, "step": 13160 }, { "grad_norm": 0.8946903347969055, "learning_rate": 9.818662520253618e-05, "loss": 0.1087, "step": 13170 }, { "grad_norm": 0.6688480973243713, "learning_rate": 9.818220995093126e-05, "loss": 0.1138, "step": 13180 }, { "grad_norm": 1.0933171510696411, "learning_rate": 9.817778943019228e-05, "loss": 0.1023, "step": 13190 }, { "grad_norm": 0.6086375117301941, "learning_rate": 9.81733636408027e-05, "loss": 0.1029, "step": 13200 }, { "grad_norm": 0.5307818055152893, "learning_rate": 9.816893258324649e-05, "loss": 0.097, "step": 13210 }, { "grad_norm": 0.6608577370643616, "learning_rate": 9.816449625800823e-05, "loss": 0.1051, "step": 13220 }, { "grad_norm": 1.0327222347259521, "learning_rate": 9.816005466557308e-05, "loss": 0.0912, "step": 13230 }, { "grad_norm": 0.9993724226951599, "learning_rate": 9.815560780642674e-05, "loss": 0.0766, "step": 13240 }, { "grad_norm": 0.6955808997154236, "learning_rate": 9.815115568105555e-05, "loss": 0.0957, "step": 13250 }, { "grad_norm": 0.9139559864997864, "learning_rate": 9.814669828994638e-05, "loss": 0.0834, "step": 13260 }, { "grad_norm": 0.6451410055160522, "learning_rate": 9.814223563358665e-05, "loss": 0.0735, "step": 13270 }, { "grad_norm": 0.5708401203155518, "learning_rate": 9.813776771246443e-05, "loss": 0.0862, "step": 13280 }, { "grad_norm": 0.6422336101531982, "learning_rate": 9.813329452706829e-05, "loss": 0.0821, "step": 13290 }, { "grad_norm": 0.6409863829612732, "learning_rate": 9.812881607788744e-05, "loss": 0.12, "step": 13300 }, { "grad_norm": 0.7015243768692017, "learning_rate": 9.812433236541163e-05, "loss": 0.091, "step": 13310 }, { "grad_norm": 0.7381022572517395, "learning_rate": 9.811984339013116e-05, "loss": 0.0836, "step": 13320 }, { "grad_norm": 0.6604214906692505, "learning_rate": 9.811534915253698e-05, "loss": 0.0864, "step": 13330 }, { "grad_norm": 0.5981302261352539, "learning_rate": 9.811084965312056e-05, "loss": 0.0762, "step": 13340 }, { "grad_norm": 0.7292526364326477, "learning_rate": 9.810634489237396e-05, "loss": 0.1081, "step": 13350 }, { "grad_norm": 0.6028213500976562, "learning_rate": 9.81018348707898e-05, "loss": 0.0954, "step": 13360 }, { "grad_norm": 0.7686637043952942, "learning_rate": 9.809731958886131e-05, "loss": 0.103, "step": 13370 }, { "grad_norm": 0.5400387048721313, "learning_rate": 9.809279904708224e-05, "loss": 0.0902, "step": 13380 }, { "grad_norm": 0.5722435712814331, "learning_rate": 9.808827324594699e-05, "loss": 0.0885, "step": 13390 }, { "grad_norm": 0.6572107672691345, "learning_rate": 9.808374218595046e-05, "loss": 0.0969, "step": 13400 }, { "grad_norm": 0.6157988905906677, "learning_rate": 9.80792058675882e-05, "loss": 0.0995, "step": 13410 }, { "grad_norm": 1.0347334146499634, "learning_rate": 9.807466429135627e-05, "loss": 0.1018, "step": 13420 }, { "grad_norm": 0.6262720227241516, "learning_rate": 9.807011745775132e-05, "loss": 0.1103, "step": 13430 }, { "grad_norm": 0.7824971079826355, "learning_rate": 9.806556536727061e-05, "loss": 0.0942, "step": 13440 }, { "grad_norm": 0.6562498807907104, "learning_rate": 9.806100802041193e-05, "loss": 0.0846, "step": 13450 }, { "grad_norm": 0.7569496631622314, "learning_rate": 9.805644541767368e-05, "loss": 0.0998, "step": 13460 }, { "grad_norm": 1.263023853302002, "learning_rate": 9.805187755955478e-05, "loss": 0.1362, "step": 13470 }, { "grad_norm": 0.6535708904266357, "learning_rate": 9.804730444655483e-05, "loss": 0.0917, "step": 13480 }, { "grad_norm": 0.7588937282562256, "learning_rate": 9.804272607917388e-05, "loss": 0.081, "step": 13490 }, { "grad_norm": 0.7620049118995667, "learning_rate": 9.803814245791265e-05, "loss": 0.0809, "step": 13500 }, { "grad_norm": 0.795418918132782, "learning_rate": 9.803355358327239e-05, "loss": 0.0894, "step": 13510 }, { "grad_norm": 0.7063590884208679, "learning_rate": 9.802895945575492e-05, "loss": 0.0986, "step": 13520 }, { "grad_norm": 0.7876195907592773, "learning_rate": 9.802436007586266e-05, "loss": 0.0948, "step": 13530 }, { "grad_norm": 0.7564787864685059, "learning_rate": 9.801975544409858e-05, "loss": 0.0695, "step": 13540 }, { "grad_norm": 1.1554003953933716, "learning_rate": 9.801514556096625e-05, "loss": 0.1021, "step": 13550 }, { "grad_norm": 0.7451618313789368, "learning_rate": 9.801053042696977e-05, "loss": 0.1027, "step": 13560 }, { "grad_norm": 0.6802493929862976, "learning_rate": 9.800591004261388e-05, "loss": 0.0884, "step": 13570 }, { "grad_norm": 0.5353127717971802, "learning_rate": 9.800128440840385e-05, "loss": 0.0782, "step": 13580 }, { "grad_norm": 0.6877938508987427, "learning_rate": 9.799665352484552e-05, "loss": 0.0847, "step": 13590 }, { "grad_norm": 0.791941225528717, "learning_rate": 9.799201739244532e-05, "loss": 0.1253, "step": 13600 }, { "grad_norm": 0.7019888162612915, "learning_rate": 9.798737601171025e-05, "loss": 0.0754, "step": 13610 }, { "grad_norm": 0.7820659279823303, "learning_rate": 9.79827293831479e-05, "loss": 0.0797, "step": 13620 }, { "grad_norm": 0.8592237234115601, "learning_rate": 9.797807750726638e-05, "loss": 0.0751, "step": 13630 }, { "grad_norm": 0.812949001789093, "learning_rate": 9.797342038457446e-05, "loss": 0.1005, "step": 13640 }, { "grad_norm": 0.7160244584083557, "learning_rate": 9.796875801558141e-05, "loss": 0.083, "step": 13650 }, { "grad_norm": 0.7743610739707947, "learning_rate": 9.79640904007971e-05, "loss": 0.0803, "step": 13660 }, { "grad_norm": 0.6093307137489319, "learning_rate": 9.795941754073199e-05, "loss": 0.1088, "step": 13670 }, { "grad_norm": 0.7842146754264832, "learning_rate": 9.795473943589705e-05, "loss": 0.0939, "step": 13680 }, { "grad_norm": 1.098933458328247, "learning_rate": 9.795005608680394e-05, "loss": 0.1026, "step": 13690 }, { "grad_norm": 1.207817554473877, "learning_rate": 9.794536749396477e-05, "loss": 0.0953, "step": 13700 }, { "grad_norm": 0.9271324276924133, "learning_rate": 9.79406736578923e-05, "loss": 0.1013, "step": 13710 }, { "grad_norm": 0.7941787838935852, "learning_rate": 9.793597457909984e-05, "loss": 0.0724, "step": 13720 }, { "grad_norm": 0.6633531451225281, "learning_rate": 9.793127025810127e-05, "loss": 0.1275, "step": 13730 }, { "grad_norm": 1.0042647123336792, "learning_rate": 9.792656069541104e-05, "loss": 0.1011, "step": 13740 }, { "grad_norm": 0.8199745416641235, "learning_rate": 9.79218458915442e-05, "loss": 0.0919, "step": 13750 }, { "grad_norm": 0.8827410936355591, "learning_rate": 9.791712584701634e-05, "loss": 0.1218, "step": 13760 }, { "grad_norm": 0.9247709512710571, "learning_rate": 9.791240056234364e-05, "loss": 0.1059, "step": 13770 }, { "grad_norm": 0.6830347180366516, "learning_rate": 9.790767003804283e-05, "loss": 0.0951, "step": 13780 }, { "grad_norm": 0.632236659526825, "learning_rate": 9.790293427463126e-05, "loss": 0.0989, "step": 13790 }, { "grad_norm": 0.8703123331069946, "learning_rate": 9.789819327262684e-05, "loss": 0.1253, "step": 13800 }, { "grad_norm": 0.6749424934387207, "learning_rate": 9.7893447032548e-05, "loss": 0.1018, "step": 13810 }, { "grad_norm": 0.6394045948982239, "learning_rate": 9.78886955549138e-05, "loss": 0.0994, "step": 13820 }, { "grad_norm": 0.7327377200126648, "learning_rate": 9.788393884024387e-05, "loss": 0.1005, "step": 13830 }, { "grad_norm": 0.655753493309021, "learning_rate": 9.787917688905836e-05, "loss": 0.1085, "step": 13840 }, { "grad_norm": 0.954761803150177, "learning_rate": 9.787440970187807e-05, "loss": 0.097, "step": 13850 }, { "grad_norm": 1.0181405544281006, "learning_rate": 9.786963727922429e-05, "loss": 0.0972, "step": 13860 }, { "grad_norm": 0.9422725439071655, "learning_rate": 9.786485962161897e-05, "loss": 0.0896, "step": 13870 }, { "grad_norm": 0.6877397298812866, "learning_rate": 9.786007672958455e-05, "loss": 0.0705, "step": 13880 }, { "grad_norm": 0.8890173435211182, "learning_rate": 9.78552886036441e-05, "loss": 0.0731, "step": 13890 }, { "grad_norm": 0.7549246549606323, "learning_rate": 9.785049524432124e-05, "loss": 0.0753, "step": 13900 }, { "grad_norm": 0.8692585229873657, "learning_rate": 9.784569665214016e-05, "loss": 0.1086, "step": 13910 }, { "grad_norm": 0.7869105935096741, "learning_rate": 9.784089282762563e-05, "loss": 0.0893, "step": 13920 }, { "grad_norm": 0.4656275510787964, "learning_rate": 9.7836083771303e-05, "loss": 0.0968, "step": 13930 }, { "grad_norm": 0.6939019560813904, "learning_rate": 9.783126948369817e-05, "loss": 0.0899, "step": 13940 }, { "grad_norm": 0.6785926222801208, "learning_rate": 9.78264499653376e-05, "loss": 0.0764, "step": 13950 }, { "grad_norm": 0.9625083208084106, "learning_rate": 9.782162521674838e-05, "loss": 0.1254, "step": 13960 }, { "grad_norm": 1.1585646867752075, "learning_rate": 9.781679523845812e-05, "loss": 0.0773, "step": 13970 }, { "grad_norm": 0.8543418645858765, "learning_rate": 9.781196003099502e-05, "loss": 0.0866, "step": 13980 }, { "grad_norm": 1.1404697895050049, "learning_rate": 9.780711959488786e-05, "loss": 0.1121, "step": 13990 }, { "grad_norm": 0.787520706653595, "learning_rate": 9.780227393066599e-05, "loss": 0.0958, "step": 14000 }, { "grad_norm": 0.6832594275474548, "learning_rate": 9.77974230388593e-05, "loss": 0.0983, "step": 14010 }, { "grad_norm": 0.964546799659729, "learning_rate": 9.779256691999829e-05, "loss": 0.1261, "step": 14020 }, { "grad_norm": 0.6835670471191406, "learning_rate": 9.778770557461403e-05, "loss": 0.0881, "step": 14030 }, { "grad_norm": 0.7891148328781128, "learning_rate": 9.778283900323812e-05, "loss": 0.0914, "step": 14040 }, { "grad_norm": 0.632830798625946, "learning_rate": 9.777796720640277e-05, "loss": 0.137, "step": 14050 }, { "grad_norm": 0.7452606558799744, "learning_rate": 9.777309018464078e-05, "loss": 0.0982, "step": 14060 }, { "grad_norm": 0.8482068777084351, "learning_rate": 9.776820793848547e-05, "loss": 0.077, "step": 14070 }, { "grad_norm": 0.8735761046409607, "learning_rate": 9.776332046847075e-05, "loss": 0.0942, "step": 14080 }, { "grad_norm": 0.6125011444091797, "learning_rate": 9.775842777513111e-05, "loss": 0.0778, "step": 14090 }, { "grad_norm": 0.5569607019424438, "learning_rate": 9.775352985900163e-05, "loss": 0.0865, "step": 14100 }, { "grad_norm": 0.7017377614974976, "learning_rate": 9.774862672061791e-05, "loss": 0.0769, "step": 14110 }, { "grad_norm": 0.672940731048584, "learning_rate": 9.774371836051616e-05, "loss": 0.0872, "step": 14120 }, { "grad_norm": 0.7954258322715759, "learning_rate": 9.773880477923315e-05, "loss": 0.0848, "step": 14130 }, { "grad_norm": 0.7738628387451172, "learning_rate": 9.773388597730623e-05, "loss": 0.0854, "step": 14140 }, { "grad_norm": 0.7556672096252441, "learning_rate": 9.77289619552733e-05, "loss": 0.1047, "step": 14150 }, { "grad_norm": 0.5185940861701965, "learning_rate": 9.772403271367285e-05, "loss": 0.1022, "step": 14160 }, { "grad_norm": 0.8286740183830261, "learning_rate": 9.771909825304396e-05, "loss": 0.1064, "step": 14170 }, { "grad_norm": 1.0280590057373047, "learning_rate": 9.771415857392619e-05, "loss": 0.1138, "step": 14180 }, { "grad_norm": 1.0760300159454346, "learning_rate": 9.770921367685978e-05, "loss": 0.0906, "step": 14190 }, { "grad_norm": 1.0332658290863037, "learning_rate": 9.770426356238551e-05, "loss": 0.096, "step": 14200 }, { "grad_norm": 0.7728298902511597, "learning_rate": 9.769930823104469e-05, "loss": 0.1107, "step": 14210 }, { "grad_norm": 0.6033651828765869, "learning_rate": 9.769434768337926e-05, "loss": 0.0868, "step": 14220 }, { "grad_norm": 0.9126642346382141, "learning_rate": 9.768938191993164e-05, "loss": 0.0771, "step": 14230 }, { "grad_norm": 0.9075986742973328, "learning_rate": 9.768441094124494e-05, "loss": 0.0751, "step": 14240 }, { "grad_norm": 0.7574477195739746, "learning_rate": 9.767943474786275e-05, "loss": 0.0784, "step": 14250 }, { "grad_norm": 0.6810993552207947, "learning_rate": 9.767445334032923e-05, "loss": 0.0782, "step": 14260 }, { "grad_norm": 0.7724669575691223, "learning_rate": 9.766946671918919e-05, "loss": 0.0868, "step": 14270 }, { "grad_norm": 0.6759118437767029, "learning_rate": 9.766447488498796e-05, "loss": 0.0963, "step": 14280 }, { "grad_norm": 0.738978922367096, "learning_rate": 9.765947783827139e-05, "loss": 0.1043, "step": 14290 }, { "grad_norm": 0.8376222252845764, "learning_rate": 9.765447557958599e-05, "loss": 0.1113, "step": 14300 }, { "grad_norm": 0.6272358894348145, "learning_rate": 9.764946810947879e-05, "loss": 0.0881, "step": 14310 }, { "grad_norm": 1.2075897455215454, "learning_rate": 9.764445542849738e-05, "loss": 0.1231, "step": 14320 }, { "grad_norm": 1.2999303340911865, "learning_rate": 9.763943753718998e-05, "loss": 0.1047, "step": 14330 }, { "grad_norm": 0.7986828684806824, "learning_rate": 9.76344144361053e-05, "loss": 0.0933, "step": 14340 }, { "grad_norm": 0.5831336975097656, "learning_rate": 9.762938612579269e-05, "loss": 0.0707, "step": 14350 }, { "grad_norm": 0.8723888993263245, "learning_rate": 9.762435260680202e-05, "loss": 0.0951, "step": 14360 }, { "grad_norm": 1.0068252086639404, "learning_rate": 9.761931387968373e-05, "loss": 0.1086, "step": 14370 }, { "grad_norm": 0.8590936660766602, "learning_rate": 9.76142699449889e-05, "loss": 0.0918, "step": 14380 }, { "grad_norm": 1.039564609527588, "learning_rate": 9.760922080326908e-05, "loss": 0.0821, "step": 14390 }, { "grad_norm": 0.9650319814682007, "learning_rate": 9.760416645507644e-05, "loss": 0.0782, "step": 14400 }, { "grad_norm": 1.0618482828140259, "learning_rate": 9.759910690096375e-05, "loss": 0.0735, "step": 14410 }, { "grad_norm": 0.7881273627281189, "learning_rate": 9.759404214148429e-05, "loss": 0.0985, "step": 14420 }, { "grad_norm": 0.8882729411125183, "learning_rate": 9.758897217719191e-05, "loss": 0.0913, "step": 14430 }, { "grad_norm": 0.6640774607658386, "learning_rate": 9.758389700864113e-05, "loss": 0.1074, "step": 14440 }, { "grad_norm": 0.6839725971221924, "learning_rate": 9.757881663638688e-05, "loss": 0.0915, "step": 14450 }, { "grad_norm": 0.6271737217903137, "learning_rate": 9.757373106098478e-05, "loss": 0.0843, "step": 14460 }, { "grad_norm": 0.891478419303894, "learning_rate": 9.756864028299097e-05, "loss": 0.0948, "step": 14470 }, { "grad_norm": 0.7414129376411438, "learning_rate": 9.75635443029622e-05, "loss": 0.0939, "step": 14480 }, { "grad_norm": 0.792766809463501, "learning_rate": 9.755844312145572e-05, "loss": 0.1003, "step": 14490 }, { "grad_norm": 0.7106525301933289, "learning_rate": 9.755333673902941e-05, "loss": 0.0902, "step": 14500 }, { "grad_norm": 0.6268448233604431, "learning_rate": 9.75482251562417e-05, "loss": 0.0887, "step": 14510 }, { "grad_norm": 0.8742836713790894, "learning_rate": 9.754310837365155e-05, "loss": 0.0793, "step": 14520 }, { "grad_norm": 0.6784353256225586, "learning_rate": 9.753798639181856e-05, "loss": 0.0872, "step": 14530 }, { "grad_norm": 0.5345059037208557, "learning_rate": 9.753285921130286e-05, "loss": 0.0683, "step": 14540 }, { "grad_norm": 0.9010978937149048, "learning_rate": 9.752772683266512e-05, "loss": 0.0886, "step": 14550 }, { "grad_norm": 0.8202176094055176, "learning_rate": 9.752258925646665e-05, "loss": 0.0848, "step": 14560 }, { "grad_norm": 0.5720468759536743, "learning_rate": 9.751744648326926e-05, "loss": 0.0844, "step": 14570 }, { "grad_norm": 0.8742696046829224, "learning_rate": 9.751229851363536e-05, "loss": 0.0948, "step": 14580 }, { "grad_norm": 0.759870707988739, "learning_rate": 9.750714534812793e-05, "loss": 0.0815, "step": 14590 }, { "grad_norm": 0.8812637329101562, "learning_rate": 9.750198698731053e-05, "loss": 0.0887, "step": 14600 }, { "grad_norm": 0.785050630569458, "learning_rate": 9.749682343174722e-05, "loss": 0.1047, "step": 14610 }, { "grad_norm": 0.586763322353363, "learning_rate": 9.749165468200272e-05, "loss": 0.0988, "step": 14620 }, { "grad_norm": 0.9524194598197937, "learning_rate": 9.748648073864229e-05, "loss": 0.0832, "step": 14630 }, { "grad_norm": 1.3600037097930908, "learning_rate": 9.748130160223168e-05, "loss": 0.0862, "step": 14640 }, { "grad_norm": 0.8185498714447021, "learning_rate": 9.747611727333734e-05, "loss": 0.0842, "step": 14650 }, { "grad_norm": 0.6504607796669006, "learning_rate": 9.74709277525262e-05, "loss": 0.0832, "step": 14660 }, { "grad_norm": 0.8380109071731567, "learning_rate": 9.746573304036576e-05, "loss": 0.0908, "step": 14670 }, { "grad_norm": 0.8722838759422302, "learning_rate": 9.746053313742412e-05, "loss": 0.089, "step": 14680 }, { "grad_norm": 0.6691676378250122, "learning_rate": 9.745532804426994e-05, "loss": 0.075, "step": 14690 }, { "grad_norm": 0.6305601000785828, "learning_rate": 9.745011776147242e-05, "loss": 0.0883, "step": 14700 }, { "grad_norm": 0.7039867639541626, "learning_rate": 9.744490228960138e-05, "loss": 0.0855, "step": 14710 }, { "grad_norm": 0.9128538370132446, "learning_rate": 9.743968162922713e-05, "loss": 0.0952, "step": 14720 }, { "grad_norm": 0.6551442742347717, "learning_rate": 9.743445578092064e-05, "loss": 0.0874, "step": 14730 }, { "grad_norm": 0.7093432545661926, "learning_rate": 9.742922474525338e-05, "loss": 0.1055, "step": 14740 }, { "grad_norm": 0.8544536828994751, "learning_rate": 9.742398852279741e-05, "loss": 0.0766, "step": 14750 }, { "grad_norm": 1.1888675689697266, "learning_rate": 9.741874711412535e-05, "loss": 0.1002, "step": 14760 }, { "grad_norm": 0.5569395422935486, "learning_rate": 9.741350051981042e-05, "loss": 0.0943, "step": 14770 }, { "grad_norm": 0.5675539970397949, "learning_rate": 9.740824874042633e-05, "loss": 0.0896, "step": 14780 }, { "grad_norm": 0.704679548740387, "learning_rate": 9.740299177654746e-05, "loss": 0.0865, "step": 14790 }, { "grad_norm": 0.7617425322532654, "learning_rate": 9.739772962874867e-05, "loss": 0.0993, "step": 14800 }, { "grad_norm": 0.6205788850784302, "learning_rate": 9.739246229760541e-05, "loss": 0.0961, "step": 14810 }, { "grad_norm": 0.6135985255241394, "learning_rate": 9.738718978369376e-05, "loss": 0.0957, "step": 14820 }, { "grad_norm": 0.8110267519950867, "learning_rate": 9.738191208759025e-05, "loss": 0.0948, "step": 14830 }, { "grad_norm": 0.8167597651481628, "learning_rate": 9.73766292098721e-05, "loss": 0.1174, "step": 14840 }, { "grad_norm": 0.7087992429733276, "learning_rate": 9.737134115111699e-05, "loss": 0.1105, "step": 14850 }, { "grad_norm": 0.5194075107574463, "learning_rate": 9.736604791190323e-05, "loss": 0.0889, "step": 14860 }, { "grad_norm": 0.5620258450508118, "learning_rate": 9.73607494928097e-05, "loss": 0.0936, "step": 14870 }, { "grad_norm": 0.46651142835617065, "learning_rate": 9.735544589441581e-05, "loss": 0.083, "step": 14880 }, { "grad_norm": 0.8106685876846313, "learning_rate": 9.735013711730154e-05, "loss": 0.082, "step": 14890 }, { "grad_norm": 0.8012251257896423, "learning_rate": 9.734482316204747e-05, "loss": 0.0831, "step": 14900 }, { "grad_norm": 1.069024682044983, "learning_rate": 9.733950402923473e-05, "loss": 0.1089, "step": 14910 }, { "grad_norm": 0.9361466765403748, "learning_rate": 9.7334179719445e-05, "loss": 0.0932, "step": 14920 }, { "grad_norm": 0.668752133846283, "learning_rate": 9.732885023326053e-05, "loss": 0.1025, "step": 14930 }, { "grad_norm": 0.5285760164260864, "learning_rate": 9.732351557126418e-05, "loss": 0.0966, "step": 14940 }, { "grad_norm": 0.9842762351036072, "learning_rate": 9.731817573403929e-05, "loss": 0.1055, "step": 14950 }, { "grad_norm": 1.01425039768219, "learning_rate": 9.731283072216985e-05, "loss": 0.0954, "step": 14960 }, { "grad_norm": 0.9985899329185486, "learning_rate": 9.730748053624039e-05, "loss": 0.0977, "step": 14970 }, { "grad_norm": 1.1155871152877808, "learning_rate": 9.730212517683598e-05, "loss": 0.1012, "step": 14980 }, { "grad_norm": 0.4851040840148926, "learning_rate": 9.729676464454228e-05, "loss": 0.0871, "step": 14990 }, { "grad_norm": 0.7776557803153992, "learning_rate": 9.72913989399455e-05, "loss": 0.0941, "step": 15000 }, { "grad_norm": 0.7896170020103455, "learning_rate": 9.728602806363242e-05, "loss": 0.0665, "step": 15010 }, { "grad_norm": 0.8393715023994446, "learning_rate": 9.728065201619043e-05, "loss": 0.0847, "step": 15020 }, { "grad_norm": 0.8014973402023315, "learning_rate": 9.727527079820742e-05, "loss": 0.0823, "step": 15030 }, { "grad_norm": 0.5507867932319641, "learning_rate": 9.726988441027186e-05, "loss": 0.0807, "step": 15040 }, { "grad_norm": 0.5697792172431946, "learning_rate": 9.726449285297281e-05, "loss": 0.0657, "step": 15050 }, { "grad_norm": 0.6578269600868225, "learning_rate": 9.72590961268999e-05, "loss": 0.0919, "step": 15060 }, { "grad_norm": 0.5813797116279602, "learning_rate": 9.725369423264328e-05, "loss": 0.0761, "step": 15070 }, { "grad_norm": 1.070248007774353, "learning_rate": 9.72482871707937e-05, "loss": 0.1021, "step": 15080 }, { "grad_norm": 1.121172308921814, "learning_rate": 9.724287494194247e-05, "loss": 0.0859, "step": 15090 }, { "grad_norm": 0.8152540326118469, "learning_rate": 9.723745754668147e-05, "loss": 0.1027, "step": 15100 }, { "grad_norm": 0.8830323815345764, "learning_rate": 9.723203498560313e-05, "loss": 0.0789, "step": 15110 }, { "grad_norm": 0.6277168393135071, "learning_rate": 9.722660725930046e-05, "loss": 0.098, "step": 15120 }, { "grad_norm": 0.9142907857894897, "learning_rate": 9.722117436836702e-05, "loss": 0.1057, "step": 15130 }, { "grad_norm": 0.7146975994110107, "learning_rate": 9.721573631339696e-05, "loss": 0.085, "step": 15140 }, { "grad_norm": 0.7576457858085632, "learning_rate": 9.721029309498494e-05, "loss": 0.0741, "step": 15150 }, { "grad_norm": 0.9228873252868652, "learning_rate": 9.720484471372627e-05, "loss": 0.1183, "step": 15160 }, { "grad_norm": 0.9629100561141968, "learning_rate": 9.719939117021673e-05, "loss": 0.1172, "step": 15170 }, { "grad_norm": 0.7641439437866211, "learning_rate": 9.719393246505275e-05, "loss": 0.0921, "step": 15180 }, { "grad_norm": 0.7151393294334412, "learning_rate": 9.718846859883128e-05, "loss": 0.1279, "step": 15190 }, { "grad_norm": 0.6644872426986694, "learning_rate": 9.718299957214982e-05, "loss": 0.0843, "step": 15200 }, { "grad_norm": 0.9538901448249817, "learning_rate": 9.717752538560646e-05, "loss": 0.112, "step": 15210 }, { "grad_norm": 0.6198346614837646, "learning_rate": 9.717204603979986e-05, "loss": 0.0782, "step": 15220 }, { "grad_norm": 0.6930460929870605, "learning_rate": 9.716656153532922e-05, "loss": 0.0999, "step": 15230 }, { "grad_norm": 0.8066372871398926, "learning_rate": 9.716107187279434e-05, "loss": 0.0782, "step": 15240 }, { "grad_norm": 0.723742663860321, "learning_rate": 9.715557705279555e-05, "loss": 0.0899, "step": 15250 }, { "grad_norm": 0.6872299909591675, "learning_rate": 9.715007707593372e-05, "loss": 0.0903, "step": 15260 }, { "grad_norm": 0.6168155074119568, "learning_rate": 9.714457194281036e-05, "loss": 0.073, "step": 15270 }, { "grad_norm": 0.5143690705299377, "learning_rate": 9.713906165402751e-05, "loss": 0.095, "step": 15280 }, { "grad_norm": 0.9198920130729675, "learning_rate": 9.713354621018774e-05, "loss": 0.0926, "step": 15290 }, { "grad_norm": 0.8038325309753418, "learning_rate": 9.712802561189422e-05, "loss": 0.0768, "step": 15300 }, { "grad_norm": 0.8061556220054626, "learning_rate": 9.712249985975069e-05, "loss": 0.0764, "step": 15310 }, { "grad_norm": 1.0310231447219849, "learning_rate": 9.71169689543614e-05, "loss": 0.0821, "step": 15320 }, { "grad_norm": 0.6428630352020264, "learning_rate": 9.711143289633123e-05, "loss": 0.0925, "step": 15330 }, { "grad_norm": 0.9140827059745789, "learning_rate": 9.710589168626561e-05, "loss": 0.0921, "step": 15340 }, { "grad_norm": 0.527445375919342, "learning_rate": 9.710034532477048e-05, "loss": 0.0802, "step": 15350 }, { "grad_norm": 0.6695477366447449, "learning_rate": 9.709479381245239e-05, "loss": 0.0864, "step": 15360 }, { "grad_norm": 0.7230620384216309, "learning_rate": 9.708923714991847e-05, "loss": 0.0937, "step": 15370 }, { "grad_norm": 0.877384603023529, "learning_rate": 9.708367533777638e-05, "loss": 0.0906, "step": 15380 }, { "grad_norm": 0.5975368618965149, "learning_rate": 9.707810837663431e-05, "loss": 0.0698, "step": 15390 }, { "grad_norm": 0.540107786655426, "learning_rate": 9.707253626710113e-05, "loss": 0.0761, "step": 15400 }, { "grad_norm": 0.7544283270835876, "learning_rate": 9.706695900978613e-05, "loss": 0.0766, "step": 15410 }, { "grad_norm": 0.7511806488037109, "learning_rate": 9.706137660529926e-05, "loss": 0.1133, "step": 15420 }, { "grad_norm": 0.5734120011329651, "learning_rate": 9.705578905425101e-05, "loss": 0.0949, "step": 15430 }, { "grad_norm": 1.0206373929977417, "learning_rate": 9.705019635725241e-05, "loss": 0.0833, "step": 15440 }, { "grad_norm": 0.5202481746673584, "learning_rate": 9.704459851491508e-05, "loss": 0.0972, "step": 15450 }, { "grad_norm": 0.9769009947776794, "learning_rate": 9.703899552785118e-05, "loss": 0.0891, "step": 15460 }, { "grad_norm": 0.5819336175918579, "learning_rate": 9.703338739667346e-05, "loss": 0.0896, "step": 15470 }, { "grad_norm": 0.8925672173500061, "learning_rate": 9.70277741219952e-05, "loss": 0.0929, "step": 15480 }, { "grad_norm": 0.9350314736366272, "learning_rate": 9.702215570443027e-05, "loss": 0.0703, "step": 15490 }, { "grad_norm": 0.6962978839874268, "learning_rate": 9.701653214459309e-05, "loss": 0.0632, "step": 15500 }, { "grad_norm": 0.7682257890701294, "learning_rate": 9.701090344309865e-05, "loss": 0.0755, "step": 15510 }, { "grad_norm": 0.6514864563941956, "learning_rate": 9.700526960056247e-05, "loss": 0.0796, "step": 15520 }, { "grad_norm": 0.6528998613357544, "learning_rate": 9.699963061760068e-05, "loss": 0.0862, "step": 15530 }, { "grad_norm": 0.7777257561683655, "learning_rate": 9.699398649482997e-05, "loss": 0.086, "step": 15540 }, { "grad_norm": 0.7812777161598206, "learning_rate": 9.698833723286753e-05, "loss": 0.0909, "step": 15550 }, { "grad_norm": 0.9194461703300476, "learning_rate": 9.698268283233118e-05, "loss": 0.0933, "step": 15560 }, { "grad_norm": 0.8624570965766907, "learning_rate": 9.697702329383929e-05, "loss": 0.078, "step": 15570 }, { "grad_norm": 0.81453937292099, "learning_rate": 9.697135861801074e-05, "loss": 0.0854, "step": 15580 }, { "grad_norm": 0.725713312625885, "learning_rate": 9.696568880546505e-05, "loss": 0.0686, "step": 15590 }, { "grad_norm": 0.9844354391098022, "learning_rate": 9.696001385682223e-05, "loss": 0.0883, "step": 15600 }, { "grad_norm": 0.6945556998252869, "learning_rate": 9.695433377270291e-05, "loss": 0.083, "step": 15610 }, { "grad_norm": 0.6631295084953308, "learning_rate": 9.694864855372824e-05, "loss": 0.0775, "step": 15620 }, { "grad_norm": 0.7470214366912842, "learning_rate": 9.694295820051995e-05, "loss": 0.0733, "step": 15630 }, { "grad_norm": 0.7154128551483154, "learning_rate": 9.693726271370032e-05, "loss": 0.0825, "step": 15640 }, { "grad_norm": 0.801540195941925, "learning_rate": 9.693156209389221e-05, "loss": 0.0874, "step": 15650 }, { "grad_norm": 0.6605640053749084, "learning_rate": 9.692585634171905e-05, "loss": 0.0914, "step": 15660 }, { "grad_norm": 0.7361282706260681, "learning_rate": 9.692014545780476e-05, "loss": 0.081, "step": 15670 }, { "grad_norm": 0.6166872382164001, "learning_rate": 9.691442944277393e-05, "loss": 0.0735, "step": 15680 }, { "grad_norm": 0.6518673300743103, "learning_rate": 9.690870829725162e-05, "loss": 0.0876, "step": 15690 }, { "grad_norm": 1.5816023349761963, "learning_rate": 9.69029820218635e-05, "loss": 0.0825, "step": 15700 }, { "grad_norm": 0.8058804273605347, "learning_rate": 9.689725061723579e-05, "loss": 0.0904, "step": 15710 }, { "grad_norm": 0.8150656223297119, "learning_rate": 9.689151408399527e-05, "loss": 0.1002, "step": 15720 }, { "grad_norm": 0.5645583868026733, "learning_rate": 9.688577242276924e-05, "loss": 0.087, "step": 15730 }, { "grad_norm": 0.7283293604850769, "learning_rate": 9.688002563418566e-05, "loss": 0.113, "step": 15740 }, { "grad_norm": 0.5971599221229553, "learning_rate": 9.687427371887293e-05, "loss": 0.1015, "step": 15750 }, { "grad_norm": 0.6920323967933655, "learning_rate": 9.686851667746012e-05, "loss": 0.0842, "step": 15760 }, { "grad_norm": 0.47025030851364136, "learning_rate": 9.686275451057677e-05, "loss": 0.0639, "step": 15770 }, { "grad_norm": 0.565360426902771, "learning_rate": 9.685698721885308e-05, "loss": 0.0724, "step": 15780 }, { "grad_norm": 0.571317195892334, "learning_rate": 9.68512148029197e-05, "loss": 0.0857, "step": 15790 }, { "grad_norm": 0.9415205717086792, "learning_rate": 9.684543726340791e-05, "loss": 0.0902, "step": 15800 }, { "grad_norm": 0.8274579048156738, "learning_rate": 9.683965460094952e-05, "loss": 0.103, "step": 15810 }, { "grad_norm": 0.6501867771148682, "learning_rate": 9.683386681617694e-05, "loss": 0.08, "step": 15820 }, { "grad_norm": 0.8274012804031372, "learning_rate": 9.68280739097231e-05, "loss": 0.1052, "step": 15830 }, { "grad_norm": 0.9016930460929871, "learning_rate": 9.682227588222148e-05, "loss": 0.1125, "step": 15840 }, { "grad_norm": 0.45885512232780457, "learning_rate": 9.681647273430618e-05, "loss": 0.0855, "step": 15850 }, { "grad_norm": 0.743261992931366, "learning_rate": 9.681066446661182e-05, "loss": 0.0816, "step": 15860 }, { "grad_norm": 0.8339987993240356, "learning_rate": 9.680485107977357e-05, "loss": 0.0664, "step": 15870 }, { "grad_norm": 0.6364918351173401, "learning_rate": 9.679903257442716e-05, "loss": 0.0925, "step": 15880 }, { "grad_norm": 0.5337017774581909, "learning_rate": 9.679320895120891e-05, "loss": 0.0781, "step": 15890 }, { "grad_norm": 1.1017043590545654, "learning_rate": 9.67873802107557e-05, "loss": 0.0857, "step": 15900 }, { "grad_norm": 0.9187553524971008, "learning_rate": 9.67815463537049e-05, "loss": 0.0825, "step": 15910 }, { "grad_norm": 0.685505211353302, "learning_rate": 9.677570738069457e-05, "loss": 0.0876, "step": 15920 }, { "grad_norm": 0.6564694046974182, "learning_rate": 9.676986329236318e-05, "loss": 0.0736, "step": 15930 }, { "grad_norm": 0.8154304623603821, "learning_rate": 9.676401408934987e-05, "loss": 0.0808, "step": 15940 }, { "grad_norm": 0.8245432376861572, "learning_rate": 9.675815977229428e-05, "loss": 0.0798, "step": 15950 }, { "grad_norm": 0.7530726790428162, "learning_rate": 9.675230034183664e-05, "loss": 0.0873, "step": 15960 }, { "grad_norm": 0.7204634547233582, "learning_rate": 9.674643579861773e-05, "loss": 0.0672, "step": 15970 }, { "grad_norm": 1.0257023572921753, "learning_rate": 9.674056614327886e-05, "loss": 0.0891, "step": 15980 }, { "grad_norm": 0.5802040100097656, "learning_rate": 9.673469137646198e-05, "loss": 0.0845, "step": 15990 }, { "grad_norm": 0.6902487277984619, "learning_rate": 9.67288114988095e-05, "loss": 0.0729, "step": 16000 }, { "grad_norm": 0.75093674659729, "learning_rate": 9.672292651096447e-05, "loss": 0.0709, "step": 16010 }, { "grad_norm": 0.6239150762557983, "learning_rate": 9.671703641357042e-05, "loss": 0.0731, "step": 16020 }, { "grad_norm": 0.7786989212036133, "learning_rate": 9.67111412072715e-05, "loss": 0.0877, "step": 16030 }, { "grad_norm": 0.7429212331771851, "learning_rate": 9.670524089271242e-05, "loss": 0.0696, "step": 16040 }, { "grad_norm": 0.6006675362586975, "learning_rate": 9.669933547053842e-05, "loss": 0.0657, "step": 16050 }, { "grad_norm": 0.8770735859870911, "learning_rate": 9.669342494139531e-05, "loss": 0.085, "step": 16060 }, { "grad_norm": 0.8651040196418762, "learning_rate": 9.668750930592943e-05, "loss": 0.0698, "step": 16070 }, { "grad_norm": 0.5812219381332397, "learning_rate": 9.668158856478775e-05, "loss": 0.0666, "step": 16080 }, { "grad_norm": 0.4829676151275635, "learning_rate": 9.66756627186177e-05, "loss": 0.073, "step": 16090 }, { "grad_norm": 0.5468246340751648, "learning_rate": 9.666973176806737e-05, "loss": 0.0801, "step": 16100 }, { "grad_norm": 0.6281709671020508, "learning_rate": 9.666379571378534e-05, "loss": 0.0789, "step": 16110 }, { "grad_norm": 1.0386362075805664, "learning_rate": 9.665785455642076e-05, "loss": 0.0707, "step": 16120 }, { "grad_norm": 1.0743014812469482, "learning_rate": 9.665190829662337e-05, "loss": 0.0906, "step": 16130 }, { "grad_norm": 0.9224532246589661, "learning_rate": 9.664595693504342e-05, "loss": 0.0736, "step": 16140 }, { "grad_norm": 0.7720491290092468, "learning_rate": 9.664000047233175e-05, "loss": 0.0982, "step": 16150 }, { "grad_norm": 0.7038226127624512, "learning_rate": 9.663403890913976e-05, "loss": 0.0831, "step": 16160 }, { "grad_norm": 0.6858058571815491, "learning_rate": 9.662807224611938e-05, "loss": 0.0917, "step": 16170 }, { "grad_norm": 0.8304696083068848, "learning_rate": 9.662210048392311e-05, "loss": 0.0739, "step": 16180 }, { "grad_norm": 0.6956483125686646, "learning_rate": 9.661612362320405e-05, "loss": 0.1091, "step": 16190 }, { "grad_norm": 0.5675424933433533, "learning_rate": 9.661014166461579e-05, "loss": 0.0736, "step": 16200 }, { "grad_norm": 0.48882803320884705, "learning_rate": 9.66041546088125e-05, "loss": 0.1079, "step": 16210 }, { "grad_norm": 1.1574528217315674, "learning_rate": 9.659816245644895e-05, "loss": 0.0847, "step": 16220 }, { "grad_norm": 0.6381628513336182, "learning_rate": 9.65921652081804e-05, "loss": 0.0828, "step": 16230 }, { "grad_norm": 0.5429662466049194, "learning_rate": 9.658616286466271e-05, "loss": 0.0797, "step": 16240 }, { "grad_norm": 0.8996042013168335, "learning_rate": 9.65801554265523e-05, "loss": 0.0814, "step": 16250 }, { "grad_norm": 0.8401418328285217, "learning_rate": 9.657414289450612e-05, "loss": 0.1158, "step": 16260 }, { "grad_norm": 0.7700481414794922, "learning_rate": 9.656812526918171e-05, "loss": 0.109, "step": 16270 }, { "grad_norm": 0.7057241797447205, "learning_rate": 9.656210255123712e-05, "loss": 0.0757, "step": 16280 }, { "grad_norm": 0.6519992351531982, "learning_rate": 9.6556074741331e-05, "loss": 0.0706, "step": 16290 }, { "grad_norm": 0.7713102698326111, "learning_rate": 9.655004184012256e-05, "loss": 0.0925, "step": 16300 }, { "grad_norm": 0.7053150534629822, "learning_rate": 9.654400384827152e-05, "loss": 0.0993, "step": 16310 }, { "grad_norm": 0.5866265892982483, "learning_rate": 9.653796076643818e-05, "loss": 0.0761, "step": 16320 }, { "grad_norm": 0.7138978838920593, "learning_rate": 9.653191259528344e-05, "loss": 0.0893, "step": 16330 }, { "grad_norm": 0.837959349155426, "learning_rate": 9.65258593354687e-05, "loss": 0.0891, "step": 16340 }, { "grad_norm": 0.569046676158905, "learning_rate": 9.651980098765591e-05, "loss": 0.0797, "step": 16350 }, { "grad_norm": 0.5277349352836609, "learning_rate": 9.651373755250765e-05, "loss": 0.0808, "step": 16360 }, { "grad_norm": 0.7941372990608215, "learning_rate": 9.650766903068697e-05, "loss": 0.0696, "step": 16370 }, { "grad_norm": 0.7768509984016418, "learning_rate": 9.650159542285753e-05, "loss": 0.0971, "step": 16380 }, { "grad_norm": 0.7181211113929749, "learning_rate": 9.649551672968353e-05, "loss": 0.0821, "step": 16390 }, { "grad_norm": 0.6109887957572937, "learning_rate": 9.648943295182973e-05, "loss": 0.0873, "step": 16400 }, { "grad_norm": 0.5998690128326416, "learning_rate": 9.648334408996144e-05, "loss": 0.0769, "step": 16410 }, { "grad_norm": 0.5698604583740234, "learning_rate": 9.647725014474452e-05, "loss": 0.0633, "step": 16420 }, { "grad_norm": 0.8739631772041321, "learning_rate": 9.64711511168454e-05, "loss": 0.0928, "step": 16430 }, { "grad_norm": 0.672294020652771, "learning_rate": 9.646504700693108e-05, "loss": 0.0705, "step": 16440 }, { "grad_norm": 0.9953305125236511, "learning_rate": 9.645893781566907e-05, "loss": 0.0896, "step": 16450 }, { "grad_norm": 0.6927785873413086, "learning_rate": 9.645282354372744e-05, "loss": 0.0896, "step": 16460 }, { "grad_norm": 1.1207712888717651, "learning_rate": 9.644670419177491e-05, "loss": 0.1114, "step": 16470 }, { "grad_norm": 0.6648139357566833, "learning_rate": 9.644057976048062e-05, "loss": 0.0868, "step": 16480 }, { "grad_norm": 0.9572907090187073, "learning_rate": 9.643445025051435e-05, "loss": 0.0829, "step": 16490 }, { "grad_norm": 0.8022940754890442, "learning_rate": 9.642831566254641e-05, "loss": 0.0845, "step": 16500 }, { "grad_norm": 0.41541051864624023, "learning_rate": 9.642217599724769e-05, "loss": 0.0706, "step": 16510 }, { "grad_norm": 0.8102608323097229, "learning_rate": 9.64160312552896e-05, "loss": 0.0812, "step": 16520 }, { "grad_norm": 0.9172006249427795, "learning_rate": 9.64098814373441e-05, "loss": 0.0899, "step": 16530 }, { "grad_norm": 0.86594158411026, "learning_rate": 9.640372654408374e-05, "loss": 0.076, "step": 16540 }, { "grad_norm": 0.5887507796287537, "learning_rate": 9.639756657618162e-05, "loss": 0.0821, "step": 16550 }, { "grad_norm": 0.8554551601409912, "learning_rate": 9.639140153431138e-05, "loss": 0.1038, "step": 16560 }, { "grad_norm": 0.9577121138572693, "learning_rate": 9.638523141914721e-05, "loss": 0.066, "step": 16570 }, { "grad_norm": 0.6331597566604614, "learning_rate": 9.637905623136388e-05, "loss": 0.076, "step": 16580 }, { "grad_norm": 0.8303258419036865, "learning_rate": 9.637287597163669e-05, "loss": 0.0964, "step": 16590 }, { "grad_norm": 0.7653710246086121, "learning_rate": 9.63666906406415e-05, "loss": 0.0968, "step": 16600 }, { "grad_norm": 0.8311349153518677, "learning_rate": 9.636050023905473e-05, "loss": 0.0789, "step": 16610 }, { "grad_norm": 0.6832036972045898, "learning_rate": 9.635430476755336e-05, "loss": 0.0781, "step": 16620 }, { "grad_norm": 0.6950039863586426, "learning_rate": 9.63481042268149e-05, "loss": 0.0688, "step": 16630 }, { "grad_norm": 0.6285008788108826, "learning_rate": 9.634189861751745e-05, "loss": 0.0568, "step": 16640 }, { "grad_norm": 0.6196058392524719, "learning_rate": 9.633568794033967e-05, "loss": 0.0702, "step": 16650 }, { "grad_norm": 0.5506609082221985, "learning_rate": 9.63294721959607e-05, "loss": 0.0734, "step": 16660 }, { "grad_norm": 0.7032768726348877, "learning_rate": 9.63232513850603e-05, "loss": 0.0698, "step": 16670 }, { "grad_norm": 0.6784831881523132, "learning_rate": 9.631702550831878e-05, "loss": 0.0785, "step": 16680 }, { "grad_norm": 0.6263952255249023, "learning_rate": 9.631079456641698e-05, "loss": 0.0881, "step": 16690 }, { "grad_norm": 0.6099335551261902, "learning_rate": 9.630455856003632e-05, "loss": 0.0697, "step": 16700 }, { "grad_norm": 0.49718883633613586, "learning_rate": 9.629831748985876e-05, "loss": 0.0775, "step": 16710 }, { "grad_norm": 0.8597530126571655, "learning_rate": 9.629207135656679e-05, "loss": 0.0795, "step": 16720 }, { "grad_norm": 0.6855043768882751, "learning_rate": 9.628582016084353e-05, "loss": 0.0544, "step": 16730 }, { "grad_norm": 0.5921860337257385, "learning_rate": 9.627956390337254e-05, "loss": 0.0746, "step": 16740 }, { "grad_norm": 0.551287829875946, "learning_rate": 9.627330258483802e-05, "loss": 0.0769, "step": 16750 }, { "grad_norm": 0.7328455448150635, "learning_rate": 9.62670362059247e-05, "loss": 0.0807, "step": 16760 }, { "grad_norm": 0.7321504950523376, "learning_rate": 9.626076476731786e-05, "loss": 0.0716, "step": 16770 }, { "grad_norm": 0.5285640954971313, "learning_rate": 9.625448826970336e-05, "loss": 0.0804, "step": 16780 }, { "grad_norm": 0.7271410226821899, "learning_rate": 9.624820671376755e-05, "loss": 0.0699, "step": 16790 }, { "grad_norm": 0.5992560982704163, "learning_rate": 9.62419201001974e-05, "loss": 0.0801, "step": 16800 }, { "grad_norm": 0.6514492034912109, "learning_rate": 9.623562842968037e-05, "loss": 0.0864, "step": 16810 }, { "grad_norm": 0.5540494918823242, "learning_rate": 9.622933170290454e-05, "loss": 0.082, "step": 16820 }, { "grad_norm": 0.7100444436073303, "learning_rate": 9.622302992055849e-05, "loss": 0.0664, "step": 16830 }, { "grad_norm": 0.9823916554450989, "learning_rate": 9.62167230833314e-05, "loss": 0.0711, "step": 16840 }, { "grad_norm": 0.6673741936683655, "learning_rate": 9.621041119191295e-05, "loss": 0.0618, "step": 16850 }, { "grad_norm": 0.7334871292114258, "learning_rate": 9.620409424699342e-05, "loss": 0.0843, "step": 16860 }, { "grad_norm": 0.6284693479537964, "learning_rate": 9.619777224926359e-05, "loss": 0.0777, "step": 16870 }, { "grad_norm": 0.6316901445388794, "learning_rate": 9.619144519941485e-05, "loss": 0.0686, "step": 16880 }, { "grad_norm": 0.8276066780090332, "learning_rate": 9.618511309813912e-05, "loss": 0.0799, "step": 16890 }, { "grad_norm": 0.5252935886383057, "learning_rate": 9.617877594612886e-05, "loss": 0.0664, "step": 16900 }, { "grad_norm": 0.6312295198440552, "learning_rate": 9.617243374407707e-05, "loss": 0.0719, "step": 16910 }, { "grad_norm": 0.47135311365127563, "learning_rate": 9.616608649267736e-05, "loss": 0.0824, "step": 16920 }, { "grad_norm": 0.4661588668823242, "learning_rate": 9.615973419262385e-05, "loss": 0.0739, "step": 16930 }, { "grad_norm": 0.725653350353241, "learning_rate": 9.615337684461119e-05, "loss": 0.0987, "step": 16940 }, { "grad_norm": 0.7078697085380554, "learning_rate": 9.614701444933465e-05, "loss": 0.0792, "step": 16950 }, { "grad_norm": 0.6233832240104675, "learning_rate": 9.614064700748997e-05, "loss": 0.0734, "step": 16960 }, { "grad_norm": 0.8205381631851196, "learning_rate": 9.613427451977352e-05, "loss": 0.0949, "step": 16970 }, { "grad_norm": 0.6759244799613953, "learning_rate": 9.612789698688216e-05, "loss": 0.079, "step": 16980 }, { "grad_norm": 0.7205877304077148, "learning_rate": 9.612151440951334e-05, "loss": 0.0905, "step": 16990 }, { "grad_norm": 0.6688286066055298, "learning_rate": 9.611512678836506e-05, "loss": 0.0861, "step": 17000 }, { "grad_norm": 0.5794441103935242, "learning_rate": 9.610873412413584e-05, "loss": 0.0829, "step": 17010 }, { "grad_norm": 0.9211238026618958, "learning_rate": 9.610233641752476e-05, "loss": 0.0756, "step": 17020 }, { "grad_norm": 0.7049955725669861, "learning_rate": 9.609593366923151e-05, "loss": 0.0673, "step": 17030 }, { "grad_norm": 0.6828753352165222, "learning_rate": 9.608952587995625e-05, "loss": 0.0696, "step": 17040 }, { "grad_norm": 1.2455030679702759, "learning_rate": 9.608311305039972e-05, "loss": 0.0775, "step": 17050 }, { "grad_norm": 0.6250420808792114, "learning_rate": 9.607669518126326e-05, "loss": 0.0869, "step": 17060 }, { "grad_norm": 0.5173908472061157, "learning_rate": 9.607027227324866e-05, "loss": 0.0774, "step": 17070 }, { "grad_norm": 0.5747040510177612, "learning_rate": 9.606384432705837e-05, "loss": 0.076, "step": 17080 }, { "grad_norm": 0.5882430672645569, "learning_rate": 9.60574113433953e-05, "loss": 0.0719, "step": 17090 }, { "grad_norm": 0.8107605576515198, "learning_rate": 9.6050973322963e-05, "loss": 0.0752, "step": 17100 }, { "grad_norm": 0.7580996155738831, "learning_rate": 9.604453026646547e-05, "loss": 0.075, "step": 17110 }, { "grad_norm": 0.781166136264801, "learning_rate": 9.603808217460735e-05, "loss": 0.0865, "step": 17120 }, { "grad_norm": 0.7879104614257812, "learning_rate": 9.603162904809377e-05, "loss": 0.1225, "step": 17130 }, { "grad_norm": 0.7091190218925476, "learning_rate": 9.602517088763045e-05, "loss": 0.0817, "step": 17140 }, { "grad_norm": 0.4500369429588318, "learning_rate": 9.601870769392365e-05, "loss": 0.0663, "step": 17150 }, { "grad_norm": 0.6354905366897583, "learning_rate": 9.601223946768017e-05, "loss": 0.1006, "step": 17160 }, { "grad_norm": 0.8812594413757324, "learning_rate": 9.600576620960734e-05, "loss": 0.0947, "step": 17170 }, { "grad_norm": 0.5297068953514099, "learning_rate": 9.599928792041308e-05, "loss": 0.0739, "step": 17180 }, { "grad_norm": 0.5751383900642395, "learning_rate": 9.599280460080587e-05, "loss": 0.0629, "step": 17190 }, { "grad_norm": 0.7540225386619568, "learning_rate": 9.59863162514947e-05, "loss": 0.081, "step": 17200 }, { "grad_norm": 0.4657314419746399, "learning_rate": 9.597982287318911e-05, "loss": 0.0672, "step": 17210 }, { "grad_norm": 0.8516983389854431, "learning_rate": 9.597332446659923e-05, "loss": 0.0661, "step": 17220 }, { "grad_norm": 1.245824933052063, "learning_rate": 9.59668210324357e-05, "loss": 0.0786, "step": 17230 }, { "grad_norm": 0.6203577518463135, "learning_rate": 9.596031257140974e-05, "loss": 0.0928, "step": 17240 }, { "grad_norm": 0.6371249556541443, "learning_rate": 9.59537990842331e-05, "loss": 0.0843, "step": 17250 }, { "grad_norm": 0.608639121055603, "learning_rate": 9.594728057161806e-05, "loss": 0.088, "step": 17260 }, { "grad_norm": 0.7696142792701721, "learning_rate": 9.594075703427752e-05, "loss": 0.0817, "step": 17270 }, { "grad_norm": 0.833187460899353, "learning_rate": 9.593422847292486e-05, "loss": 0.094, "step": 17280 }, { "grad_norm": 0.4925008714199066, "learning_rate": 9.592769488827402e-05, "loss": 0.0644, "step": 17290 }, { "grad_norm": 1.0556622743606567, "learning_rate": 9.592115628103952e-05, "loss": 0.0883, "step": 17300 }, { "grad_norm": 0.6555972099304199, "learning_rate": 9.591461265193643e-05, "loss": 0.0669, "step": 17310 }, { "grad_norm": 0.43959081172943115, "learning_rate": 9.590806400168032e-05, "loss": 0.0719, "step": 17320 }, { "grad_norm": 0.4975508153438568, "learning_rate": 9.590151033098735e-05, "loss": 0.0622, "step": 17330 }, { "grad_norm": 0.6394393444061279, "learning_rate": 9.589495164057423e-05, "loss": 0.0736, "step": 17340 }, { "grad_norm": 0.8782398104667664, "learning_rate": 9.58883879311582e-05, "loss": 0.0656, "step": 17350 }, { "grad_norm": 0.7271684408187866, "learning_rate": 9.588181920345705e-05, "loss": 0.086, "step": 17360 }, { "grad_norm": 0.7450304627418518, "learning_rate": 9.587524545818913e-05, "loss": 0.0582, "step": 17370 }, { "grad_norm": 0.37860098481178284, "learning_rate": 9.586866669607335e-05, "loss": 0.0645, "step": 17380 }, { "grad_norm": 0.724405825138092, "learning_rate": 9.586208291782915e-05, "loss": 0.0804, "step": 17390 }, { "grad_norm": 0.7280540466308594, "learning_rate": 9.58554941241765e-05, "loss": 0.0753, "step": 17400 }, { "grad_norm": 0.5771257281303406, "learning_rate": 9.584890031583596e-05, "loss": 0.0842, "step": 17410 }, { "grad_norm": 1.0553585290908813, "learning_rate": 9.584230149352861e-05, "loss": 0.0971, "step": 17420 }, { "grad_norm": 0.4523969292640686, "learning_rate": 9.58356976579761e-05, "loss": 0.0653, "step": 17430 }, { "grad_norm": 0.9791759848594666, "learning_rate": 9.58290888099006e-05, "loss": 0.1251, "step": 17440 }, { "grad_norm": 0.8876686096191406, "learning_rate": 9.582247495002486e-05, "loss": 0.073, "step": 17450 }, { "grad_norm": 0.44192472100257874, "learning_rate": 9.581585607907214e-05, "loss": 0.0671, "step": 17460 }, { "grad_norm": 0.7370651364326477, "learning_rate": 9.580923219776628e-05, "loss": 0.0789, "step": 17470 }, { "grad_norm": 0.8327777981758118, "learning_rate": 9.580260330683167e-05, "loss": 0.0741, "step": 17480 }, { "grad_norm": 0.7215842008590698, "learning_rate": 9.579596940699322e-05, "loss": 0.0733, "step": 17490 }, { "grad_norm": 0.4946309030056, "learning_rate": 9.578933049897643e-05, "loss": 0.0804, "step": 17500 }, { "grad_norm": 0.5996469259262085, "learning_rate": 9.578268658350728e-05, "loss": 0.0648, "step": 17510 }, { "grad_norm": 0.7183849811553955, "learning_rate": 9.577603766131235e-05, "loss": 0.0745, "step": 17520 }, { "grad_norm": 0.8313084244728088, "learning_rate": 9.576938373311878e-05, "loss": 0.0866, "step": 17530 }, { "grad_norm": 0.700528085231781, "learning_rate": 9.576272479965421e-05, "loss": 0.0911, "step": 17540 }, { "grad_norm": 0.798818826675415, "learning_rate": 9.575606086164687e-05, "loss": 0.0706, "step": 17550 }, { "grad_norm": 0.5213891863822937, "learning_rate": 9.57493919198255e-05, "loss": 0.1012, "step": 17560 }, { "grad_norm": 0.5649237036705017, "learning_rate": 9.57427179749194e-05, "loss": 0.0808, "step": 17570 }, { "grad_norm": 0.7133964896202087, "learning_rate": 9.573603902765846e-05, "loss": 0.062, "step": 17580 }, { "grad_norm": 0.8436857461929321, "learning_rate": 9.572935507877304e-05, "loss": 0.0713, "step": 17590 }, { "grad_norm": 0.5890178680419922, "learning_rate": 9.57226661289941e-05, "loss": 0.0686, "step": 17600 }, { "grad_norm": 0.7022961974143982, "learning_rate": 9.571597217905315e-05, "loss": 0.0689, "step": 17610 }, { "grad_norm": 0.7274775505065918, "learning_rate": 9.57092732296822e-05, "loss": 0.0748, "step": 17620 }, { "grad_norm": 0.6803616285324097, "learning_rate": 9.570256928161385e-05, "loss": 0.0661, "step": 17630 }, { "grad_norm": 0.9025002121925354, "learning_rate": 9.569586033558126e-05, "loss": 0.0792, "step": 17640 }, { "grad_norm": 0.5892335772514343, "learning_rate": 9.568914639231807e-05, "loss": 0.0773, "step": 17650 }, { "grad_norm": 0.7845051884651184, "learning_rate": 9.568242745255852e-05, "loss": 0.0815, "step": 17660 }, { "grad_norm": 0.7236023545265198, "learning_rate": 9.567570351703739e-05, "loss": 0.0787, "step": 17670 }, { "grad_norm": 0.7821173071861267, "learning_rate": 9.566897458649001e-05, "loss": 0.0854, "step": 17680 }, { "grad_norm": 0.8757914304733276, "learning_rate": 9.566224066165221e-05, "loss": 0.0683, "step": 17690 }, { "grad_norm": 0.6393228769302368, "learning_rate": 9.565550174326043e-05, "loss": 0.0681, "step": 17700 }, { "grad_norm": 0.6091574430465698, "learning_rate": 9.564875783205162e-05, "loss": 0.0818, "step": 17710 }, { "grad_norm": 0.6589227914810181, "learning_rate": 9.564200892876328e-05, "loss": 0.0742, "step": 17720 }, { "grad_norm": 0.6795088052749634, "learning_rate": 9.563525503413348e-05, "loss": 0.0787, "step": 17730 }, { "grad_norm": 0.9276701211929321, "learning_rate": 9.562849614890079e-05, "loss": 0.1273, "step": 17740 }, { "grad_norm": 0.6105020642280579, "learning_rate": 9.562173227380436e-05, "loss": 0.0949, "step": 17750 }, { "grad_norm": 0.4467310905456543, "learning_rate": 9.561496340958389e-05, "loss": 0.0776, "step": 17760 }, { "grad_norm": 0.6605092287063599, "learning_rate": 9.560818955697959e-05, "loss": 0.0972, "step": 17770 }, { "grad_norm": 0.5154673457145691, "learning_rate": 9.560141071673228e-05, "loss": 0.0565, "step": 17780 }, { "grad_norm": 0.6189737915992737, "learning_rate": 9.559462688958323e-05, "loss": 0.0837, "step": 17790 }, { "grad_norm": 0.7747674584388733, "learning_rate": 9.558783807627434e-05, "loss": 0.0781, "step": 17800 }, { "grad_norm": 0.6917162537574768, "learning_rate": 9.558104427754801e-05, "loss": 0.0748, "step": 17810 }, { "grad_norm": 0.5889813303947449, "learning_rate": 9.557424549414722e-05, "loss": 0.0677, "step": 17820 }, { "grad_norm": 1.025632381439209, "learning_rate": 9.556744172681546e-05, "loss": 0.0924, "step": 17830 }, { "grad_norm": 0.6811608672142029, "learning_rate": 9.556063297629677e-05, "loss": 0.0828, "step": 17840 }, { "grad_norm": 0.6856418251991272, "learning_rate": 9.555381924333578e-05, "loss": 0.0709, "step": 17850 }, { "grad_norm": 0.7240629196166992, "learning_rate": 9.554700052867758e-05, "loss": 0.0796, "step": 17860 }, { "grad_norm": 0.6817694306373596, "learning_rate": 9.554017683306789e-05, "loss": 0.0662, "step": 17870 }, { "grad_norm": 0.7717880010604858, "learning_rate": 9.553334815725294e-05, "loss": 0.0965, "step": 17880 }, { "grad_norm": 0.5586130619049072, "learning_rate": 9.552651450197949e-05, "loss": 0.0741, "step": 17890 }, { "grad_norm": 1.0530272722244263, "learning_rate": 9.551967586799486e-05, "loss": 0.0749, "step": 17900 }, { "grad_norm": 0.5436874628067017, "learning_rate": 9.551283225604692e-05, "loss": 0.0538, "step": 17910 }, { "grad_norm": 0.42085450887680054, "learning_rate": 9.550598366688406e-05, "loss": 0.0706, "step": 17920 }, { "grad_norm": 0.3639465868473053, "learning_rate": 9.549913010125526e-05, "loss": 0.0708, "step": 17930 }, { "grad_norm": 0.5276150107383728, "learning_rate": 9.549227155990999e-05, "loss": 0.0596, "step": 17940 }, { "grad_norm": 0.6106404066085815, "learning_rate": 9.548540804359828e-05, "loss": 0.0787, "step": 17950 }, { "grad_norm": 0.8989389538764954, "learning_rate": 9.547853955307077e-05, "loss": 0.0732, "step": 17960 }, { "grad_norm": 0.6341236233711243, "learning_rate": 9.547166608907853e-05, "loss": 0.1015, "step": 17970 }, { "grad_norm": 0.6248663067817688, "learning_rate": 9.546478765237326e-05, "loss": 0.1004, "step": 17980 }, { "grad_norm": 0.6103467345237732, "learning_rate": 9.545790424370715e-05, "loss": 0.0845, "step": 17990 }, { "grad_norm": 0.6755425333976746, "learning_rate": 9.5451015863833e-05, "loss": 0.0682, "step": 18000 }, { "grad_norm": 0.7840602397918701, "learning_rate": 9.544412251350408e-05, "loss": 0.0614, "step": 18010 }, { "grad_norm": 0.7221390604972839, "learning_rate": 9.543722419347422e-05, "loss": 0.0655, "step": 18020 }, { "grad_norm": 0.4433000385761261, "learning_rate": 9.543032090449788e-05, "loss": 0.0674, "step": 18030 }, { "grad_norm": 0.5910397171974182, "learning_rate": 9.542341264732992e-05, "loss": 0.0886, "step": 18040 }, { "grad_norm": 0.7327244281768799, "learning_rate": 9.541649942272585e-05, "loss": 0.0922, "step": 18050 }, { "grad_norm": 0.5992644429206848, "learning_rate": 9.54095812314417e-05, "loss": 0.0767, "step": 18060 }, { "grad_norm": 0.6467081904411316, "learning_rate": 9.540265807423401e-05, "loss": 0.0695, "step": 18070 }, { "grad_norm": 0.49391376972198486, "learning_rate": 9.53957299518599e-05, "loss": 0.0838, "step": 18080 }, { "grad_norm": 0.543493926525116, "learning_rate": 9.5388796865077e-05, "loss": 0.0801, "step": 18090 }, { "grad_norm": 0.6139529347419739, "learning_rate": 9.538185881464353e-05, "loss": 0.0762, "step": 18100 }, { "grad_norm": 0.691111147403717, "learning_rate": 9.537491580131821e-05, "loss": 0.0812, "step": 18110 }, { "grad_norm": 0.5540621280670166, "learning_rate": 9.53679678258603e-05, "loss": 0.0883, "step": 18120 }, { "grad_norm": 0.642581045627594, "learning_rate": 9.536101488902966e-05, "loss": 0.0548, "step": 18130 }, { "grad_norm": 0.6240785121917725, "learning_rate": 9.535405699158663e-05, "loss": 0.071, "step": 18140 }, { "grad_norm": 0.5315576195716858, "learning_rate": 9.53470941342921e-05, "loss": 0.0789, "step": 18150 }, { "grad_norm": 0.635154128074646, "learning_rate": 9.534012631790756e-05, "loss": 0.0684, "step": 18160 }, { "grad_norm": 0.5911870002746582, "learning_rate": 9.533315354319494e-05, "loss": 0.0892, "step": 18170 }, { "grad_norm": 0.9090456366539001, "learning_rate": 9.532617581091682e-05, "loss": 0.0888, "step": 18180 }, { "grad_norm": 0.7803037166595459, "learning_rate": 9.531919312183629e-05, "loss": 0.0922, "step": 18190 }, { "grad_norm": 0.631151020526886, "learning_rate": 9.531220547671688e-05, "loss": 0.0834, "step": 18200 }, { "grad_norm": 0.8014813661575317, "learning_rate": 9.530521287632285e-05, "loss": 0.0908, "step": 18210 }, { "grad_norm": 1.002347469329834, "learning_rate": 9.529821532141884e-05, "loss": 0.0773, "step": 18220 }, { "grad_norm": 0.557913601398468, "learning_rate": 9.52912128127701e-05, "loss": 0.0734, "step": 18230 }, { "grad_norm": 0.6328341364860535, "learning_rate": 9.528420535114244e-05, "loss": 0.0889, "step": 18240 }, { "grad_norm": 0.6265480518341064, "learning_rate": 9.527719293730215e-05, "loss": 0.0874, "step": 18250 }, { "grad_norm": 0.6530383229255676, "learning_rate": 9.527017557201611e-05, "loss": 0.0687, "step": 18260 }, { "grad_norm": 0.7248877882957458, "learning_rate": 9.526315325605176e-05, "loss": 0.0687, "step": 18270 }, { "grad_norm": 0.7122926712036133, "learning_rate": 9.525612599017699e-05, "loss": 0.069, "step": 18280 }, { "grad_norm": 0.9474102854728699, "learning_rate": 9.524909377516033e-05, "loss": 0.0984, "step": 18290 }, { "grad_norm": 0.678633987903595, "learning_rate": 9.524205661177081e-05, "loss": 0.0815, "step": 18300 }, { "grad_norm": 0.8227855563163757, "learning_rate": 9.523501450077801e-05, "loss": 0.0638, "step": 18310 }, { "grad_norm": 0.664366602897644, "learning_rate": 9.522796744295202e-05, "loss": 0.0761, "step": 18320 }, { "grad_norm": 0.6012675166130066, "learning_rate": 9.522091543906352e-05, "loss": 0.0792, "step": 18330 }, { "grad_norm": 0.883476197719574, "learning_rate": 9.521385848988369e-05, "loss": 0.0815, "step": 18340 }, { "grad_norm": 0.626694917678833, "learning_rate": 9.520679659618428e-05, "loss": 0.0678, "step": 18350 }, { "grad_norm": 0.5808990597724915, "learning_rate": 9.519972975873754e-05, "loss": 0.0594, "step": 18360 }, { "grad_norm": 0.6863714456558228, "learning_rate": 9.519265797831633e-05, "loss": 0.0627, "step": 18370 }, { "grad_norm": 0.59295254945755, "learning_rate": 9.518558125569399e-05, "loss": 0.0559, "step": 18380 }, { "grad_norm": 0.9713706374168396, "learning_rate": 9.517849959164442e-05, "loss": 0.0826, "step": 18390 }, { "grad_norm": 0.5458658933639526, "learning_rate": 9.517141298694205e-05, "loss": 0.0688, "step": 18400 }, { "grad_norm": 0.9748808741569519, "learning_rate": 9.516432144236188e-05, "loss": 0.0838, "step": 18410 }, { "grad_norm": 1.1862159967422485, "learning_rate": 9.515722495867941e-05, "loss": 0.0946, "step": 18420 }, { "grad_norm": 0.5928769707679749, "learning_rate": 9.515012353667072e-05, "loss": 0.0915, "step": 18430 }, { "grad_norm": 0.5303327441215515, "learning_rate": 9.51430171771124e-05, "loss": 0.0788, "step": 18440 }, { "grad_norm": 0.6174435019493103, "learning_rate": 9.513590588078159e-05, "loss": 0.0784, "step": 18450 }, { "grad_norm": 0.942877471446991, "learning_rate": 9.512878964845597e-05, "loss": 0.0813, "step": 18460 }, { "grad_norm": 0.5707474946975708, "learning_rate": 9.512166848091377e-05, "loss": 0.0622, "step": 18470 }, { "grad_norm": 0.5911072492599487, "learning_rate": 9.511454237893376e-05, "loss": 0.0882, "step": 18480 }, { "grad_norm": 0.5689173340797424, "learning_rate": 9.51074113432952e-05, "loss": 0.0716, "step": 18490 }, { "grad_norm": 0.5377030372619629, "learning_rate": 9.510027537477797e-05, "loss": 0.0775, "step": 18500 }, { "grad_norm": 0.5671442747116089, "learning_rate": 9.509313447416242e-05, "loss": 0.0782, "step": 18510 }, { "grad_norm": 0.5596413612365723, "learning_rate": 9.508598864222949e-05, "loss": 0.0647, "step": 18520 }, { "grad_norm": 0.6849573254585266, "learning_rate": 9.507883787976062e-05, "loss": 0.0776, "step": 18530 }, { "grad_norm": 0.7005646824836731, "learning_rate": 9.507168218753781e-05, "loss": 0.0844, "step": 18540 }, { "grad_norm": 0.8313305974006653, "learning_rate": 9.506452156634362e-05, "loss": 0.062, "step": 18550 }, { "grad_norm": 0.5666658282279968, "learning_rate": 9.505735601696109e-05, "loss": 0.0837, "step": 18560 }, { "grad_norm": 0.92655348777771, "learning_rate": 9.505018554017385e-05, "loss": 0.0848, "step": 18570 }, { "grad_norm": 1.0326743125915527, "learning_rate": 9.504301013676604e-05, "loss": 0.09, "step": 18580 }, { "grad_norm": 0.7119969725608826, "learning_rate": 9.503582980752238e-05, "loss": 0.0654, "step": 18590 }, { "grad_norm": 0.573836088180542, "learning_rate": 9.502864455322809e-05, "loss": 0.1028, "step": 18600 }, { "grad_norm": 0.8509664535522461, "learning_rate": 9.502145437466891e-05, "loss": 0.0795, "step": 18610 }, { "grad_norm": 0.6579105854034424, "learning_rate": 9.501425927263116e-05, "loss": 0.0724, "step": 18620 }, { "grad_norm": 0.7696194052696228, "learning_rate": 9.500705924790172e-05, "loss": 0.0882, "step": 18630 }, { "grad_norm": 0.8129510879516602, "learning_rate": 9.499985430126794e-05, "loss": 0.0993, "step": 18640 }, { "grad_norm": 0.7076505422592163, "learning_rate": 9.499264443351775e-05, "loss": 0.0818, "step": 18650 }, { "grad_norm": 1.0670984983444214, "learning_rate": 9.498542964543961e-05, "loss": 0.0765, "step": 18660 }, { "grad_norm": 0.6687481999397278, "learning_rate": 9.497820993782252e-05, "loss": 0.0952, "step": 18670 }, { "grad_norm": 0.7037619948387146, "learning_rate": 9.497098531145601e-05, "loss": 0.0732, "step": 18680 }, { "grad_norm": 0.7430578470230103, "learning_rate": 9.496375576713017e-05, "loss": 0.0733, "step": 18690 }, { "grad_norm": 0.7133407592773438, "learning_rate": 9.49565213056356e-05, "loss": 0.0696, "step": 18700 }, { "grad_norm": 0.6650049090385437, "learning_rate": 9.494928192776342e-05, "loss": 0.0732, "step": 18710 }, { "grad_norm": 0.7136136889457703, "learning_rate": 9.494203763430538e-05, "loss": 0.0856, "step": 18720 }, { "grad_norm": 1.252959132194519, "learning_rate": 9.493478842605366e-05, "loss": 0.0887, "step": 18730 }, { "grad_norm": 0.5712947845458984, "learning_rate": 9.492753430380105e-05, "loss": 0.0929, "step": 18740 }, { "grad_norm": 0.6499808430671692, "learning_rate": 9.492027526834083e-05, "loss": 0.1088, "step": 18750 }, { "grad_norm": 0.5838018655776978, "learning_rate": 9.491301132046684e-05, "loss": 0.0657, "step": 18760 }, { "grad_norm": 0.8351432085037231, "learning_rate": 9.490574246097345e-05, "loss": 0.0696, "step": 18770 }, { "grad_norm": 1.0275287628173828, "learning_rate": 9.48984686906556e-05, "loss": 0.0671, "step": 18780 }, { "grad_norm": 0.5664291381835938, "learning_rate": 9.489119001030871e-05, "loss": 0.0689, "step": 18790 }, { "grad_norm": 1.1319270133972168, "learning_rate": 9.488390642072878e-05, "loss": 0.0891, "step": 18800 }, { "grad_norm": 0.7361190915107727, "learning_rate": 9.48766179227123e-05, "loss": 0.0846, "step": 18810 }, { "grad_norm": 0.800218939781189, "learning_rate": 9.486932451705636e-05, "loss": 0.0707, "step": 18820 }, { "grad_norm": 0.7916070222854614, "learning_rate": 9.486202620455857e-05, "loss": 0.0669, "step": 18830 }, { "grad_norm": 0.8391336798667908, "learning_rate": 9.485472298601704e-05, "loss": 0.0769, "step": 18840 }, { "grad_norm": 0.6352712512016296, "learning_rate": 9.484741486223043e-05, "loss": 0.0748, "step": 18850 }, { "grad_norm": 1.018701434135437, "learning_rate": 9.484010183399797e-05, "loss": 0.0692, "step": 18860 }, { "grad_norm": 0.7049585580825806, "learning_rate": 9.483278390211938e-05, "loss": 0.0861, "step": 18870 }, { "grad_norm": 0.5664845705032349, "learning_rate": 9.482546106739496e-05, "loss": 0.071, "step": 18880 }, { "grad_norm": 0.6712239980697632, "learning_rate": 9.48181333306255e-05, "loss": 0.0685, "step": 18890 }, { "grad_norm": 0.9234093427658081, "learning_rate": 9.481080069261237e-05, "loss": 0.0639, "step": 18900 }, { "grad_norm": 0.7003942728042603, "learning_rate": 9.480346315415745e-05, "loss": 0.0697, "step": 18910 }, { "grad_norm": 0.7742511630058289, "learning_rate": 9.479612071606314e-05, "loss": 0.0667, "step": 18920 }, { "grad_norm": 0.6754580736160278, "learning_rate": 9.478877337913244e-05, "loss": 0.0965, "step": 18930 }, { "grad_norm": 0.5237807035446167, "learning_rate": 9.478142114416881e-05, "loss": 0.1064, "step": 18940 }, { "grad_norm": 1.0406609773635864, "learning_rate": 9.47740640119763e-05, "loss": 0.0796, "step": 18950 }, { "grad_norm": 0.7837470173835754, "learning_rate": 9.476670198335947e-05, "loss": 0.0704, "step": 18960 }, { "grad_norm": 0.642741322517395, "learning_rate": 9.47593350591234e-05, "loss": 0.0599, "step": 18970 }, { "grad_norm": 0.4747580885887146, "learning_rate": 9.475196324007376e-05, "loss": 0.0706, "step": 18980 }, { "grad_norm": 0.4482223391532898, "learning_rate": 9.474458652701669e-05, "loss": 0.0662, "step": 18990 }, { "grad_norm": 0.4645480215549469, "learning_rate": 9.473720492075892e-05, "loss": 0.0641, "step": 19000 }, { "grad_norm": 0.5220770835876465, "learning_rate": 9.472981842210768e-05, "loss": 0.0628, "step": 19010 }, { "grad_norm": 0.682507336139679, "learning_rate": 9.472242703187074e-05, "loss": 0.0835, "step": 19020 }, { "grad_norm": 0.5180927515029907, "learning_rate": 9.471503075085643e-05, "loss": 0.0639, "step": 19030 }, { "grad_norm": 0.76950603723526, "learning_rate": 9.470762957987359e-05, "loss": 0.0853, "step": 19040 }, { "grad_norm": 0.5670573711395264, "learning_rate": 9.470022351973158e-05, "loss": 0.0671, "step": 19050 }, { "grad_norm": 0.7418026924133301, "learning_rate": 9.469281257124034e-05, "loss": 0.0723, "step": 19060 }, { "grad_norm": 0.8072322607040405, "learning_rate": 9.46853967352103e-05, "loss": 0.0785, "step": 19070 }, { "grad_norm": 0.8806436061859131, "learning_rate": 9.467797601245246e-05, "loss": 0.0938, "step": 19080 }, { "grad_norm": 0.6550869941711426, "learning_rate": 9.467055040377834e-05, "loss": 0.0765, "step": 19090 }, { "grad_norm": 0.7768909335136414, "learning_rate": 9.466311990999999e-05, "loss": 0.0666, "step": 19100 }, { "grad_norm": 0.867767333984375, "learning_rate": 9.465568453193e-05, "loss": 0.0828, "step": 19110 }, { "grad_norm": 0.6396932601928711, "learning_rate": 9.464824427038148e-05, "loss": 0.0732, "step": 19120 }, { "grad_norm": 0.9430219531059265, "learning_rate": 9.46407991261681e-05, "loss": 0.0837, "step": 19130 }, { "grad_norm": 0.8041993379592896, "learning_rate": 9.463334910010404e-05, "loss": 0.0936, "step": 19140 }, { "grad_norm": 0.7081626057624817, "learning_rate": 9.462589419300403e-05, "loss": 0.0663, "step": 19150 }, { "grad_norm": 0.5789684057235718, "learning_rate": 9.461843440568333e-05, "loss": 0.0675, "step": 19160 }, { "grad_norm": 0.6683184504508972, "learning_rate": 9.461096973895773e-05, "loss": 0.0618, "step": 19170 }, { "grad_norm": 0.5228683948516846, "learning_rate": 9.460350019364355e-05, "loss": 0.075, "step": 19180 }, { "grad_norm": 0.45297375321388245, "learning_rate": 9.459602577055764e-05, "loss": 0.0585, "step": 19190 }, { "grad_norm": 0.9006586074829102, "learning_rate": 9.45885464705174e-05, "loss": 0.0744, "step": 19200 }, { "grad_norm": 0.6545486450195312, "learning_rate": 9.458106229434076e-05, "loss": 0.0582, "step": 19210 }, { "grad_norm": 0.7851908802986145, "learning_rate": 9.457357324284617e-05, "loss": 0.0785, "step": 19220 }, { "grad_norm": 0.5170843601226807, "learning_rate": 9.456607931685262e-05, "loss": 0.0776, "step": 19230 }, { "grad_norm": 0.6555768251419067, "learning_rate": 9.455858051717965e-05, "loss": 0.0823, "step": 19240 }, { "grad_norm": 0.6648879051208496, "learning_rate": 9.45510768446473e-05, "loss": 0.0779, "step": 19250 }, { "grad_norm": 0.5170975923538208, "learning_rate": 9.454356830007618e-05, "loss": 0.0603, "step": 19260 }, { "grad_norm": 0.5455427169799805, "learning_rate": 9.45360548842874e-05, "loss": 0.0499, "step": 19270 }, { "grad_norm": 0.6231252551078796, "learning_rate": 9.452853659810261e-05, "loss": 0.0791, "step": 19280 }, { "grad_norm": 0.7828670144081116, "learning_rate": 9.452101344234401e-05, "loss": 0.0866, "step": 19290 }, { "grad_norm": 0.5383355021476746, "learning_rate": 9.451348541783431e-05, "loss": 0.0601, "step": 19300 }, { "grad_norm": 0.7344184517860413, "learning_rate": 9.450595252539678e-05, "loss": 0.0719, "step": 19310 }, { "grad_norm": 0.6279488801956177, "learning_rate": 9.449841476585518e-05, "loss": 0.0642, "step": 19320 }, { "grad_norm": 0.627905011177063, "learning_rate": 9.449087214003384e-05, "loss": 0.0827, "step": 19330 }, { "grad_norm": 0.7220600247383118, "learning_rate": 9.448332464875765e-05, "loss": 0.0842, "step": 19340 }, { "grad_norm": 0.5067951083183289, "learning_rate": 9.447577229285192e-05, "loss": 0.0697, "step": 19350 }, { "grad_norm": 0.6881146430969238, "learning_rate": 9.446821507314261e-05, "loss": 0.0615, "step": 19360 }, { "grad_norm": 0.7248697876930237, "learning_rate": 9.446065299045617e-05, "loss": 0.0566, "step": 19370 }, { "grad_norm": 0.9432387351989746, "learning_rate": 9.445308604561955e-05, "loss": 0.0713, "step": 19380 }, { "grad_norm": 0.7336834669113159, "learning_rate": 9.444551423946028e-05, "loss": 0.065, "step": 19390 }, { "grad_norm": 0.6588897705078125, "learning_rate": 9.443793757280638e-05, "loss": 0.0707, "step": 19400 }, { "grad_norm": 0.6019229292869568, "learning_rate": 9.443035604648646e-05, "loss": 0.0798, "step": 19410 }, { "grad_norm": 0.646888256072998, "learning_rate": 9.44227696613296e-05, "loss": 0.0584, "step": 19420 }, { "grad_norm": 0.4340895414352417, "learning_rate": 9.441517841816542e-05, "loss": 0.0646, "step": 19430 }, { "grad_norm": 0.9365910887718201, "learning_rate": 9.440758231782413e-05, "loss": 0.0773, "step": 19440 }, { "grad_norm": 0.4547099769115448, "learning_rate": 9.439998136113639e-05, "loss": 0.1007, "step": 19450 }, { "grad_norm": 0.5951218008995056, "learning_rate": 9.439237554893344e-05, "loss": 0.0797, "step": 19460 }, { "grad_norm": 1.1497056484222412, "learning_rate": 9.438476488204705e-05, "loss": 0.0742, "step": 19470 }, { "grad_norm": 0.8016380667686462, "learning_rate": 9.43771493613095e-05, "loss": 0.0821, "step": 19480 }, { "grad_norm": 0.5840169191360474, "learning_rate": 9.436952898755362e-05, "loss": 0.0784, "step": 19490 }, { "grad_norm": 0.7773677706718445, "learning_rate": 9.436190376161276e-05, "loss": 0.0602, "step": 19500 }, { "grad_norm": 0.6560947895050049, "learning_rate": 9.43542736843208e-05, "loss": 0.0826, "step": 19510 }, { "grad_norm": 0.6009579300880432, "learning_rate": 9.434663875651216e-05, "loss": 0.0782, "step": 19520 }, { "grad_norm": 0.8344559073448181, "learning_rate": 9.433899897902177e-05, "loss": 0.0751, "step": 19530 }, { "grad_norm": 0.8894550204277039, "learning_rate": 9.433135435268511e-05, "loss": 0.0746, "step": 19540 }, { "grad_norm": 0.7074988484382629, "learning_rate": 9.432370487833819e-05, "loss": 0.0957, "step": 19550 }, { "grad_norm": 0.9496901035308838, "learning_rate": 9.431605055681756e-05, "loss": 0.067, "step": 19560 }, { "grad_norm": 0.4565162658691406, "learning_rate": 9.430839138896026e-05, "loss": 0.0754, "step": 19570 }, { "grad_norm": 0.5934494137763977, "learning_rate": 9.43007273756039e-05, "loss": 0.0606, "step": 19580 }, { "grad_norm": 0.8041598200798035, "learning_rate": 9.429305851758658e-05, "loss": 0.0613, "step": 19590 }, { "grad_norm": 0.5990252494812012, "learning_rate": 9.428538481574699e-05, "loss": 0.0662, "step": 19600 }, { "grad_norm": 0.6165305376052856, "learning_rate": 9.42777062709243e-05, "loss": 0.0646, "step": 19610 }, { "grad_norm": 0.9616354703903198, "learning_rate": 9.427002288395821e-05, "loss": 0.0793, "step": 19620 }, { "grad_norm": 0.7124348878860474, "learning_rate": 9.426233465568898e-05, "loss": 0.1026, "step": 19630 }, { "grad_norm": 0.5521982312202454, "learning_rate": 9.42546415869574e-05, "loss": 0.0688, "step": 19640 }, { "grad_norm": 0.6784673929214478, "learning_rate": 9.424694367860473e-05, "loss": 0.084, "step": 19650 }, { "grad_norm": 0.6443589925765991, "learning_rate": 9.423924093147284e-05, "loss": 0.0648, "step": 19660 }, { "grad_norm": 1.0140706300735474, "learning_rate": 9.423153334640407e-05, "loss": 0.0953, "step": 19670 }, { "grad_norm": 0.7281126379966736, "learning_rate": 9.42238209242413e-05, "loss": 0.0789, "step": 19680 }, { "grad_norm": 0.863281786441803, "learning_rate": 9.421610366582798e-05, "loss": 0.0938, "step": 19690 }, { "grad_norm": 0.6621860265731812, "learning_rate": 9.420838157200803e-05, "loss": 0.0736, "step": 19700 }, { "grad_norm": 0.46812543272972107, "learning_rate": 9.420065464362594e-05, "loss": 0.0656, "step": 19710 }, { "grad_norm": 0.5114164352416992, "learning_rate": 9.419292288152673e-05, "loss": 0.0666, "step": 19720 }, { "grad_norm": 0.6782495975494385, "learning_rate": 9.418518628655588e-05, "loss": 0.0756, "step": 19730 }, { "grad_norm": 0.7192097902297974, "learning_rate": 9.417744485955951e-05, "loss": 0.064, "step": 19740 }, { "grad_norm": 0.8385806679725647, "learning_rate": 9.41696986013842e-05, "loss": 0.0845, "step": 19750 }, { "grad_norm": 0.7790148854255676, "learning_rate": 9.416194751287705e-05, "loss": 0.0906, "step": 19760 }, { "grad_norm": 1.0929700136184692, "learning_rate": 9.415419159488572e-05, "loss": 0.0693, "step": 19770 }, { "grad_norm": 0.4788677990436554, "learning_rate": 9.414643084825837e-05, "loss": 0.0691, "step": 19780 }, { "grad_norm": 0.7158330678939819, "learning_rate": 9.413866527384372e-05, "loss": 0.078, "step": 19790 }, { "grad_norm": 0.5625753998756409, "learning_rate": 9.4130894872491e-05, "loss": 0.0731, "step": 19800 }, { "grad_norm": 0.6700866222381592, "learning_rate": 9.412311964504998e-05, "loss": 0.0741, "step": 19810 }, { "grad_norm": 0.5703208446502686, "learning_rate": 9.411533959237091e-05, "loss": 0.077, "step": 19820 }, { "grad_norm": 0.5488616228103638, "learning_rate": 9.410755471530464e-05, "loss": 0.0832, "step": 19830 }, { "grad_norm": 0.5154468417167664, "learning_rate": 9.40997650147025e-05, "loss": 0.0905, "step": 19840 }, { "grad_norm": 0.7220809459686279, "learning_rate": 9.409197049141637e-05, "loss": 0.0644, "step": 19850 }, { "grad_norm": 0.6359589099884033, "learning_rate": 9.408417114629863e-05, "loss": 0.0638, "step": 19860 }, { "grad_norm": 0.6413829922676086, "learning_rate": 9.40763669802022e-05, "loss": 0.0604, "step": 19870 }, { "grad_norm": 0.5893056988716125, "learning_rate": 9.406855799398056e-05, "loss": 0.0696, "step": 19880 }, { "grad_norm": 0.6834655404090881, "learning_rate": 9.406074418848767e-05, "loss": 0.0686, "step": 19890 }, { "grad_norm": 0.6837335228919983, "learning_rate": 9.405292556457805e-05, "loss": 0.0808, "step": 19900 }, { "grad_norm": 0.677278459072113, "learning_rate": 9.404510212310671e-05, "loss": 0.0661, "step": 19910 }, { "grad_norm": 0.5838308334350586, "learning_rate": 9.403727386492924e-05, "loss": 0.0558, "step": 19920 }, { "grad_norm": 0.7088913917541504, "learning_rate": 9.40294407909017e-05, "loss": 0.0701, "step": 19930 }, { "grad_norm": 0.6565207839012146, "learning_rate": 9.40216029018807e-05, "loss": 0.0739, "step": 19940 }, { "grad_norm": 0.6821344494819641, "learning_rate": 9.401376019872338e-05, "loss": 0.0708, "step": 19950 }, { "grad_norm": 0.7214193344116211, "learning_rate": 9.400591268228746e-05, "loss": 0.0628, "step": 19960 }, { "grad_norm": 0.557472825050354, "learning_rate": 9.399806035343106e-05, "loss": 0.0759, "step": 19970 }, { "grad_norm": 0.4388158321380615, "learning_rate": 9.399020321301294e-05, "loss": 0.0617, "step": 19980 }, { "grad_norm": 0.6353409290313721, "learning_rate": 9.398234126189234e-05, "loss": 0.0613, "step": 19990 }, { "grad_norm": 0.9256353378295898, "learning_rate": 9.397447450092902e-05, "loss": 0.0812, "step": 20000 }, { "grad_norm": 0.5277315378189087, "learning_rate": 9.39666029309833e-05, "loss": 0.0815, "step": 20010 }, { "grad_norm": 0.6480581760406494, "learning_rate": 9.395872655291596e-05, "loss": 0.0598, "step": 20020 }, { "grad_norm": 0.6098353862762451, "learning_rate": 9.395084536758838e-05, "loss": 0.0625, "step": 20030 }, { "grad_norm": 0.49343299865722656, "learning_rate": 9.394295937586243e-05, "loss": 0.0724, "step": 20040 }, { "grad_norm": 0.6623957753181458, "learning_rate": 9.393506857860052e-05, "loss": 0.067, "step": 20050 }, { "grad_norm": 0.6522277593612671, "learning_rate": 9.392717297666555e-05, "loss": 0.0667, "step": 20060 }, { "grad_norm": 0.6562954783439636, "learning_rate": 9.391927257092101e-05, "loss": 0.0835, "step": 20070 }, { "grad_norm": 0.48688364028930664, "learning_rate": 9.391136736223085e-05, "loss": 0.0705, "step": 20080 }, { "grad_norm": 0.6847462058067322, "learning_rate": 9.390345735145956e-05, "loss": 0.079, "step": 20090 }, { "grad_norm": 0.6428698897361755, "learning_rate": 9.389554253947219e-05, "loss": 0.0642, "step": 20100 }, { "grad_norm": 0.4947228729724884, "learning_rate": 9.388762292713428e-05, "loss": 0.0763, "step": 20110 }, { "grad_norm": 0.526415228843689, "learning_rate": 9.38796985153119e-05, "loss": 0.0635, "step": 20120 }, { "grad_norm": 0.47035956382751465, "learning_rate": 9.387176930487169e-05, "loss": 0.0513, "step": 20130 }, { "grad_norm": 0.500321090221405, "learning_rate": 9.386383529668072e-05, "loss": 0.0491, "step": 20140 }, { "grad_norm": 0.7681546211242676, "learning_rate": 9.385589649160669e-05, "loss": 0.0674, "step": 20150 }, { "grad_norm": 0.4898097515106201, "learning_rate": 9.384795289051775e-05, "loss": 0.0665, "step": 20160 }, { "grad_norm": 0.4500886797904968, "learning_rate": 9.384000449428261e-05, "loss": 0.0679, "step": 20170 }, { "grad_norm": 0.8239960670471191, "learning_rate": 9.383205130377048e-05, "loss": 0.0778, "step": 20180 }, { "grad_norm": 0.755725085735321, "learning_rate": 9.382409331985114e-05, "loss": 0.0636, "step": 20190 }, { "grad_norm": 0.6015438437461853, "learning_rate": 9.381613054339482e-05, "loss": 0.083, "step": 20200 }, { "grad_norm": 0.5327343940734863, "learning_rate": 9.380816297527235e-05, "loss": 0.0612, "step": 20210 }, { "grad_norm": 0.5154967904090881, "learning_rate": 9.380019061635506e-05, "loss": 0.0613, "step": 20220 }, { "grad_norm": 0.8363960385322571, "learning_rate": 9.379221346751474e-05, "loss": 0.0609, "step": 20230 }, { "grad_norm": 0.621385395526886, "learning_rate": 9.378423152962382e-05, "loss": 0.0814, "step": 20240 }, { "grad_norm": 0.4866364300251007, "learning_rate": 9.377624480355517e-05, "loss": 0.0541, "step": 20250 }, { "grad_norm": 0.7953694462776184, "learning_rate": 9.376825329018219e-05, "loss": 0.0801, "step": 20260 }, { "grad_norm": 0.4820706844329834, "learning_rate": 9.376025699037884e-05, "loss": 0.0597, "step": 20270 }, { "grad_norm": 0.7765020132064819, "learning_rate": 9.37522559050196e-05, "loss": 0.0606, "step": 20280 }, { "grad_norm": 0.4716947674751282, "learning_rate": 9.37442500349794e-05, "loss": 0.0802, "step": 20290 }, { "grad_norm": 0.5268313884735107, "learning_rate": 9.373623938113381e-05, "loss": 0.0846, "step": 20300 }, { "grad_norm": 0.5630910396575928, "learning_rate": 9.372822394435883e-05, "loss": 0.0686, "step": 20310 }, { "grad_norm": 0.5865086317062378, "learning_rate": 9.372020372553102e-05, "loss": 0.0697, "step": 20320 }, { "grad_norm": 0.45547494292259216, "learning_rate": 9.371217872552746e-05, "loss": 0.0615, "step": 20330 }, { "grad_norm": 0.6056851744651794, "learning_rate": 9.370414894522576e-05, "loss": 0.0785, "step": 20340 }, { "grad_norm": 0.8465441465377808, "learning_rate": 9.369611438550406e-05, "loss": 0.0687, "step": 20350 }, { "grad_norm": 0.8069586157798767, "learning_rate": 9.368807504724095e-05, "loss": 0.0821, "step": 20360 }, { "grad_norm": 0.9458903670310974, "learning_rate": 9.368003093131565e-05, "loss": 0.0903, "step": 20370 }, { "grad_norm": 0.8382983207702637, "learning_rate": 9.367198203860785e-05, "loss": 0.1005, "step": 20380 }, { "grad_norm": 0.6527132391929626, "learning_rate": 9.366392836999774e-05, "loss": 0.0783, "step": 20390 }, { "grad_norm": 0.6217461228370667, "learning_rate": 9.365586992636607e-05, "loss": 0.0755, "step": 20400 }, { "grad_norm": 0.657349705696106, "learning_rate": 9.364780670859412e-05, "loss": 0.0707, "step": 20410 }, { "grad_norm": 0.5906326770782471, "learning_rate": 9.363973871756364e-05, "loss": 0.0645, "step": 20420 }, { "grad_norm": 0.6483169794082642, "learning_rate": 9.363166595415696e-05, "loss": 0.0631, "step": 20430 }, { "grad_norm": 0.5064703226089478, "learning_rate": 9.362358841925686e-05, "loss": 0.0673, "step": 20440 }, { "grad_norm": 0.9301286935806274, "learning_rate": 9.361550611374674e-05, "loss": 0.0645, "step": 20450 }, { "grad_norm": 1.1171391010284424, "learning_rate": 9.360741903851043e-05, "loss": 0.0807, "step": 20460 }, { "grad_norm": 0.8455895185470581, "learning_rate": 9.359932719443236e-05, "loss": 0.0736, "step": 20470 }, { "grad_norm": 0.7977123260498047, "learning_rate": 9.35912305823974e-05, "loss": 0.067, "step": 20480 }, { "grad_norm": 0.4711049199104309, "learning_rate": 9.358312920329101e-05, "loss": 0.0651, "step": 20490 }, { "grad_norm": 0.7093694806098938, "learning_rate": 9.357502305799914e-05, "loss": 0.0658, "step": 20500 }, { "grad_norm": 0.5943201780319214, "learning_rate": 9.356691214740824e-05, "loss": 0.053, "step": 20510 }, { "grad_norm": 0.597891092300415, "learning_rate": 9.355879647240535e-05, "loss": 0.0618, "step": 20520 }, { "grad_norm": 0.8547161817550659, "learning_rate": 9.355067603387798e-05, "loss": 0.0629, "step": 20530 }, { "grad_norm": 0.4886120557785034, "learning_rate": 9.354255083271412e-05, "loss": 0.0693, "step": 20540 }, { "grad_norm": 0.6908599734306335, "learning_rate": 9.353442086980239e-05, "loss": 0.0851, "step": 20550 }, { "grad_norm": 0.978649377822876, "learning_rate": 9.352628614603185e-05, "loss": 0.0905, "step": 20560 }, { "grad_norm": 0.7981446385383606, "learning_rate": 9.351814666229209e-05, "loss": 0.1055, "step": 20570 }, { "grad_norm": 0.4966045022010803, "learning_rate": 9.351000241947324e-05, "loss": 0.059, "step": 20580 }, { "grad_norm": 0.720550537109375, "learning_rate": 9.350185341846594e-05, "loss": 0.0767, "step": 20590 }, { "grad_norm": 0.7338313460350037, "learning_rate": 9.349369966016134e-05, "loss": 0.0767, "step": 20600 }, { "grad_norm": 0.49836498498916626, "learning_rate": 9.348554114545117e-05, "loss": 0.0531, "step": 20610 }, { "grad_norm": 0.7041211724281311, "learning_rate": 9.347737787522758e-05, "loss": 0.0601, "step": 20620 }, { "grad_norm": 0.6660584211349487, "learning_rate": 9.346920985038332e-05, "loss": 0.0757, "step": 20630 }, { "grad_norm": 0.7479256987571716, "learning_rate": 9.346103707181162e-05, "loss": 0.062, "step": 20640 }, { "grad_norm": 0.6866773366928101, "learning_rate": 9.345285954040626e-05, "loss": 0.0639, "step": 20650 }, { "grad_norm": 0.6550526022911072, "learning_rate": 9.34446772570615e-05, "loss": 0.0701, "step": 20660 }, { "grad_norm": 0.568451464176178, "learning_rate": 9.343649022267214e-05, "loss": 0.0459, "step": 20670 }, { "grad_norm": 0.5948096513748169, "learning_rate": 9.342829843813353e-05, "loss": 0.063, "step": 20680 }, { "grad_norm": 0.7023323774337769, "learning_rate": 9.342010190434149e-05, "loss": 0.0731, "step": 20690 }, { "grad_norm": 0.6038087010383606, "learning_rate": 9.34119006221924e-05, "loss": 0.0752, "step": 20700 }, { "grad_norm": 0.6946277618408203, "learning_rate": 9.340369459258313e-05, "loss": 0.0658, "step": 20710 }, { "grad_norm": 0.8837080597877502, "learning_rate": 9.339548381641106e-05, "loss": 0.0627, "step": 20720 }, { "grad_norm": 0.6244024038314819, "learning_rate": 9.338726829457413e-05, "loss": 0.0615, "step": 20730 }, { "grad_norm": 0.43551039695739746, "learning_rate": 9.337904802797078e-05, "loss": 0.0545, "step": 20740 }, { "grad_norm": 0.7951133847236633, "learning_rate": 9.337082301749993e-05, "loss": 0.0661, "step": 20750 }, { "grad_norm": 0.6330073475837708, "learning_rate": 9.336259326406109e-05, "loss": 0.0753, "step": 20760 }, { "grad_norm": 0.6072526574134827, "learning_rate": 9.335435876855427e-05, "loss": 0.0589, "step": 20770 }, { "grad_norm": 1.0788524150848389, "learning_rate": 9.334611953187994e-05, "loss": 0.0674, "step": 20780 }, { "grad_norm": 0.8213870525360107, "learning_rate": 9.333787555493914e-05, "loss": 0.0733, "step": 20790 }, { "grad_norm": 0.5934633016586304, "learning_rate": 9.332962683863345e-05, "loss": 0.0569, "step": 20800 }, { "grad_norm": 0.6668462157249451, "learning_rate": 9.332137338386489e-05, "loss": 0.0716, "step": 20810 }, { "grad_norm": 0.7812184691429138, "learning_rate": 9.33131151915361e-05, "loss": 0.0666, "step": 20820 }, { "grad_norm": 0.6573068499565125, "learning_rate": 9.330485226255012e-05, "loss": 0.0812, "step": 20830 }, { "grad_norm": 0.6548001766204834, "learning_rate": 9.329658459781061e-05, "loss": 0.0797, "step": 20840 }, { "grad_norm": 0.6049080491065979, "learning_rate": 9.328831219822172e-05, "loss": 0.0645, "step": 20850 }, { "grad_norm": 0.5765124559402466, "learning_rate": 9.328003506468808e-05, "loss": 0.0746, "step": 20860 }, { "grad_norm": 1.103316068649292, "learning_rate": 9.327175319811488e-05, "loss": 0.0581, "step": 20870 }, { "grad_norm": 0.5546075701713562, "learning_rate": 9.326346659940781e-05, "loss": 0.0683, "step": 20880 }, { "grad_norm": 0.8457241058349609, "learning_rate": 9.325517526947308e-05, "loss": 0.0663, "step": 20890 }, { "grad_norm": 0.6957375407218933, "learning_rate": 9.32468792092174e-05, "loss": 0.0798, "step": 20900 }, { "grad_norm": 0.7080093026161194, "learning_rate": 9.323857841954803e-05, "loss": 0.0756, "step": 20910 }, { "grad_norm": 0.8519935607910156, "learning_rate": 9.323027290137276e-05, "loss": 0.0669, "step": 20920 }, { "grad_norm": 0.5943576693534851, "learning_rate": 9.322196265559981e-05, "loss": 0.0727, "step": 20930 }, { "grad_norm": 0.8394585251808167, "learning_rate": 9.321364768313803e-05, "loss": 0.0623, "step": 20940 }, { "grad_norm": 0.40489834547042847, "learning_rate": 9.32053279848967e-05, "loss": 0.0615, "step": 20950 }, { "grad_norm": 0.5827263593673706, "learning_rate": 9.319700356178567e-05, "loss": 0.069, "step": 20960 }, { "grad_norm": 0.7171061635017395, "learning_rate": 9.318867441471527e-05, "loss": 0.0607, "step": 20970 }, { "grad_norm": 0.724163293838501, "learning_rate": 9.318034054459637e-05, "loss": 0.0685, "step": 20980 }, { "grad_norm": 0.6458821296691895, "learning_rate": 9.317200195234034e-05, "loss": 0.0651, "step": 20990 }, { "grad_norm": 0.6794959306716919, "learning_rate": 9.316365863885909e-05, "loss": 0.0633, "step": 21000 }, { "grad_norm": 0.7218928933143616, "learning_rate": 9.315531060506502e-05, "loss": 0.0738, "step": 21010 }, { "grad_norm": 0.7362763285636902, "learning_rate": 9.314695785187108e-05, "loss": 0.0822, "step": 21020 }, { "grad_norm": 0.6330583095550537, "learning_rate": 9.313860038019069e-05, "loss": 0.0689, "step": 21030 }, { "grad_norm": 0.5771594643592834, "learning_rate": 9.313023819093782e-05, "loss": 0.0581, "step": 21040 }, { "grad_norm": 0.8374747037887573, "learning_rate": 9.312187128502695e-05, "loss": 0.0803, "step": 21050 }, { "grad_norm": 0.6199407577514648, "learning_rate": 9.311349966337307e-05, "loss": 0.0659, "step": 21060 }, { "grad_norm": 0.7632842063903809, "learning_rate": 9.310512332689169e-05, "loss": 0.0594, "step": 21070 }, { "grad_norm": 0.6503345966339111, "learning_rate": 9.309674227649883e-05, "loss": 0.0643, "step": 21080 }, { "grad_norm": 0.6333267688751221, "learning_rate": 9.308835651311103e-05, "loss": 0.0685, "step": 21090 }, { "grad_norm": 1.2127634286880493, "learning_rate": 9.307996603764533e-05, "loss": 0.076, "step": 21100 }, { "grad_norm": 0.48798948526382446, "learning_rate": 9.307157085101932e-05, "loss": 0.0629, "step": 21110 }, { "grad_norm": 0.5585361123085022, "learning_rate": 9.306317095415109e-05, "loss": 0.0649, "step": 21120 }, { "grad_norm": 0.7350255250930786, "learning_rate": 9.305476634795922e-05, "loss": 0.0712, "step": 21130 }, { "grad_norm": 0.489985853433609, "learning_rate": 9.304635703336284e-05, "loss": 0.0552, "step": 21140 }, { "grad_norm": 0.7781439423561096, "learning_rate": 9.303794301128157e-05, "loss": 0.0721, "step": 21150 }, { "grad_norm": 0.2997679114341736, "learning_rate": 9.302952428263555e-05, "loss": 0.0622, "step": 21160 }, { "grad_norm": 0.7135196924209595, "learning_rate": 9.302110084834545e-05, "loss": 0.073, "step": 21170 }, { "grad_norm": 0.6414480805397034, "learning_rate": 9.301267270933245e-05, "loss": 0.0693, "step": 21180 }, { "grad_norm": 0.7487824559211731, "learning_rate": 9.300423986651823e-05, "loss": 0.0672, "step": 21190 }, { "grad_norm": 0.5533331632614136, "learning_rate": 9.299580232082501e-05, "loss": 0.0737, "step": 21200 }, { "grad_norm": 0.7707563042640686, "learning_rate": 9.298736007317547e-05, "loss": 0.1058, "step": 21210 }, { "grad_norm": 0.7741224765777588, "learning_rate": 9.297891312449288e-05, "loss": 0.0687, "step": 21220 }, { "grad_norm": 0.7646015286445618, "learning_rate": 9.297046147570094e-05, "loss": 0.0771, "step": 21230 }, { "grad_norm": 0.7793241143226624, "learning_rate": 9.296200512772396e-05, "loss": 0.0859, "step": 21240 }, { "grad_norm": 0.6035143136978149, "learning_rate": 9.295354408148668e-05, "loss": 0.0517, "step": 21250 }, { "grad_norm": 0.7328526377677917, "learning_rate": 9.294507833791441e-05, "loss": 0.0552, "step": 21260 }, { "grad_norm": 0.5406626462936401, "learning_rate": 9.293660789793295e-05, "loss": 0.0592, "step": 21270 }, { "grad_norm": 0.4497734308242798, "learning_rate": 9.292813276246858e-05, "loss": 0.0651, "step": 21280 }, { "grad_norm": 0.6221569776535034, "learning_rate": 9.291965293244816e-05, "loss": 0.0833, "step": 21290 }, { "grad_norm": 0.5060731172561646, "learning_rate": 9.291116840879904e-05, "loss": 0.0621, "step": 21300 }, { "grad_norm": 0.8561057448387146, "learning_rate": 9.290267919244904e-05, "loss": 0.0836, "step": 21310 }, { "grad_norm": 0.6860404014587402, "learning_rate": 9.289418528432655e-05, "loss": 0.0623, "step": 21320 }, { "grad_norm": 0.847003698348999, "learning_rate": 9.288568668536045e-05, "loss": 0.0744, "step": 21330 }, { "grad_norm": 0.5440164804458618, "learning_rate": 9.287718339648013e-05, "loss": 0.0648, "step": 21340 }, { "grad_norm": 0.3462541997432709, "learning_rate": 9.28686754186155e-05, "loss": 0.0531, "step": 21350 }, { "grad_norm": 0.6173624396324158, "learning_rate": 9.286016275269698e-05, "loss": 0.0683, "step": 21360 }, { "grad_norm": 0.49404123425483704, "learning_rate": 9.285164539965551e-05, "loss": 0.0696, "step": 21370 }, { "grad_norm": 0.38882145285606384, "learning_rate": 9.284312336042251e-05, "loss": 0.0448, "step": 21380 }, { "grad_norm": 0.7210344076156616, "learning_rate": 9.283459663592996e-05, "loss": 0.0674, "step": 21390 }, { "grad_norm": 0.6405932903289795, "learning_rate": 9.282606522711033e-05, "loss": 0.1178, "step": 21400 }, { "grad_norm": 0.5778178572654724, "learning_rate": 9.281752913489657e-05, "loss": 0.0874, "step": 21410 }, { "grad_norm": 1.3827402591705322, "learning_rate": 9.280898836022222e-05, "loss": 0.079, "step": 21420 }, { "grad_norm": 0.8565956950187683, "learning_rate": 9.280044290402126e-05, "loss": 0.0732, "step": 21430 }, { "grad_norm": 0.790266215801239, "learning_rate": 9.279189276722821e-05, "loss": 0.0803, "step": 21440 }, { "grad_norm": 0.6222357153892517, "learning_rate": 9.278333795077812e-05, "loss": 0.0652, "step": 21450 }, { "grad_norm": 0.5489981770515442, "learning_rate": 9.27747784556065e-05, "loss": 0.0482, "step": 21460 }, { "grad_norm": 0.6510983109474182, "learning_rate": 9.276621428264942e-05, "loss": 0.0664, "step": 21470 }, { "grad_norm": 0.6983004212379456, "learning_rate": 9.275764543284345e-05, "loss": 0.0523, "step": 21480 }, { "grad_norm": 0.6345512866973877, "learning_rate": 9.274907190712566e-05, "loss": 0.0739, "step": 21490 }, { "grad_norm": 0.805196225643158, "learning_rate": 9.274049370643363e-05, "loss": 0.0655, "step": 21500 }, { "grad_norm": 0.6808098554611206, "learning_rate": 9.273191083170547e-05, "loss": 0.0816, "step": 21510 }, { "grad_norm": 0.5600252151489258, "learning_rate": 9.27233232838798e-05, "loss": 0.0626, "step": 21520 }, { "grad_norm": 0.4688268303871155, "learning_rate": 9.27147310638957e-05, "loss": 0.0556, "step": 21530 }, { "grad_norm": 0.8594339489936829, "learning_rate": 9.270613417269286e-05, "loss": 0.0869, "step": 21540 }, { "grad_norm": 0.6407378911972046, "learning_rate": 9.269753261121138e-05, "loss": 0.0627, "step": 21550 }, { "grad_norm": 0.6601231098175049, "learning_rate": 9.268892638039194e-05, "loss": 0.0831, "step": 21560 }, { "grad_norm": 0.7517446875572205, "learning_rate": 9.268031548117569e-05, "loss": 0.0631, "step": 21570 }, { "grad_norm": 0.5630958676338196, "learning_rate": 9.26716999145043e-05, "loss": 0.0748, "step": 21580 }, { "grad_norm": 0.4768742322921753, "learning_rate": 9.266307968131998e-05, "loss": 0.0715, "step": 21590 }, { "grad_norm": 0.5742219686508179, "learning_rate": 9.26544547825654e-05, "loss": 0.0753, "step": 21600 }, { "grad_norm": 0.5169433951377869, "learning_rate": 9.264582521918376e-05, "loss": 0.0603, "step": 21610 }, { "grad_norm": 0.9329814910888672, "learning_rate": 9.263719099211881e-05, "loss": 0.0649, "step": 21620 }, { "grad_norm": 0.6815803647041321, "learning_rate": 9.262855210231476e-05, "loss": 0.0778, "step": 21630 }, { "grad_norm": 0.6849743723869324, "learning_rate": 9.261990855071633e-05, "loss": 0.0605, "step": 21640 }, { "grad_norm": 0.8793437480926514, "learning_rate": 9.261126033826878e-05, "loss": 0.0776, "step": 21650 }, { "grad_norm": 0.481534481048584, "learning_rate": 9.260260746591786e-05, "loss": 0.0626, "step": 21660 }, { "grad_norm": 0.7647196054458618, "learning_rate": 9.259394993460985e-05, "loss": 0.1053, "step": 21670 }, { "grad_norm": 0.395953893661499, "learning_rate": 9.258528774529151e-05, "loss": 0.0526, "step": 21680 }, { "grad_norm": 0.7965511679649353, "learning_rate": 9.257662089891013e-05, "loss": 0.0605, "step": 21690 }, { "grad_norm": 0.6802946329116821, "learning_rate": 9.25679493964135e-05, "loss": 0.0651, "step": 21700 }, { "grad_norm": 0.7031601667404175, "learning_rate": 9.255927323874994e-05, "loss": 0.0586, "step": 21710 }, { "grad_norm": 0.5374963879585266, "learning_rate": 9.255059242686822e-05, "loss": 0.0668, "step": 21720 }, { "grad_norm": 0.7910909652709961, "learning_rate": 9.254190696171769e-05, "loss": 0.0676, "step": 21730 }, { "grad_norm": 0.9036080837249756, "learning_rate": 9.25332168442482e-05, "loss": 0.0786, "step": 21740 }, { "grad_norm": 0.4070473611354828, "learning_rate": 9.252452207541004e-05, "loss": 0.0945, "step": 21750 }, { "grad_norm": 0.5068053007125854, "learning_rate": 9.251582265615409e-05, "loss": 0.063, "step": 21760 }, { "grad_norm": 0.6998125910758972, "learning_rate": 9.250711858743169e-05, "loss": 0.0715, "step": 21770 }, { "grad_norm": 1.374099612236023, "learning_rate": 9.24984098701947e-05, "loss": 0.0802, "step": 21780 }, { "grad_norm": 0.8777163028717041, "learning_rate": 9.248969650539552e-05, "loss": 0.0931, "step": 21790 }, { "grad_norm": 0.5948116183280945, "learning_rate": 9.2480978493987e-05, "loss": 0.0726, "step": 21800 }, { "grad_norm": 0.6302632093429565, "learning_rate": 9.247225583692256e-05, "loss": 0.0708, "step": 21810 }, { "grad_norm": 0.4255381226539612, "learning_rate": 9.246352853515607e-05, "loss": 0.0721, "step": 21820 }, { "grad_norm": 0.5181814432144165, "learning_rate": 9.245479658964194e-05, "loss": 0.0642, "step": 21830 }, { "grad_norm": 0.6620521545410156, "learning_rate": 9.244606000133507e-05, "loss": 0.0673, "step": 21840 }, { "grad_norm": 0.45525217056274414, "learning_rate": 9.24373187711909e-05, "loss": 0.0778, "step": 21850 }, { "grad_norm": 0.8264591693878174, "learning_rate": 9.242857290016537e-05, "loss": 0.0824, "step": 21860 }, { "grad_norm": 0.7267661094665527, "learning_rate": 9.241982238921488e-05, "loss": 0.065, "step": 21870 }, { "grad_norm": 0.74007648229599, "learning_rate": 9.24110672392964e-05, "loss": 0.081, "step": 21880 }, { "grad_norm": 0.5320773124694824, "learning_rate": 9.240230745136737e-05, "loss": 0.0624, "step": 21890 }, { "grad_norm": 1.0289782285690308, "learning_rate": 9.239354302638575e-05, "loss": 0.0839, "step": 21900 }, { "grad_norm": 1.0451452732086182, "learning_rate": 9.238477396531e-05, "loss": 0.0635, "step": 21910 }, { "grad_norm": 0.5228879451751709, "learning_rate": 9.23760002690991e-05, "loss": 0.0781, "step": 21920 }, { "grad_norm": 0.6643615365028381, "learning_rate": 9.236722193871252e-05, "loss": 0.0734, "step": 21930 }, { "grad_norm": 0.501960813999176, "learning_rate": 9.235843897511023e-05, "loss": 0.0705, "step": 21940 }, { "grad_norm": 0.6557395458221436, "learning_rate": 9.234965137925276e-05, "loss": 0.0827, "step": 21950 }, { "grad_norm": 0.4243808388710022, "learning_rate": 9.234085915210108e-05, "loss": 0.055, "step": 21960 }, { "grad_norm": 0.7003279328346252, "learning_rate": 9.23320622946167e-05, "loss": 0.0636, "step": 21970 }, { "grad_norm": 0.5870492458343506, "learning_rate": 9.232326080776163e-05, "loss": 0.0634, "step": 21980 }, { "grad_norm": 0.8256074786186218, "learning_rate": 9.23144546924984e-05, "loss": 0.062, "step": 21990 }, { "grad_norm": 0.7804890275001526, "learning_rate": 9.230564394979e-05, "loss": 0.0735, "step": 22000 }, { "grad_norm": 0.6844106912612915, "learning_rate": 9.22968285806e-05, "loss": 0.0764, "step": 22010 }, { "grad_norm": 0.7597667574882507, "learning_rate": 9.228800858589242e-05, "loss": 0.0618, "step": 22020 }, { "grad_norm": 0.7145110368728638, "learning_rate": 9.227918396663179e-05, "loss": 0.0721, "step": 22030 }, { "grad_norm": 0.8700379729270935, "learning_rate": 9.227035472378319e-05, "loss": 0.0774, "step": 22040 }, { "grad_norm": 0.47112905979156494, "learning_rate": 9.226152085831213e-05, "loss": 0.0683, "step": 22050 }, { "grad_norm": 0.8206712007522583, "learning_rate": 9.22526823711847e-05, "loss": 0.07, "step": 22060 }, { "grad_norm": 0.5935795903205872, "learning_rate": 9.224383926336745e-05, "loss": 0.0592, "step": 22070 }, { "grad_norm": 0.5620988011360168, "learning_rate": 9.223499153582744e-05, "loss": 0.0722, "step": 22080 }, { "grad_norm": 0.8616687059402466, "learning_rate": 9.222613918953226e-05, "loss": 0.0655, "step": 22090 }, { "grad_norm": 0.7958621978759766, "learning_rate": 9.221728222544999e-05, "loss": 0.0586, "step": 22100 }, { "grad_norm": 0.8249159455299377, "learning_rate": 9.22084206445492e-05, "loss": 0.0675, "step": 22110 }, { "grad_norm": 0.7576160430908203, "learning_rate": 9.2199554447799e-05, "loss": 0.0905, "step": 22120 }, { "grad_norm": 0.41845640540122986, "learning_rate": 9.219068363616897e-05, "loss": 0.0565, "step": 22130 }, { "grad_norm": 0.9312517046928406, "learning_rate": 9.218180821062919e-05, "loss": 0.0637, "step": 22140 }, { "grad_norm": 0.6939126253128052, "learning_rate": 9.21729281721503e-05, "loss": 0.0541, "step": 22150 }, { "grad_norm": 0.6183778047561646, "learning_rate": 9.216404352170339e-05, "loss": 0.07, "step": 22160 }, { "grad_norm": 0.4154135584831238, "learning_rate": 9.215515426026007e-05, "loss": 0.0503, "step": 22170 }, { "grad_norm": 0.6992517113685608, "learning_rate": 9.214626038879246e-05, "loss": 0.0712, "step": 22180 }, { "grad_norm": 0.5527384877204895, "learning_rate": 9.21373619082732e-05, "loss": 0.061, "step": 22190 }, { "grad_norm": 0.431938111782074, "learning_rate": 9.212845881967535e-05, "loss": 0.0799, "step": 22200 }, { "grad_norm": 0.4399835765361786, "learning_rate": 9.211955112397262e-05, "loss": 0.0552, "step": 22210 }, { "grad_norm": 0.6523917317390442, "learning_rate": 9.211063882213909e-05, "loss": 0.0797, "step": 22220 }, { "grad_norm": 0.7454336285591125, "learning_rate": 9.210172191514942e-05, "loss": 0.0612, "step": 22230 }, { "grad_norm": 0.7331944704055786, "learning_rate": 9.209280040397874e-05, "loss": 0.069, "step": 22240 }, { "grad_norm": 0.5205386281013489, "learning_rate": 9.208387428960268e-05, "loss": 0.0772, "step": 22250 }, { "grad_norm": 0.6818968057632446, "learning_rate": 9.20749435729974e-05, "loss": 0.0637, "step": 22260 }, { "grad_norm": 0.5920016169548035, "learning_rate": 9.206600825513957e-05, "loss": 0.055, "step": 22270 }, { "grad_norm": 0.4699495732784271, "learning_rate": 9.20570683370063e-05, "loss": 0.0585, "step": 22280 }, { "grad_norm": 0.5228455662727356, "learning_rate": 9.204812381957528e-05, "loss": 0.0483, "step": 22290 }, { "grad_norm": 0.4606221318244934, "learning_rate": 9.203917470382465e-05, "loss": 0.0585, "step": 22300 }, { "grad_norm": 0.66716068983078, "learning_rate": 9.203022099073309e-05, "loss": 0.0652, "step": 22310 }, { "grad_norm": 0.770646333694458, "learning_rate": 9.202126268127976e-05, "loss": 0.0638, "step": 22320 }, { "grad_norm": 0.55039381980896, "learning_rate": 9.20122997764443e-05, "loss": 0.0512, "step": 22330 }, { "grad_norm": 0.8704005479812622, "learning_rate": 9.200333227720692e-05, "loss": 0.0643, "step": 22340 }, { "grad_norm": 0.9771534204483032, "learning_rate": 9.199436018454826e-05, "loss": 0.0692, "step": 22350 }, { "grad_norm": 0.7392579913139343, "learning_rate": 9.198538349944952e-05, "loss": 0.0605, "step": 22360 }, { "grad_norm": 0.9413859844207764, "learning_rate": 9.197640222289234e-05, "loss": 0.0748, "step": 22370 }, { "grad_norm": 1.16837477684021, "learning_rate": 9.196741635585895e-05, "loss": 0.0636, "step": 22380 }, { "grad_norm": 0.8234198689460754, "learning_rate": 9.195842589933199e-05, "loss": 0.0801, "step": 22390 }, { "grad_norm": 0.9466294050216675, "learning_rate": 9.194943085429466e-05, "loss": 0.0734, "step": 22400 }, { "grad_norm": 0.4336681365966797, "learning_rate": 9.194043122173065e-05, "loss": 0.0627, "step": 22410 }, { "grad_norm": 0.53693026304245, "learning_rate": 9.193142700262413e-05, "loss": 0.0763, "step": 22420 }, { "grad_norm": 0.6142138838768005, "learning_rate": 9.192241819795979e-05, "loss": 0.0602, "step": 22430 }, { "grad_norm": 0.6155266761779785, "learning_rate": 9.191340480872284e-05, "loss": 0.0682, "step": 22440 }, { "grad_norm": 0.5574639439582825, "learning_rate": 9.190438683589895e-05, "loss": 0.0829, "step": 22450 }, { "grad_norm": 0.876837968826294, "learning_rate": 9.189536428047432e-05, "loss": 0.0636, "step": 22460 }, { "grad_norm": 0.696510910987854, "learning_rate": 9.188633714343564e-05, "loss": 0.0636, "step": 22470 }, { "grad_norm": 0.6962957382202148, "learning_rate": 9.18773054257701e-05, "loss": 0.0625, "step": 22480 }, { "grad_norm": 0.9393858909606934, "learning_rate": 9.18682691284654e-05, "loss": 0.0592, "step": 22490 }, { "grad_norm": 0.6542838215827942, "learning_rate": 9.185922825250974e-05, "loss": 0.0624, "step": 22500 }, { "grad_norm": 0.9957099556922913, "learning_rate": 9.185018279889181e-05, "loss": 0.0536, "step": 22510 }, { "grad_norm": 0.5954249501228333, "learning_rate": 9.184113276860082e-05, "loss": 0.0753, "step": 22520 }, { "grad_norm": 0.8189108371734619, "learning_rate": 9.183207816262645e-05, "loss": 0.0641, "step": 22530 }, { "grad_norm": 0.8419960737228394, "learning_rate": 9.182301898195891e-05, "loss": 0.0606, "step": 22540 }, { "grad_norm": 0.6132464408874512, "learning_rate": 9.181395522758889e-05, "loss": 0.0626, "step": 22550 }, { "grad_norm": 0.840308666229248, "learning_rate": 9.180488690050759e-05, "loss": 0.0928, "step": 22560 }, { "grad_norm": 0.7627367377281189, "learning_rate": 9.179581400170671e-05, "loss": 0.0782, "step": 22570 }, { "grad_norm": 0.5554839968681335, "learning_rate": 9.178673653217845e-05, "loss": 0.0677, "step": 22580 }, { "grad_norm": 0.7630854845046997, "learning_rate": 9.177765449291551e-05, "loss": 0.0638, "step": 22590 }, { "grad_norm": 0.5388658046722412, "learning_rate": 9.176856788491109e-05, "loss": 0.0682, "step": 22600 }, { "grad_norm": 0.5412654876708984, "learning_rate": 9.175947670915887e-05, "loss": 0.0714, "step": 22610 }, { "grad_norm": 0.6970741748809814, "learning_rate": 9.175038096665309e-05, "loss": 0.0784, "step": 22620 }, { "grad_norm": 0.6488847732543945, "learning_rate": 9.17412806583884e-05, "loss": 0.057, "step": 22630 }, { "grad_norm": 0.6649056673049927, "learning_rate": 9.173217578536002e-05, "loss": 0.0643, "step": 22640 }, { "grad_norm": 0.6357215642929077, "learning_rate": 9.172306634856362e-05, "loss": 0.0524, "step": 22650 }, { "grad_norm": 0.5224043726921082, "learning_rate": 9.171395234899545e-05, "loss": 0.0637, "step": 22660 }, { "grad_norm": 0.7341600060462952, "learning_rate": 9.170483378765214e-05, "loss": 0.0528, "step": 22670 }, { "grad_norm": 0.6549896001815796, "learning_rate": 9.169571066553091e-05, "loss": 0.0526, "step": 22680 }, { "grad_norm": 0.6148679852485657, "learning_rate": 9.168658298362946e-05, "loss": 0.0825, "step": 22690 }, { "grad_norm": 0.8138225078582764, "learning_rate": 9.167745074294598e-05, "loss": 0.0788, "step": 22700 }, { "grad_norm": 0.6711809635162354, "learning_rate": 9.166831394447913e-05, "loss": 0.0598, "step": 22710 }, { "grad_norm": 0.7543305158615112, "learning_rate": 9.165917258922812e-05, "loss": 0.0718, "step": 22720 }, { "grad_norm": 0.5909258723258972, "learning_rate": 9.165002667819262e-05, "loss": 0.0581, "step": 22730 }, { "grad_norm": 0.5558348298072815, "learning_rate": 9.164087621237282e-05, "loss": 0.0589, "step": 22740 }, { "grad_norm": 0.47499093413352966, "learning_rate": 9.163172119276942e-05, "loss": 0.0655, "step": 22750 }, { "grad_norm": 0.44947269558906555, "learning_rate": 9.162256162038358e-05, "loss": 0.0635, "step": 22760 }, { "grad_norm": 0.6758842468261719, "learning_rate": 9.161339749621698e-05, "loss": 0.0879, "step": 22770 }, { "grad_norm": 0.586625337600708, "learning_rate": 9.160422882127177e-05, "loss": 0.0691, "step": 22780 }, { "grad_norm": 0.7599595785140991, "learning_rate": 9.159505559655069e-05, "loss": 0.0601, "step": 22790 }, { "grad_norm": 0.9523987174034119, "learning_rate": 9.158587782305684e-05, "loss": 0.0781, "step": 22800 }, { "grad_norm": 0.6905624270439148, "learning_rate": 9.157669550179391e-05, "loss": 0.0546, "step": 22810 }, { "grad_norm": 0.6507166624069214, "learning_rate": 9.156750863376609e-05, "loss": 0.064, "step": 22820 }, { "grad_norm": 0.6958456635475159, "learning_rate": 9.155831721997801e-05, "loss": 0.1007, "step": 22830 }, { "grad_norm": 0.512998104095459, "learning_rate": 9.154912126143484e-05, "loss": 0.0582, "step": 22840 }, { "grad_norm": 0.600456178188324, "learning_rate": 9.153992075914224e-05, "loss": 0.0623, "step": 22850 }, { "grad_norm": 0.5575225353240967, "learning_rate": 9.153071571410635e-05, "loss": 0.0586, "step": 22860 }, { "grad_norm": 0.5463181734085083, "learning_rate": 9.152150612733384e-05, "loss": 0.0649, "step": 22870 }, { "grad_norm": 0.693956196308136, "learning_rate": 9.151229199983184e-05, "loss": 0.0604, "step": 22880 }, { "grad_norm": 0.8800989985466003, "learning_rate": 9.150307333260802e-05, "loss": 0.0612, "step": 22890 }, { "grad_norm": 0.7637301087379456, "learning_rate": 9.149385012667048e-05, "loss": 0.0801, "step": 22900 }, { "grad_norm": 0.6915245056152344, "learning_rate": 9.148462238302788e-05, "loss": 0.0957, "step": 22910 }, { "grad_norm": 0.5938403606414795, "learning_rate": 9.147539010268936e-05, "loss": 0.0674, "step": 22920 }, { "grad_norm": 0.9505122900009155, "learning_rate": 9.14661532866645e-05, "loss": 0.0591, "step": 22930 }, { "grad_norm": 0.7260526418685913, "learning_rate": 9.145691193596348e-05, "loss": 0.0583, "step": 22940 }, { "grad_norm": 0.9404434561729431, "learning_rate": 9.144766605159691e-05, "loss": 0.0594, "step": 22950 }, { "grad_norm": 0.6660282611846924, "learning_rate": 9.14384156345759e-05, "loss": 0.0677, "step": 22960 }, { "grad_norm": 0.7782296538352966, "learning_rate": 9.142916068591204e-05, "loss": 0.0707, "step": 22970 }, { "grad_norm": 0.704304039478302, "learning_rate": 9.141990120661746e-05, "loss": 0.0695, "step": 22980 }, { "grad_norm": 0.6009057760238647, "learning_rate": 9.141063719770475e-05, "loss": 0.0488, "step": 22990 }, { "grad_norm": 0.8448660969734192, "learning_rate": 9.140136866018704e-05, "loss": 0.0677, "step": 23000 }, { "grad_norm": 0.5349968671798706, "learning_rate": 9.139209559507788e-05, "loss": 0.0502, "step": 23010 }, { "grad_norm": 1.3334448337554932, "learning_rate": 9.13828180033914e-05, "loss": 0.0687, "step": 23020 }, { "grad_norm": 0.5881000757217407, "learning_rate": 9.137353588614212e-05, "loss": 0.0511, "step": 23030 }, { "grad_norm": 0.4724942445755005, "learning_rate": 9.136424924434519e-05, "loss": 0.0604, "step": 23040 }, { "grad_norm": 0.4593295454978943, "learning_rate": 9.135495807901615e-05, "loss": 0.0558, "step": 23050 }, { "grad_norm": 0.47979241609573364, "learning_rate": 9.134566239117108e-05, "loss": 0.0682, "step": 23060 }, { "grad_norm": 0.7914753556251526, "learning_rate": 9.13363621818265e-05, "loss": 0.0645, "step": 23070 }, { "grad_norm": 0.43168899416923523, "learning_rate": 9.132705745199953e-05, "loss": 0.0588, "step": 23080 }, { "grad_norm": 0.590835690498352, "learning_rate": 9.131774820270768e-05, "loss": 0.0676, "step": 23090 }, { "grad_norm": 0.9251859188079834, "learning_rate": 9.130843443496901e-05, "loss": 0.0821, "step": 23100 }, { "grad_norm": 0.8632352352142334, "learning_rate": 9.129911614980206e-05, "loss": 0.0685, "step": 23110 }, { "grad_norm": 0.688255250453949, "learning_rate": 9.128979334822584e-05, "loss": 0.0644, "step": 23120 }, { "grad_norm": 0.84086674451828, "learning_rate": 9.128046603125992e-05, "loss": 0.0624, "step": 23130 }, { "grad_norm": 1.0998644828796387, "learning_rate": 9.12711341999243e-05, "loss": 0.0537, "step": 23140 }, { "grad_norm": 0.5848618149757385, "learning_rate": 9.12617978552395e-05, "loss": 0.0647, "step": 23150 }, { "grad_norm": 0.636387050151825, "learning_rate": 9.12524569982265e-05, "loss": 0.0772, "step": 23160 }, { "grad_norm": 0.7811180949211121, "learning_rate": 9.124311162990684e-05, "loss": 0.0664, "step": 23170 }, { "grad_norm": 0.7811311483383179, "learning_rate": 9.12337617513025e-05, "loss": 0.063, "step": 23180 }, { "grad_norm": 0.4157959222793579, "learning_rate": 9.122440736343596e-05, "loss": 0.0646, "step": 23190 }, { "grad_norm": 0.9182662963867188, "learning_rate": 9.12150484673302e-05, "loss": 0.0628, "step": 23200 }, { "grad_norm": 0.5941351056098938, "learning_rate": 9.120568506400873e-05, "loss": 0.0975, "step": 23210 }, { "grad_norm": 0.9328193664550781, "learning_rate": 9.119631715449548e-05, "loss": 0.0752, "step": 23220 }, { "grad_norm": 0.6612783074378967, "learning_rate": 9.118694473981493e-05, "loss": 0.0798, "step": 23230 }, { "grad_norm": 0.9656277298927307, "learning_rate": 9.117756782099203e-05, "loss": 0.0756, "step": 23240 }, { "grad_norm": 0.7415328621864319, "learning_rate": 9.11681863990522e-05, "loss": 0.071, "step": 23250 }, { "grad_norm": 0.7339551448822021, "learning_rate": 9.115880047502142e-05, "loss": 0.1044, "step": 23260 }, { "grad_norm": 0.8234159350395203, "learning_rate": 9.114941004992609e-05, "loss": 0.0671, "step": 23270 }, { "grad_norm": 0.8419848680496216, "learning_rate": 9.114001512479317e-05, "loss": 0.0823, "step": 23280 }, { "grad_norm": 0.5416736006736755, "learning_rate": 9.113061570065003e-05, "loss": 0.0751, "step": 23290 }, { "grad_norm": 0.7302373647689819, "learning_rate": 9.112121177852459e-05, "loss": 0.0714, "step": 23300 }, { "grad_norm": 0.6167539358139038, "learning_rate": 9.111180335944527e-05, "loss": 0.0637, "step": 23310 }, { "grad_norm": 0.4859737753868103, "learning_rate": 9.110239044444093e-05, "loss": 0.077, "step": 23320 }, { "grad_norm": 0.9934116005897522, "learning_rate": 9.109297303454099e-05, "loss": 0.0804, "step": 23330 }, { "grad_norm": 0.4652949869632721, "learning_rate": 9.108355113077526e-05, "loss": 0.0505, "step": 23340 }, { "grad_norm": 0.6209690570831299, "learning_rate": 9.107412473417419e-05, "loss": 0.0769, "step": 23350 }, { "grad_norm": 0.4016770124435425, "learning_rate": 9.106469384576858e-05, "loss": 0.0547, "step": 23360 }, { "grad_norm": 0.4962335228919983, "learning_rate": 9.105525846658978e-05, "loss": 0.0647, "step": 23370 }, { "grad_norm": 0.730532705783844, "learning_rate": 9.104581859766965e-05, "loss": 0.0578, "step": 23380 }, { "grad_norm": 0.7000329494476318, "learning_rate": 9.10363742400405e-05, "loss": 0.0755, "step": 23390 }, { "grad_norm": 0.5898085236549377, "learning_rate": 9.102692539473518e-05, "loss": 0.0652, "step": 23400 }, { "grad_norm": 1.208073377609253, "learning_rate": 9.101747206278697e-05, "loss": 0.0765, "step": 23410 }, { "grad_norm": 0.638715386390686, "learning_rate": 9.100801424522968e-05, "loss": 0.0578, "step": 23420 }, { "grad_norm": 0.6057698726654053, "learning_rate": 9.099855194309762e-05, "loss": 0.0724, "step": 23430 }, { "grad_norm": 0.8928183913230896, "learning_rate": 9.098908515742554e-05, "loss": 0.068, "step": 23440 }, { "grad_norm": 0.8096154928207397, "learning_rate": 9.097961388924873e-05, "loss": 0.0643, "step": 23450 }, { "grad_norm": 0.8206090927124023, "learning_rate": 9.097013813960298e-05, "loss": 0.0704, "step": 23460 }, { "grad_norm": 0.6431007385253906, "learning_rate": 9.09606579095245e-05, "loss": 0.0699, "step": 23470 }, { "grad_norm": 0.6411623954772949, "learning_rate": 9.095117320005008e-05, "loss": 0.0726, "step": 23480 }, { "grad_norm": 0.7363631129264832, "learning_rate": 9.094168401221691e-05, "loss": 0.0555, "step": 23490 }, { "grad_norm": 0.5619300603866577, "learning_rate": 9.093219034706273e-05, "loss": 0.0913, "step": 23500 }, { "grad_norm": 0.46712616086006165, "learning_rate": 9.092269220562577e-05, "loss": 0.0703, "step": 23510 }, { "grad_norm": 1.1559200286865234, "learning_rate": 9.09131895889447e-05, "loss": 0.066, "step": 23520 }, { "grad_norm": 0.599989116191864, "learning_rate": 9.090368249805873e-05, "loss": 0.0794, "step": 23530 }, { "grad_norm": 0.6798908114433289, "learning_rate": 9.089417093400754e-05, "loss": 0.0681, "step": 23540 }, { "grad_norm": 0.6959998607635498, "learning_rate": 9.088465489783131e-05, "loss": 0.0655, "step": 23550 }, { "grad_norm": 0.9796645641326904, "learning_rate": 9.087513439057068e-05, "loss": 0.0684, "step": 23560 }, { "grad_norm": 0.5023664236068726, "learning_rate": 9.08656094132668e-05, "loss": 0.0549, "step": 23570 }, { "grad_norm": 0.8425535559654236, "learning_rate": 9.085607996696134e-05, "loss": 0.0734, "step": 23580 }, { "grad_norm": 0.7205772995948792, "learning_rate": 9.084654605269639e-05, "loss": 0.0806, "step": 23590 }, { "grad_norm": 0.5663341879844666, "learning_rate": 9.083700767151457e-05, "loss": 0.0667, "step": 23600 }, { "grad_norm": 0.7528102397918701, "learning_rate": 9.082746482445898e-05, "loss": 0.0761, "step": 23610 }, { "grad_norm": 0.5089343190193176, "learning_rate": 9.081791751257325e-05, "loss": 0.0589, "step": 23620 }, { "grad_norm": 0.7109576463699341, "learning_rate": 9.080836573690142e-05, "loss": 0.0664, "step": 23630 }, { "grad_norm": 0.7260392308235168, "learning_rate": 9.079880949848805e-05, "loss": 0.0675, "step": 23640 }, { "grad_norm": 0.5574293732643127, "learning_rate": 9.078924879837822e-05, "loss": 0.0583, "step": 23650 }, { "grad_norm": 0.8391837477684021, "learning_rate": 9.077968363761747e-05, "loss": 0.1014, "step": 23660 }, { "grad_norm": 0.4931475818157196, "learning_rate": 9.077011401725182e-05, "loss": 0.0505, "step": 23670 }, { "grad_norm": 0.8088796138763428, "learning_rate": 9.07605399383278e-05, "loss": 0.0617, "step": 23680 }, { "grad_norm": 0.6794323325157166, "learning_rate": 9.075096140189243e-05, "loss": 0.0664, "step": 23690 }, { "grad_norm": 0.6424477696418762, "learning_rate": 9.074137840899318e-05, "loss": 0.0647, "step": 23700 }, { "grad_norm": 0.8732838034629822, "learning_rate": 9.073179096067804e-05, "loss": 0.0884, "step": 23710 }, { "grad_norm": 0.6536474227905273, "learning_rate": 9.072219905799549e-05, "loss": 0.0624, "step": 23720 }, { "grad_norm": 0.7098111510276794, "learning_rate": 9.071260270199447e-05, "loss": 0.0652, "step": 23730 }, { "grad_norm": 0.8677237033843994, "learning_rate": 9.070300189372441e-05, "loss": 0.0626, "step": 23740 }, { "grad_norm": 0.46448051929473877, "learning_rate": 9.069339663423528e-05, "loss": 0.0942, "step": 23750 }, { "grad_norm": 0.7387637495994568, "learning_rate": 9.068378692457747e-05, "loss": 0.0644, "step": 23760 }, { "grad_norm": 0.49123695492744446, "learning_rate": 9.067417276580189e-05, "loss": 0.0525, "step": 23770 }, { "grad_norm": 0.8878092765808105, "learning_rate": 9.066455415895993e-05, "loss": 0.0666, "step": 23780 }, { "grad_norm": 0.7218703031539917, "learning_rate": 9.065493110510346e-05, "loss": 0.0578, "step": 23790 }, { "grad_norm": 1.0473780632019043, "learning_rate": 9.064530360528484e-05, "loss": 0.0611, "step": 23800 }, { "grad_norm": 0.6416307091712952, "learning_rate": 9.063567166055695e-05, "loss": 0.0648, "step": 23810 }, { "grad_norm": 0.641293466091156, "learning_rate": 9.062603527197308e-05, "loss": 0.0558, "step": 23820 }, { "grad_norm": 0.5800656080245972, "learning_rate": 9.06163944405871e-05, "loss": 0.0555, "step": 23830 }, { "grad_norm": 0.525452733039856, "learning_rate": 9.060674916745327e-05, "loss": 0.0638, "step": 23840 }, { "grad_norm": 0.7784101963043213, "learning_rate": 9.05970994536264e-05, "loss": 0.0648, "step": 23850 }, { "grad_norm": 0.5378946661949158, "learning_rate": 9.05874453001618e-05, "loss": 0.0576, "step": 23860 }, { "grad_norm": 0.8003596067428589, "learning_rate": 9.057778670811517e-05, "loss": 0.0617, "step": 23870 }, { "grad_norm": 0.6583264470100403, "learning_rate": 9.056812367854281e-05, "loss": 0.0616, "step": 23880 }, { "grad_norm": 0.685175359249115, "learning_rate": 9.055845621250143e-05, "loss": 0.0536, "step": 23890 }, { "grad_norm": 0.8145262598991394, "learning_rate": 9.054878431104825e-05, "loss": 0.0567, "step": 23900 }, { "grad_norm": 0.7206791639328003, "learning_rate": 9.0539107975241e-05, "loss": 0.0555, "step": 23910 }, { "grad_norm": 0.8585787415504456, "learning_rate": 9.052942720613784e-05, "loss": 0.0754, "step": 23920 }, { "grad_norm": 0.6848154067993164, "learning_rate": 9.051974200479745e-05, "loss": 0.0764, "step": 23930 }, { "grad_norm": 0.5644232034683228, "learning_rate": 9.051005237227901e-05, "loss": 0.0636, "step": 23940 }, { "grad_norm": 0.6744145154953003, "learning_rate": 9.050035830964215e-05, "loss": 0.0761, "step": 23950 }, { "grad_norm": 0.818295955657959, "learning_rate": 9.049065981794698e-05, "loss": 0.0757, "step": 23960 }, { "grad_norm": 0.6438801884651184, "learning_rate": 9.048095689825414e-05, "loss": 0.0632, "step": 23970 }, { "grad_norm": 0.9889392852783203, "learning_rate": 9.047124955162472e-05, "loss": 0.068, "step": 23980 }, { "grad_norm": 0.4530384838581085, "learning_rate": 9.046153777912028e-05, "loss": 0.0855, "step": 23990 }, { "grad_norm": 0.8981839418411255, "learning_rate": 9.045182158180292e-05, "loss": 0.0667, "step": 24000 }, { "grad_norm": 0.6424546241760254, "learning_rate": 9.044210096073516e-05, "loss": 0.0646, "step": 24010 }, { "grad_norm": 0.6575665473937988, "learning_rate": 9.043237591698004e-05, "loss": 0.0584, "step": 24020 }, { "grad_norm": 0.707368791103363, "learning_rate": 9.04226464516011e-05, "loss": 0.0596, "step": 24030 }, { "grad_norm": 0.7310400009155273, "learning_rate": 9.041291256566229e-05, "loss": 0.0699, "step": 24040 }, { "grad_norm": 0.7429332137107849, "learning_rate": 9.040317426022814e-05, "loss": 0.0573, "step": 24050 }, { "grad_norm": 0.7353325486183167, "learning_rate": 9.03934315363636e-05, "loss": 0.0576, "step": 24060 }, { "grad_norm": 0.7505713105201721, "learning_rate": 9.038368439513409e-05, "loss": 0.0671, "step": 24070 }, { "grad_norm": 0.7181944251060486, "learning_rate": 9.03739328376056e-05, "loss": 0.0706, "step": 24080 }, { "grad_norm": 0.6639470458030701, "learning_rate": 9.036417686484451e-05, "loss": 0.0529, "step": 24090 }, { "grad_norm": 0.6106414198875427, "learning_rate": 9.035441647791773e-05, "loss": 0.0792, "step": 24100 }, { "grad_norm": 0.48896732926368713, "learning_rate": 9.034465167789263e-05, "loss": 0.0555, "step": 24110 }, { "grad_norm": 0.585541307926178, "learning_rate": 9.033488246583706e-05, "loss": 0.0597, "step": 24120 }, { "grad_norm": 0.7296209931373596, "learning_rate": 9.032510884281941e-05, "loss": 0.0766, "step": 24130 }, { "grad_norm": 0.6197822093963623, "learning_rate": 9.031533080990848e-05, "loss": 0.085, "step": 24140 }, { "grad_norm": 0.6341955065727234, "learning_rate": 9.030554836817358e-05, "loss": 0.0668, "step": 24150 }, { "grad_norm": 0.6780478358268738, "learning_rate": 9.029576151868451e-05, "loss": 0.072, "step": 24160 }, { "grad_norm": 0.46869921684265137, "learning_rate": 9.028597026251155e-05, "loss": 0.0573, "step": 24170 }, { "grad_norm": 0.749461829662323, "learning_rate": 9.027617460072547e-05, "loss": 0.0634, "step": 24180 }, { "grad_norm": 0.7517626881599426, "learning_rate": 9.026637453439745e-05, "loss": 0.0635, "step": 24190 }, { "grad_norm": 0.584732174873352, "learning_rate": 9.025657006459927e-05, "loss": 0.079, "step": 24200 }, { "grad_norm": 0.6655400395393372, "learning_rate": 9.024676119240311e-05, "loss": 0.0624, "step": 24210 }, { "grad_norm": 0.9703962802886963, "learning_rate": 9.023694791888166e-05, "loss": 0.0659, "step": 24220 }, { "grad_norm": 0.7433769702911377, "learning_rate": 9.022713024510808e-05, "loss": 0.0514, "step": 24230 }, { "grad_norm": 0.5952943563461304, "learning_rate": 9.021730817215601e-05, "loss": 0.0611, "step": 24240 }, { "grad_norm": 0.9283833503723145, "learning_rate": 9.02074817010996e-05, "loss": 0.0633, "step": 24250 }, { "grad_norm": 0.7150196433067322, "learning_rate": 9.019765083301342e-05, "loss": 0.0665, "step": 24260 }, { "grad_norm": 0.7181621193885803, "learning_rate": 9.01878155689726e-05, "loss": 0.0651, "step": 24270 }, { "grad_norm": 0.6158758401870728, "learning_rate": 9.017797591005268e-05, "loss": 0.0549, "step": 24280 }, { "grad_norm": 0.6585310101509094, "learning_rate": 9.016813185732972e-05, "loss": 0.0642, "step": 24290 }, { "grad_norm": 0.6659776568412781, "learning_rate": 9.015828341188027e-05, "loss": 0.0619, "step": 24300 }, { "grad_norm": 0.5005221366882324, "learning_rate": 9.01484305747813e-05, "loss": 0.0521, "step": 24310 }, { "grad_norm": 0.8580551147460938, "learning_rate": 9.013857334711033e-05, "loss": 0.051, "step": 24320 }, { "grad_norm": 0.8401547074317932, "learning_rate": 9.012871172994534e-05, "loss": 0.0573, "step": 24330 }, { "grad_norm": 0.5233162641525269, "learning_rate": 9.011884572436476e-05, "loss": 0.0555, "step": 24340 }, { "grad_norm": 0.6383221745491028, "learning_rate": 9.010897533144754e-05, "loss": 0.0692, "step": 24350 }, { "grad_norm": 0.7774398326873779, "learning_rate": 9.009910055227306e-05, "loss": 0.0881, "step": 24360 }, { "grad_norm": 0.8058040142059326, "learning_rate": 9.008922138792124e-05, "loss": 0.0606, "step": 24370 }, { "grad_norm": 0.7294807434082031, "learning_rate": 9.007933783947244e-05, "loss": 0.0557, "step": 24380 }, { "grad_norm": 0.8332152366638184, "learning_rate": 9.006944990800752e-05, "loss": 0.0748, "step": 24390 }, { "grad_norm": 0.6870911121368408, "learning_rate": 9.005955759460779e-05, "loss": 0.0511, "step": 24400 }, { "grad_norm": 0.6162073016166687, "learning_rate": 9.004966090035508e-05, "loss": 0.0984, "step": 24410 }, { "grad_norm": 0.6803450584411621, "learning_rate": 9.003975982633166e-05, "loss": 0.0546, "step": 24420 }, { "grad_norm": 0.7720158696174622, "learning_rate": 9.00298543736203e-05, "loss": 0.0581, "step": 24430 }, { "grad_norm": 0.5255580544471741, "learning_rate": 9.001994454330427e-05, "loss": 0.0657, "step": 24440 }, { "grad_norm": 0.5880337357521057, "learning_rate": 9.001003033646727e-05, "loss": 0.0612, "step": 24450 }, { "grad_norm": 0.4981938302516937, "learning_rate": 9.00001117541935e-05, "loss": 0.0552, "step": 24460 }, { "grad_norm": 0.5410014986991882, "learning_rate": 8.999018879756764e-05, "loss": 0.0548, "step": 24470 }, { "grad_norm": 0.5419895648956299, "learning_rate": 8.998026146767487e-05, "loss": 0.0533, "step": 24480 }, { "grad_norm": 0.5413182377815247, "learning_rate": 8.99703297656008e-05, "loss": 0.095, "step": 24490 }, { "grad_norm": 0.6550939679145813, "learning_rate": 8.996039369243156e-05, "loss": 0.0576, "step": 24500 }, { "grad_norm": 0.4858931601047516, "learning_rate": 8.995045324925378e-05, "loss": 0.0819, "step": 24510 }, { "grad_norm": 0.4785076677799225, "learning_rate": 8.994050843715448e-05, "loss": 0.0637, "step": 24520 }, { "grad_norm": 0.5795877575874329, "learning_rate": 8.993055925722121e-05, "loss": 0.0638, "step": 24530 }, { "grad_norm": 0.9629910588264465, "learning_rate": 8.992060571054202e-05, "loss": 0.0635, "step": 24540 }, { "grad_norm": 1.0495637655258179, "learning_rate": 8.991064779820542e-05, "loss": 0.0785, "step": 24550 }, { "grad_norm": 0.5861210823059082, "learning_rate": 8.990068552130036e-05, "loss": 0.0631, "step": 24560 }, { "grad_norm": 0.6286914944648743, "learning_rate": 8.989071888091634e-05, "loss": 0.0833, "step": 24570 }, { "grad_norm": 0.7486906051635742, "learning_rate": 8.988074787814329e-05, "loss": 0.063, "step": 24580 }, { "grad_norm": 0.5242853164672852, "learning_rate": 8.987077251407158e-05, "loss": 0.0597, "step": 24590 }, { "grad_norm": 0.5678353905677795, "learning_rate": 8.986079278979216e-05, "loss": 0.0622, "step": 24600 }, { "grad_norm": 0.5418411493301392, "learning_rate": 8.985080870639635e-05, "loss": 0.0665, "step": 24610 }, { "grad_norm": 0.8488007187843323, "learning_rate": 8.984082026497603e-05, "loss": 0.0563, "step": 24620 }, { "grad_norm": 0.4960850775241852, "learning_rate": 8.98308274666235e-05, "loss": 0.07, "step": 24630 }, { "grad_norm": 0.6929486393928528, "learning_rate": 8.982083031243155e-05, "loss": 0.0655, "step": 24640 }, { "grad_norm": 0.7567822933197021, "learning_rate": 8.98108288034935e-05, "loss": 0.0566, "step": 24650 }, { "grad_norm": 0.6338359713554382, "learning_rate": 8.980082294090305e-05, "loss": 0.0599, "step": 24660 }, { "grad_norm": 0.5246230363845825, "learning_rate": 8.979081272575443e-05, "loss": 0.0641, "step": 24670 }, { "grad_norm": 0.8579446077346802, "learning_rate": 8.978079815914236e-05, "loss": 0.0816, "step": 24680 }, { "grad_norm": 0.5597295761108398, "learning_rate": 8.977077924216202e-05, "loss": 0.0498, "step": 24690 }, { "grad_norm": 0.5707972049713135, "learning_rate": 8.976075597590905e-05, "loss": 0.055, "step": 24700 }, { "grad_norm": 0.5062350630760193, "learning_rate": 8.975072836147958e-05, "loss": 0.0848, "step": 24710 }, { "grad_norm": 0.4785536229610443, "learning_rate": 8.974069639997025e-05, "loss": 0.0472, "step": 24720 }, { "grad_norm": 0.7539515495300293, "learning_rate": 8.973066009247808e-05, "loss": 0.0605, "step": 24730 }, { "grad_norm": 0.4827932119369507, "learning_rate": 8.972061944010066e-05, "loss": 0.0604, "step": 24740 }, { "grad_norm": 0.9260759949684143, "learning_rate": 8.971057444393603e-05, "loss": 0.0616, "step": 24750 }, { "grad_norm": 0.7664723992347717, "learning_rate": 8.970052510508268e-05, "loss": 0.0674, "step": 24760 }, { "grad_norm": 0.7338983416557312, "learning_rate": 8.969047142463959e-05, "loss": 0.0675, "step": 24770 }, { "grad_norm": 0.5097230076789856, "learning_rate": 8.968041340370621e-05, "loss": 0.0613, "step": 24780 }, { "grad_norm": 0.6566788554191589, "learning_rate": 8.96703510433825e-05, "loss": 0.0702, "step": 24790 }, { "grad_norm": 0.6694848537445068, "learning_rate": 8.966028434476883e-05, "loss": 0.0642, "step": 24800 }, { "grad_norm": 0.5032482147216797, "learning_rate": 8.96502133089661e-05, "loss": 0.0614, "step": 24810 }, { "grad_norm": 0.4669320285320282, "learning_rate": 8.964013793707564e-05, "loss": 0.0714, "step": 24820 }, { "grad_norm": 0.44156360626220703, "learning_rate": 8.963005823019932e-05, "loss": 0.0499, "step": 24830 }, { "grad_norm": 0.7757032513618469, "learning_rate": 8.961997418943939e-05, "loss": 0.0531, "step": 24840 }, { "grad_norm": 0.9081595540046692, "learning_rate": 8.960988581589865e-05, "loss": 0.0633, "step": 24850 }, { "grad_norm": 0.7449161410331726, "learning_rate": 8.959979311068037e-05, "loss": 0.0848, "step": 24860 }, { "grad_norm": 0.9306915402412415, "learning_rate": 8.958969607488823e-05, "loss": 0.0882, "step": 24870 }, { "grad_norm": 0.6741805076599121, "learning_rate": 8.957959470962647e-05, "loss": 0.0642, "step": 24880 }, { "grad_norm": 0.668030321598053, "learning_rate": 8.956948901599971e-05, "loss": 0.0638, "step": 24890 }, { "grad_norm": 0.678317666053772, "learning_rate": 8.955937899511315e-05, "loss": 0.064, "step": 24900 }, { "grad_norm": 0.44022390246391296, "learning_rate": 8.954926464807238e-05, "loss": 0.0484, "step": 24910 }, { "grad_norm": 0.5923506617546082, "learning_rate": 8.953914597598347e-05, "loss": 0.087, "step": 24920 }, { "grad_norm": 0.4135302007198334, "learning_rate": 8.952902297995303e-05, "loss": 0.0623, "step": 24930 }, { "grad_norm": 0.7872441411018372, "learning_rate": 8.951889566108804e-05, "loss": 0.0582, "step": 24940 }, { "grad_norm": 0.6047751903533936, "learning_rate": 8.950876402049606e-05, "loss": 0.0456, "step": 24950 }, { "grad_norm": 0.6373208165168762, "learning_rate": 8.949862805928504e-05, "loss": 0.0577, "step": 24960 }, { "grad_norm": 0.8113589286804199, "learning_rate": 8.948848777856343e-05, "loss": 0.0541, "step": 24970 }, { "grad_norm": 0.5133482813835144, "learning_rate": 8.947834317944017e-05, "loss": 0.0668, "step": 24980 }, { "grad_norm": 0.46956539154052734, "learning_rate": 8.946819426302466e-05, "loss": 0.0741, "step": 24990 }, { "grad_norm": 0.6398188471794128, "learning_rate": 8.945804103042676e-05, "loss": 0.0528, "step": 25000 }, { "grad_norm": 0.8350642919540405, "learning_rate": 8.944788348275681e-05, "loss": 0.0726, "step": 25010 }, { "grad_norm": 1.2943702936172485, "learning_rate": 8.943772162112565e-05, "loss": 0.0896, "step": 25020 }, { "grad_norm": 0.4798538386821747, "learning_rate": 8.942755544664454e-05, "loss": 0.0769, "step": 25030 }, { "grad_norm": 0.6562946438789368, "learning_rate": 8.941738496042525e-05, "loss": 0.079, "step": 25040 }, { "grad_norm": 0.6160877346992493, "learning_rate": 8.940721016357999e-05, "loss": 0.0787, "step": 25050 }, { "grad_norm": 0.7209897637367249, "learning_rate": 8.939703105722148e-05, "loss": 0.0772, "step": 25060 }, { "grad_norm": 0.9392722249031067, "learning_rate": 8.93868476424629e-05, "loss": 0.0659, "step": 25070 }, { "grad_norm": 0.6363813281059265, "learning_rate": 8.937665992041786e-05, "loss": 0.0747, "step": 25080 }, { "grad_norm": 0.9134360551834106, "learning_rate": 8.93664678922005e-05, "loss": 0.0783, "step": 25090 }, { "grad_norm": 0.7172461152076721, "learning_rate": 8.93562715589254e-05, "loss": 0.068, "step": 25100 }, { "grad_norm": 0.5489200353622437, "learning_rate": 8.934607092170762e-05, "loss": 0.0765, "step": 25110 }, { "grad_norm": 0.7275508642196655, "learning_rate": 8.933586598166266e-05, "loss": 0.0644, "step": 25120 }, { "grad_norm": 0.6946519017219543, "learning_rate": 8.932565673990655e-05, "loss": 0.0621, "step": 25130 }, { "grad_norm": 0.7024709582328796, "learning_rate": 8.931544319755574e-05, "loss": 0.0707, "step": 25140 }, { "grad_norm": 0.7229775190353394, "learning_rate": 8.930522535572718e-05, "loss": 0.063, "step": 25150 }, { "grad_norm": 0.6615909337997437, "learning_rate": 8.929500321553826e-05, "loss": 0.0558, "step": 25160 }, { "grad_norm": 0.8475241661071777, "learning_rate": 8.928477677810686e-05, "loss": 0.0794, "step": 25170 }, { "grad_norm": 0.4571586549282074, "learning_rate": 8.927454604455137e-05, "loss": 0.0615, "step": 25180 }, { "grad_norm": 0.8811157941818237, "learning_rate": 8.926431101599053e-05, "loss": 0.0714, "step": 25190 }, { "grad_norm": 0.7182996273040771, "learning_rate": 8.925407169354369e-05, "loss": 0.0751, "step": 25200 }, { "grad_norm": 0.830603301525116, "learning_rate": 8.92438280783306e-05, "loss": 0.0716, "step": 25210 }, { "grad_norm": 0.6894407868385315, "learning_rate": 8.923358017147146e-05, "loss": 0.0654, "step": 25220 }, { "grad_norm": 0.6888595819473267, "learning_rate": 8.922332797408697e-05, "loss": 0.0547, "step": 25230 }, { "grad_norm": 0.5827785730361938, "learning_rate": 8.921307148729831e-05, "loss": 0.0659, "step": 25240 }, { "grad_norm": 0.43079525232315063, "learning_rate": 8.920281071222712e-05, "loss": 0.0471, "step": 25250 }, { "grad_norm": 0.8017891049385071, "learning_rate": 8.919254564999548e-05, "loss": 0.0623, "step": 25260 }, { "grad_norm": 0.5953913331031799, "learning_rate": 8.918227630172598e-05, "loss": 0.067, "step": 25270 }, { "grad_norm": 0.6348592042922974, "learning_rate": 8.917200266854165e-05, "loss": 0.0878, "step": 25280 }, { "grad_norm": 0.6576595902442932, "learning_rate": 8.9161724751566e-05, "loss": 0.058, "step": 25290 }, { "grad_norm": 0.5059027671813965, "learning_rate": 8.915144255192302e-05, "loss": 0.0642, "step": 25300 }, { "grad_norm": 0.5715941786766052, "learning_rate": 8.914115607073714e-05, "loss": 0.0541, "step": 25310 }, { "grad_norm": 0.6209056377410889, "learning_rate": 8.913086530913327e-05, "loss": 0.0789, "step": 25320 }, { "grad_norm": 0.5844419598579407, "learning_rate": 8.912057026823681e-05, "loss": 0.0588, "step": 25330 }, { "grad_norm": 0.7624979615211487, "learning_rate": 8.91102709491736e-05, "loss": 0.0529, "step": 25340 }, { "grad_norm": 0.5179833173751831, "learning_rate": 8.909996735306996e-05, "loss": 0.0571, "step": 25350 }, { "grad_norm": 0.7304280996322632, "learning_rate": 8.908965948105268e-05, "loss": 0.0802, "step": 25360 }, { "grad_norm": 0.5667169690132141, "learning_rate": 8.907934733424901e-05, "loss": 0.0514, "step": 25370 }, { "grad_norm": 0.6839239597320557, "learning_rate": 8.906903091378666e-05, "loss": 0.069, "step": 25380 }, { "grad_norm": 1.0235018730163574, "learning_rate": 8.905871022079384e-05, "loss": 0.0666, "step": 25390 }, { "grad_norm": 0.7368860244750977, "learning_rate": 8.90483852563992e-05, "loss": 0.0582, "step": 25400 }, { "grad_norm": 0.9292412400245667, "learning_rate": 8.903805602173185e-05, "loss": 0.0589, "step": 25410 }, { "grad_norm": 0.7935663461685181, "learning_rate": 8.902772251792137e-05, "loss": 0.0664, "step": 25420 }, { "grad_norm": 0.6191421747207642, "learning_rate": 8.901738474609786e-05, "loss": 0.0601, "step": 25430 }, { "grad_norm": 0.4822882115840912, "learning_rate": 8.900704270739179e-05, "loss": 0.0568, "step": 25440 }, { "grad_norm": 0.4957992732524872, "learning_rate": 8.89966964029342e-05, "loss": 0.0569, "step": 25450 }, { "grad_norm": 0.6494478583335876, "learning_rate": 8.898634583385652e-05, "loss": 0.0704, "step": 25460 }, { "grad_norm": 0.6804738640785217, "learning_rate": 8.897599100129065e-05, "loss": 0.0628, "step": 25470 }, { "grad_norm": 0.6247587203979492, "learning_rate": 8.896563190636903e-05, "loss": 0.0611, "step": 25480 }, { "grad_norm": 0.6760870814323425, "learning_rate": 8.895526855022448e-05, "loss": 0.0691, "step": 25490 }, { "grad_norm": 0.4723418951034546, "learning_rate": 8.894490093399033e-05, "loss": 0.0521, "step": 25500 }, { "grad_norm": 0.490833580493927, "learning_rate": 8.893452905880035e-05, "loss": 0.0796, "step": 25510 }, { "grad_norm": 0.6552218198776245, "learning_rate": 8.892415292578883e-05, "loss": 0.0781, "step": 25520 }, { "grad_norm": 0.5590020418167114, "learning_rate": 8.891377253609046e-05, "loss": 0.0834, "step": 25530 }, { "grad_norm": 0.5238093137741089, "learning_rate": 8.890338789084043e-05, "loss": 0.0573, "step": 25540 }, { "grad_norm": 0.8654107451438904, "learning_rate": 8.88929989911744e-05, "loss": 0.0717, "step": 25550 }, { "grad_norm": 0.7310697436332703, "learning_rate": 8.888260583822847e-05, "loss": 0.0691, "step": 25560 }, { "grad_norm": 0.8601935505867004, "learning_rate": 8.887220843313921e-05, "loss": 0.0752, "step": 25570 }, { "grad_norm": 0.5654410719871521, "learning_rate": 8.88618067770437e-05, "loss": 0.0566, "step": 25580 }, { "grad_norm": 0.7051360011100769, "learning_rate": 8.885140087107942e-05, "loss": 0.0625, "step": 25590 }, { "grad_norm": 0.8893915414810181, "learning_rate": 8.884099071638436e-05, "loss": 0.0659, "step": 25600 }, { "grad_norm": 0.615837574005127, "learning_rate": 8.883057631409695e-05, "loss": 0.0665, "step": 25610 }, { "grad_norm": 0.42727699875831604, "learning_rate": 8.882015766535608e-05, "loss": 0.0622, "step": 25620 }, { "grad_norm": 0.6052654385566711, "learning_rate": 8.880973477130115e-05, "loss": 0.0637, "step": 25630 }, { "grad_norm": 0.4960401654243469, "learning_rate": 8.879930763307197e-05, "loss": 0.0569, "step": 25640 }, { "grad_norm": 1.0042595863342285, "learning_rate": 8.878887625180884e-05, "loss": 0.0662, "step": 25650 }, { "grad_norm": 0.8665496706962585, "learning_rate": 8.877844062865253e-05, "loss": 0.0696, "step": 25660 }, { "grad_norm": 0.5248518586158752, "learning_rate": 8.876800076474424e-05, "loss": 0.0531, "step": 25670 }, { "grad_norm": 0.46433156728744507, "learning_rate": 8.875755666122568e-05, "loss": 0.0567, "step": 25680 }, { "grad_norm": 0.5506015419960022, "learning_rate": 8.8747108319239e-05, "loss": 0.0487, "step": 25690 }, { "grad_norm": 0.39078906178474426, "learning_rate": 8.87366557399268e-05, "loss": 0.0523, "step": 25700 }, { "grad_norm": 0.6113458275794983, "learning_rate": 8.872619892443217e-05, "loss": 0.072, "step": 25710 }, { "grad_norm": 0.35400402545928955, "learning_rate": 8.871573787389865e-05, "loss": 0.0674, "step": 25720 }, { "grad_norm": 0.494435578584671, "learning_rate": 8.870527258947024e-05, "loss": 0.05, "step": 25730 }, { "grad_norm": 0.6917860507965088, "learning_rate": 8.869480307229143e-05, "loss": 0.0624, "step": 25740 }, { "grad_norm": 0.7795498371124268, "learning_rate": 8.868432932350712e-05, "loss": 0.0536, "step": 25750 }, { "grad_norm": 0.5286410450935364, "learning_rate": 8.867385134426272e-05, "loss": 0.0559, "step": 25760 }, { "grad_norm": 0.6708279848098755, "learning_rate": 8.866336913570407e-05, "loss": 0.0581, "step": 25770 }, { "grad_norm": 0.6142275929450989, "learning_rate": 8.865288269897751e-05, "loss": 0.06, "step": 25780 }, { "grad_norm": 0.5021533966064453, "learning_rate": 8.864239203522981e-05, "loss": 0.0721, "step": 25790 }, { "grad_norm": 0.5349394679069519, "learning_rate": 8.863189714560822e-05, "loss": 0.0811, "step": 25800 }, { "grad_norm": 0.6557825803756714, "learning_rate": 8.862139803126043e-05, "loss": 0.0691, "step": 25810 }, { "grad_norm": 0.8920680284500122, "learning_rate": 8.861089469333463e-05, "loss": 0.0622, "step": 25820 }, { "grad_norm": 0.4949592649936676, "learning_rate": 8.860038713297944e-05, "loss": 0.0554, "step": 25830 }, { "grad_norm": 0.602886974811554, "learning_rate": 8.858987535134394e-05, "loss": 0.0607, "step": 25840 }, { "grad_norm": 0.7104763388633728, "learning_rate": 8.857935934957769e-05, "loss": 0.0588, "step": 25850 }, { "grad_norm": 0.8207781314849854, "learning_rate": 8.856883912883071e-05, "loss": 0.0714, "step": 25860 }, { "grad_norm": 1.0519353151321411, "learning_rate": 8.855831469025346e-05, "loss": 0.0628, "step": 25870 }, { "grad_norm": 0.6451332569122314, "learning_rate": 8.854778603499689e-05, "loss": 0.073, "step": 25880 }, { "grad_norm": 0.5686650276184082, "learning_rate": 8.85372531642124e-05, "loss": 0.0616, "step": 25890 }, { "grad_norm": 0.5703813433647156, "learning_rate": 8.852671607905185e-05, "loss": 0.0468, "step": 25900 }, { "grad_norm": 0.46801966428756714, "learning_rate": 8.851617478066754e-05, "loss": 0.0555, "step": 25910 }, { "grad_norm": 0.691646158695221, "learning_rate": 8.850562927021227e-05, "loss": 0.0557, "step": 25920 }, { "grad_norm": 0.6021662950515747, "learning_rate": 8.849507954883928e-05, "loss": 0.0623, "step": 25930 }, { "grad_norm": 0.631039023399353, "learning_rate": 8.848452561770226e-05, "loss": 0.0646, "step": 25940 }, { "grad_norm": 0.448377788066864, "learning_rate": 8.847396747795538e-05, "loss": 0.0541, "step": 25950 }, { "grad_norm": 0.4512673318386078, "learning_rate": 8.846340513075327e-05, "loss": 0.061, "step": 25960 }, { "grad_norm": 0.7021973729133606, "learning_rate": 8.845283857725099e-05, "loss": 0.0762, "step": 25970 }, { "grad_norm": 0.47010862827301025, "learning_rate": 8.844226781860409e-05, "loss": 0.0431, "step": 25980 }, { "grad_norm": 0.4761680066585541, "learning_rate": 8.84316928559686e-05, "loss": 0.056, "step": 25990 }, { "grad_norm": 0.6437444090843201, "learning_rate": 8.842111369050094e-05, "loss": 0.047, "step": 26000 }, { "grad_norm": 0.8303787708282471, "learning_rate": 8.841053032335808e-05, "loss": 0.0717, "step": 26010 }, { "grad_norm": 0.663316011428833, "learning_rate": 8.839994275569735e-05, "loss": 0.0891, "step": 26020 }, { "grad_norm": 0.7904129028320312, "learning_rate": 8.838935098867662e-05, "loss": 0.0612, "step": 26030 }, { "grad_norm": 0.739872395992279, "learning_rate": 8.837875502345418e-05, "loss": 0.0707, "step": 26040 }, { "grad_norm": 0.9522392749786377, "learning_rate": 8.83681548611888e-05, "loss": 0.0874, "step": 26050 }, { "grad_norm": 0.5526230335235596, "learning_rate": 8.835755050303969e-05, "loss": 0.0671, "step": 26060 }, { "grad_norm": 0.7041542530059814, "learning_rate": 8.834694195016653e-05, "loss": 0.0453, "step": 26070 }, { "grad_norm": 0.4835246205329895, "learning_rate": 8.833632920372942e-05, "loss": 0.0535, "step": 26080 }, { "grad_norm": 0.755855917930603, "learning_rate": 8.832571226488903e-05, "loss": 0.0529, "step": 26090 }, { "grad_norm": 0.7134770154953003, "learning_rate": 8.831509113480634e-05, "loss": 0.0461, "step": 26100 }, { "grad_norm": 0.6035339832305908, "learning_rate": 8.83044658146429e-05, "loss": 0.0554, "step": 26110 }, { "grad_norm": 0.7898473143577576, "learning_rate": 8.829383630556067e-05, "loss": 0.0602, "step": 26120 }, { "grad_norm": 0.7142320275306702, "learning_rate": 8.828320260872207e-05, "loss": 0.0585, "step": 26130 }, { "grad_norm": 0.7397618293762207, "learning_rate": 8.827256472529e-05, "loss": 0.0904, "step": 26140 }, { "grad_norm": 0.7084391713142395, "learning_rate": 8.826192265642778e-05, "loss": 0.0572, "step": 26150 }, { "grad_norm": 0.9553273916244507, "learning_rate": 8.825127640329923e-05, "loss": 0.0634, "step": 26160 }, { "grad_norm": 0.5528642535209656, "learning_rate": 8.824062596706861e-05, "loss": 0.0545, "step": 26170 }, { "grad_norm": 0.7385661602020264, "learning_rate": 8.822997134890062e-05, "loss": 0.0735, "step": 26180 }, { "grad_norm": 0.6420140862464905, "learning_rate": 8.821931254996044e-05, "loss": 0.0611, "step": 26190 }, { "grad_norm": 0.5169084072113037, "learning_rate": 8.82086495714137e-05, "loss": 0.054, "step": 26200 }, { "grad_norm": 0.6534458994865417, "learning_rate": 8.81979824144265e-05, "loss": 0.0571, "step": 26210 }, { "grad_norm": 0.6775082945823669, "learning_rate": 8.818731108016536e-05, "loss": 0.0594, "step": 26220 }, { "grad_norm": 0.5258061289787292, "learning_rate": 8.81766355697973e-05, "loss": 0.0557, "step": 26230 }, { "grad_norm": 0.8326359987258911, "learning_rate": 8.816595588448977e-05, "loss": 0.0625, "step": 26240 }, { "grad_norm": 0.6117682456970215, "learning_rate": 8.81552720254107e-05, "loss": 0.0771, "step": 26250 }, { "grad_norm": 0.5651735067367554, "learning_rate": 8.814458399372842e-05, "loss": 0.0525, "step": 26260 }, { "grad_norm": 0.4825117588043213, "learning_rate": 8.813389179061181e-05, "loss": 0.0518, "step": 26270 }, { "grad_norm": 0.5434389710426331, "learning_rate": 8.812319541723012e-05, "loss": 0.0642, "step": 26280 }, { "grad_norm": 0.5207276344299316, "learning_rate": 8.811249487475309e-05, "loss": 0.0531, "step": 26290 }, { "grad_norm": 0.4272240102291107, "learning_rate": 8.810179016435092e-05, "loss": 0.059, "step": 26300 }, { "grad_norm": 0.39768850803375244, "learning_rate": 8.809108128719428e-05, "loss": 0.0475, "step": 26310 }, { "grad_norm": 0.3533796966075897, "learning_rate": 8.808036824445424e-05, "loss": 0.0434, "step": 26320 }, { "grad_norm": 0.7694326639175415, "learning_rate": 8.806965103730238e-05, "loss": 0.0605, "step": 26330 }, { "grad_norm": 0.5781317949295044, "learning_rate": 8.805892966691074e-05, "loss": 0.0462, "step": 26340 }, { "grad_norm": 0.6355881690979004, "learning_rate": 8.804820413445175e-05, "loss": 0.0632, "step": 26350 }, { "grad_norm": 0.38662558794021606, "learning_rate": 8.803747444109837e-05, "loss": 0.0447, "step": 26360 }, { "grad_norm": 0.5071911811828613, "learning_rate": 8.802674058802399e-05, "loss": 0.0452, "step": 26370 }, { "grad_norm": 0.5624143481254578, "learning_rate": 8.801600257640241e-05, "loss": 0.0473, "step": 26380 }, { "grad_norm": 0.5936595797538757, "learning_rate": 8.800526040740795e-05, "loss": 0.0455, "step": 26390 }, { "grad_norm": 0.6568562388420105, "learning_rate": 8.799451408221535e-05, "loss": 0.0845, "step": 26400 }, { "grad_norm": 0.5901906490325928, "learning_rate": 8.798376360199982e-05, "loss": 0.0491, "step": 26410 }, { "grad_norm": 0.7614646553993225, "learning_rate": 8.797300896793701e-05, "loss": 0.0485, "step": 26420 }, { "grad_norm": 0.456586092710495, "learning_rate": 8.796225018120302e-05, "loss": 0.06, "step": 26430 }, { "grad_norm": 0.5686904788017273, "learning_rate": 8.795148724297444e-05, "loss": 0.0466, "step": 26440 }, { "grad_norm": 0.5291309952735901, "learning_rate": 8.794072015442825e-05, "loss": 0.0482, "step": 26450 }, { "grad_norm": 0.7118139863014221, "learning_rate": 8.792994891674198e-05, "loss": 0.0451, "step": 26460 }, { "grad_norm": 0.913690984249115, "learning_rate": 8.79191735310935e-05, "loss": 0.0852, "step": 26470 }, { "grad_norm": 0.663625180721283, "learning_rate": 8.790839399866122e-05, "loss": 0.0539, "step": 26480 }, { "grad_norm": 0.461821585893631, "learning_rate": 8.789761032062397e-05, "loss": 0.0525, "step": 26490 }, { "grad_norm": 0.5010414719581604, "learning_rate": 8.788682249816103e-05, "loss": 0.056, "step": 26500 }, { "grad_norm": 0.6716000437736511, "learning_rate": 8.787603053245215e-05, "loss": 0.0542, "step": 26510 }, { "grad_norm": 0.9217151403427124, "learning_rate": 8.78652344246775e-05, "loss": 0.084, "step": 26520 }, { "grad_norm": 0.8050244450569153, "learning_rate": 8.785443417601776e-05, "loss": 0.0723, "step": 26530 }, { "grad_norm": 0.6791256070137024, "learning_rate": 8.784362978765401e-05, "loss": 0.0601, "step": 26540 }, { "grad_norm": 0.7569360136985779, "learning_rate": 8.783282126076779e-05, "loss": 0.0579, "step": 26550 }, { "grad_norm": 0.5697029232978821, "learning_rate": 8.782200859654112e-05, "loss": 0.049, "step": 26560 }, { "grad_norm": 0.8220356106758118, "learning_rate": 8.781119179615646e-05, "loss": 0.0493, "step": 26570 }, { "grad_norm": 0.5408744812011719, "learning_rate": 8.780037086079674e-05, "loss": 0.0583, "step": 26580 }, { "grad_norm": 0.5243074893951416, "learning_rate": 8.778954579164527e-05, "loss": 0.044, "step": 26590 }, { "grad_norm": 0.5349138379096985, "learning_rate": 8.777871658988588e-05, "loss": 0.051, "step": 26600 }, { "grad_norm": 0.7644124627113342, "learning_rate": 8.776788325670285e-05, "loss": 0.0582, "step": 26610 }, { "grad_norm": 0.44271260499954224, "learning_rate": 8.775704579328089e-05, "loss": 0.0731, "step": 26620 }, { "grad_norm": 0.524868369102478, "learning_rate": 8.774620420080517e-05, "loss": 0.0686, "step": 26630 }, { "grad_norm": 0.715099573135376, "learning_rate": 8.773535848046131e-05, "loss": 0.0729, "step": 26640 }, { "grad_norm": 0.5685938000679016, "learning_rate": 8.772450863343538e-05, "loss": 0.058, "step": 26650 }, { "grad_norm": 0.7431249022483826, "learning_rate": 8.77136546609139e-05, "loss": 0.0526, "step": 26660 }, { "grad_norm": 0.42486631870269775, "learning_rate": 8.770279656408385e-05, "loss": 0.0701, "step": 26670 }, { "grad_norm": 0.5884155631065369, "learning_rate": 8.769193434413265e-05, "loss": 0.0431, "step": 26680 }, { "grad_norm": 0.6075438857078552, "learning_rate": 8.76810680022482e-05, "loss": 0.0592, "step": 26690 }, { "grad_norm": 0.7959399819374084, "learning_rate": 8.767019753961878e-05, "loss": 0.0503, "step": 26700 }, { "grad_norm": 0.5923810005187988, "learning_rate": 8.765932295743321e-05, "loss": 0.0583, "step": 26710 }, { "grad_norm": 0.40992099046707153, "learning_rate": 8.764844425688068e-05, "loss": 0.0575, "step": 26720 }, { "grad_norm": 0.8853257298469543, "learning_rate": 8.763756143915092e-05, "loss": 0.0623, "step": 26730 }, { "grad_norm": 0.7775055766105652, "learning_rate": 8.7626674505434e-05, "loss": 0.0717, "step": 26740 }, { "grad_norm": 0.672703206539154, "learning_rate": 8.761578345692053e-05, "loss": 0.051, "step": 26750 }, { "grad_norm": 0.6234543323516846, "learning_rate": 8.760488829480156e-05, "loss": 0.0661, "step": 26760 }, { "grad_norm": 0.7819085121154785, "learning_rate": 8.759398902026854e-05, "loss": 0.0642, "step": 26770 }, { "grad_norm": 0.8409708738327026, "learning_rate": 8.758308563451339e-05, "loss": 0.0667, "step": 26780 }, { "grad_norm": 0.6258885860443115, "learning_rate": 8.75721781387285e-05, "loss": 0.0589, "step": 26790 }, { "grad_norm": 0.3679754137992859, "learning_rate": 8.75612665341067e-05, "loss": 0.0553, "step": 26800 }, { "grad_norm": 1.2781765460968018, "learning_rate": 8.755035082184126e-05, "loss": 0.0639, "step": 26810 }, { "grad_norm": 0.4649830460548401, "learning_rate": 8.753943100312592e-05, "loss": 0.0486, "step": 26820 }, { "grad_norm": 0.8061181306838989, "learning_rate": 8.752850707915484e-05, "loss": 0.0652, "step": 26830 }, { "grad_norm": 0.6995047330856323, "learning_rate": 8.751757905112264e-05, "loss": 0.0498, "step": 26840 }, { "grad_norm": 0.6646412014961243, "learning_rate": 8.75066469202244e-05, "loss": 0.0569, "step": 26850 }, { "grad_norm": 0.4776745140552521, "learning_rate": 8.749571068765567e-05, "loss": 0.0494, "step": 26860 }, { "grad_norm": 0.38926321268081665, "learning_rate": 8.748477035461238e-05, "loss": 0.0564, "step": 26870 }, { "grad_norm": 1.4244260787963867, "learning_rate": 8.747382592229095e-05, "loss": 0.0657, "step": 26880 }, { "grad_norm": 0.6213458776473999, "learning_rate": 8.746287739188828e-05, "loss": 0.0585, "step": 26890 }, { "grad_norm": 0.6400648951530457, "learning_rate": 8.745192476460165e-05, "loss": 0.0823, "step": 26900 }, { "grad_norm": 0.43265819549560547, "learning_rate": 8.744096804162882e-05, "loss": 0.0888, "step": 26910 }, { "grad_norm": 1.1142995357513428, "learning_rate": 8.743000722416804e-05, "loss": 0.0592, "step": 26920 }, { "grad_norm": 0.5545693635940552, "learning_rate": 8.741904231341793e-05, "loss": 0.0769, "step": 26930 }, { "grad_norm": 0.7545728087425232, "learning_rate": 8.740807331057762e-05, "loss": 0.0887, "step": 26940 }, { "grad_norm": 0.6277825832366943, "learning_rate": 8.739710021684667e-05, "loss": 0.0605, "step": 26950 }, { "grad_norm": 0.8009586334228516, "learning_rate": 8.738612303342503e-05, "loss": 0.0743, "step": 26960 }, { "grad_norm": 0.8552682995796204, "learning_rate": 8.73751417615132e-05, "loss": 0.0808, "step": 26970 }, { "grad_norm": 0.6584646701812744, "learning_rate": 8.736415640231208e-05, "loss": 0.0606, "step": 26980 }, { "grad_norm": 0.5242036581039429, "learning_rate": 8.735316695702297e-05, "loss": 0.04, "step": 26990 }, { "grad_norm": 0.6643736362457275, "learning_rate": 8.734217342684769e-05, "loss": 0.0558, "step": 27000 }, { "grad_norm": 0.9307712316513062, "learning_rate": 8.733117581298847e-05, "loss": 0.0643, "step": 27010 }, { "grad_norm": 0.4486657977104187, "learning_rate": 8.732017411664796e-05, "loss": 0.0509, "step": 27020 }, { "grad_norm": 1.1744493246078491, "learning_rate": 8.730916833902936e-05, "loss": 0.0901, "step": 27030 }, { "grad_norm": 0.4469805955886841, "learning_rate": 8.729815848133618e-05, "loss": 0.0737, "step": 27040 }, { "grad_norm": 0.48502713441848755, "learning_rate": 8.728714454477247e-05, "loss": 0.0784, "step": 27050 }, { "grad_norm": 0.5814555883407593, "learning_rate": 8.727612653054269e-05, "loss": 0.0669, "step": 27060 }, { "grad_norm": 0.5269268155097961, "learning_rate": 8.726510443985176e-05, "loss": 0.0869, "step": 27070 }, { "grad_norm": 0.4452076554298401, "learning_rate": 8.725407827390503e-05, "loss": 0.0488, "step": 27080 }, { "grad_norm": 0.5372799038887024, "learning_rate": 8.724304803390833e-05, "loss": 0.0502, "step": 27090 }, { "grad_norm": 0.4569877088069916, "learning_rate": 8.723201372106788e-05, "loss": 0.0567, "step": 27100 }, { "grad_norm": 0.33642029762268066, "learning_rate": 8.722097533659038e-05, "loss": 0.074, "step": 27110 }, { "grad_norm": 0.6904877424240112, "learning_rate": 8.720993288168299e-05, "loss": 0.0622, "step": 27120 }, { "grad_norm": 0.453225702047348, "learning_rate": 8.719888635755327e-05, "loss": 0.0554, "step": 27130 }, { "grad_norm": 0.7176165580749512, "learning_rate": 8.718783576540928e-05, "loss": 0.0452, "step": 27140 }, { "grad_norm": 0.659501314163208, "learning_rate": 8.717678110645948e-05, "loss": 0.0488, "step": 27150 }, { "grad_norm": 0.5802245736122131, "learning_rate": 8.716572238191279e-05, "loss": 0.055, "step": 27160 }, { "grad_norm": 0.730253279209137, "learning_rate": 8.715465959297857e-05, "loss": 0.0753, "step": 27170 }, { "grad_norm": 0.7002321481704712, "learning_rate": 8.714359274086665e-05, "loss": 0.0509, "step": 27180 }, { "grad_norm": 0.5651087164878845, "learning_rate": 8.713252182678726e-05, "loss": 0.0901, "step": 27190 }, { "grad_norm": 0.9493050575256348, "learning_rate": 8.712144685195112e-05, "loss": 0.0626, "step": 27200 }, { "grad_norm": 0.6742130517959595, "learning_rate": 8.711036781756936e-05, "loss": 0.0516, "step": 27210 }, { "grad_norm": 0.403969943523407, "learning_rate": 8.709928472485357e-05, "loss": 0.0472, "step": 27220 }, { "grad_norm": 0.5583747625350952, "learning_rate": 8.708819757501579e-05, "loss": 0.044, "step": 27230 }, { "grad_norm": 0.6949424743652344, "learning_rate": 8.707710636926846e-05, "loss": 0.0793, "step": 27240 }, { "grad_norm": 0.4945354163646698, "learning_rate": 8.706601110882455e-05, "loss": 0.0483, "step": 27250 }, { "grad_norm": 1.0634726285934448, "learning_rate": 8.705491179489738e-05, "loss": 0.0552, "step": 27260 }, { "grad_norm": 0.49055057764053345, "learning_rate": 8.704380842870077e-05, "loss": 0.0415, "step": 27270 }, { "grad_norm": 0.5106921792030334, "learning_rate": 8.703270101144895e-05, "loss": 0.0434, "step": 27280 }, { "grad_norm": 0.5174590945243835, "learning_rate": 8.702158954435664e-05, "loss": 0.0549, "step": 27290 }, { "grad_norm": 0.465176522731781, "learning_rate": 8.701047402863896e-05, "loss": 0.0709, "step": 27300 }, { "grad_norm": 0.42574501037597656, "learning_rate": 8.699935446551148e-05, "loss": 0.0709, "step": 27310 }, { "grad_norm": 0.6015035510063171, "learning_rate": 8.698823085619022e-05, "loss": 0.0641, "step": 27320 }, { "grad_norm": 0.34928637742996216, "learning_rate": 8.697710320189166e-05, "loss": 0.0562, "step": 27330 }, { "grad_norm": 0.5278319120407104, "learning_rate": 8.696597150383268e-05, "loss": 0.0709, "step": 27340 }, { "grad_norm": 0.7498689293861389, "learning_rate": 8.695483576323063e-05, "loss": 0.0552, "step": 27350 }, { "grad_norm": 0.5729643702507019, "learning_rate": 8.69436959813033e-05, "loss": 0.0465, "step": 27360 }, { "grad_norm": 0.5887826085090637, "learning_rate": 8.693255215926892e-05, "loss": 0.0585, "step": 27370 }, { "grad_norm": 0.6263673901557922, "learning_rate": 8.692140429834617e-05, "loss": 0.0506, "step": 27380 }, { "grad_norm": 0.5770047903060913, "learning_rate": 8.691025239975415e-05, "loss": 0.065, "step": 27390 }, { "grad_norm": 0.6888108849525452, "learning_rate": 8.689909646471243e-05, "loss": 0.0676, "step": 27400 }, { "grad_norm": 0.7826325297355652, "learning_rate": 8.688793649444099e-05, "loss": 0.0733, "step": 27410 }, { "grad_norm": 0.5283001065254211, "learning_rate": 8.687677249016029e-05, "loss": 0.061, "step": 27420 }, { "grad_norm": 0.5986341834068298, "learning_rate": 8.686560445309118e-05, "loss": 0.0631, "step": 27430 }, { "grad_norm": 0.43830063939094543, "learning_rate": 8.685443238445499e-05, "loss": 0.069, "step": 27440 }, { "grad_norm": 0.7635226845741272, "learning_rate": 8.68432562854735e-05, "loss": 0.0574, "step": 27450 }, { "grad_norm": 0.6012451648712158, "learning_rate": 8.683207615736887e-05, "loss": 0.0536, "step": 27460 }, { "grad_norm": 0.6505747437477112, "learning_rate": 8.682089200136379e-05, "loss": 0.0493, "step": 27470 }, { "grad_norm": 0.4741673171520233, "learning_rate": 8.680970381868132e-05, "loss": 0.0556, "step": 27480 }, { "grad_norm": 0.7530203461647034, "learning_rate": 8.679851161054498e-05, "loss": 0.0657, "step": 27490 }, { "grad_norm": 0.7349664568901062, "learning_rate": 8.678731537817873e-05, "loss": 0.0526, "step": 27500 }, { "grad_norm": 0.5479376912117004, "learning_rate": 8.677611512280697e-05, "loss": 0.0646, "step": 27510 }, { "grad_norm": 0.6550157070159912, "learning_rate": 8.676491084565457e-05, "loss": 0.0914, "step": 27520 }, { "grad_norm": 0.7092416882514954, "learning_rate": 8.675370254794678e-05, "loss": 0.0503, "step": 27530 }, { "grad_norm": 0.5239381194114685, "learning_rate": 8.674249023090935e-05, "loss": 0.0531, "step": 27540 }, { "grad_norm": 0.4636121392250061, "learning_rate": 8.673127389576843e-05, "loss": 0.0601, "step": 27550 }, { "grad_norm": 0.752644419670105, "learning_rate": 8.67200535437506e-05, "loss": 0.0596, "step": 27560 }, { "grad_norm": 0.6443470120429993, "learning_rate": 8.670882917608296e-05, "loss": 0.0708, "step": 27570 }, { "grad_norm": 0.8444588780403137, "learning_rate": 8.669760079399292e-05, "loss": 0.0589, "step": 27580 }, { "grad_norm": 0.6167011260986328, "learning_rate": 8.668636839870845e-05, "loss": 0.0536, "step": 27590 }, { "grad_norm": 0.4469539225101471, "learning_rate": 8.667513199145789e-05, "loss": 0.062, "step": 27600 }, { "grad_norm": 0.5086582899093628, "learning_rate": 8.666389157347002e-05, "loss": 0.0729, "step": 27610 }, { "grad_norm": 0.6969667077064514, "learning_rate": 8.66526471459741e-05, "loss": 0.0634, "step": 27620 }, { "grad_norm": 0.6734515428543091, "learning_rate": 8.66413987101998e-05, "loss": 0.0554, "step": 27630 }, { "grad_norm": 0.38199788331985474, "learning_rate": 8.663014626737723e-05, "loss": 0.0738, "step": 27640 }, { "grad_norm": 0.803598940372467, "learning_rate": 8.661888981873691e-05, "loss": 0.0823, "step": 27650 }, { "grad_norm": 1.196663737297058, "learning_rate": 8.660762936550988e-05, "loss": 0.0694, "step": 27660 }, { "grad_norm": 0.8179227113723755, "learning_rate": 8.659636490892753e-05, "loss": 0.0711, "step": 27670 }, { "grad_norm": 0.6992975473403931, "learning_rate": 8.658509645022174e-05, "loss": 0.0651, "step": 27680 }, { "grad_norm": 0.43260565400123596, "learning_rate": 8.657382399062481e-05, "loss": 0.043, "step": 27690 }, { "grad_norm": 0.7214030027389526, "learning_rate": 8.656254753136946e-05, "loss": 0.0595, "step": 27700 }, { "grad_norm": 0.5455387234687805, "learning_rate": 8.655126707368891e-05, "loss": 0.0632, "step": 27710 }, { "grad_norm": 0.5294789671897888, "learning_rate": 8.653998261881672e-05, "loss": 0.0502, "step": 27720 }, { "grad_norm": 0.4891185462474823, "learning_rate": 8.652869416798699e-05, "loss": 0.0836, "step": 27730 }, { "grad_norm": 0.5931333899497986, "learning_rate": 8.651740172243417e-05, "loss": 0.0532, "step": 27740 }, { "grad_norm": 0.4800463318824768, "learning_rate": 8.65061052833932e-05, "loss": 0.0573, "step": 27750 }, { "grad_norm": 0.6357388496398926, "learning_rate": 8.649480485209945e-05, "loss": 0.0493, "step": 27760 }, { "grad_norm": 0.6304186582565308, "learning_rate": 8.64835004297887e-05, "loss": 0.0604, "step": 27770 }, { "grad_norm": 0.7946129441261292, "learning_rate": 8.64721920176972e-05, "loss": 0.0969, "step": 27780 }, { "grad_norm": 0.6462501287460327, "learning_rate": 8.646087961706164e-05, "loss": 0.0473, "step": 27790 }, { "grad_norm": 0.7601629495620728, "learning_rate": 8.644956322911908e-05, "loss": 0.0651, "step": 27800 }, { "grad_norm": 1.1164376735687256, "learning_rate": 8.643824285510709e-05, "loss": 0.0763, "step": 27810 }, { "grad_norm": 0.6571993827819824, "learning_rate": 8.642691849626364e-05, "loss": 0.0599, "step": 27820 }, { "grad_norm": 0.5680731534957886, "learning_rate": 8.641559015382717e-05, "loss": 0.0869, "step": 27830 }, { "grad_norm": 0.5903797149658203, "learning_rate": 8.640425782903649e-05, "loss": 0.0502, "step": 27840 }, { "grad_norm": 0.8294682502746582, "learning_rate": 8.639292152313091e-05, "loss": 0.0468, "step": 27850 }, { "grad_norm": 0.6826444864273071, "learning_rate": 8.638158123735015e-05, "loss": 0.0636, "step": 27860 }, { "grad_norm": 0.5606412887573242, "learning_rate": 8.637023697293436e-05, "loss": 0.0624, "step": 27870 }, { "grad_norm": 0.6046707630157471, "learning_rate": 8.635888873112414e-05, "loss": 0.047, "step": 27880 }, { "grad_norm": 0.5315872430801392, "learning_rate": 8.634753651316052e-05, "loss": 0.0602, "step": 27890 }, { "grad_norm": 0.5521880984306335, "learning_rate": 8.633618032028496e-05, "loss": 0.0665, "step": 27900 }, { "grad_norm": 0.5487511157989502, "learning_rate": 8.632482015373934e-05, "loss": 0.0586, "step": 27910 }, { "grad_norm": 0.6941881775856018, "learning_rate": 8.6313456014766e-05, "loss": 0.0672, "step": 27920 }, { "grad_norm": 0.606529176235199, "learning_rate": 8.630208790460771e-05, "loss": 0.0923, "step": 27930 }, { "grad_norm": 0.809227466583252, "learning_rate": 8.629071582450768e-05, "loss": 0.0895, "step": 27940 }, { "grad_norm": 0.6171464323997498, "learning_rate": 8.62793397757095e-05, "loss": 0.0659, "step": 27950 }, { "grad_norm": 0.7156135439872742, "learning_rate": 8.626795975945729e-05, "loss": 0.0672, "step": 27960 }, { "grad_norm": 0.9832422137260437, "learning_rate": 8.625657577699551e-05, "loss": 0.075, "step": 27970 }, { "grad_norm": 0.617347240447998, "learning_rate": 8.624518782956914e-05, "loss": 0.0496, "step": 27980 }, { "grad_norm": 0.3974236249923706, "learning_rate": 8.62337959184235e-05, "loss": 0.0699, "step": 27990 }, { "grad_norm": 0.6245284080505371, "learning_rate": 8.622240004480441e-05, "loss": 0.0569, "step": 28000 }, { "grad_norm": 1.015311598777771, "learning_rate": 8.621100020995814e-05, "loss": 0.0646, "step": 28010 }, { "grad_norm": 0.3634939193725586, "learning_rate": 8.619959641513132e-05, "loss": 0.0577, "step": 28020 }, { "grad_norm": 0.5852189064025879, "learning_rate": 8.618818866157105e-05, "loss": 0.0673, "step": 28030 }, { "grad_norm": 0.6755378842353821, "learning_rate": 8.617677695052487e-05, "loss": 0.06, "step": 28040 }, { "grad_norm": 0.5224453210830688, "learning_rate": 8.616536128324078e-05, "loss": 0.056, "step": 28050 }, { "grad_norm": 0.44333863258361816, "learning_rate": 8.615394166096712e-05, "loss": 0.0629, "step": 28060 }, { "grad_norm": 0.8174710869789124, "learning_rate": 8.614251808495279e-05, "loss": 0.0556, "step": 28070 }, { "grad_norm": 0.45557403564453125, "learning_rate": 8.6131090556447e-05, "loss": 0.0479, "step": 28080 }, { "grad_norm": 1.0046114921569824, "learning_rate": 8.611965907669947e-05, "loss": 0.0551, "step": 28090 }, { "grad_norm": 0.5207643508911133, "learning_rate": 8.610822364696034e-05, "loss": 0.0503, "step": 28100 }, { "grad_norm": 0.8543736934661865, "learning_rate": 8.609678426848015e-05, "loss": 0.059, "step": 28110 }, { "grad_norm": 0.5915364623069763, "learning_rate": 8.60853409425099e-05, "loss": 0.0582, "step": 28120 }, { "grad_norm": 0.48590126633644104, "learning_rate": 8.607389367030104e-05, "loss": 0.0671, "step": 28130 }, { "grad_norm": 0.765120804309845, "learning_rate": 8.606244245310538e-05, "loss": 0.0587, "step": 28140 }, { "grad_norm": 0.810741126537323, "learning_rate": 8.605098729217525e-05, "loss": 0.07, "step": 28150 }, { "grad_norm": 0.6951033473014832, "learning_rate": 8.603952818876335e-05, "loss": 0.06, "step": 28160 }, { "grad_norm": 0.6115968823432922, "learning_rate": 8.602806514412281e-05, "loss": 0.0627, "step": 28170 }, { "grad_norm": 0.6445460319519043, "learning_rate": 8.601659815950726e-05, "loss": 0.0497, "step": 28180 }, { "grad_norm": 0.7698470950126648, "learning_rate": 8.600512723617067e-05, "loss": 0.0626, "step": 28190 }, { "grad_norm": 0.6560800671577454, "learning_rate": 8.59936523753675e-05, "loss": 0.0563, "step": 28200 }, { "grad_norm": 0.5745024681091309, "learning_rate": 8.598217357835264e-05, "loss": 0.0472, "step": 28210 }, { "grad_norm": 0.9345596432685852, "learning_rate": 8.597069084638135e-05, "loss": 0.067, "step": 28220 }, { "grad_norm": 0.7616354823112488, "learning_rate": 8.595920418070939e-05, "loss": 0.0426, "step": 28230 }, { "grad_norm": 0.5811212062835693, "learning_rate": 8.594771358259295e-05, "loss": 0.0477, "step": 28240 }, { "grad_norm": 0.6720995306968689, "learning_rate": 8.593621905328858e-05, "loss": 0.0515, "step": 28250 }, { "grad_norm": 0.5712747573852539, "learning_rate": 8.592472059405333e-05, "loss": 0.071, "step": 28260 }, { "grad_norm": 0.8550304174423218, "learning_rate": 8.591321820614464e-05, "loss": 0.0747, "step": 28270 }, { "grad_norm": 0.710439920425415, "learning_rate": 8.590171189082041e-05, "loss": 0.0575, "step": 28280 }, { "grad_norm": 0.4763299524784088, "learning_rate": 8.589020164933894e-05, "loss": 0.0429, "step": 28290 }, { "grad_norm": 0.823810875415802, "learning_rate": 8.587868748295898e-05, "loss": 0.074, "step": 28300 }, { "grad_norm": 0.9160043597221375, "learning_rate": 8.586716939293971e-05, "loss": 0.0832, "step": 28310 }, { "grad_norm": 0.8368088006973267, "learning_rate": 8.58556473805407e-05, "loss": 0.0728, "step": 28320 }, { "grad_norm": 0.6117218732833862, "learning_rate": 8.584412144702202e-05, "loss": 0.0585, "step": 28330 }, { "grad_norm": 0.7815631031990051, "learning_rate": 8.58325915936441e-05, "loss": 0.0608, "step": 28340 }, { "grad_norm": 0.5344248414039612, "learning_rate": 8.582105782166783e-05, "loss": 0.0591, "step": 28350 }, { "grad_norm": 0.5924983620643616, "learning_rate": 8.580952013235455e-05, "loss": 0.0499, "step": 28360 }, { "grad_norm": 0.8513888716697693, "learning_rate": 8.579797852696596e-05, "loss": 0.0572, "step": 28370 }, { "grad_norm": 0.6217703223228455, "learning_rate": 8.578643300676428e-05, "loss": 0.0557, "step": 28380 }, { "grad_norm": 0.6925026178359985, "learning_rate": 8.577488357301209e-05, "loss": 0.059, "step": 28390 }, { "grad_norm": 0.5812597870826721, "learning_rate": 8.576333022697242e-05, "loss": 0.0452, "step": 28400 }, { "grad_norm": 0.7273154258728027, "learning_rate": 8.575177296990873e-05, "loss": 0.0619, "step": 28410 }, { "grad_norm": 0.8401070833206177, "learning_rate": 8.574021180308489e-05, "loss": 0.0611, "step": 28420 }, { "grad_norm": 0.810310959815979, "learning_rate": 8.572864672776523e-05, "loss": 0.096, "step": 28430 }, { "grad_norm": 0.5337298512458801, "learning_rate": 8.571707774521447e-05, "loss": 0.0427, "step": 28440 }, { "grad_norm": 0.6230490207672119, "learning_rate": 8.57055048566978e-05, "loss": 0.0615, "step": 28450 }, { "grad_norm": 0.702769935131073, "learning_rate": 8.569392806348078e-05, "loss": 0.0412, "step": 28460 }, { "grad_norm": 0.4449405074119568, "learning_rate": 8.568234736682947e-05, "loss": 0.0524, "step": 28470 }, { "grad_norm": 0.6632161736488342, "learning_rate": 8.567076276801029e-05, "loss": 0.056, "step": 28480 }, { "grad_norm": 0.5122987627983093, "learning_rate": 8.565917426829013e-05, "loss": 0.0686, "step": 28490 }, { "grad_norm": 0.5691812634468079, "learning_rate": 8.564758186893628e-05, "loss": 0.058, "step": 28500 }, { "grad_norm": 0.7513211965560913, "learning_rate": 8.563598557121649e-05, "loss": 0.0595, "step": 28510 }, { "grad_norm": 0.5120598673820496, "learning_rate": 8.562438537639888e-05, "loss": 0.0623, "step": 28520 }, { "grad_norm": 0.7480705380439758, "learning_rate": 8.561278128575206e-05, "loss": 0.0739, "step": 28530 }, { "grad_norm": 0.621911346912384, "learning_rate": 8.5601173300545e-05, "loss": 0.0649, "step": 28540 }, { "grad_norm": 0.751550018787384, "learning_rate": 8.558956142204717e-05, "loss": 0.0558, "step": 28550 }, { "grad_norm": 0.7275345325469971, "learning_rate": 8.55779456515284e-05, "loss": 0.0623, "step": 28560 }, { "grad_norm": 0.9023893475532532, "learning_rate": 8.556632599025898e-05, "loss": 0.0684, "step": 28570 }, { "grad_norm": 0.5464584827423096, "learning_rate": 8.555470243950964e-05, "loss": 0.0621, "step": 28580 }, { "grad_norm": 0.4834330081939697, "learning_rate": 8.554307500055148e-05, "loss": 0.0513, "step": 28590 }, { "grad_norm": 0.656776487827301, "learning_rate": 8.553144367465609e-05, "loss": 0.0673, "step": 28600 }, { "grad_norm": 0.5364528298377991, "learning_rate": 8.551980846309544e-05, "loss": 0.0503, "step": 28610 }, { "grad_norm": 0.5746975541114807, "learning_rate": 8.550816936714193e-05, "loss": 0.0509, "step": 28620 }, { "grad_norm": 0.5166106224060059, "learning_rate": 8.549652638806841e-05, "loss": 0.0871, "step": 28630 }, { "grad_norm": 0.5428603291511536, "learning_rate": 8.548487952714812e-05, "loss": 0.0457, "step": 28640 }, { "grad_norm": 0.7247041463851929, "learning_rate": 8.547322878565478e-05, "loss": 0.0584, "step": 28650 }, { "grad_norm": 0.5548688173294067, "learning_rate": 8.546157416486245e-05, "loss": 0.0426, "step": 28660 }, { "grad_norm": 0.6391072273254395, "learning_rate": 8.54499156660457e-05, "loss": 0.0602, "step": 28670 }, { "grad_norm": 0.6170901656150818, "learning_rate": 8.543825329047947e-05, "loss": 0.0574, "step": 28680 }, { "grad_norm": 0.7281854152679443, "learning_rate": 8.542658703943913e-05, "loss": 0.0547, "step": 28690 }, { "grad_norm": 0.8760702610015869, "learning_rate": 8.541491691420051e-05, "loss": 0.0786, "step": 28700 }, { "grad_norm": 0.6939191222190857, "learning_rate": 8.54032429160398e-05, "loss": 0.048, "step": 28710 }, { "grad_norm": 0.692887544631958, "learning_rate": 8.539156504623369e-05, "loss": 0.0419, "step": 28720 }, { "grad_norm": 1.6488093137741089, "learning_rate": 8.537988330605923e-05, "loss": 0.0551, "step": 28730 }, { "grad_norm": 0.6836596131324768, "learning_rate": 8.536819769679393e-05, "loss": 0.0425, "step": 28740 }, { "grad_norm": 0.7152780294418335, "learning_rate": 8.53565082197157e-05, "loss": 0.0512, "step": 28750 }, { "grad_norm": 0.49102166295051575, "learning_rate": 8.534481487610289e-05, "loss": 0.0534, "step": 28760 }, { "grad_norm": 0.8135676980018616, "learning_rate": 8.533311766723428e-05, "loss": 0.0587, "step": 28770 }, { "grad_norm": 0.5554731488227844, "learning_rate": 8.532141659438901e-05, "loss": 0.0738, "step": 28780 }, { "grad_norm": 0.5060629844665527, "learning_rate": 8.530971165884675e-05, "loss": 0.0609, "step": 28790 }, { "grad_norm": 0.54930180311203, "learning_rate": 8.529800286188752e-05, "loss": 0.0485, "step": 28800 }, { "grad_norm": 0.5405973196029663, "learning_rate": 8.528629020479175e-05, "loss": 0.0547, "step": 28810 }, { "grad_norm": 0.504706621170044, "learning_rate": 8.527457368884033e-05, "loss": 0.0562, "step": 28820 }, { "grad_norm": 0.6233529448509216, "learning_rate": 8.526285331531458e-05, "loss": 0.0521, "step": 28830 }, { "grad_norm": 0.5550420880317688, "learning_rate": 8.525112908549621e-05, "loss": 0.0571, "step": 28840 }, { "grad_norm": 0.5582935214042664, "learning_rate": 8.523940100066735e-05, "loss": 0.0518, "step": 28850 }, { "grad_norm": 0.8264774680137634, "learning_rate": 8.52276690621106e-05, "loss": 0.0671, "step": 28860 }, { "grad_norm": 0.7974614500999451, "learning_rate": 8.521593327110889e-05, "loss": 0.0589, "step": 28870 }, { "grad_norm": 0.4210436940193176, "learning_rate": 8.520419362894569e-05, "loss": 0.0488, "step": 28880 }, { "grad_norm": 0.5038926005363464, "learning_rate": 8.51924501369048e-05, "loss": 0.0562, "step": 28890 }, { "grad_norm": 0.5165967345237732, "learning_rate": 8.518070279627047e-05, "loss": 0.0515, "step": 28900 }, { "grad_norm": 0.5617037415504456, "learning_rate": 8.516895160832737e-05, "loss": 0.0626, "step": 28910 }, { "grad_norm": 0.5818395018577576, "learning_rate": 8.515719657436061e-05, "loss": 0.0492, "step": 28920 }, { "grad_norm": 0.8567428588867188, "learning_rate": 8.514543769565568e-05, "loss": 0.054, "step": 28930 }, { "grad_norm": 0.5123169422149658, "learning_rate": 8.513367497349853e-05, "loss": 0.0768, "step": 28940 }, { "grad_norm": 0.5636535286903381, "learning_rate": 8.51219084091755e-05, "loss": 0.0624, "step": 28950 }, { "grad_norm": 0.5629149079322815, "learning_rate": 8.511013800397338e-05, "loss": 0.0637, "step": 28960 }, { "grad_norm": 0.7399353981018066, "learning_rate": 8.509836375917937e-05, "loss": 0.061, "step": 28970 }, { "grad_norm": 0.5035538077354431, "learning_rate": 8.508658567608104e-05, "loss": 0.0477, "step": 28980 }, { "grad_norm": 0.7030888795852661, "learning_rate": 8.507480375596647e-05, "loss": 0.0717, "step": 28990 }, { "grad_norm": 0.4934810400009155, "learning_rate": 8.506301800012408e-05, "loss": 0.0603, "step": 29000 }, { "grad_norm": 0.6449507474899292, "learning_rate": 8.505122840984278e-05, "loss": 0.0549, "step": 29010 }, { "grad_norm": 0.4850868284702301, "learning_rate": 8.503943498641182e-05, "loss": 0.0459, "step": 29020 }, { "grad_norm": 0.424283504486084, "learning_rate": 8.502763773112095e-05, "loss": 0.0535, "step": 29030 }, { "grad_norm": 0.6011964082717896, "learning_rate": 8.501583664526026e-05, "loss": 0.0496, "step": 29040 }, { "grad_norm": 0.5124248266220093, "learning_rate": 8.500403173012032e-05, "loss": 0.087, "step": 29050 }, { "grad_norm": 0.5820369720458984, "learning_rate": 8.499222298699211e-05, "loss": 0.0466, "step": 29060 }, { "grad_norm": 0.4627508223056793, "learning_rate": 8.498041041716701e-05, "loss": 0.0541, "step": 29070 }, { "grad_norm": 0.8257958889007568, "learning_rate": 8.496859402193681e-05, "loss": 0.0556, "step": 29080 }, { "grad_norm": 0.540972888469696, "learning_rate": 8.495677380259374e-05, "loss": 0.0515, "step": 29090 }, { "grad_norm": 0.5949788689613342, "learning_rate": 8.494494976043045e-05, "loss": 0.0564, "step": 29100 }, { "grad_norm": 0.5719385147094727, "learning_rate": 8.493312189673998e-05, "loss": 0.0544, "step": 29110 }, { "grad_norm": 0.6329962015151978, "learning_rate": 8.492129021281584e-05, "loss": 0.0624, "step": 29120 }, { "grad_norm": 0.5077531933784485, "learning_rate": 8.490945470995188e-05, "loss": 0.0519, "step": 29130 }, { "grad_norm": 0.5588739514350891, "learning_rate": 8.489761538944247e-05, "loss": 0.0483, "step": 29140 }, { "grad_norm": 0.728256344795227, "learning_rate": 8.48857722525823e-05, "loss": 0.077, "step": 29150 }, { "grad_norm": 0.6725552678108215, "learning_rate": 8.487392530066652e-05, "loss": 0.0564, "step": 29160 }, { "grad_norm": 0.6907811164855957, "learning_rate": 8.486207453499069e-05, "loss": 0.0663, "step": 29170 }, { "grad_norm": 0.6952424049377441, "learning_rate": 8.485021995685082e-05, "loss": 0.0713, "step": 29180 }, { "grad_norm": 0.7561842799186707, "learning_rate": 8.483836156754328e-05, "loss": 0.0535, "step": 29190 }, { "grad_norm": 0.5272150635719299, "learning_rate": 8.482649936836491e-05, "loss": 0.0786, "step": 29200 }, { "grad_norm": 0.3722020387649536, "learning_rate": 8.481463336061293e-05, "loss": 0.0398, "step": 29210 }, { "grad_norm": 0.7060693502426147, "learning_rate": 8.480276354558496e-05, "loss": 0.0446, "step": 29220 }, { "grad_norm": 0.7044903635978699, "learning_rate": 8.479088992457913e-05, "loss": 0.0558, "step": 29230 }, { "grad_norm": 0.5704482793807983, "learning_rate": 8.477901249889387e-05, "loss": 0.0584, "step": 29240 }, { "grad_norm": 0.5656274557113647, "learning_rate": 8.47671312698281e-05, "loss": 0.0404, "step": 29250 }, { "grad_norm": 0.6154388785362244, "learning_rate": 8.475524623868112e-05, "loss": 0.058, "step": 29260 }, { "grad_norm": 0.3908239006996155, "learning_rate": 8.474335740675266e-05, "loss": 0.0479, "step": 29270 }, { "grad_norm": 0.4688337743282318, "learning_rate": 8.473146477534289e-05, "loss": 0.0405, "step": 29280 }, { "grad_norm": 0.5202878713607788, "learning_rate": 8.471956834575232e-05, "loss": 0.0518, "step": 29290 }, { "grad_norm": 0.6009300947189331, "learning_rate": 8.470766811928197e-05, "loss": 0.0544, "step": 29300 }, { "grad_norm": 0.44399839639663696, "learning_rate": 8.469576409723323e-05, "loss": 0.0506, "step": 29310 }, { "grad_norm": 0.7353965640068054, "learning_rate": 8.468385628090788e-05, "loss": 0.053, "step": 29320 }, { "grad_norm": 0.5318849086761475, "learning_rate": 8.467194467160815e-05, "loss": 0.0618, "step": 29330 }, { "grad_norm": 0.7403227090835571, "learning_rate": 8.466002927063667e-05, "loss": 0.0545, "step": 29340 }, { "grad_norm": 0.6111747622489929, "learning_rate": 8.464811007929651e-05, "loss": 0.052, "step": 29350 }, { "grad_norm": 0.6199637055397034, "learning_rate": 8.463618709889114e-05, "loss": 0.0536, "step": 29360 }, { "grad_norm": 0.7282902002334595, "learning_rate": 8.462426033072442e-05, "loss": 0.055, "step": 29370 }, { "grad_norm": 0.7434869408607483, "learning_rate": 8.461232977610061e-05, "loss": 0.0545, "step": 29380 }, { "grad_norm": 0.440280944108963, "learning_rate": 8.46003954363245e-05, "loss": 0.0521, "step": 29390 }, { "grad_norm": 0.4603901505470276, "learning_rate": 8.458845731270115e-05, "loss": 0.0629, "step": 29400 }, { "grad_norm": 0.466966837644577, "learning_rate": 8.45765154065361e-05, "loss": 0.0437, "step": 29410 }, { "grad_norm": 1.0875639915466309, "learning_rate": 8.456456971913532e-05, "loss": 0.0818, "step": 29420 }, { "grad_norm": 0.4054872691631317, "learning_rate": 8.455262025180517e-05, "loss": 0.0472, "step": 29430 }, { "grad_norm": 0.5068404674530029, "learning_rate": 8.454066700585242e-05, "loss": 0.062, "step": 29440 }, { "grad_norm": 0.5405072569847107, "learning_rate": 8.452870998258423e-05, "loss": 0.0618, "step": 29450 }, { "grad_norm": 0.3947611153125763, "learning_rate": 8.451674918330825e-05, "loss": 0.0704, "step": 29460 }, { "grad_norm": 0.6294853687286377, "learning_rate": 8.450478460933246e-05, "loss": 0.082, "step": 29470 }, { "grad_norm": 0.5677049160003662, "learning_rate": 8.449281626196532e-05, "loss": 0.0541, "step": 29480 }, { "grad_norm": 0.6910837888717651, "learning_rate": 8.448084414251564e-05, "loss": 0.0482, "step": 29490 }, { "grad_norm": 0.7112698554992676, "learning_rate": 8.446886825229271e-05, "loss": 0.0652, "step": 29500 }, { "grad_norm": 0.5149466395378113, "learning_rate": 8.445688859260615e-05, "loss": 0.0582, "step": 29510 }, { "grad_norm": 0.5135365128517151, "learning_rate": 8.444490516476606e-05, "loss": 0.0804, "step": 29520 }, { "grad_norm": 0.5226494073867798, "learning_rate": 8.443291797008293e-05, "loss": 0.0568, "step": 29530 }, { "grad_norm": 0.6049647927284241, "learning_rate": 8.442092700986765e-05, "loss": 0.045, "step": 29540 }, { "grad_norm": 0.8335763812065125, "learning_rate": 8.440893228543156e-05, "loss": 0.0485, "step": 29550 }, { "grad_norm": 0.43999770283699036, "learning_rate": 8.439693379808638e-05, "loss": 0.055, "step": 29560 }, { "grad_norm": 0.6009343266487122, "learning_rate": 8.43849315491442e-05, "loss": 0.0502, "step": 29570 }, { "grad_norm": 0.419799268245697, "learning_rate": 8.437292553991763e-05, "loss": 0.0462, "step": 29580 }, { "grad_norm": 0.43394842743873596, "learning_rate": 8.436091577171959e-05, "loss": 0.0494, "step": 29590 }, { "grad_norm": 0.7916247844696045, "learning_rate": 8.434890224586347e-05, "loss": 0.0618, "step": 29600 }, { "grad_norm": 0.5005068778991699, "learning_rate": 8.433688496366303e-05, "loss": 0.0567, "step": 29610 }, { "grad_norm": 0.6382719874382019, "learning_rate": 8.432486392643248e-05, "loss": 0.0433, "step": 29620 }, { "grad_norm": 0.7588242292404175, "learning_rate": 8.431283913548643e-05, "loss": 0.0744, "step": 29630 }, { "grad_norm": 1.0431877374649048, "learning_rate": 8.430081059213985e-05, "loss": 0.0554, "step": 29640 }, { "grad_norm": 0.9021137952804565, "learning_rate": 8.428877829770823e-05, "loss": 0.0945, "step": 29650 }, { "grad_norm": 0.8010504841804504, "learning_rate": 8.427674225350735e-05, "loss": 0.0498, "step": 29660 }, { "grad_norm": 0.4667436480522156, "learning_rate": 8.426470246085347e-05, "loss": 0.0952, "step": 29670 }, { "grad_norm": 0.65386563539505, "learning_rate": 8.425265892106324e-05, "loss": 0.0611, "step": 29680 }, { "grad_norm": 0.860553503036499, "learning_rate": 8.424061163545374e-05, "loss": 0.0557, "step": 29690 }, { "grad_norm": 0.646673858165741, "learning_rate": 8.422856060534243e-05, "loss": 0.1015, "step": 29700 }, { "grad_norm": 0.5842317342758179, "learning_rate": 8.421650583204718e-05, "loss": 0.0564, "step": 29710 }, { "grad_norm": 1.4893299341201782, "learning_rate": 8.420444731688633e-05, "loss": 0.0628, "step": 29720 }, { "grad_norm": 1.699898362159729, "learning_rate": 8.419238506117852e-05, "loss": 0.0597, "step": 29730 }, { "grad_norm": 0.7809450030326843, "learning_rate": 8.418031906624289e-05, "loss": 0.0749, "step": 29740 }, { "grad_norm": 0.8760849237442017, "learning_rate": 8.416824933339898e-05, "loss": 0.0527, "step": 29750 }, { "grad_norm": 0.7736906409263611, "learning_rate": 8.415617586396667e-05, "loss": 0.0787, "step": 29760 }, { "grad_norm": 0.766496479511261, "learning_rate": 8.414409865926632e-05, "loss": 0.052, "step": 29770 }, { "grad_norm": 0.5118948221206665, "learning_rate": 8.413201772061867e-05, "loss": 0.0516, "step": 29780 }, { "grad_norm": 0.6453195214271545, "learning_rate": 8.411993304934488e-05, "loss": 0.0556, "step": 29790 }, { "grad_norm": 0.7074663639068604, "learning_rate": 8.410784464676654e-05, "loss": 0.0484, "step": 29800 }, { "grad_norm": 0.7574697136878967, "learning_rate": 8.409575251420556e-05, "loss": 0.06, "step": 29810 }, { "grad_norm": 0.7136794924736023, "learning_rate": 8.408365665298435e-05, "loss": 0.0562, "step": 29820 }, { "grad_norm": 0.7394798994064331, "learning_rate": 8.40715570644257e-05, "loss": 0.0562, "step": 29830 }, { "grad_norm": 0.540984034538269, "learning_rate": 8.40594537498528e-05, "loss": 0.059, "step": 29840 }, { "grad_norm": 0.4863586127758026, "learning_rate": 8.404734671058924e-05, "loss": 0.0471, "step": 29850 }, { "grad_norm": 0.6824972629547119, "learning_rate": 8.403523594795902e-05, "loss": 0.0431, "step": 29860 }, { "grad_norm": 0.8617722988128662, "learning_rate": 8.402312146328659e-05, "loss": 0.0698, "step": 29870 }, { "grad_norm": 0.5865499973297119, "learning_rate": 8.401100325789675e-05, "loss": 0.0544, "step": 29880 }, { "grad_norm": 0.5131210684776306, "learning_rate": 8.399888133311472e-05, "loss": 0.0481, "step": 29890 }, { "grad_norm": 0.7644587755203247, "learning_rate": 8.398675569026613e-05, "loss": 0.0995, "step": 29900 }, { "grad_norm": 0.7333899736404419, "learning_rate": 8.397462633067705e-05, "loss": 0.0529, "step": 29910 }, { "grad_norm": 0.5628916621208191, "learning_rate": 8.396249325567392e-05, "loss": 0.0638, "step": 29920 }, { "grad_norm": 0.7034074664115906, "learning_rate": 8.395035646658357e-05, "loss": 0.0469, "step": 29930 }, { "grad_norm": 0.49886205792427063, "learning_rate": 8.39382159647333e-05, "loss": 0.0686, "step": 29940 }, { "grad_norm": 0.8505822420120239, "learning_rate": 8.392607175145075e-05, "loss": 0.0513, "step": 29950 }, { "grad_norm": 0.5670628547668457, "learning_rate": 8.3913923828064e-05, "loss": 0.0456, "step": 29960 }, { "grad_norm": 0.7146621346473694, "learning_rate": 8.390177219590152e-05, "loss": 0.0533, "step": 29970 }, { "grad_norm": 0.5333665013313293, "learning_rate": 8.388961685629222e-05, "loss": 0.0502, "step": 29980 }, { "grad_norm": 0.842754602432251, "learning_rate": 8.387745781056536e-05, "loss": 0.063, "step": 29990 }, { "grad_norm": 0.8068184852600098, "learning_rate": 8.386529506005065e-05, "loss": 0.0625, "step": 30000 }, { "grad_norm": 0.6430418491363525, "learning_rate": 8.38531286060782e-05, "loss": 0.0404, "step": 30010 }, { "grad_norm": 0.32072845101356506, "learning_rate": 8.384095844997849e-05, "loss": 0.0391, "step": 30020 }, { "grad_norm": 0.5808405876159668, "learning_rate": 8.382878459308245e-05, "loss": 0.0399, "step": 30030 }, { "grad_norm": 0.4286409914493561, "learning_rate": 8.381660703672138e-05, "loss": 0.0611, "step": 30040 }, { "grad_norm": 0.5144693851470947, "learning_rate": 8.380442578222702e-05, "loss": 0.0475, "step": 30050 }, { "grad_norm": 0.6451444029808044, "learning_rate": 8.379224083093146e-05, "loss": 0.0477, "step": 30060 }, { "grad_norm": 0.6495888829231262, "learning_rate": 8.378005218416727e-05, "loss": 0.0907, "step": 30070 }, { "grad_norm": 0.4826519787311554, "learning_rate": 8.376785984326735e-05, "loss": 0.0382, "step": 30080 }, { "grad_norm": 0.6512212157249451, "learning_rate": 8.375566380956506e-05, "loss": 0.0493, "step": 30090 }, { "grad_norm": 0.570907711982727, "learning_rate": 8.374346408439411e-05, "loss": 0.0798, "step": 30100 }, { "grad_norm": 1.3307335376739502, "learning_rate": 8.373126066908868e-05, "loss": 0.0729, "step": 30110 }, { "grad_norm": 0.46390050649642944, "learning_rate": 8.371905356498326e-05, "loss": 0.0558, "step": 30120 }, { "grad_norm": 0.7628745436668396, "learning_rate": 8.370684277341288e-05, "loss": 0.0426, "step": 30130 }, { "grad_norm": 0.7271339893341064, "learning_rate": 8.369462829571282e-05, "loss": 0.0626, "step": 30140 }, { "grad_norm": 0.6875180006027222, "learning_rate": 8.36824101332189e-05, "loss": 0.0447, "step": 30150 }, { "grad_norm": 0.6897113919258118, "learning_rate": 8.367018828726721e-05, "loss": 0.0577, "step": 30160 }, { "grad_norm": 0.5618499517440796, "learning_rate": 8.365796275919438e-05, "loss": 0.068, "step": 30170 }, { "grad_norm": 0.6999192237854004, "learning_rate": 8.364573355033734e-05, "loss": 0.0616, "step": 30180 }, { "grad_norm": 0.9657220244407654, "learning_rate": 8.363350066203346e-05, "loss": 0.0632, "step": 30190 }, { "grad_norm": 0.3454291522502899, "learning_rate": 8.362126409562053e-05, "loss": 0.0441, "step": 30200 }, { "grad_norm": 0.3711743652820587, "learning_rate": 8.360902385243667e-05, "loss": 0.0433, "step": 30210 }, { "grad_norm": 0.8009222745895386, "learning_rate": 8.359677993382052e-05, "loss": 0.0571, "step": 30220 }, { "grad_norm": 0.48401719331741333, "learning_rate": 8.358453234111103e-05, "loss": 0.0573, "step": 30230 }, { "grad_norm": 0.5005865097045898, "learning_rate": 8.357228107564756e-05, "loss": 0.0609, "step": 30240 }, { "grad_norm": 0.3814983069896698, "learning_rate": 8.356002613876993e-05, "loss": 0.0522, "step": 30250 }, { "grad_norm": 0.3946830630302429, "learning_rate": 8.35477675318183e-05, "loss": 0.0451, "step": 30260 }, { "grad_norm": 0.8562687039375305, "learning_rate": 8.353550525613323e-05, "loss": 0.0453, "step": 30270 }, { "grad_norm": 0.6086758375167847, "learning_rate": 8.352323931305572e-05, "loss": 0.0439, "step": 30280 }, { "grad_norm": 0.7286983728408813, "learning_rate": 8.351096970392717e-05, "loss": 0.0481, "step": 30290 }, { "grad_norm": 0.5348806977272034, "learning_rate": 8.349869643008937e-05, "loss": 0.0528, "step": 30300 }, { "grad_norm": 0.661064624786377, "learning_rate": 8.348641949288449e-05, "loss": 0.0573, "step": 30310 }, { "grad_norm": 0.8681427836418152, "learning_rate": 8.34741388936551e-05, "loss": 0.0732, "step": 30320 }, { "grad_norm": 0.44329437613487244, "learning_rate": 8.346185463374423e-05, "loss": 0.0616, "step": 30330 }, { "grad_norm": 0.6700881719589233, "learning_rate": 8.344956671449524e-05, "loss": 0.07, "step": 30340 }, { "grad_norm": 0.5990056395530701, "learning_rate": 8.343727513725192e-05, "loss": 0.0428, "step": 30350 }, { "grad_norm": 0.5907198786735535, "learning_rate": 8.342497990335847e-05, "loss": 0.0571, "step": 30360 }, { "grad_norm": 0.855341374874115, "learning_rate": 8.341268101415946e-05, "loss": 0.0536, "step": 30370 }, { "grad_norm": 0.6473464965820312, "learning_rate": 8.34003784709999e-05, "loss": 0.0426, "step": 30380 }, { "grad_norm": 0.7513547539710999, "learning_rate": 8.338807227522516e-05, "loss": 0.06, "step": 30390 }, { "grad_norm": 0.6631538271903992, "learning_rate": 8.337576242818103e-05, "loss": 0.0535, "step": 30400 }, { "grad_norm": 0.4629865884780884, "learning_rate": 8.336344893121372e-05, "loss": 0.0604, "step": 30410 }, { "grad_norm": 0.6074305772781372, "learning_rate": 8.335113178566977e-05, "loss": 0.0465, "step": 30420 }, { "grad_norm": 0.7378824353218079, "learning_rate": 8.33388109928962e-05, "loss": 0.0578, "step": 30430 }, { "grad_norm": 0.7588924169540405, "learning_rate": 8.33264865542404e-05, "loss": 0.0556, "step": 30440 }, { "grad_norm": 0.5536662340164185, "learning_rate": 8.331415847105013e-05, "loss": 0.0516, "step": 30450 }, { "grad_norm": 0.433270126581192, "learning_rate": 8.330182674467357e-05, "loss": 0.0436, "step": 30460 }, { "grad_norm": 0.6031191349029541, "learning_rate": 8.32894913764593e-05, "loss": 0.0468, "step": 30470 }, { "grad_norm": 0.26867279410362244, "learning_rate": 8.327715236775633e-05, "loss": 0.0543, "step": 30480 }, { "grad_norm": 0.42437082529067993, "learning_rate": 8.326480971991398e-05, "loss": 0.0418, "step": 30490 }, { "grad_norm": 0.6797770857810974, "learning_rate": 8.325246343428206e-05, "loss": 0.0843, "step": 30500 }, { "grad_norm": 0.7369707226753235, "learning_rate": 8.324011351221072e-05, "loss": 0.0581, "step": 30510 }, { "grad_norm": 0.4570421576499939, "learning_rate": 8.322775995505057e-05, "loss": 0.0514, "step": 30520 }, { "grad_norm": 0.9234622716903687, "learning_rate": 8.321540276415254e-05, "loss": 0.0487, "step": 30530 }, { "grad_norm": 0.638795793056488, "learning_rate": 8.320304194086798e-05, "loss": 0.0604, "step": 30540 }, { "grad_norm": 0.8034037351608276, "learning_rate": 8.31906774865487e-05, "loss": 0.0519, "step": 30550 }, { "grad_norm": 0.5198103189468384, "learning_rate": 8.317830940254682e-05, "loss": 0.0414, "step": 30560 }, { "grad_norm": 0.6164689660072327, "learning_rate": 8.316593769021491e-05, "loss": 0.0616, "step": 30570 }, { "grad_norm": 0.6742591261863708, "learning_rate": 8.315356235090592e-05, "loss": 0.0594, "step": 30580 }, { "grad_norm": 0.7187947630882263, "learning_rate": 8.314118338597319e-05, "loss": 0.0528, "step": 30590 }, { "grad_norm": 1.558611512184143, "learning_rate": 8.312880079677048e-05, "loss": 0.0673, "step": 30600 }, { "grad_norm": 0.5134990811347961, "learning_rate": 8.311641458465191e-05, "loss": 0.0789, "step": 30610 }, { "grad_norm": 0.6772116422653198, "learning_rate": 8.310402475097205e-05, "loss": 0.0514, "step": 30620 }, { "grad_norm": 0.8313131928443909, "learning_rate": 8.309163129708581e-05, "loss": 0.0665, "step": 30630 }, { "grad_norm": 0.8090296983718872, "learning_rate": 8.307923422434852e-05, "loss": 0.0476, "step": 30640 }, { "grad_norm": 0.49224382638931274, "learning_rate": 8.30668335341159e-05, "loss": 0.0449, "step": 30650 }, { "grad_norm": 0.7026625871658325, "learning_rate": 8.305442922774408e-05, "loss": 0.0536, "step": 30660 }, { "grad_norm": 0.4752962291240692, "learning_rate": 8.304202130658959e-05, "loss": 0.0695, "step": 30670 }, { "grad_norm": 0.5976340770721436, "learning_rate": 8.302960977200931e-05, "loss": 0.0652, "step": 30680 }, { "grad_norm": 0.9221054911613464, "learning_rate": 8.301719462536058e-05, "loss": 0.0601, "step": 30690 }, { "grad_norm": 0.5795983076095581, "learning_rate": 8.300477586800108e-05, "loss": 0.0699, "step": 30700 }, { "grad_norm": 0.7327424883842468, "learning_rate": 8.299235350128892e-05, "loss": 0.0505, "step": 30710 }, { "grad_norm": 0.4593263268470764, "learning_rate": 8.297992752658258e-05, "loss": 0.0501, "step": 30720 }, { "grad_norm": 0.43932849168777466, "learning_rate": 8.296749794524094e-05, "loss": 0.073, "step": 30730 }, { "grad_norm": 0.5610955357551575, "learning_rate": 8.295506475862332e-05, "loss": 0.0447, "step": 30740 }, { "grad_norm": 0.5725947022438049, "learning_rate": 8.294262796808933e-05, "loss": 0.0517, "step": 30750 }, { "grad_norm": 0.41426485776901245, "learning_rate": 8.293018757499909e-05, "loss": 0.0365, "step": 30760 }, { "grad_norm": 0.4045019745826721, "learning_rate": 8.291774358071305e-05, "loss": 0.0437, "step": 30770 }, { "grad_norm": 1.1170737743377686, "learning_rate": 8.290529598659205e-05, "loss": 0.059, "step": 30780 }, { "grad_norm": 0.4841304421424866, "learning_rate": 8.289284479399738e-05, "loss": 0.0711, "step": 30790 }, { "grad_norm": 0.46308159828186035, "learning_rate": 8.288039000429064e-05, "loss": 0.0497, "step": 30800 }, { "grad_norm": 0.7971171140670776, "learning_rate": 8.286793161883388e-05, "loss": 0.0521, "step": 30810 }, { "grad_norm": 0.7538214325904846, "learning_rate": 8.285546963898954e-05, "loss": 0.0504, "step": 30820 }, { "grad_norm": 0.6601231694221497, "learning_rate": 8.284300406612044e-05, "loss": 0.0596, "step": 30830 }, { "grad_norm": 0.5134977102279663, "learning_rate": 8.283053490158978e-05, "loss": 0.0536, "step": 30840 }, { "grad_norm": 0.5816752910614014, "learning_rate": 8.28180621467612e-05, "loss": 0.064, "step": 30850 }, { "grad_norm": 0.5770457983016968, "learning_rate": 8.280558580299867e-05, "loss": 0.0849, "step": 30860 }, { "grad_norm": 1.1030124425888062, "learning_rate": 8.279310587166661e-05, "loss": 0.0583, "step": 30870 }, { "grad_norm": 1.660538673400879, "learning_rate": 8.278062235412978e-05, "loss": 0.0534, "step": 30880 }, { "grad_norm": 0.7898409366607666, "learning_rate": 8.276813525175339e-05, "loss": 0.0791, "step": 30890 }, { "grad_norm": 0.42419400811195374, "learning_rate": 8.2755644565903e-05, "loss": 0.0555, "step": 30900 }, { "grad_norm": 0.6362043619155884, "learning_rate": 8.274315029794454e-05, "loss": 0.0547, "step": 30910 }, { "grad_norm": 0.6274636387825012, "learning_rate": 8.273065244924443e-05, "loss": 0.056, "step": 30920 }, { "grad_norm": 0.41795825958251953, "learning_rate": 8.271815102116936e-05, "loss": 0.0469, "step": 30930 }, { "grad_norm": 0.39934656023979187, "learning_rate": 8.270564601508648e-05, "loss": 0.0435, "step": 30940 }, { "grad_norm": 0.4468233287334442, "learning_rate": 8.269313743236333e-05, "loss": 0.051, "step": 30950 }, { "grad_norm": 0.8803958296775818, "learning_rate": 8.268062527436783e-05, "loss": 0.0734, "step": 30960 }, { "grad_norm": 0.44917017221450806, "learning_rate": 8.266810954246828e-05, "loss": 0.0548, "step": 30970 }, { "grad_norm": 0.5425060987472534, "learning_rate": 8.265559023803338e-05, "loss": 0.0544, "step": 30980 }, { "grad_norm": 0.4803340435028076, "learning_rate": 8.264306736243223e-05, "loss": 0.0471, "step": 30990 }, { "grad_norm": 0.6427192687988281, "learning_rate": 8.263054091703432e-05, "loss": 0.0609, "step": 31000 }, { "grad_norm": 0.6435278058052063, "learning_rate": 8.26180109032095e-05, "loss": 0.0607, "step": 31010 }, { "grad_norm": 0.5923286080360413, "learning_rate": 8.260547732232807e-05, "loss": 0.0495, "step": 31020 }, { "grad_norm": 0.7357374429702759, "learning_rate": 8.259294017576063e-05, "loss": 0.0512, "step": 31030 }, { "grad_norm": 0.5783547163009644, "learning_rate": 8.258039946487828e-05, "loss": 0.0472, "step": 31040 }, { "grad_norm": 0.5469175577163696, "learning_rate": 8.256785519105241e-05, "loss": 0.0453, "step": 31050 }, { "grad_norm": 0.49231866002082825, "learning_rate": 8.255530735565488e-05, "loss": 0.0488, "step": 31060 }, { "grad_norm": 0.5711551904678345, "learning_rate": 8.254275596005787e-05, "loss": 0.0548, "step": 31070 }, { "grad_norm": 0.8437358140945435, "learning_rate": 8.2530201005634e-05, "loss": 0.0411, "step": 31080 }, { "grad_norm": 0.5082220435142517, "learning_rate": 8.251764249375626e-05, "loss": 0.0493, "step": 31090 }, { "grad_norm": 0.44232016801834106, "learning_rate": 8.250508042579803e-05, "loss": 0.0408, "step": 31100 }, { "grad_norm": 0.4788110554218292, "learning_rate": 8.249251480313307e-05, "loss": 0.0658, "step": 31110 }, { "grad_norm": 0.6837171316146851, "learning_rate": 8.247994562713555e-05, "loss": 0.0465, "step": 31120 }, { "grad_norm": 0.5357500910758972, "learning_rate": 8.246737289917998e-05, "loss": 0.0522, "step": 31130 }, { "grad_norm": 0.5016770362854004, "learning_rate": 8.245479662064135e-05, "loss": 0.0623, "step": 31140 }, { "grad_norm": 0.5016173720359802, "learning_rate": 8.244221679289496e-05, "loss": 0.0506, "step": 31150 }, { "grad_norm": 0.5599626898765564, "learning_rate": 8.242963341731652e-05, "loss": 0.0628, "step": 31160 }, { "grad_norm": 1.013882040977478, "learning_rate": 8.24170464952821e-05, "loss": 0.0473, "step": 31170 }, { "grad_norm": 0.6007261276245117, "learning_rate": 8.240445602816824e-05, "loss": 0.0494, "step": 31180 }, { "grad_norm": 0.5345967411994934, "learning_rate": 8.239186201735179e-05, "loss": 0.056, "step": 31190 }, { "grad_norm": 0.7514596581459045, "learning_rate": 8.237926446420998e-05, "loss": 0.0486, "step": 31200 }, { "grad_norm": 0.6060565710067749, "learning_rate": 8.236666337012052e-05, "loss": 0.0504, "step": 31210 }, { "grad_norm": 0.7605944275856018, "learning_rate": 8.23540587364614e-05, "loss": 0.0555, "step": 31220 }, { "grad_norm": 0.81788569688797, "learning_rate": 8.234145056461107e-05, "loss": 0.0444, "step": 31230 }, { "grad_norm": 0.6564434766769409, "learning_rate": 8.232883885594832e-05, "loss": 0.0437, "step": 31240 }, { "grad_norm": 0.5499234199523926, "learning_rate": 8.231622361185236e-05, "loss": 0.0515, "step": 31250 }, { "grad_norm": 0.6468026638031006, "learning_rate": 8.230360483370278e-05, "loss": 0.0561, "step": 31260 }, { "grad_norm": 0.48954102396965027, "learning_rate": 8.229098252287953e-05, "loss": 0.0528, "step": 31270 }, { "grad_norm": 0.5193198323249817, "learning_rate": 8.2278356680763e-05, "loss": 0.0516, "step": 31280 }, { "grad_norm": 0.6640859842300415, "learning_rate": 8.22657273087339e-05, "loss": 0.0535, "step": 31290 }, { "grad_norm": 0.5208341479301453, "learning_rate": 8.225309440817336e-05, "loss": 0.0694, "step": 31300 }, { "grad_norm": 0.7011538743972778, "learning_rate": 8.224045798046293e-05, "loss": 0.0536, "step": 31310 }, { "grad_norm": 0.5414659976959229, "learning_rate": 8.22278180269845e-05, "loss": 0.0624, "step": 31320 }, { "grad_norm": 0.654827356338501, "learning_rate": 8.221517454912031e-05, "loss": 0.0443, "step": 31330 }, { "grad_norm": 0.6835029125213623, "learning_rate": 8.220252754825308e-05, "loss": 0.0604, "step": 31340 }, { "grad_norm": 0.8266622424125671, "learning_rate": 8.218987702576586e-05, "loss": 0.0484, "step": 31350 }, { "grad_norm": 1.7763872146606445, "learning_rate": 8.217722298304207e-05, "loss": 0.0517, "step": 31360 }, { "grad_norm": 0.4870441257953644, "learning_rate": 8.216456542146557e-05, "loss": 0.0629, "step": 31370 }, { "grad_norm": 0.7264660596847534, "learning_rate": 8.215190434242055e-05, "loss": 0.0615, "step": 31380 }, { "grad_norm": 0.6011906266212463, "learning_rate": 8.213923974729161e-05, "loss": 0.0492, "step": 31390 }, { "grad_norm": 0.6965083479881287, "learning_rate": 8.212657163746373e-05, "loss": 0.0528, "step": 31400 }, { "grad_norm": 0.7466484904289246, "learning_rate": 8.211390001432227e-05, "loss": 0.0526, "step": 31410 }, { "grad_norm": 0.5973432660102844, "learning_rate": 8.210122487925297e-05, "loss": 0.0559, "step": 31420 }, { "grad_norm": 0.4777005612850189, "learning_rate": 8.208854623364202e-05, "loss": 0.0471, "step": 31430 }, { "grad_norm": 0.6741523742675781, "learning_rate": 8.207586407887589e-05, "loss": 0.0566, "step": 31440 }, { "grad_norm": 0.8755640387535095, "learning_rate": 8.206317841634148e-05, "loss": 0.0634, "step": 31450 }, { "grad_norm": 0.8138473033905029, "learning_rate": 8.205048924742609e-05, "loss": 0.0608, "step": 31460 }, { "grad_norm": 0.4104393720626831, "learning_rate": 8.203779657351738e-05, "loss": 0.0689, "step": 31470 }, { "grad_norm": 0.5113550424575806, "learning_rate": 8.20251003960034e-05, "loss": 0.0504, "step": 31480 }, { "grad_norm": 0.5004459023475647, "learning_rate": 8.201240071627258e-05, "loss": 0.0591, "step": 31490 }, { "grad_norm": 0.6525952219963074, "learning_rate": 8.199969753571377e-05, "loss": 0.0695, "step": 31500 }, { "grad_norm": 0.8689609169960022, "learning_rate": 8.198699085571615e-05, "loss": 0.0502, "step": 31510 }, { "grad_norm": 0.6184221506118774, "learning_rate": 8.197428067766928e-05, "loss": 0.0572, "step": 31520 }, { "grad_norm": 0.5776312947273254, "learning_rate": 8.196156700296316e-05, "loss": 0.048, "step": 31530 }, { "grad_norm": 0.5451358556747437, "learning_rate": 8.194884983298814e-05, "loss": 0.0656, "step": 31540 }, { "grad_norm": 0.68075031042099, "learning_rate": 8.193612916913491e-05, "loss": 0.0679, "step": 31550 }, { "grad_norm": 0.43241459131240845, "learning_rate": 8.192340501279463e-05, "loss": 0.048, "step": 31560 }, { "grad_norm": 0.4301852881908417, "learning_rate": 8.191067736535876e-05, "loss": 0.0741, "step": 31570 }, { "grad_norm": 0.6021920442581177, "learning_rate": 8.18979462282192e-05, "loss": 0.0548, "step": 31580 }, { "grad_norm": 0.6405321955680847, "learning_rate": 8.188521160276819e-05, "loss": 0.0517, "step": 31590 }, { "grad_norm": 0.44253283739089966, "learning_rate": 8.187247349039837e-05, "loss": 0.0466, "step": 31600 }, { "grad_norm": 0.4352651536464691, "learning_rate": 8.185973189250278e-05, "loss": 0.0543, "step": 31610 }, { "grad_norm": 0.6041138172149658, "learning_rate": 8.184698681047482e-05, "loss": 0.0544, "step": 31620 }, { "grad_norm": 0.9169513583183289, "learning_rate": 8.183423824570827e-05, "loss": 0.0496, "step": 31630 }, { "grad_norm": 0.9022266864776611, "learning_rate": 8.182148619959725e-05, "loss": 0.0559, "step": 31640 }, { "grad_norm": 0.45563313364982605, "learning_rate": 8.180873067353636e-05, "loss": 0.0516, "step": 31650 }, { "grad_norm": 0.49717801809310913, "learning_rate": 8.179597166892052e-05, "loss": 0.0421, "step": 31660 }, { "grad_norm": 0.7537761330604553, "learning_rate": 8.178320918714501e-05, "loss": 0.041, "step": 31670 }, { "grad_norm": 0.5719937682151794, "learning_rate": 8.177044322960554e-05, "loss": 0.0562, "step": 31680 }, { "grad_norm": 0.42181164026260376, "learning_rate": 8.175767379769816e-05, "loss": 0.0489, "step": 31690 }, { "grad_norm": 0.8118376135826111, "learning_rate": 8.174490089281932e-05, "loss": 0.0546, "step": 31700 }, { "grad_norm": 0.5840182304382324, "learning_rate": 8.173212451636584e-05, "loss": 0.0532, "step": 31710 }, { "grad_norm": 0.6062561273574829, "learning_rate": 8.171934466973493e-05, "loss": 0.0524, "step": 31720 }, { "grad_norm": 0.5878864526748657, "learning_rate": 8.170656135432418e-05, "loss": 0.0555, "step": 31730 }, { "grad_norm": 0.6113609671592712, "learning_rate": 8.169377457153155e-05, "loss": 0.0589, "step": 31740 }, { "grad_norm": 0.6896016597747803, "learning_rate": 8.168098432275539e-05, "loss": 0.0651, "step": 31750 }, { "grad_norm": 0.726345419883728, "learning_rate": 8.166819060939442e-05, "loss": 0.0515, "step": 31760 }, { "grad_norm": 0.7690410614013672, "learning_rate": 8.165539343284772e-05, "loss": 0.0717, "step": 31770 }, { "grad_norm": 0.7366175651550293, "learning_rate": 8.16425927945148e-05, "loss": 0.0628, "step": 31780 }, { "grad_norm": 0.5972748398780823, "learning_rate": 8.162978869579551e-05, "loss": 0.0468, "step": 31790 }, { "grad_norm": 0.6633315682411194, "learning_rate": 8.161698113809007e-05, "loss": 0.0444, "step": 31800 }, { "grad_norm": 0.43071097135543823, "learning_rate": 8.160417012279911e-05, "loss": 0.0487, "step": 31810 }, { "grad_norm": 0.6914149522781372, "learning_rate": 8.159135565132363e-05, "loss": 0.0407, "step": 31820 }, { "grad_norm": 0.892390787601471, "learning_rate": 8.157853772506498e-05, "loss": 0.0497, "step": 31830 }, { "grad_norm": 0.5900867581367493, "learning_rate": 8.156571634542494e-05, "loss": 0.0743, "step": 31840 }, { "grad_norm": 0.6341433525085449, "learning_rate": 8.15528915138056e-05, "loss": 0.0536, "step": 31850 }, { "grad_norm": 0.4790830612182617, "learning_rate": 8.154006323160949e-05, "loss": 0.0548, "step": 31860 }, { "grad_norm": 0.5713433623313904, "learning_rate": 8.152723150023949e-05, "loss": 0.0476, "step": 31870 }, { "grad_norm": 0.9243682026863098, "learning_rate": 8.151439632109886e-05, "loss": 0.0493, "step": 31880 }, { "grad_norm": 0.7607223391532898, "learning_rate": 8.150155769559122e-05, "loss": 0.0537, "step": 31890 }, { "grad_norm": 0.8040353655815125, "learning_rate": 8.148871562512058e-05, "loss": 0.0459, "step": 31900 }, { "grad_norm": 0.37669607996940613, "learning_rate": 8.147587011109136e-05, "loss": 0.0619, "step": 31910 }, { "grad_norm": 0.5622105002403259, "learning_rate": 8.14630211549083e-05, "loss": 0.0483, "step": 31920 }, { "grad_norm": 0.6126827597618103, "learning_rate": 8.145016875797655e-05, "loss": 0.0586, "step": 31930 }, { "grad_norm": 0.8316585421562195, "learning_rate": 8.143731292170164e-05, "loss": 0.0609, "step": 31940 }, { "grad_norm": 0.51405930519104, "learning_rate": 8.142445364748944e-05, "loss": 0.0456, "step": 31950 }, { "grad_norm": 0.7742401361465454, "learning_rate": 8.141159093674624e-05, "loss": 0.0526, "step": 31960 }, { "grad_norm": 0.6239575743675232, "learning_rate": 8.139872479087869e-05, "loss": 0.0664, "step": 31970 }, { "grad_norm": 0.6187298893928528, "learning_rate": 8.13858552112938e-05, "loss": 0.0685, "step": 31980 }, { "grad_norm": 0.5140230655670166, "learning_rate": 8.137298219939895e-05, "loss": 0.052, "step": 31990 }, { "grad_norm": 0.6258172988891602, "learning_rate": 8.136010575660196e-05, "loss": 0.0565, "step": 32000 }, { "grad_norm": 0.72130286693573, "learning_rate": 8.134722588431095e-05, "loss": 0.062, "step": 32010 }, { "grad_norm": 0.5641573071479797, "learning_rate": 8.133434258393444e-05, "loss": 0.0479, "step": 32020 }, { "grad_norm": 0.4868684709072113, "learning_rate": 8.132145585688134e-05, "loss": 0.0488, "step": 32030 }, { "grad_norm": 0.4720939099788666, "learning_rate": 8.130856570456093e-05, "loss": 0.0473, "step": 32040 }, { "grad_norm": 0.609568178653717, "learning_rate": 8.129567212838283e-05, "loss": 0.0623, "step": 32050 }, { "grad_norm": 0.9277063012123108, "learning_rate": 8.128277512975708e-05, "loss": 0.0505, "step": 32060 }, { "grad_norm": 0.8237242698669434, "learning_rate": 8.126987471009408e-05, "loss": 0.0537, "step": 32070 }, { "grad_norm": 0.4258837401866913, "learning_rate": 8.125697087080459e-05, "loss": 0.0475, "step": 32080 }, { "grad_norm": 0.4348285496234894, "learning_rate": 8.124406361329976e-05, "loss": 0.0414, "step": 32090 }, { "grad_norm": 0.3911788761615753, "learning_rate": 8.123115293899111e-05, "loss": 0.0616, "step": 32100 }, { "grad_norm": 0.7153493762016296, "learning_rate": 8.121823884929055e-05, "loss": 0.0502, "step": 32110 }, { "grad_norm": 0.6625062823295593, "learning_rate": 8.12053213456103e-05, "loss": 0.0505, "step": 32120 }, { "grad_norm": 0.5569779872894287, "learning_rate": 8.119240042936303e-05, "loss": 0.0482, "step": 32130 }, { "grad_norm": 0.9886958599090576, "learning_rate": 8.117947610196175e-05, "loss": 0.0538, "step": 32140 }, { "grad_norm": 0.6692584753036499, "learning_rate": 8.116654836481982e-05, "loss": 0.0494, "step": 32150 }, { "grad_norm": 0.6030862927436829, "learning_rate": 8.115361721935106e-05, "loss": 0.0536, "step": 32160 }, { "grad_norm": 0.431770384311676, "learning_rate": 8.114068266696953e-05, "loss": 0.055, "step": 32170 }, { "grad_norm": 0.6389089226722717, "learning_rate": 8.112774470908978e-05, "loss": 0.0494, "step": 32180 }, { "grad_norm": 0.6792505979537964, "learning_rate": 8.111480334712665e-05, "loss": 0.0504, "step": 32190 }, { "grad_norm": 0.5106507539749146, "learning_rate": 8.110185858249542e-05, "loss": 0.0484, "step": 32200 }, { "grad_norm": 0.5270494222640991, "learning_rate": 8.108891041661168e-05, "loss": 0.0578, "step": 32210 }, { "grad_norm": 0.6897917985916138, "learning_rate": 8.107595885089146e-05, "loss": 0.0432, "step": 32220 }, { "grad_norm": 0.3901243209838867, "learning_rate": 8.106300388675108e-05, "loss": 0.0565, "step": 32230 }, { "grad_norm": 0.651096522808075, "learning_rate": 8.105004552560731e-05, "loss": 0.0513, "step": 32240 }, { "grad_norm": 0.4358106553554535, "learning_rate": 8.103708376887724e-05, "loss": 0.0392, "step": 32250 }, { "grad_norm": 0.5829225182533264, "learning_rate": 8.102411861797836e-05, "loss": 0.0576, "step": 32260 }, { "grad_norm": 0.6147562861442566, "learning_rate": 8.101115007432851e-05, "loss": 0.0533, "step": 32270 }, { "grad_norm": 0.6446709036827087, "learning_rate": 8.09981781393459e-05, "loss": 0.0588, "step": 32280 }, { "grad_norm": 0.6738800406455994, "learning_rate": 8.098520281444915e-05, "loss": 0.0694, "step": 32290 }, { "grad_norm": 0.7129114270210266, "learning_rate": 8.09722241010572e-05, "loss": 0.0539, "step": 32300 }, { "grad_norm": 0.719082236289978, "learning_rate": 8.095924200058939e-05, "loss": 0.0685, "step": 32310 }, { "grad_norm": 0.47991809248924255, "learning_rate": 8.094625651446541e-05, "loss": 0.0499, "step": 32320 }, { "grad_norm": 0.38368380069732666, "learning_rate": 8.093326764410536e-05, "loss": 0.0451, "step": 32330 }, { "grad_norm": 0.4853332042694092, "learning_rate": 8.092027539092966e-05, "loss": 0.0486, "step": 32340 }, { "grad_norm": 0.5338862538337708, "learning_rate": 8.090727975635913e-05, "loss": 0.0455, "step": 32350 }, { "grad_norm": 0.6516401171684265, "learning_rate": 8.089428074181497e-05, "loss": 0.0516, "step": 32360 }, { "grad_norm": 0.4540645182132721, "learning_rate": 8.088127834871871e-05, "loss": 0.0607, "step": 32370 }, { "grad_norm": 0.6408059597015381, "learning_rate": 8.086827257849226e-05, "loss": 0.0517, "step": 32380 }, { "grad_norm": 0.7454990148544312, "learning_rate": 8.085526343255795e-05, "loss": 0.0496, "step": 32390 }, { "grad_norm": 0.6414504647254944, "learning_rate": 8.084225091233842e-05, "loss": 0.0505, "step": 32400 }, { "grad_norm": 0.5384784936904907, "learning_rate": 8.082923501925668e-05, "loss": 0.0653, "step": 32410 }, { "grad_norm": 0.7289745807647705, "learning_rate": 8.081621575473617e-05, "loss": 0.0566, "step": 32420 }, { "grad_norm": 0.6229825615882874, "learning_rate": 8.080319312020064e-05, "loss": 0.0467, "step": 32430 }, { "grad_norm": 0.4812510311603546, "learning_rate": 8.079016711707421e-05, "loss": 0.0438, "step": 32440 }, { "grad_norm": 0.47640806436538696, "learning_rate": 8.077713774678139e-05, "loss": 0.0526, "step": 32450 }, { "grad_norm": 0.6025669574737549, "learning_rate": 8.076410501074707e-05, "loss": 0.0604, "step": 32460 }, { "grad_norm": 0.4654606580734253, "learning_rate": 8.075106891039647e-05, "loss": 0.0631, "step": 32470 }, { "grad_norm": 0.5700111389160156, "learning_rate": 8.073802944715523e-05, "loss": 0.0583, "step": 32480 }, { "grad_norm": 0.8269456028938293, "learning_rate": 8.072498662244929e-05, "loss": 0.0566, "step": 32490 }, { "grad_norm": 0.6512296795845032, "learning_rate": 8.0711940437705e-05, "loss": 0.0586, "step": 32500 }, { "grad_norm": 0.45872965455055237, "learning_rate": 8.06988908943491e-05, "loss": 0.0439, "step": 32510 }, { "grad_norm": 0.8108077645301819, "learning_rate": 8.068583799380863e-05, "loss": 0.0632, "step": 32520 }, { "grad_norm": 0.5256950855255127, "learning_rate": 8.067278173751104e-05, "loss": 0.0494, "step": 32530 }, { "grad_norm": 0.7657178044319153, "learning_rate": 8.065972212688417e-05, "loss": 0.0566, "step": 32540 }, { "grad_norm": 0.46297821402549744, "learning_rate": 8.064665916335618e-05, "loss": 0.0693, "step": 32550 }, { "grad_norm": 0.5548151135444641, "learning_rate": 8.063359284835564e-05, "loss": 0.0466, "step": 32560 }, { "grad_norm": 0.5481754541397095, "learning_rate": 8.062052318331142e-05, "loss": 0.0587, "step": 32570 }, { "grad_norm": 0.48007214069366455, "learning_rate": 8.060745016965283e-05, "loss": 0.0567, "step": 32580 }, { "grad_norm": 0.6329478025436401, "learning_rate": 8.059437380880952e-05, "loss": 0.0418, "step": 32590 }, { "grad_norm": 0.6187704801559448, "learning_rate": 8.058129410221146e-05, "loss": 0.0473, "step": 32600 }, { "grad_norm": 0.5811978578567505, "learning_rate": 8.056821105128908e-05, "loss": 0.0507, "step": 32610 }, { "grad_norm": 0.5117042660713196, "learning_rate": 8.05551246574731e-05, "loss": 0.0544, "step": 32620 }, { "grad_norm": 0.6280187964439392, "learning_rate": 8.05420349221946e-05, "loss": 0.0512, "step": 32630 }, { "grad_norm": 0.5744190812110901, "learning_rate": 8.05289418468851e-05, "loss": 0.0482, "step": 32640 }, { "grad_norm": 0.3568062484264374, "learning_rate": 8.051584543297642e-05, "loss": 0.0709, "step": 32650 }, { "grad_norm": 0.402236670255661, "learning_rate": 8.050274568190074e-05, "loss": 0.0443, "step": 32660 }, { "grad_norm": 0.4938729703426361, "learning_rate": 8.048964259509067e-05, "loss": 0.0436, "step": 32670 }, { "grad_norm": 0.29288312792778015, "learning_rate": 8.047653617397914e-05, "loss": 0.0562, "step": 32680 }, { "grad_norm": 0.6923588514328003, "learning_rate": 8.046342641999941e-05, "loss": 0.0444, "step": 32690 }, { "grad_norm": 0.5175601243972778, "learning_rate": 8.045031333458517e-05, "loss": 0.0437, "step": 32700 }, { "grad_norm": 0.5238311290740967, "learning_rate": 8.043719691917047e-05, "loss": 0.0535, "step": 32710 }, { "grad_norm": 0.5841000080108643, "learning_rate": 8.042407717518966e-05, "loss": 0.0523, "step": 32720 }, { "grad_norm": 0.5251265168190002, "learning_rate": 8.041095410407751e-05, "loss": 0.0585, "step": 32730 }, { "grad_norm": 0.45275115966796875, "learning_rate": 8.039782770726913e-05, "loss": 0.0753, "step": 32740 }, { "grad_norm": 0.817834198474884, "learning_rate": 8.038469798620004e-05, "loss": 0.0614, "step": 32750 }, { "grad_norm": 0.42809659242630005, "learning_rate": 8.037156494230604e-05, "loss": 0.0466, "step": 32760 }, { "grad_norm": 0.5678197145462036, "learning_rate": 8.035842857702338e-05, "loss": 0.0623, "step": 32770 }, { "grad_norm": 0.5155614614486694, "learning_rate": 8.03452888917886e-05, "loss": 0.0722, "step": 32780 }, { "grad_norm": 0.694980800151825, "learning_rate": 8.033214588803866e-05, "loss": 0.0595, "step": 32790 }, { "grad_norm": 0.7451290488243103, "learning_rate": 8.031899956721083e-05, "loss": 0.0665, "step": 32800 }, { "grad_norm": 0.6457003355026245, "learning_rate": 8.030584993074282e-05, "loss": 0.0575, "step": 32810 }, { "grad_norm": 0.7917545437812805, "learning_rate": 8.02926969800726e-05, "loss": 0.0632, "step": 32820 }, { "grad_norm": 0.7033556699752808, "learning_rate": 8.027954071663859e-05, "loss": 0.0493, "step": 32830 }, { "grad_norm": 0.7377550601959229, "learning_rate": 8.026638114187954e-05, "loss": 0.0568, "step": 32840 }, { "grad_norm": 0.7067922353744507, "learning_rate": 8.025321825723456e-05, "loss": 0.0671, "step": 32850 }, { "grad_norm": 1.0961096286773682, "learning_rate": 8.02400520641431e-05, "loss": 0.0695, "step": 32860 }, { "grad_norm": 0.42589643597602844, "learning_rate": 8.022688256404501e-05, "loss": 0.0575, "step": 32870 }, { "grad_norm": 0.5916790962219238, "learning_rate": 8.02137097583805e-05, "loss": 0.0621, "step": 32880 }, { "grad_norm": 0.553495466709137, "learning_rate": 8.02005336485901e-05, "loss": 0.0497, "step": 32890 }, { "grad_norm": 0.8663196563720703, "learning_rate": 8.018735423611476e-05, "loss": 0.0545, "step": 32900 }, { "grad_norm": 1.2254139184951782, "learning_rate": 8.017417152239574e-05, "loss": 0.0547, "step": 32910 }, { "grad_norm": 0.747564435005188, "learning_rate": 8.01609855088747e-05, "loss": 0.0693, "step": 32920 }, { "grad_norm": 0.5905075669288635, "learning_rate": 8.014779619699362e-05, "loss": 0.0414, "step": 32930 }, { "grad_norm": 0.41944193840026855, "learning_rate": 8.013460358819489e-05, "loss": 0.0422, "step": 32940 }, { "grad_norm": 0.32764673233032227, "learning_rate": 8.01214076839212e-05, "loss": 0.0497, "step": 32950 }, { "grad_norm": 0.6853904724121094, "learning_rate": 8.010820848561565e-05, "loss": 0.07, "step": 32960 }, { "grad_norm": 0.5726180672645569, "learning_rate": 8.009500599472171e-05, "loss": 0.054, "step": 32970 }, { "grad_norm": 0.5036649107933044, "learning_rate": 8.008180021268314e-05, "loss": 0.0507, "step": 32980 }, { "grad_norm": 0.6016293168067932, "learning_rate": 8.006859114094414e-05, "loss": 0.0424, "step": 32990 }, { "grad_norm": 0.7294667363166809, "learning_rate": 8.005537878094921e-05, "loss": 0.053, "step": 33000 }, { "grad_norm": 0.6358551383018494, "learning_rate": 8.004216313414323e-05, "loss": 0.0522, "step": 33010 }, { "grad_norm": 0.7061586976051331, "learning_rate": 8.002894420197149e-05, "loss": 0.0477, "step": 33020 }, { "grad_norm": 0.5064224004745483, "learning_rate": 8.001572198587954e-05, "loss": 0.0544, "step": 33030 }, { "grad_norm": 0.7085022926330566, "learning_rate": 8.000249648731338e-05, "loss": 0.0524, "step": 33040 }, { "grad_norm": 0.9490669965744019, "learning_rate": 7.998926770771928e-05, "loss": 0.045, "step": 33050 }, { "grad_norm": 0.47702473402023315, "learning_rate": 7.997603564854397e-05, "loss": 0.0567, "step": 33060 }, { "grad_norm": 0.5509864091873169, "learning_rate": 7.996280031123448e-05, "loss": 0.0515, "step": 33070 }, { "grad_norm": 0.7872182130813599, "learning_rate": 7.994956169723818e-05, "loss": 0.0623, "step": 33080 }, { "grad_norm": 0.5400022268295288, "learning_rate": 7.993631980800285e-05, "loss": 0.0444, "step": 33090 }, { "grad_norm": 0.4538843333721161, "learning_rate": 7.992307464497659e-05, "loss": 0.0478, "step": 33100 }, { "grad_norm": 0.5554863214492798, "learning_rate": 7.990982620960787e-05, "loss": 0.0625, "step": 33110 }, { "grad_norm": 0.5979050397872925, "learning_rate": 7.989657450334554e-05, "loss": 0.0599, "step": 33120 }, { "grad_norm": 0.41812148690223694, "learning_rate": 7.988331952763877e-05, "loss": 0.0541, "step": 33130 }, { "grad_norm": 0.5825703144073486, "learning_rate": 7.987006128393709e-05, "loss": 0.0522, "step": 33140 }, { "grad_norm": 0.4154801368713379, "learning_rate": 7.985679977369043e-05, "loss": 0.0596, "step": 33150 }, { "grad_norm": 0.4563461244106293, "learning_rate": 7.984353499834902e-05, "loss": 0.0456, "step": 33160 }, { "grad_norm": 0.4169352650642395, "learning_rate": 7.983026695936351e-05, "loss": 0.0521, "step": 33170 }, { "grad_norm": 0.7006558775901794, "learning_rate": 7.981699565818486e-05, "loss": 0.0486, "step": 33180 }, { "grad_norm": 0.5213529467582703, "learning_rate": 7.980372109626437e-05, "loss": 0.0478, "step": 33190 }, { "grad_norm": 0.8421850800514221, "learning_rate": 7.979044327505375e-05, "loss": 0.0618, "step": 33200 }, { "grad_norm": 0.812343418598175, "learning_rate": 7.977716219600506e-05, "loss": 0.0514, "step": 33210 }, { "grad_norm": 0.5962663292884827, "learning_rate": 7.97638778605707e-05, "loss": 0.0444, "step": 33220 }, { "grad_norm": 0.57368004322052, "learning_rate": 7.975059027020338e-05, "loss": 0.0523, "step": 33230 }, { "grad_norm": 0.7383473515510559, "learning_rate": 7.973729942635623e-05, "loss": 0.0481, "step": 33240 }, { "grad_norm": 0.4851733148097992, "learning_rate": 7.972400533048273e-05, "loss": 0.0345, "step": 33250 }, { "grad_norm": 0.4364774525165558, "learning_rate": 7.97107079840367e-05, "loss": 0.0465, "step": 33260 }, { "grad_norm": 0.3211192190647125, "learning_rate": 7.969740738847231e-05, "loss": 0.0534, "step": 33270 }, { "grad_norm": 0.5548050403594971, "learning_rate": 7.968410354524411e-05, "loss": 0.0551, "step": 33280 }, { "grad_norm": 0.586448609828949, "learning_rate": 7.967079645580697e-05, "loss": 0.0558, "step": 33290 }, { "grad_norm": 0.6926543712615967, "learning_rate": 7.965748612161612e-05, "loss": 0.0534, "step": 33300 }, { "grad_norm": 0.5256012678146362, "learning_rate": 7.96441725441272e-05, "loss": 0.0482, "step": 33310 }, { "grad_norm": 0.4203701913356781, "learning_rate": 7.963085572479614e-05, "loss": 0.0498, "step": 33320 }, { "grad_norm": 0.5518797039985657, "learning_rate": 7.961753566507924e-05, "loss": 0.0465, "step": 33330 }, { "grad_norm": 0.6120755672454834, "learning_rate": 7.960421236643316e-05, "loss": 0.0719, "step": 33340 }, { "grad_norm": 0.43645480275154114, "learning_rate": 7.959088583031496e-05, "loss": 0.0523, "step": 33350 }, { "grad_norm": 0.7316871881484985, "learning_rate": 7.957755605818194e-05, "loss": 0.0511, "step": 33360 }, { "grad_norm": 0.42741090059280396, "learning_rate": 7.956422305149185e-05, "loss": 0.0433, "step": 33370 }, { "grad_norm": 0.6482309103012085, "learning_rate": 7.95508868117028e-05, "loss": 0.0556, "step": 33380 }, { "grad_norm": 0.5547244548797607, "learning_rate": 7.953754734027318e-05, "loss": 0.0499, "step": 33390 }, { "grad_norm": 0.5213826298713684, "learning_rate": 7.952420463866182e-05, "loss": 0.0542, "step": 33400 }, { "grad_norm": 0.6291033029556274, "learning_rate": 7.951085870832782e-05, "loss": 0.0585, "step": 33410 }, { "grad_norm": 0.446343332529068, "learning_rate": 7.949750955073067e-05, "loss": 0.0433, "step": 33420 }, { "grad_norm": 0.8500809669494629, "learning_rate": 7.948415716733022e-05, "loss": 0.0636, "step": 33430 }, { "grad_norm": 0.5858516097068787, "learning_rate": 7.947080155958669e-05, "loss": 0.0524, "step": 33440 }, { "grad_norm": 0.518081545829773, "learning_rate": 7.94574427289606e-05, "loss": 0.0366, "step": 33450 }, { "grad_norm": 0.486316055059433, "learning_rate": 7.944408067691284e-05, "loss": 0.0751, "step": 33460 }, { "grad_norm": 0.3880402445793152, "learning_rate": 7.943071540490473e-05, "loss": 0.0451, "step": 33470 }, { "grad_norm": 0.5677867531776428, "learning_rate": 7.94173469143978e-05, "loss": 0.0547, "step": 33480 }, { "grad_norm": 0.5127423405647278, "learning_rate": 7.940397520685406e-05, "loss": 0.042, "step": 33490 }, { "grad_norm": 0.8173666000366211, "learning_rate": 7.939060028373577e-05, "loss": 0.0675, "step": 33500 }, { "grad_norm": 0.7144628167152405, "learning_rate": 7.937722214650565e-05, "loss": 0.0466, "step": 33510 }, { "grad_norm": 0.7427124977111816, "learning_rate": 7.936384079662666e-05, "loss": 0.0468, "step": 33520 }, { "grad_norm": 0.45661604404449463, "learning_rate": 7.93504562355622e-05, "loss": 0.0424, "step": 33530 }, { "grad_norm": 0.620171070098877, "learning_rate": 7.933706846477599e-05, "loss": 0.0399, "step": 33540 }, { "grad_norm": 0.7781606316566467, "learning_rate": 7.932367748573206e-05, "loss": 0.0526, "step": 33550 }, { "grad_norm": 0.9584630727767944, "learning_rate": 7.931028329989485e-05, "loss": 0.0644, "step": 33560 }, { "grad_norm": 0.6404216885566711, "learning_rate": 7.929688590872913e-05, "loss": 0.0446, "step": 33570 }, { "grad_norm": 0.4266243875026703, "learning_rate": 7.928348531370003e-05, "loss": 0.0465, "step": 33580 }, { "grad_norm": 0.6545534133911133, "learning_rate": 7.927008151627297e-05, "loss": 0.052, "step": 33590 }, { "grad_norm": 0.7196189165115356, "learning_rate": 7.925667451791383e-05, "loss": 0.0617, "step": 33600 }, { "grad_norm": 0.3518139719963074, "learning_rate": 7.924326432008874e-05, "loss": 0.0327, "step": 33610 }, { "grad_norm": 0.7985785007476807, "learning_rate": 7.922985092426422e-05, "loss": 0.0603, "step": 33620 }, { "grad_norm": 0.5085580348968506, "learning_rate": 7.921643433190717e-05, "loss": 0.0503, "step": 33630 }, { "grad_norm": 0.47604942321777344, "learning_rate": 7.920301454448478e-05, "loss": 0.0404, "step": 33640 }, { "grad_norm": 0.4717019200325012, "learning_rate": 7.918959156346461e-05, "loss": 0.0483, "step": 33650 }, { "grad_norm": 0.5267356634140015, "learning_rate": 7.91761653903146e-05, "loss": 0.0444, "step": 33660 }, { "grad_norm": 0.4614465534687042, "learning_rate": 7.916273602650302e-05, "loss": 0.0415, "step": 33670 }, { "grad_norm": 0.727062463760376, "learning_rate": 7.914930347349847e-05, "loss": 0.0434, "step": 33680 }, { "grad_norm": 0.7517411112785339, "learning_rate": 7.913586773276992e-05, "loss": 0.0462, "step": 33690 }, { "grad_norm": 0.678608775138855, "learning_rate": 7.912242880578667e-05, "loss": 0.0507, "step": 33700 }, { "grad_norm": 0.4737692177295685, "learning_rate": 7.910898669401839e-05, "loss": 0.0431, "step": 33710 }, { "grad_norm": 0.37918052077293396, "learning_rate": 7.909554139893511e-05, "loss": 0.0666, "step": 33720 }, { "grad_norm": 0.5344062447547913, "learning_rate": 7.908209292200715e-05, "loss": 0.047, "step": 33730 }, { "grad_norm": 0.7689208388328552, "learning_rate": 7.906864126470523e-05, "loss": 0.053, "step": 33740 }, { "grad_norm": 0.4326581358909607, "learning_rate": 7.905518642850041e-05, "loss": 0.048, "step": 33750 }, { "grad_norm": 0.7387319207191467, "learning_rate": 7.904172841486409e-05, "loss": 0.0534, "step": 33760 }, { "grad_norm": 0.6396524310112, "learning_rate": 7.902826722526801e-05, "loss": 0.0536, "step": 33770 }, { "grad_norm": 0.8462465405464172, "learning_rate": 7.901480286118427e-05, "loss": 0.0509, "step": 33780 }, { "grad_norm": 0.5272505283355713, "learning_rate": 7.900133532408531e-05, "loss": 0.0494, "step": 33790 }, { "grad_norm": 0.7163825035095215, "learning_rate": 7.898786461544395e-05, "loss": 0.0467, "step": 33800 }, { "grad_norm": 0.7622002363204956, "learning_rate": 7.897439073673325e-05, "loss": 0.0462, "step": 33810 }, { "grad_norm": 0.44866329431533813, "learning_rate": 7.896091368942677e-05, "loss": 0.045, "step": 33820 }, { "grad_norm": 0.6616137027740479, "learning_rate": 7.894743347499832e-05, "loss": 0.0596, "step": 33830 }, { "grad_norm": 0.7899677157402039, "learning_rate": 7.893395009492203e-05, "loss": 0.054, "step": 33840 }, { "grad_norm": 0.7008536458015442, "learning_rate": 7.892046355067248e-05, "loss": 0.0593, "step": 33850 }, { "grad_norm": 0.7167180180549622, "learning_rate": 7.890697384372451e-05, "loss": 0.0458, "step": 33860 }, { "grad_norm": 0.8007446527481079, "learning_rate": 7.889348097555336e-05, "loss": 0.0595, "step": 33870 }, { "grad_norm": 0.5068817734718323, "learning_rate": 7.887998494763455e-05, "loss": 0.0398, "step": 33880 }, { "grad_norm": 1.0144085884094238, "learning_rate": 7.886648576144404e-05, "loss": 0.0635, "step": 33890 }, { "grad_norm": 0.45964816212654114, "learning_rate": 7.885298341845802e-05, "loss": 0.0549, "step": 33900 }, { "grad_norm": 0.5548744201660156, "learning_rate": 7.883947792015311e-05, "loss": 0.0493, "step": 33910 }, { "grad_norm": 0.8057962656021118, "learning_rate": 7.882596926800628e-05, "loss": 0.0631, "step": 33920 }, { "grad_norm": 0.6641196012496948, "learning_rate": 7.881245746349477e-05, "loss": 0.0524, "step": 33930 }, { "grad_norm": 0.6205506920814514, "learning_rate": 7.879894250809623e-05, "loss": 0.047, "step": 33940 }, { "grad_norm": 0.7419583797454834, "learning_rate": 7.878542440328865e-05, "loss": 0.0537, "step": 33950 }, { "grad_norm": 0.5601464509963989, "learning_rate": 7.877190315055031e-05, "loss": 0.0476, "step": 33960 }, { "grad_norm": 0.5365039110183716, "learning_rate": 7.875837875135991e-05, "loss": 0.0494, "step": 33970 }, { "grad_norm": 0.5910404920578003, "learning_rate": 7.874485120719646e-05, "loss": 0.0627, "step": 33980 }, { "grad_norm": 0.45472782850265503, "learning_rate": 7.873132051953928e-05, "loss": 0.0454, "step": 33990 }, { "grad_norm": 1.0370997190475464, "learning_rate": 7.87177866898681e-05, "loss": 0.0617, "step": 34000 }, { "grad_norm": 0.4656742811203003, "learning_rate": 7.870424971966294e-05, "loss": 0.0545, "step": 34010 }, { "grad_norm": 0.6286762952804565, "learning_rate": 7.869070961040419e-05, "loss": 0.077, "step": 34020 }, { "grad_norm": 0.5258335471153259, "learning_rate": 7.867716636357257e-05, "loss": 0.0613, "step": 34030 }, { "grad_norm": 0.9149165153503418, "learning_rate": 7.866361998064915e-05, "loss": 0.0679, "step": 34040 }, { "grad_norm": 0.5444478392601013, "learning_rate": 7.865007046311534e-05, "loss": 0.055, "step": 34050 }, { "grad_norm": 0.4773537814617157, "learning_rate": 7.86365178124529e-05, "loss": 0.0432, "step": 34060 }, { "grad_norm": 0.4042905569076538, "learning_rate": 7.862296203014394e-05, "loss": 0.0469, "step": 34070 }, { "grad_norm": 0.35495254397392273, "learning_rate": 7.860940311767088e-05, "loss": 0.0396, "step": 34080 }, { "grad_norm": 0.480380117893219, "learning_rate": 7.85958410765165e-05, "loss": 0.0439, "step": 34090 }, { "grad_norm": 0.6093246936798096, "learning_rate": 7.858227590816394e-05, "loss": 0.0487, "step": 34100 }, { "grad_norm": 0.7468807101249695, "learning_rate": 7.856870761409664e-05, "loss": 0.0465, "step": 34110 }, { "grad_norm": 1.090312123298645, "learning_rate": 7.855513619579846e-05, "loss": 0.0635, "step": 34120 }, { "grad_norm": 0.6336459517478943, "learning_rate": 7.85415616547535e-05, "loss": 0.055, "step": 34130 }, { "grad_norm": 0.44229963421821594, "learning_rate": 7.852798399244627e-05, "loss": 0.0445, "step": 34140 }, { "grad_norm": 0.624804675579071, "learning_rate": 7.851440321036161e-05, "loss": 0.0511, "step": 34150 }, { "grad_norm": 0.5697391629219055, "learning_rate": 7.850081930998468e-05, "loss": 0.037, "step": 34160 }, { "grad_norm": 0.37584853172302246, "learning_rate": 7.8487232292801e-05, "loss": 0.0359, "step": 34170 }, { "grad_norm": 0.5390157699584961, "learning_rate": 7.847364216029642e-05, "loss": 0.0504, "step": 34180 }, { "grad_norm": 0.608066976070404, "learning_rate": 7.846004891395716e-05, "loss": 0.0455, "step": 34190 }, { "grad_norm": 0.4231899678707123, "learning_rate": 7.844645255526972e-05, "loss": 0.0444, "step": 34200 }, { "grad_norm": 0.5939666628837585, "learning_rate": 7.843285308572101e-05, "loss": 0.0493, "step": 34210 }, { "grad_norm": 0.40857353806495667, "learning_rate": 7.841925050679823e-05, "loss": 0.0615, "step": 34220 }, { "grad_norm": 0.7651355266571045, "learning_rate": 7.840564481998895e-05, "loss": 0.0475, "step": 34230 }, { "grad_norm": 0.7567111849784851, "learning_rate": 7.839203602678105e-05, "loss": 0.0452, "step": 34240 }, { "grad_norm": 0.4378606379032135, "learning_rate": 7.837842412866279e-05, "loss": 0.0415, "step": 34250 }, { "grad_norm": 0.8315716981887817, "learning_rate": 7.836480912712272e-05, "loss": 0.0492, "step": 34260 }, { "grad_norm": 0.7830004692077637, "learning_rate": 7.835119102364976e-05, "loss": 0.0533, "step": 34270 }, { "grad_norm": 0.34764355421066284, "learning_rate": 7.83375698197332e-05, "loss": 0.0353, "step": 34280 }, { "grad_norm": 0.6251844763755798, "learning_rate": 7.83239455168626e-05, "loss": 0.0567, "step": 34290 }, { "grad_norm": 0.5704159736633301, "learning_rate": 7.83103181165279e-05, "loss": 0.0461, "step": 34300 }, { "grad_norm": 0.33582788705825806, "learning_rate": 7.829668762021937e-05, "loss": 0.0354, "step": 34310 }, { "grad_norm": 0.3874717056751251, "learning_rate": 7.828305402942764e-05, "loss": 0.0514, "step": 34320 }, { "grad_norm": 0.502220630645752, "learning_rate": 7.826941734564363e-05, "loss": 0.0655, "step": 34330 }, { "grad_norm": 0.6765967011451721, "learning_rate": 7.825577757035865e-05, "loss": 0.0415, "step": 34340 }, { "grad_norm": 0.5771852731704712, "learning_rate": 7.824213470506431e-05, "loss": 0.041, "step": 34350 }, { "grad_norm": 0.6333300471305847, "learning_rate": 7.822848875125257e-05, "loss": 0.0447, "step": 34360 }, { "grad_norm": 0.6661006212234497, "learning_rate": 7.821483971041576e-05, "loss": 0.0453, "step": 34370 }, { "grad_norm": 0.6004821062088013, "learning_rate": 7.820118758404649e-05, "loss": 0.0412, "step": 34380 }, { "grad_norm": 0.9684802293777466, "learning_rate": 7.818753237363776e-05, "loss": 0.0517, "step": 34390 }, { "grad_norm": 0.6044334769248962, "learning_rate": 7.817387408068286e-05, "loss": 0.0568, "step": 34400 }, { "grad_norm": 0.6918192505836487, "learning_rate": 7.816021270667544e-05, "loss": 0.0466, "step": 34410 }, { "grad_norm": 0.5781479477882385, "learning_rate": 7.81465482531095e-05, "loss": 0.0503, "step": 34420 }, { "grad_norm": 0.5077358484268188, "learning_rate": 7.813288072147938e-05, "loss": 0.0595, "step": 34430 }, { "grad_norm": 1.0185649394989014, "learning_rate": 7.811921011327972e-05, "loss": 0.0489, "step": 34440 }, { "grad_norm": 0.5756243467330933, "learning_rate": 7.810553643000549e-05, "loss": 0.0478, "step": 34450 }, { "grad_norm": 0.39619702100753784, "learning_rate": 7.809185967315206e-05, "loss": 0.041, "step": 34460 }, { "grad_norm": 0.7106523513793945, "learning_rate": 7.80781798442151e-05, "loss": 0.0463, "step": 34470 }, { "grad_norm": 0.5659753084182739, "learning_rate": 7.806449694469061e-05, "loss": 0.0583, "step": 34480 }, { "grad_norm": 0.4569985568523407, "learning_rate": 7.805081097607492e-05, "loss": 0.0769, "step": 34490 }, { "grad_norm": 0.5740454792976379, "learning_rate": 7.803712193986474e-05, "loss": 0.0488, "step": 34500 }, { "grad_norm": 1.0111689567565918, "learning_rate": 7.802342983755702e-05, "loss": 0.053, "step": 34510 }, { "grad_norm": 0.5288112163543701, "learning_rate": 7.800973467064918e-05, "loss": 0.0387, "step": 34520 }, { "grad_norm": 0.6837865114212036, "learning_rate": 7.799603644063884e-05, "loss": 0.0449, "step": 34530 }, { "grad_norm": 0.6201534271240234, "learning_rate": 7.798233514902405e-05, "loss": 0.0461, "step": 34540 }, { "grad_norm": 0.5882416367530823, "learning_rate": 7.796863079730318e-05, "loss": 0.0536, "step": 34550 }, { "grad_norm": 0.5591309666633606, "learning_rate": 7.795492338697488e-05, "loss": 0.0481, "step": 34560 }, { "grad_norm": 0.5452646017074585, "learning_rate": 7.794121291953819e-05, "loss": 0.0408, "step": 34570 }, { "grad_norm": 0.5575910806655884, "learning_rate": 7.792749939649246e-05, "loss": 0.0656, "step": 34580 }, { "grad_norm": 0.8886428475379944, "learning_rate": 7.79137828193374e-05, "loss": 0.0547, "step": 34590 }, { "grad_norm": 0.4956223964691162, "learning_rate": 7.790006318957301e-05, "loss": 0.0658, "step": 34600 }, { "grad_norm": 0.4919802248477936, "learning_rate": 7.788634050869965e-05, "loss": 0.0611, "step": 34610 }, { "grad_norm": 0.422446072101593, "learning_rate": 7.787261477821803e-05, "loss": 0.0396, "step": 34620 }, { "grad_norm": 0.6439512968063354, "learning_rate": 7.785888599962916e-05, "loss": 0.0436, "step": 34630 }, { "grad_norm": 0.5296176075935364, "learning_rate": 7.784515417443439e-05, "loss": 0.0521, "step": 34640 }, { "grad_norm": 0.7884268164634705, "learning_rate": 7.783141930413545e-05, "loss": 0.0547, "step": 34650 }, { "grad_norm": 0.6608907580375671, "learning_rate": 7.78176813902343e-05, "loss": 0.0523, "step": 34660 }, { "grad_norm": 0.6071703433990479, "learning_rate": 7.780394043423336e-05, "loss": 0.0584, "step": 34670 }, { "grad_norm": 0.6292412877082825, "learning_rate": 7.77901964376353e-05, "loss": 0.045, "step": 34680 }, { "grad_norm": 0.8762920498847961, "learning_rate": 7.777644940194316e-05, "loss": 0.0486, "step": 34690 }, { "grad_norm": 0.3674548268318176, "learning_rate": 7.776269932866023e-05, "loss": 0.0417, "step": 34700 }, { "grad_norm": 0.5413514375686646, "learning_rate": 7.774894621929026e-05, "loss": 0.0512, "step": 34710 }, { "grad_norm": 0.6962745785713196, "learning_rate": 7.773519007533725e-05, "loss": 0.0559, "step": 34720 }, { "grad_norm": 0.5606333017349243, "learning_rate": 7.772143089830556e-05, "loss": 0.0409, "step": 34730 }, { "grad_norm": 0.7296273708343506, "learning_rate": 7.770766868969985e-05, "loss": 0.0588, "step": 34740 }, { "grad_norm": 0.7602560520172119, "learning_rate": 7.769390345102518e-05, "loss": 0.0413, "step": 34750 }, { "grad_norm": 0.5219868421554565, "learning_rate": 7.768013518378683e-05, "loss": 0.0393, "step": 34760 }, { "grad_norm": 1.2366973161697388, "learning_rate": 7.766636388949053e-05, "loss": 0.0517, "step": 34770 }, { "grad_norm": 0.28287768363952637, "learning_rate": 7.765258956964229e-05, "loss": 0.0349, "step": 34780 }, { "grad_norm": 0.5155012011528015, "learning_rate": 7.76388122257484e-05, "loss": 0.0572, "step": 34790 }, { "grad_norm": 0.6407374143600464, "learning_rate": 7.762503185931558e-05, "loss": 0.063, "step": 34800 }, { "grad_norm": 0.6009171009063721, "learning_rate": 7.76112484718508e-05, "loss": 0.0383, "step": 34810 }, { "grad_norm": 0.39695411920547485, "learning_rate": 7.75974620648614e-05, "loss": 0.0444, "step": 34820 }, { "grad_norm": 0.5148819088935852, "learning_rate": 7.758367263985503e-05, "loss": 0.0506, "step": 34830 }, { "grad_norm": 0.5601354837417603, "learning_rate": 7.75698801983397e-05, "loss": 0.0575, "step": 34840 }, { "grad_norm": 0.7413053512573242, "learning_rate": 7.755608474182372e-05, "loss": 0.0513, "step": 34850 }, { "grad_norm": 0.42576199769973755, "learning_rate": 7.754228627181574e-05, "loss": 0.0429, "step": 34860 }, { "grad_norm": 0.4844992756843567, "learning_rate": 7.752848478982476e-05, "loss": 0.0412, "step": 34870 }, { "grad_norm": 0.5211899876594543, "learning_rate": 7.751468029736006e-05, "loss": 0.0675, "step": 34880 }, { "grad_norm": 0.5960580706596375, "learning_rate": 7.750087279593129e-05, "loss": 0.0527, "step": 34890 }, { "grad_norm": 0.6911504864692688, "learning_rate": 7.748706228704843e-05, "loss": 0.0542, "step": 34900 }, { "grad_norm": 0.4891565442085266, "learning_rate": 7.747324877222176e-05, "loss": 0.0398, "step": 34910 }, { "grad_norm": 0.6912614703178406, "learning_rate": 7.745943225296188e-05, "loss": 0.0593, "step": 34920 }, { "grad_norm": 0.48266538977622986, "learning_rate": 7.744561273077981e-05, "loss": 0.0506, "step": 34930 }, { "grad_norm": 0.44230011105537415, "learning_rate": 7.743179020718678e-05, "loss": 0.0386, "step": 34940 }, { "grad_norm": 0.6311309337615967, "learning_rate": 7.741796468369443e-05, "loss": 0.043, "step": 34950 }, { "grad_norm": 0.4645148515701294, "learning_rate": 7.740413616181466e-05, "loss": 0.0483, "step": 34960 }, { "grad_norm": 0.3544218838214874, "learning_rate": 7.739030464305978e-05, "loss": 0.0448, "step": 34970 }, { "grad_norm": 0.7499890923500061, "learning_rate": 7.737647012894235e-05, "loss": 0.0492, "step": 34980 }, { "grad_norm": 0.5179836750030518, "learning_rate": 7.736263262097532e-05, "loss": 0.0463, "step": 34990 }, { "grad_norm": 0.46468424797058105, "learning_rate": 7.734879212067192e-05, "loss": 0.0361, "step": 35000 }, { "grad_norm": 0.6404867768287659, "learning_rate": 7.733494862954573e-05, "loss": 0.0598, "step": 35010 }, { "grad_norm": 0.43852803111076355, "learning_rate": 7.732110214911066e-05, "loss": 0.0501, "step": 35020 }, { "grad_norm": 0.5122090578079224, "learning_rate": 7.730725268088094e-05, "loss": 0.0444, "step": 35030 }, { "grad_norm": 0.6165454387664795, "learning_rate": 7.729340022637112e-05, "loss": 0.0495, "step": 35040 }, { "grad_norm": 0.6634222865104675, "learning_rate": 7.727954478709607e-05, "loss": 0.066, "step": 35050 }, { "grad_norm": 0.4802394509315491, "learning_rate": 7.726568636457103e-05, "loss": 0.0523, "step": 35060 }, { "grad_norm": 0.5261974334716797, "learning_rate": 7.725182496031153e-05, "loss": 0.0659, "step": 35070 }, { "grad_norm": 0.48941704630851746, "learning_rate": 7.723796057583341e-05, "loss": 0.0515, "step": 35080 }, { "grad_norm": 0.37222886085510254, "learning_rate": 7.722409321265287e-05, "loss": 0.0444, "step": 35090 }, { "grad_norm": 0.5702667236328125, "learning_rate": 7.721022287228645e-05, "loss": 0.0494, "step": 35100 }, { "grad_norm": 0.5879408121109009, "learning_rate": 7.719634955625094e-05, "loss": 0.0609, "step": 35110 }, { "grad_norm": 0.575186014175415, "learning_rate": 7.718247326606355e-05, "loss": 0.0639, "step": 35120 }, { "grad_norm": 0.5879302024841309, "learning_rate": 7.716859400324175e-05, "loss": 0.0481, "step": 35130 }, { "grad_norm": 0.7145262360572815, "learning_rate": 7.715471176930335e-05, "loss": 0.0628, "step": 35140 }, { "grad_norm": 0.41855981945991516, "learning_rate": 7.714082656576651e-05, "loss": 0.0537, "step": 35150 }, { "grad_norm": 1.225601077079773, "learning_rate": 7.712693839414968e-05, "loss": 0.0568, "step": 35160 }, { "grad_norm": 0.6961718201637268, "learning_rate": 7.711304725597164e-05, "loss": 0.0634, "step": 35170 }, { "grad_norm": 0.4737505614757538, "learning_rate": 7.709915315275151e-05, "loss": 0.063, "step": 35180 }, { "grad_norm": 0.49565231800079346, "learning_rate": 7.708525608600876e-05, "loss": 0.0401, "step": 35190 }, { "grad_norm": 0.703769862651825, "learning_rate": 7.707135605726311e-05, "loss": 0.08, "step": 35200 }, { "grad_norm": 1.0064431428909302, "learning_rate": 7.705745306803466e-05, "loss": 0.0693, "step": 35210 }, { "grad_norm": 0.781176745891571, "learning_rate": 7.704354711984383e-05, "loss": 0.06, "step": 35220 }, { "grad_norm": 0.7422366738319397, "learning_rate": 7.702963821421133e-05, "loss": 0.0565, "step": 35230 }, { "grad_norm": 0.7026157975196838, "learning_rate": 7.701572635265826e-05, "loss": 0.0549, "step": 35240 }, { "grad_norm": 0.47025126218795776, "learning_rate": 7.700181153670596e-05, "loss": 0.0504, "step": 35250 }, { "grad_norm": 0.5732898116111755, "learning_rate": 7.698789376787616e-05, "loss": 0.0538, "step": 35260 }, { "grad_norm": 0.5324406027793884, "learning_rate": 7.697397304769085e-05, "loss": 0.0512, "step": 35270 }, { "grad_norm": 0.5575951933860779, "learning_rate": 7.696004937767241e-05, "loss": 0.0536, "step": 35280 }, { "grad_norm": 0.5705363154411316, "learning_rate": 7.694612275934352e-05, "loss": 0.0545, "step": 35290 }, { "grad_norm": 0.5553750395774841, "learning_rate": 7.693219319422714e-05, "loss": 0.0454, "step": 35300 }, { "grad_norm": 0.9133810997009277, "learning_rate": 7.69182606838466e-05, "loss": 0.0454, "step": 35310 }, { "grad_norm": 0.5888324975967407, "learning_rate": 7.690432522972558e-05, "loss": 0.0745, "step": 35320 }, { "grad_norm": 0.3911738991737366, "learning_rate": 7.689038683338796e-05, "loss": 0.0516, "step": 35330 }, { "grad_norm": 0.5264691114425659, "learning_rate": 7.687644549635808e-05, "loss": 0.056, "step": 35340 }, { "grad_norm": 0.4144309461116791, "learning_rate": 7.686250122016053e-05, "loss": 0.0613, "step": 35350 }, { "grad_norm": 0.8302633166313171, "learning_rate": 7.684855400632022e-05, "loss": 0.0433, "step": 35360 }, { "grad_norm": 0.4059644937515259, "learning_rate": 7.683460385636243e-05, "loss": 0.0506, "step": 35370 }, { "grad_norm": 1.1328870058059692, "learning_rate": 7.68206507718127e-05, "loss": 0.0474, "step": 35380 }, { "grad_norm": 0.3703603744506836, "learning_rate": 7.680669475419693e-05, "loss": 0.0569, "step": 35390 }, { "grad_norm": 0.6204279661178589, "learning_rate": 7.679273580504132e-05, "loss": 0.0428, "step": 35400 }, { "grad_norm": 0.7324540019035339, "learning_rate": 7.67787739258724e-05, "loss": 0.051, "step": 35410 }, { "grad_norm": 0.8445518016815186, "learning_rate": 7.676480911821705e-05, "loss": 0.0571, "step": 35420 }, { "grad_norm": 0.674835205078125, "learning_rate": 7.675084138360238e-05, "loss": 0.0505, "step": 35430 }, { "grad_norm": 0.6127839088439941, "learning_rate": 7.673687072355592e-05, "loss": 0.0511, "step": 35440 }, { "grad_norm": 0.5326679348945618, "learning_rate": 7.67228971396055e-05, "loss": 0.0485, "step": 35450 }, { "grad_norm": 0.5019058585166931, "learning_rate": 7.670892063327922e-05, "loss": 0.0431, "step": 35460 }, { "grad_norm": 0.5160215497016907, "learning_rate": 7.669494120610552e-05, "loss": 0.0436, "step": 35470 }, { "grad_norm": 0.40830913186073303, "learning_rate": 7.66809588596132e-05, "loss": 0.0648, "step": 35480 }, { "grad_norm": 0.5016390085220337, "learning_rate": 7.666697359533132e-05, "loss": 0.0567, "step": 35490 }, { "grad_norm": 0.6388654708862305, "learning_rate": 7.665298541478932e-05, "loss": 0.0831, "step": 35500 }, { "grad_norm": 0.7041410803794861, "learning_rate": 7.66389943195169e-05, "loss": 0.0553, "step": 35510 }, { "grad_norm": 0.7933346629142761, "learning_rate": 7.66250003110441e-05, "loss": 0.0724, "step": 35520 }, { "grad_norm": 0.42810994386672974, "learning_rate": 7.66110033909013e-05, "loss": 0.0374, "step": 35530 }, { "grad_norm": 0.4779262840747833, "learning_rate": 7.659700356061918e-05, "loss": 0.0412, "step": 35540 }, { "grad_norm": 0.6179508566856384, "learning_rate": 7.658300082172875e-05, "loss": 0.0493, "step": 35550 }, { "grad_norm": 0.39080435037612915, "learning_rate": 7.65689951757613e-05, "loss": 0.063, "step": 35560 }, { "grad_norm": 0.6063286662101746, "learning_rate": 7.655498662424849e-05, "loss": 0.0306, "step": 35570 }, { "grad_norm": 0.6799598932266235, "learning_rate": 7.654097516872227e-05, "loss": 0.0425, "step": 35580 }, { "grad_norm": 0.48175087571144104, "learning_rate": 7.65269608107149e-05, "loss": 0.0576, "step": 35590 }, { "grad_norm": 0.45580506324768066, "learning_rate": 7.6512943551759e-05, "loss": 0.039, "step": 35600 }, { "grad_norm": 0.5222275853157043, "learning_rate": 7.649892339338743e-05, "loss": 0.0441, "step": 35610 }, { "grad_norm": 0.479539155960083, "learning_rate": 7.648490033713344e-05, "loss": 0.0622, "step": 35620 }, { "grad_norm": 0.3822052478790283, "learning_rate": 7.647087438453058e-05, "loss": 0.0385, "step": 35630 }, { "grad_norm": 0.5646068453788757, "learning_rate": 7.64568455371127e-05, "loss": 0.0452, "step": 35640 }, { "grad_norm": 0.521122395992279, "learning_rate": 7.644281379641396e-05, "loss": 0.0521, "step": 35650 }, { "grad_norm": 0.47845378518104553, "learning_rate": 7.642877916396887e-05, "loss": 0.043, "step": 35660 }, { "grad_norm": 0.6043205261230469, "learning_rate": 7.641474164131221e-05, "loss": 0.0519, "step": 35670 }, { "grad_norm": 0.5010520219802856, "learning_rate": 7.640070122997913e-05, "loss": 0.0484, "step": 35680 }, { "grad_norm": 0.6381887197494507, "learning_rate": 7.638665793150505e-05, "loss": 0.0437, "step": 35690 }, { "grad_norm": 0.5462384223937988, "learning_rate": 7.637261174742574e-05, "loss": 0.0538, "step": 35700 }, { "grad_norm": 0.45894524455070496, "learning_rate": 7.635856267927725e-05, "loss": 0.0448, "step": 35710 }, { "grad_norm": 0.5655822157859802, "learning_rate": 7.634451072859597e-05, "loss": 0.0354, "step": 35720 }, { "grad_norm": 0.51292884349823, "learning_rate": 7.633045589691863e-05, "loss": 0.0375, "step": 35730 }, { "grad_norm": 0.50863116979599, "learning_rate": 7.63163981857822e-05, "loss": 0.0468, "step": 35740 }, { "grad_norm": 0.6354930400848389, "learning_rate": 7.630233759672403e-05, "loss": 0.0413, "step": 35750 }, { "grad_norm": 0.7146017551422119, "learning_rate": 7.628827413128177e-05, "loss": 0.0429, "step": 35760 }, { "grad_norm": 0.551535964012146, "learning_rate": 7.627420779099338e-05, "loss": 0.0364, "step": 35770 }, { "grad_norm": 0.6622471213340759, "learning_rate": 7.626013857739711e-05, "loss": 0.047, "step": 35780 }, { "grad_norm": 0.5340768098831177, "learning_rate": 7.624606649203158e-05, "loss": 0.0419, "step": 35790 }, { "grad_norm": 0.6062677502632141, "learning_rate": 7.623199153643569e-05, "loss": 0.0387, "step": 35800 }, { "grad_norm": 0.7181792855262756, "learning_rate": 7.621791371214863e-05, "loss": 0.044, "step": 35810 }, { "grad_norm": 0.7111291885375977, "learning_rate": 7.620383302070995e-05, "loss": 0.0549, "step": 35820 }, { "grad_norm": 0.6572160124778748, "learning_rate": 7.61897494636595e-05, "loss": 0.0512, "step": 35830 }, { "grad_norm": 0.9672219753265381, "learning_rate": 7.617566304253739e-05, "loss": 0.0431, "step": 35840 }, { "grad_norm": 0.6127946972846985, "learning_rate": 7.616157375888416e-05, "loss": 0.0514, "step": 35850 }, { "grad_norm": 0.5018458366394043, "learning_rate": 7.614748161424053e-05, "loss": 0.042, "step": 35860 }, { "grad_norm": 1.0880494117736816, "learning_rate": 7.613338661014763e-05, "loss": 0.0519, "step": 35870 }, { "grad_norm": 0.6338955163955688, "learning_rate": 7.611928874814686e-05, "loss": 0.0417, "step": 35880 }, { "grad_norm": 0.5407823920249939, "learning_rate": 7.610518802977996e-05, "loss": 0.0791, "step": 35890 }, { "grad_norm": 0.5098739862442017, "learning_rate": 7.609108445658893e-05, "loss": 0.0351, "step": 35900 }, { "grad_norm": 0.4213334918022156, "learning_rate": 7.607697803011612e-05, "loss": 0.0509, "step": 35910 }, { "grad_norm": 0.5024478435516357, "learning_rate": 7.606286875190421e-05, "loss": 0.0481, "step": 35920 }, { "grad_norm": 0.6090206503868103, "learning_rate": 7.604875662349617e-05, "loss": 0.037, "step": 35930 }, { "grad_norm": 0.4719896912574768, "learning_rate": 7.603464164643525e-05, "loss": 0.0319, "step": 35940 }, { "grad_norm": 0.5113743543624878, "learning_rate": 7.602052382226507e-05, "loss": 0.0648, "step": 35950 }, { "grad_norm": 0.6326121687889099, "learning_rate": 7.600640315252954e-05, "loss": 0.0647, "step": 35960 }, { "grad_norm": 1.556321144104004, "learning_rate": 7.599227963877284e-05, "loss": 0.0486, "step": 35970 }, { "grad_norm": 0.5564117431640625, "learning_rate": 7.597815328253954e-05, "loss": 0.0412, "step": 35980 }, { "grad_norm": 0.4130033552646637, "learning_rate": 7.596402408537443e-05, "loss": 0.0632, "step": 35990 }, { "grad_norm": 1.6953223943710327, "learning_rate": 7.59498920488227e-05, "loss": 0.0475, "step": 36000 }, { "grad_norm": 0.47293588519096375, "learning_rate": 7.593575717442979e-05, "loss": 0.0414, "step": 36010 }, { "grad_norm": 0.5155156254768372, "learning_rate": 7.592161946374147e-05, "loss": 0.0688, "step": 36020 }, { "grad_norm": 0.5197055339813232, "learning_rate": 7.590747891830381e-05, "loss": 0.0615, "step": 36030 }, { "grad_norm": 0.5517199039459229, "learning_rate": 7.58933355396632e-05, "loss": 0.0434, "step": 36040 }, { "grad_norm": 0.3950560688972473, "learning_rate": 7.587918932936636e-05, "loss": 0.066, "step": 36050 }, { "grad_norm": 0.8134890794754028, "learning_rate": 7.586504028896028e-05, "loss": 0.0535, "step": 36060 }, { "grad_norm": 0.3599063754081726, "learning_rate": 7.585088841999228e-05, "loss": 0.0508, "step": 36070 }, { "grad_norm": 0.6393517851829529, "learning_rate": 7.583673372400999e-05, "loss": 0.0469, "step": 36080 }, { "grad_norm": 0.7387639880180359, "learning_rate": 7.582257620256134e-05, "loss": 0.0435, "step": 36090 }, { "grad_norm": 0.475919634103775, "learning_rate": 7.580841585719458e-05, "loss": 0.0507, "step": 36100 }, { "grad_norm": 0.5292834043502808, "learning_rate": 7.579425268945825e-05, "loss": 0.0489, "step": 36110 }, { "grad_norm": 0.9751700162887573, "learning_rate": 7.578008670090127e-05, "loss": 0.0438, "step": 36120 }, { "grad_norm": 0.49075189232826233, "learning_rate": 7.576591789307272e-05, "loss": 0.0579, "step": 36130 }, { "grad_norm": 0.43642860651016235, "learning_rate": 7.575174626752216e-05, "loss": 0.048, "step": 36140 }, { "grad_norm": 0.6375479698181152, "learning_rate": 7.573757182579934e-05, "loss": 0.0468, "step": 36150 }, { "grad_norm": 0.5416220426559448, "learning_rate": 7.572339456945435e-05, "loss": 0.0381, "step": 36160 }, { "grad_norm": 0.4067683517932892, "learning_rate": 7.570921450003762e-05, "loss": 0.0475, "step": 36170 }, { "grad_norm": 0.39891892671585083, "learning_rate": 7.569503161909983e-05, "loss": 0.0483, "step": 36180 }, { "grad_norm": 0.5454921126365662, "learning_rate": 7.568084592819202e-05, "loss": 0.0542, "step": 36190 }, { "grad_norm": 0.7138345241546631, "learning_rate": 7.566665742886551e-05, "loss": 0.0648, "step": 36200 }, { "grad_norm": 0.4118436574935913, "learning_rate": 7.565246612267194e-05, "loss": 0.0414, "step": 36210 }, { "grad_norm": 0.7455448508262634, "learning_rate": 7.563827201116325e-05, "loss": 0.0509, "step": 36220 }, { "grad_norm": 0.5473430752754211, "learning_rate": 7.562407509589164e-05, "loss": 0.0605, "step": 36230 }, { "grad_norm": 0.5024620890617371, "learning_rate": 7.560987537840973e-05, "loss": 0.0568, "step": 36240 }, { "grad_norm": 0.40146875381469727, "learning_rate": 7.559567286027036e-05, "loss": 0.0446, "step": 36250 }, { "grad_norm": 0.6420806050300598, "learning_rate": 7.558146754302668e-05, "loss": 0.0534, "step": 36260 }, { "grad_norm": 0.6151865720748901, "learning_rate": 7.556725942823217e-05, "loss": 0.0462, "step": 36270 }, { "grad_norm": 0.3202366530895233, "learning_rate": 7.555304851744061e-05, "loss": 0.0448, "step": 36280 }, { "grad_norm": 0.45083555579185486, "learning_rate": 7.553883481220608e-05, "loss": 0.0439, "step": 36290 }, { "grad_norm": 0.4931736886501312, "learning_rate": 7.552461831408298e-05, "loss": 0.0456, "step": 36300 }, { "grad_norm": 0.5030111074447632, "learning_rate": 7.551039902462599e-05, "loss": 0.0473, "step": 36310 }, { "grad_norm": 0.42490190267562866, "learning_rate": 7.549617694539011e-05, "loss": 0.0534, "step": 36320 }, { "grad_norm": 0.42347729206085205, "learning_rate": 7.548195207793066e-05, "loss": 0.0629, "step": 36330 }, { "grad_norm": 0.5306676030158997, "learning_rate": 7.546772442380323e-05, "loss": 0.044, "step": 36340 }, { "grad_norm": 0.5452432036399841, "learning_rate": 7.545349398456376e-05, "loss": 0.052, "step": 36350 }, { "grad_norm": 0.46452638506889343, "learning_rate": 7.543926076176845e-05, "loss": 0.0387, "step": 36360 }, { "grad_norm": 0.803796648979187, "learning_rate": 7.542502475697385e-05, "loss": 0.0459, "step": 36370 }, { "grad_norm": 0.4620610475540161, "learning_rate": 7.541078597173675e-05, "loss": 0.0451, "step": 36380 }, { "grad_norm": 0.620821475982666, "learning_rate": 7.539654440761431e-05, "loss": 0.0446, "step": 36390 }, { "grad_norm": 0.52768474817276, "learning_rate": 7.538230006616395e-05, "loss": 0.0352, "step": 36400 }, { "grad_norm": 0.5875639319419861, "learning_rate": 7.536805294894342e-05, "loss": 0.0627, "step": 36410 }, { "grad_norm": 0.4059617519378662, "learning_rate": 7.535380305751076e-05, "loss": 0.0501, "step": 36420 }, { "grad_norm": 0.6797771453857422, "learning_rate": 7.533955039342431e-05, "loss": 0.046, "step": 36430 }, { "grad_norm": 0.5838540196418762, "learning_rate": 7.532529495824274e-05, "loss": 0.0468, "step": 36440 }, { "grad_norm": 0.49692457914352417, "learning_rate": 7.531103675352497e-05, "loss": 0.0458, "step": 36450 }, { "grad_norm": 1.0066620111465454, "learning_rate": 7.529677578083029e-05, "loss": 0.0543, "step": 36460 }, { "grad_norm": 0.8569527864456177, "learning_rate": 7.528251204171823e-05, "loss": 0.0567, "step": 36470 }, { "grad_norm": 0.69082111120224, "learning_rate": 7.526824553774867e-05, "loss": 0.0485, "step": 36480 }, { "grad_norm": 0.5199905037879944, "learning_rate": 7.525397627048178e-05, "loss": 0.0615, "step": 36490 }, { "grad_norm": 0.6534593105316162, "learning_rate": 7.523970424147802e-05, "loss": 0.0546, "step": 36500 }, { "grad_norm": 0.5304285883903503, "learning_rate": 7.522542945229813e-05, "loss": 0.0381, "step": 36510 }, { "grad_norm": 1.6382766962051392, "learning_rate": 7.521115190450321e-05, "loss": 0.0471, "step": 36520 }, { "grad_norm": 0.6636239886283875, "learning_rate": 7.519687159965462e-05, "loss": 0.0473, "step": 36530 }, { "grad_norm": 0.9101640582084656, "learning_rate": 7.518258853931403e-05, "loss": 0.0519, "step": 36540 }, { "grad_norm": 0.4850505590438843, "learning_rate": 7.516830272504342e-05, "loss": 0.0367, "step": 36550 }, { "grad_norm": 0.6102988719940186, "learning_rate": 7.515401415840509e-05, "loss": 0.0467, "step": 36560 }, { "grad_norm": 0.5656558275222778, "learning_rate": 7.513972284096155e-05, "loss": 0.0449, "step": 36570 }, { "grad_norm": 0.5535973310470581, "learning_rate": 7.512542877427574e-05, "loss": 0.0378, "step": 36580 }, { "grad_norm": 0.627199113368988, "learning_rate": 7.51111319599108e-05, "loss": 0.0402, "step": 36590 }, { "grad_norm": 0.6168316006660461, "learning_rate": 7.509683239943024e-05, "loss": 0.0439, "step": 36600 }, { "grad_norm": 0.6206540465354919, "learning_rate": 7.50825300943978e-05, "loss": 0.0449, "step": 36610 }, { "grad_norm": 0.8357094526290894, "learning_rate": 7.50682250463776e-05, "loss": 0.0552, "step": 36620 }, { "grad_norm": 0.689815104007721, "learning_rate": 7.5053917256934e-05, "loss": 0.0518, "step": 36630 }, { "grad_norm": 0.6602489352226257, "learning_rate": 7.503960672763166e-05, "loss": 0.0559, "step": 36640 }, { "grad_norm": 0.4826158285140991, "learning_rate": 7.502529346003559e-05, "loss": 0.053, "step": 36650 }, { "grad_norm": 0.3473629653453827, "learning_rate": 7.501097745571106e-05, "loss": 0.0437, "step": 36660 }, { "grad_norm": 0.49301469326019287, "learning_rate": 7.499665871622361e-05, "loss": 0.0548, "step": 36670 }, { "grad_norm": 0.4290088415145874, "learning_rate": 7.498233724313917e-05, "loss": 0.044, "step": 36680 }, { "grad_norm": 0.5438337922096252, "learning_rate": 7.496801303802387e-05, "loss": 0.0429, "step": 36690 }, { "grad_norm": 0.34547343850135803, "learning_rate": 7.495368610244423e-05, "loss": 0.0413, "step": 36700 }, { "grad_norm": 0.7831894159317017, "learning_rate": 7.493935643796697e-05, "loss": 0.0464, "step": 36710 }, { "grad_norm": 0.472423791885376, "learning_rate": 7.49250240461592e-05, "loss": 0.042, "step": 36720 }, { "grad_norm": 0.4425719976425171, "learning_rate": 7.491068892858826e-05, "loss": 0.0354, "step": 36730 }, { "grad_norm": 0.467157781124115, "learning_rate": 7.489635108682184e-05, "loss": 0.0458, "step": 36740 }, { "grad_norm": 0.614872395992279, "learning_rate": 7.48820105224279e-05, "loss": 0.0586, "step": 36750 }, { "grad_norm": 0.5851783752441406, "learning_rate": 7.48676672369747e-05, "loss": 0.0459, "step": 36760 }, { "grad_norm": 0.4042919874191284, "learning_rate": 7.485332123203079e-05, "loss": 0.0423, "step": 36770 }, { "grad_norm": 0.35525643825531006, "learning_rate": 7.483897250916504e-05, "loss": 0.0371, "step": 36780 }, { "grad_norm": 0.8109886646270752, "learning_rate": 7.48246210699466e-05, "loss": 0.0588, "step": 36790 }, { "grad_norm": 0.3847719430923462, "learning_rate": 7.481026691594492e-05, "loss": 0.0622, "step": 36800 }, { "grad_norm": 0.6328919529914856, "learning_rate": 7.479591004872975e-05, "loss": 0.0384, "step": 36810 }, { "grad_norm": 0.5117196440696716, "learning_rate": 7.478155046987114e-05, "loss": 0.0586, "step": 36820 }, { "grad_norm": 0.5974984765052795, "learning_rate": 7.476718818093944e-05, "loss": 0.0493, "step": 36830 }, { "grad_norm": 0.6324846148490906, "learning_rate": 7.475282318350527e-05, "loss": 0.0516, "step": 36840 }, { "grad_norm": 0.7608712315559387, "learning_rate": 7.473845547913957e-05, "loss": 0.0563, "step": 36850 }, { "grad_norm": 0.48833274841308594, "learning_rate": 7.472408506941357e-05, "loss": 0.0409, "step": 36860 }, { "grad_norm": 0.5304222106933594, "learning_rate": 7.470971195589881e-05, "loss": 0.0363, "step": 36870 }, { "grad_norm": 0.7278719544410706, "learning_rate": 7.46953361401671e-05, "loss": 0.0457, "step": 36880 }, { "grad_norm": 0.42947614192962646, "learning_rate": 7.468095762379055e-05, "loss": 0.0333, "step": 36890 }, { "grad_norm": 0.4891910254955292, "learning_rate": 7.466657640834158e-05, "loss": 0.0544, "step": 36900 }, { "grad_norm": 0.6179765462875366, "learning_rate": 7.46521924953929e-05, "loss": 0.0513, "step": 36910 }, { "grad_norm": 0.5039427280426025, "learning_rate": 7.463780588651752e-05, "loss": 0.0545, "step": 36920 }, { "grad_norm": 0.46284258365631104, "learning_rate": 7.462341658328873e-05, "loss": 0.0455, "step": 36930 }, { "grad_norm": 0.5802695751190186, "learning_rate": 7.460902458728012e-05, "loss": 0.0596, "step": 36940 }, { "grad_norm": 1.1059982776641846, "learning_rate": 7.459462990006558e-05, "loss": 0.0577, "step": 36950 }, { "grad_norm": 0.7621592283248901, "learning_rate": 7.458023252321929e-05, "loss": 0.0409, "step": 36960 }, { "grad_norm": 0.48981714248657227, "learning_rate": 7.456583245831573e-05, "loss": 0.0395, "step": 36970 }, { "grad_norm": 0.6086740493774414, "learning_rate": 7.455142970692968e-05, "loss": 0.0403, "step": 36980 }, { "grad_norm": 0.40154126286506653, "learning_rate": 7.453702427063615e-05, "loss": 0.0593, "step": 36990 }, { "grad_norm": 0.5344070792198181, "learning_rate": 7.452261615101057e-05, "loss": 0.0454, "step": 37000 }, { "grad_norm": 0.625694990158081, "learning_rate": 7.450820534962856e-05, "loss": 0.0629, "step": 37010 }, { "grad_norm": 0.38498809933662415, "learning_rate": 7.449379186806604e-05, "loss": 0.0453, "step": 37020 }, { "grad_norm": 0.5061218738555908, "learning_rate": 7.44793757078993e-05, "loss": 0.0422, "step": 37030 }, { "grad_norm": 0.7151042222976685, "learning_rate": 7.44649568707048e-05, "loss": 0.0515, "step": 37040 }, { "grad_norm": 0.4244166314601898, "learning_rate": 7.445053535805942e-05, "loss": 0.0414, "step": 37050 }, { "grad_norm": 0.49346843361854553, "learning_rate": 7.443611117154024e-05, "loss": 0.0407, "step": 37060 }, { "grad_norm": 0.529911994934082, "learning_rate": 7.442168431272469e-05, "loss": 0.0354, "step": 37070 }, { "grad_norm": 0.6629113554954529, "learning_rate": 7.440725478319044e-05, "loss": 0.0372, "step": 37080 }, { "grad_norm": 0.6430678367614746, "learning_rate": 7.439282258451553e-05, "loss": 0.0458, "step": 37090 }, { "grad_norm": 0.5048726201057434, "learning_rate": 7.43783877182782e-05, "loss": 0.0447, "step": 37100 }, { "grad_norm": 0.8943959474563599, "learning_rate": 7.436395018605707e-05, "loss": 0.053, "step": 37110 }, { "grad_norm": 1.1017314195632935, "learning_rate": 7.434950998943094e-05, "loss": 0.0685, "step": 37120 }, { "grad_norm": 0.44705191254615784, "learning_rate": 7.433506712997902e-05, "loss": 0.0485, "step": 37130 }, { "grad_norm": 0.8212897777557373, "learning_rate": 7.432062160928077e-05, "loss": 0.0712, "step": 37140 }, { "grad_norm": 0.5459153056144714, "learning_rate": 7.430617342891588e-05, "loss": 0.0373, "step": 37150 }, { "grad_norm": 0.512543261051178, "learning_rate": 7.429172259046442e-05, "loss": 0.0479, "step": 37160 }, { "grad_norm": 0.41524365544319153, "learning_rate": 7.42772690955067e-05, "loss": 0.0504, "step": 37170 }, { "grad_norm": 0.5417271852493286, "learning_rate": 7.426281294562332e-05, "loss": 0.0518, "step": 37180 }, { "grad_norm": 0.8461228013038635, "learning_rate": 7.424835414239519e-05, "loss": 0.0607, "step": 37190 }, { "grad_norm": 0.4853563904762268, "learning_rate": 7.423389268740352e-05, "loss": 0.0501, "step": 37200 }, { "grad_norm": 0.6960793733596802, "learning_rate": 7.42194285822298e-05, "loss": 0.0504, "step": 37210 }, { "grad_norm": 0.6740293502807617, "learning_rate": 7.420496182845576e-05, "loss": 0.0382, "step": 37220 }, { "grad_norm": 1.0118186473846436, "learning_rate": 7.419049242766348e-05, "loss": 0.0506, "step": 37230 }, { "grad_norm": 0.3907841444015503, "learning_rate": 7.417602038143533e-05, "loss": 0.0739, "step": 37240 }, { "grad_norm": 0.724892795085907, "learning_rate": 7.416154569135393e-05, "loss": 0.0719, "step": 37250 }, { "grad_norm": 0.46467792987823486, "learning_rate": 7.414706835900222e-05, "loss": 0.0698, "step": 37260 }, { "grad_norm": 0.5075388550758362, "learning_rate": 7.413258838596345e-05, "loss": 0.0405, "step": 37270 }, { "grad_norm": 0.6353322267532349, "learning_rate": 7.411810577382106e-05, "loss": 0.0481, "step": 37280 }, { "grad_norm": 0.7149138450622559, "learning_rate": 7.410362052415889e-05, "loss": 0.0601, "step": 37290 }, { "grad_norm": 1.6865288019180298, "learning_rate": 7.408913263856102e-05, "loss": 0.047, "step": 37300 }, { "grad_norm": 0.5110983848571777, "learning_rate": 7.407464211861183e-05, "loss": 0.0504, "step": 37310 }, { "grad_norm": 0.7024976015090942, "learning_rate": 7.406014896589597e-05, "loss": 0.0575, "step": 37320 }, { "grad_norm": 0.3713613748550415, "learning_rate": 7.404565318199838e-05, "loss": 0.0451, "step": 37330 }, { "grad_norm": 0.5647347569465637, "learning_rate": 7.403115476850432e-05, "loss": 0.0506, "step": 37340 }, { "grad_norm": 0.46767571568489075, "learning_rate": 7.401665372699932e-05, "loss": 0.0388, "step": 37350 }, { "grad_norm": 0.3895799517631531, "learning_rate": 7.400215005906916e-05, "loss": 0.0412, "step": 37360 }, { "grad_norm": 0.46190986037254333, "learning_rate": 7.398764376629996e-05, "loss": 0.0461, "step": 37370 }, { "grad_norm": 0.6013032793998718, "learning_rate": 7.39731348502781e-05, "loss": 0.0389, "step": 37380 }, { "grad_norm": 0.4405149519443512, "learning_rate": 7.395862331259028e-05, "loss": 0.0306, "step": 37390 }, { "grad_norm": 0.5240306854248047, "learning_rate": 7.394410915482341e-05, "loss": 0.0416, "step": 37400 }, { "grad_norm": 0.8152920603752136, "learning_rate": 7.392959237856476e-05, "loss": 0.0522, "step": 37410 }, { "grad_norm": 0.5962770581245422, "learning_rate": 7.391507298540187e-05, "loss": 0.0404, "step": 37420 }, { "grad_norm": 0.46787527203559875, "learning_rate": 7.390055097692257e-05, "loss": 0.0434, "step": 37430 }, { "grad_norm": 0.35217198729515076, "learning_rate": 7.388602635471494e-05, "loss": 0.0375, "step": 37440 }, { "grad_norm": 0.6792711615562439, "learning_rate": 7.387149912036738e-05, "loss": 0.0956, "step": 37450 }, { "grad_norm": 0.34384089708328247, "learning_rate": 7.385696927546854e-05, "loss": 0.0613, "step": 37460 }, { "grad_norm": 0.4987333118915558, "learning_rate": 7.384243682160743e-05, "loss": 0.044, "step": 37470 }, { "grad_norm": 0.5110102295875549, "learning_rate": 7.382790176037329e-05, "loss": 0.0375, "step": 37480 }, { "grad_norm": 0.5569661259651184, "learning_rate": 7.381336409335562e-05, "loss": 0.037, "step": 37490 }, { "grad_norm": 0.38005441427230835, "learning_rate": 7.379882382214426e-05, "loss": 0.038, "step": 37500 }, { "grad_norm": 0.41860777139663696, "learning_rate": 7.378428094832931e-05, "loss": 0.0413, "step": 37510 }, { "grad_norm": 0.7397767305374146, "learning_rate": 7.376973547350114e-05, "loss": 0.0395, "step": 37520 }, { "grad_norm": 0.4988030195236206, "learning_rate": 7.375518739925046e-05, "loss": 0.066, "step": 37530 }, { "grad_norm": 0.3266013264656067, "learning_rate": 7.374063672716818e-05, "loss": 0.064, "step": 37540 }, { "grad_norm": 0.5103148818016052, "learning_rate": 7.372608345884558e-05, "loss": 0.042, "step": 37550 }, { "grad_norm": 0.5699405670166016, "learning_rate": 7.371152759587415e-05, "loss": 0.0722, "step": 37560 }, { "grad_norm": 0.5408596992492676, "learning_rate": 7.369696913984572e-05, "loss": 0.0479, "step": 37570 }, { "grad_norm": 0.6333545446395874, "learning_rate": 7.368240809235237e-05, "loss": 0.0429, "step": 37580 }, { "grad_norm": 0.6285173892974854, "learning_rate": 7.366784445498649e-05, "loss": 0.0438, "step": 37590 }, { "grad_norm": 0.7847068309783936, "learning_rate": 7.36532782293407e-05, "loss": 0.0518, "step": 37600 }, { "grad_norm": 0.5620402097702026, "learning_rate": 7.363870941700797e-05, "loss": 0.0677, "step": 37610 }, { "grad_norm": 0.6078295707702637, "learning_rate": 7.362413801958152e-05, "loss": 0.0474, "step": 37620 }, { "grad_norm": 0.6736663579940796, "learning_rate": 7.360956403865486e-05, "loss": 0.0511, "step": 37630 }, { "grad_norm": 0.5438583493232727, "learning_rate": 7.359498747582177e-05, "loss": 0.0445, "step": 37640 }, { "grad_norm": 0.6411523222923279, "learning_rate": 7.358040833267634e-05, "loss": 0.0595, "step": 37650 }, { "grad_norm": 0.5093022584915161, "learning_rate": 7.356582661081288e-05, "loss": 0.0333, "step": 37660 }, { "grad_norm": 0.4694210886955261, "learning_rate": 7.355124231182607e-05, "loss": 0.0506, "step": 37670 }, { "grad_norm": 0.719715416431427, "learning_rate": 7.353665543731083e-05, "loss": 0.0416, "step": 37680 }, { "grad_norm": 0.4503311514854431, "learning_rate": 7.352206598886229e-05, "loss": 0.0508, "step": 37690 }, { "grad_norm": 0.6523804664611816, "learning_rate": 7.350747396807601e-05, "loss": 0.0441, "step": 37700 }, { "grad_norm": 0.45181235671043396, "learning_rate": 7.349287937654772e-05, "loss": 0.0442, "step": 37710 }, { "grad_norm": 0.7632930278778076, "learning_rate": 7.347828221587345e-05, "loss": 0.0502, "step": 37720 }, { "grad_norm": 0.7620870471000671, "learning_rate": 7.346368248764955e-05, "loss": 0.0504, "step": 37730 }, { "grad_norm": 0.7244992852210999, "learning_rate": 7.34490801934726e-05, "loss": 0.0476, "step": 37740 }, { "grad_norm": 0.5131390690803528, "learning_rate": 7.343447533493947e-05, "loss": 0.0431, "step": 37750 }, { "grad_norm": 0.6566930413246155, "learning_rate": 7.341986791364739e-05, "loss": 0.0409, "step": 37760 }, { "grad_norm": 0.6629148125648499, "learning_rate": 7.340525793119374e-05, "loss": 0.0498, "step": 37770 }, { "grad_norm": 0.6283897161483765, "learning_rate": 7.33906453891763e-05, "loss": 0.0631, "step": 37780 }, { "grad_norm": 0.40072157979011536, "learning_rate": 7.337603028919303e-05, "loss": 0.0511, "step": 37790 }, { "grad_norm": 0.524690568447113, "learning_rate": 7.336141263284225e-05, "loss": 0.0657, "step": 37800 }, { "grad_norm": 0.49759888648986816, "learning_rate": 7.334679242172249e-05, "loss": 0.0421, "step": 37810 }, { "grad_norm": 0.5697259902954102, "learning_rate": 7.33321696574326e-05, "loss": 0.0482, "step": 37820 }, { "grad_norm": 0.48133179545402527, "learning_rate": 7.331754434157174e-05, "loss": 0.0455, "step": 37830 }, { "grad_norm": 0.6686473488807678, "learning_rate": 7.330291647573928e-05, "loss": 0.0464, "step": 37840 }, { "grad_norm": 0.33420923352241516, "learning_rate": 7.32882860615349e-05, "loss": 0.0433, "step": 37850 }, { "grad_norm": 0.6900251507759094, "learning_rate": 7.327365310055858e-05, "loss": 0.0551, "step": 37860 }, { "grad_norm": 0.48749691247940063, "learning_rate": 7.325901759441055e-05, "loss": 0.0422, "step": 37870 }, { "grad_norm": 0.558250367641449, "learning_rate": 7.324437954469132e-05, "loss": 0.0379, "step": 37880 }, { "grad_norm": 0.492662250995636, "learning_rate": 7.32297389530017e-05, "loss": 0.0397, "step": 37890 }, { "grad_norm": 0.5256264209747314, "learning_rate": 7.321509582094274e-05, "loss": 0.0448, "step": 37900 }, { "grad_norm": 0.708267331123352, "learning_rate": 7.320045015011582e-05, "loss": 0.0458, "step": 37910 }, { "grad_norm": 0.5214334726333618, "learning_rate": 7.318580194212253e-05, "loss": 0.0601, "step": 37920 }, { "grad_norm": 0.5179596543312073, "learning_rate": 7.317115119856483e-05, "loss": 0.0423, "step": 37930 }, { "grad_norm": 0.6006896495819092, "learning_rate": 7.315649792104485e-05, "loss": 0.0541, "step": 37940 }, { "grad_norm": 0.8844835162162781, "learning_rate": 7.314184211116507e-05, "loss": 0.0493, "step": 37950 }, { "grad_norm": 1.2176556587219238, "learning_rate": 7.312718377052823e-05, "loss": 0.0469, "step": 37960 }, { "grad_norm": 0.5891726613044739, "learning_rate": 7.311252290073737e-05, "loss": 0.0401, "step": 37970 }, { "grad_norm": 0.4880416691303253, "learning_rate": 7.309785950339572e-05, "loss": 0.0462, "step": 37980 }, { "grad_norm": 0.6328912377357483, "learning_rate": 7.308319358010692e-05, "loss": 0.0378, "step": 37990 }, { "grad_norm": 0.4437880516052246, "learning_rate": 7.306852513247474e-05, "loss": 0.0544, "step": 38000 }, { "grad_norm": 0.7114630937576294, "learning_rate": 7.305385416210336e-05, "loss": 0.0407, "step": 38010 }, { "grad_norm": 0.6831417083740234, "learning_rate": 7.303918067059711e-05, "loss": 0.0492, "step": 38020 }, { "grad_norm": 0.5144539475440979, "learning_rate": 7.302450465956075e-05, "loss": 0.05, "step": 38030 }, { "grad_norm": 1.09745454788208, "learning_rate": 7.300982613059914e-05, "loss": 0.0421, "step": 38040 }, { "grad_norm": 0.4664342403411865, "learning_rate": 7.299514508531757e-05, "loss": 0.0321, "step": 38050 }, { "grad_norm": 0.7853325605392456, "learning_rate": 7.298046152532148e-05, "loss": 0.045, "step": 38060 }, { "grad_norm": 0.2920570969581604, "learning_rate": 7.296577545221668e-05, "loss": 0.0643, "step": 38070 }, { "grad_norm": 0.3836992383003235, "learning_rate": 7.29510868676092e-05, "loss": 0.034, "step": 38080 }, { "grad_norm": 0.5099164843559265, "learning_rate": 7.293639577310538e-05, "loss": 0.0573, "step": 38090 }, { "grad_norm": 0.8498555421829224, "learning_rate": 7.292170217031179e-05, "loss": 0.0437, "step": 38100 }, { "grad_norm": 0.3793815076351166, "learning_rate": 7.290700606083532e-05, "loss": 0.0427, "step": 38110 }, { "grad_norm": 0.567352294921875, "learning_rate": 7.28923074462831e-05, "loss": 0.0529, "step": 38120 }, { "grad_norm": 0.6244038939476013, "learning_rate": 7.287760632826255e-05, "loss": 0.0571, "step": 38130 }, { "grad_norm": 0.6519172191619873, "learning_rate": 7.286290270838139e-05, "loss": 0.0601, "step": 38140 }, { "grad_norm": 0.45087671279907227, "learning_rate": 7.284819658824756e-05, "loss": 0.0503, "step": 38150 }, { "grad_norm": 0.4663292467594147, "learning_rate": 7.283348796946929e-05, "loss": 0.0567, "step": 38160 }, { "grad_norm": 0.45555150508880615, "learning_rate": 7.28187768536551e-05, "loss": 0.0624, "step": 38170 }, { "grad_norm": 0.5493409633636475, "learning_rate": 7.280406324241379e-05, "loss": 0.0479, "step": 38180 }, { "grad_norm": 0.6806027889251709, "learning_rate": 7.278934713735441e-05, "loss": 0.0756, "step": 38190 }, { "grad_norm": 0.4646702706813812, "learning_rate": 7.277462854008629e-05, "loss": 0.0469, "step": 38200 }, { "grad_norm": 0.559615433216095, "learning_rate": 7.275990745221904e-05, "loss": 0.044, "step": 38210 }, { "grad_norm": 0.4107145369052887, "learning_rate": 7.274518387536252e-05, "loss": 0.0489, "step": 38220 }, { "grad_norm": 0.5155571699142456, "learning_rate": 7.273045781112689e-05, "loss": 0.0397, "step": 38230 }, { "grad_norm": 0.41845768690109253, "learning_rate": 7.271572926112257e-05, "loss": 0.0466, "step": 38240 }, { "grad_norm": 0.29390189051628113, "learning_rate": 7.270099822696024e-05, "loss": 0.0475, "step": 38250 }, { "grad_norm": 0.42783674597740173, "learning_rate": 7.26862647102509e-05, "loss": 0.0744, "step": 38260 }, { "grad_norm": 0.45554453134536743, "learning_rate": 7.267152871260573e-05, "loss": 0.0325, "step": 38270 }, { "grad_norm": 0.5098001956939697, "learning_rate": 7.265679023563629e-05, "loss": 0.046, "step": 38280 }, { "grad_norm": 0.5593408346176147, "learning_rate": 7.264204928095433e-05, "loss": 0.0471, "step": 38290 }, { "grad_norm": 0.47251588106155396, "learning_rate": 7.262730585017188e-05, "loss": 0.0336, "step": 38300 }, { "grad_norm": 0.5816261768341064, "learning_rate": 7.261255994490131e-05, "loss": 0.047, "step": 38310 }, { "grad_norm": 0.6598597764968872, "learning_rate": 7.259781156675517e-05, "loss": 0.05, "step": 38320 }, { "grad_norm": 0.26484861969947815, "learning_rate": 7.258306071734633e-05, "loss": 0.0364, "step": 38330 }, { "grad_norm": 0.47973179817199707, "learning_rate": 7.256830739828792e-05, "loss": 0.0452, "step": 38340 }, { "grad_norm": 0.6831160187721252, "learning_rate": 7.255355161119336e-05, "loss": 0.0538, "step": 38350 }, { "grad_norm": 0.666320264339447, "learning_rate": 7.253879335767628e-05, "loss": 0.0457, "step": 38360 }, { "grad_norm": 0.6485728025436401, "learning_rate": 7.252403263935067e-05, "loss": 0.0836, "step": 38370 }, { "grad_norm": 0.588708221912384, "learning_rate": 7.25092694578307e-05, "loss": 0.0505, "step": 38380 }, { "grad_norm": 0.4723913371562958, "learning_rate": 7.249450381473087e-05, "loss": 0.0397, "step": 38390 }, { "grad_norm": 0.9799138307571411, "learning_rate": 7.247973571166593e-05, "loss": 0.0678, "step": 38400 }, { "grad_norm": 0.47122785449028015, "learning_rate": 7.246496515025089e-05, "loss": 0.0474, "step": 38410 }, { "grad_norm": 0.6258354783058167, "learning_rate": 7.245019213210101e-05, "loss": 0.0337, "step": 38420 }, { "grad_norm": 0.5581914186477661, "learning_rate": 7.24354166588319e-05, "loss": 0.0349, "step": 38430 }, { "grad_norm": 0.9398354291915894, "learning_rate": 7.242063873205935e-05, "loss": 0.0426, "step": 38440 }, { "grad_norm": 1.2319796085357666, "learning_rate": 7.240585835339946e-05, "loss": 0.0552, "step": 38450 }, { "grad_norm": 0.5286974310874939, "learning_rate": 7.239107552446857e-05, "loss": 0.0512, "step": 38460 }, { "grad_norm": 0.555601179599762, "learning_rate": 7.237629024688334e-05, "loss": 0.0372, "step": 38470 }, { "grad_norm": 0.6065652966499329, "learning_rate": 7.236150252226064e-05, "loss": 0.0519, "step": 38480 }, { "grad_norm": 0.5326778888702393, "learning_rate": 7.234671235221765e-05, "loss": 0.0456, "step": 38490 }, { "grad_norm": 0.5511674880981445, "learning_rate": 7.233191973837179e-05, "loss": 0.0393, "step": 38500 }, { "grad_norm": 0.4957098960876465, "learning_rate": 7.231712468234078e-05, "loss": 0.0423, "step": 38510 }, { "grad_norm": 0.6140421032905579, "learning_rate": 7.230232718574254e-05, "loss": 0.0436, "step": 38520 }, { "grad_norm": 0.5464488863945007, "learning_rate": 7.228752725019535e-05, "loss": 0.0385, "step": 38530 }, { "grad_norm": 0.6363173723220825, "learning_rate": 7.227272487731769e-05, "loss": 0.042, "step": 38540 }, { "grad_norm": 0.5816105008125305, "learning_rate": 7.225792006872831e-05, "loss": 0.0373, "step": 38550 }, { "grad_norm": 0.46862873435020447, "learning_rate": 7.224311282604628e-05, "loss": 0.0317, "step": 38560 }, { "grad_norm": 0.25849589705467224, "learning_rate": 7.222830315089085e-05, "loss": 0.0492, "step": 38570 }, { "grad_norm": 0.6745702028274536, "learning_rate": 7.22134910448816e-05, "loss": 0.0675, "step": 38580 }, { "grad_norm": 0.684074342250824, "learning_rate": 7.219867650963839e-05, "loss": 0.0369, "step": 38590 }, { "grad_norm": 0.36376485228538513, "learning_rate": 7.21838595467813e-05, "loss": 0.0462, "step": 38600 }, { "grad_norm": 0.5901667475700378, "learning_rate": 7.216904015793067e-05, "loss": 0.0478, "step": 38610 }, { "grad_norm": 0.4992017447948456, "learning_rate": 7.215421834470713e-05, "loss": 0.0465, "step": 38620 }, { "grad_norm": 0.3782084882259369, "learning_rate": 7.213939410873158e-05, "loss": 0.0503, "step": 38630 }, { "grad_norm": 0.39890822768211365, "learning_rate": 7.212456745162519e-05, "loss": 0.0601, "step": 38640 }, { "grad_norm": 0.7718183994293213, "learning_rate": 7.210973837500937e-05, "loss": 0.052, "step": 38650 }, { "grad_norm": 0.7028900384902954, "learning_rate": 7.209490688050578e-05, "loss": 0.077, "step": 38660 }, { "grad_norm": 0.7165597081184387, "learning_rate": 7.208007296973641e-05, "loss": 0.0463, "step": 38670 }, { "grad_norm": 0.823478639125824, "learning_rate": 7.206523664432345e-05, "loss": 0.0829, "step": 38680 }, { "grad_norm": 0.5032960176467896, "learning_rate": 7.205039790588939e-05, "loss": 0.0562, "step": 38690 }, { "grad_norm": 0.6214709281921387, "learning_rate": 7.203555675605697e-05, "loss": 0.0492, "step": 38700 }, { "grad_norm": 0.5437238216400146, "learning_rate": 7.202071319644917e-05, "loss": 0.0341, "step": 38710 }, { "grad_norm": 0.47099611163139343, "learning_rate": 7.200586722868932e-05, "loss": 0.0398, "step": 38720 }, { "grad_norm": 1.1622395515441895, "learning_rate": 7.199101885440088e-05, "loss": 0.0457, "step": 38730 }, { "grad_norm": 0.4827169179916382, "learning_rate": 7.197616807520771e-05, "loss": 0.0364, "step": 38740 }, { "grad_norm": 0.7690921425819397, "learning_rate": 7.196131489273381e-05, "loss": 0.0461, "step": 38750 }, { "grad_norm": 0.6070765852928162, "learning_rate": 7.194645930860355e-05, "loss": 0.0713, "step": 38760 }, { "grad_norm": 0.7686904072761536, "learning_rate": 7.19316013244415e-05, "loss": 0.0467, "step": 38770 }, { "grad_norm": 0.5267509818077087, "learning_rate": 7.191674094187248e-05, "loss": 0.0451, "step": 38780 }, { "grad_norm": 0.8314331769943237, "learning_rate": 7.190187816252165e-05, "loss": 0.0463, "step": 38790 }, { "grad_norm": 0.5723344087600708, "learning_rate": 7.188701298801435e-05, "loss": 0.048, "step": 38800 }, { "grad_norm": 0.7178784608840942, "learning_rate": 7.18721454199762e-05, "loss": 0.0402, "step": 38810 }, { "grad_norm": 0.6886762976646423, "learning_rate": 7.185727546003312e-05, "loss": 0.0365, "step": 38820 }, { "grad_norm": 0.4053296744823456, "learning_rate": 7.184240310981126e-05, "loss": 0.0513, "step": 38830 }, { "grad_norm": 0.46282267570495605, "learning_rate": 7.182752837093704e-05, "loss": 0.0366, "step": 38840 }, { "grad_norm": 0.2575823962688446, "learning_rate": 7.181265124503711e-05, "loss": 0.0449, "step": 38850 }, { "grad_norm": 0.7231431603431702, "learning_rate": 7.179777173373847e-05, "loss": 0.0835, "step": 38860 }, { "grad_norm": 0.4687613844871521, "learning_rate": 7.178288983866826e-05, "loss": 0.0421, "step": 38870 }, { "grad_norm": 0.45146045088768005, "learning_rate": 7.176800556145397e-05, "loss": 0.0455, "step": 38880 }, { "grad_norm": 0.5337093472480774, "learning_rate": 7.175311890372334e-05, "loss": 0.0408, "step": 38890 }, { "grad_norm": 0.5742146372795105, "learning_rate": 7.17382298671043e-05, "loss": 0.0488, "step": 38900 }, { "grad_norm": 0.6766349077224731, "learning_rate": 7.172333845322515e-05, "loss": 0.045, "step": 38910 }, { "grad_norm": 0.5465413331985474, "learning_rate": 7.170844466371436e-05, "loss": 0.0375, "step": 38920 }, { "grad_norm": 0.6976367831230164, "learning_rate": 7.16935485002007e-05, "loss": 0.0563, "step": 38930 }, { "grad_norm": 0.520174503326416, "learning_rate": 7.167864996431319e-05, "loss": 0.0574, "step": 38940 }, { "grad_norm": 0.5841535329818726, "learning_rate": 7.166374905768111e-05, "loss": 0.0406, "step": 38950 }, { "grad_norm": 0.6051838994026184, "learning_rate": 7.164884578193401e-05, "loss": 0.0423, "step": 38960 }, { "grad_norm": 0.748991072177887, "learning_rate": 7.163394013870168e-05, "loss": 0.0518, "step": 38970 }, { "grad_norm": 0.6420962810516357, "learning_rate": 7.161903212961419e-05, "loss": 0.0638, "step": 38980 }, { "grad_norm": 0.5462857484817505, "learning_rate": 7.160412175630183e-05, "loss": 0.0502, "step": 38990 }, { "grad_norm": 0.5468903183937073, "learning_rate": 7.158920902039521e-05, "loss": 0.0413, "step": 39000 }, { "grad_norm": 0.6675649881362915, "learning_rate": 7.157429392352514e-05, "loss": 0.0537, "step": 39010 }, { "grad_norm": 0.5538516044616699, "learning_rate": 7.155937646732275e-05, "loss": 0.0508, "step": 39020 }, { "grad_norm": 0.4495801627635956, "learning_rate": 7.154445665341933e-05, "loss": 0.0359, "step": 39030 }, { "grad_norm": 0.5068318843841553, "learning_rate": 7.152953448344654e-05, "loss": 0.0422, "step": 39040 }, { "grad_norm": 0.7527812719345093, "learning_rate": 7.151460995903624e-05, "loss": 0.0403, "step": 39050 }, { "grad_norm": 0.5138373970985413, "learning_rate": 7.149968308182052e-05, "loss": 0.0448, "step": 39060 }, { "grad_norm": 0.5559006333351135, "learning_rate": 7.14847538534318e-05, "loss": 0.0412, "step": 39070 }, { "grad_norm": 0.5400506258010864, "learning_rate": 7.14698222755027e-05, "loss": 0.0451, "step": 39080 }, { "grad_norm": 0.522895336151123, "learning_rate": 7.14548883496661e-05, "loss": 0.0488, "step": 39090 }, { "grad_norm": 0.35938069224357605, "learning_rate": 7.143995207755517e-05, "loss": 0.0429, "step": 39100 }, { "grad_norm": 0.7913472056388855, "learning_rate": 7.142501346080333e-05, "loss": 0.0532, "step": 39110 }, { "grad_norm": 0.3714182376861572, "learning_rate": 7.141007250104421e-05, "loss": 0.052, "step": 39120 }, { "grad_norm": 0.5999217629432678, "learning_rate": 7.139512919991176e-05, "loss": 0.05, "step": 39130 }, { "grad_norm": 0.5086472034454346, "learning_rate": 7.138018355904014e-05, "loss": 0.0348, "step": 39140 }, { "grad_norm": 0.4688912034034729, "learning_rate": 7.13652355800638e-05, "loss": 0.0503, "step": 39150 }, { "grad_norm": 0.4549897015094757, "learning_rate": 7.135028526461743e-05, "loss": 0.0367, "step": 39160 }, { "grad_norm": 0.41848501563072205, "learning_rate": 7.133533261433594e-05, "loss": 0.0388, "step": 39170 }, { "grad_norm": 0.73745197057724, "learning_rate": 7.132037763085457e-05, "loss": 0.0343, "step": 39180 }, { "grad_norm": 0.5331742167472839, "learning_rate": 7.130542031580875e-05, "loss": 0.0502, "step": 39190 }, { "grad_norm": 0.7282891273498535, "learning_rate": 7.12904606708342e-05, "loss": 0.0494, "step": 39200 }, { "grad_norm": 0.4599485695362091, "learning_rate": 7.127549869756687e-05, "loss": 0.0472, "step": 39210 }, { "grad_norm": 0.47641393542289734, "learning_rate": 7.126053439764299e-05, "loss": 0.0424, "step": 39220 }, { "grad_norm": 0.4843176603317261, "learning_rate": 7.124556777269904e-05, "loss": 0.0519, "step": 39230 }, { "grad_norm": 0.38294467329978943, "learning_rate": 7.123059882437174e-05, "loss": 0.0365, "step": 39240 }, { "grad_norm": 0.38779184222221375, "learning_rate": 7.121562755429807e-05, "loss": 0.05, "step": 39250 }, { "grad_norm": 0.49517500400543213, "learning_rate": 7.120065396411528e-05, "loss": 0.0355, "step": 39260 }, { "grad_norm": 0.36607739329338074, "learning_rate": 7.118567805546084e-05, "loss": 0.0317, "step": 39270 }, { "grad_norm": 0.4858349859714508, "learning_rate": 7.117069982997248e-05, "loss": 0.0372, "step": 39280 }, { "grad_norm": 0.49371179938316345, "learning_rate": 7.115571928928825e-05, "loss": 0.0394, "step": 39290 }, { "grad_norm": 0.5729467868804932, "learning_rate": 7.114073643504635e-05, "loss": 0.0396, "step": 39300 }, { "grad_norm": 0.6786319613456726, "learning_rate": 7.11257512688853e-05, "loss": 0.0373, "step": 39310 }, { "grad_norm": 0.6403212547302246, "learning_rate": 7.111076379244384e-05, "loss": 0.0485, "step": 39320 }, { "grad_norm": 0.7424314618110657, "learning_rate": 7.109577400736101e-05, "loss": 0.0496, "step": 39330 }, { "grad_norm": 0.5260311365127563, "learning_rate": 7.108078191527605e-05, "loss": 0.0401, "step": 39340 }, { "grad_norm": 0.6275930404663086, "learning_rate": 7.106578751782847e-05, "loss": 0.0467, "step": 39350 }, { "grad_norm": 0.7408058047294617, "learning_rate": 7.105079081665803e-05, "loss": 0.0433, "step": 39360 }, { "grad_norm": 0.5619286298751831, "learning_rate": 7.103579181340476e-05, "loss": 0.0582, "step": 39370 }, { "grad_norm": 0.3823997974395752, "learning_rate": 7.102079050970893e-05, "loss": 0.0419, "step": 39380 }, { "grad_norm": 0.787889838218689, "learning_rate": 7.100578690721105e-05, "loss": 0.0432, "step": 39390 }, { "grad_norm": 0.7438694834709167, "learning_rate": 7.09907810075519e-05, "loss": 0.0537, "step": 39400 }, { "grad_norm": 0.7072626948356628, "learning_rate": 7.097577281237249e-05, "loss": 0.0592, "step": 39410 }, { "grad_norm": 0.476018488407135, "learning_rate": 7.09607623233141e-05, "loss": 0.051, "step": 39420 }, { "grad_norm": 0.5692206025123596, "learning_rate": 7.094574954201827e-05, "loss": 0.0387, "step": 39430 }, { "grad_norm": 0.9746919274330139, "learning_rate": 7.093073447012675e-05, "loss": 0.0541, "step": 39440 }, { "grad_norm": 0.5483725666999817, "learning_rate": 7.09157171092816e-05, "loss": 0.0463, "step": 39450 }, { "grad_norm": 0.5693802833557129, "learning_rate": 7.090069746112504e-05, "loss": 0.0473, "step": 39460 }, { "grad_norm": 1.0215210914611816, "learning_rate": 7.088567552729965e-05, "loss": 0.0535, "step": 39470 }, { "grad_norm": 0.6952365636825562, "learning_rate": 7.087065130944818e-05, "loss": 0.0561, "step": 39480 }, { "grad_norm": 0.5890967845916748, "learning_rate": 7.085562480921366e-05, "loss": 0.0383, "step": 39490 }, { "grad_norm": 0.3149339556694031, "learning_rate": 7.084059602823937e-05, "loss": 0.0357, "step": 39500 }, { "grad_norm": 0.5379436016082764, "learning_rate": 7.082556496816882e-05, "loss": 0.0492, "step": 39510 }, { "grad_norm": 0.4484158754348755, "learning_rate": 7.081053163064582e-05, "loss": 0.0583, "step": 39520 }, { "grad_norm": 0.6722437739372253, "learning_rate": 7.079549601731434e-05, "loss": 0.0569, "step": 39530 }, { "grad_norm": 0.48868343234062195, "learning_rate": 7.07804581298187e-05, "loss": 0.0365, "step": 39540 }, { "grad_norm": 0.8001405596733093, "learning_rate": 7.07654179698034e-05, "loss": 0.0618, "step": 39550 }, { "grad_norm": 0.5342220067977905, "learning_rate": 7.075037553891321e-05, "loss": 0.0419, "step": 39560 }, { "grad_norm": 0.5424968600273132, "learning_rate": 7.073533083879315e-05, "loss": 0.0512, "step": 39570 }, { "grad_norm": 0.6270466446876526, "learning_rate": 7.072028387108849e-05, "loss": 0.0544, "step": 39580 }, { "grad_norm": 0.4282774031162262, "learning_rate": 7.070523463744472e-05, "loss": 0.0537, "step": 39590 }, { "grad_norm": 0.44126203656196594, "learning_rate": 7.069018313950763e-05, "loss": 0.0467, "step": 39600 }, { "grad_norm": 0.6089631915092468, "learning_rate": 7.067512937892322e-05, "loss": 0.0514, "step": 39610 }, { "grad_norm": 0.5535975694656372, "learning_rate": 7.066007335733775e-05, "loss": 0.049, "step": 39620 }, { "grad_norm": 0.4602363407611847, "learning_rate": 7.064501507639772e-05, "loss": 0.0519, "step": 39630 }, { "grad_norm": 0.44457387924194336, "learning_rate": 7.062995453774987e-05, "loss": 0.047, "step": 39640 }, { "grad_norm": 0.6803357601165771, "learning_rate": 7.061489174304121e-05, "loss": 0.0342, "step": 39650 }, { "grad_norm": 0.5820778012275696, "learning_rate": 7.0599826693919e-05, "loss": 0.0327, "step": 39660 }, { "grad_norm": 0.769597589969635, "learning_rate": 7.05847593920307e-05, "loss": 0.0424, "step": 39670 }, { "grad_norm": 0.4008581340312958, "learning_rate": 7.056968983902406e-05, "loss": 0.0347, "step": 39680 }, { "grad_norm": 0.6173561811447144, "learning_rate": 7.055461803654706e-05, "loss": 0.0371, "step": 39690 }, { "grad_norm": 0.7394320368766785, "learning_rate": 7.053954398624794e-05, "loss": 0.0704, "step": 39700 }, { "grad_norm": 0.4497465193271637, "learning_rate": 7.052446768977518e-05, "loss": 0.0377, "step": 39710 }, { "grad_norm": 0.5656811594963074, "learning_rate": 7.050938914877748e-05, "loss": 0.0675, "step": 39720 }, { "grad_norm": 0.36162319779396057, "learning_rate": 7.049430836490382e-05, "loss": 0.0322, "step": 39730 }, { "grad_norm": 0.45931145548820496, "learning_rate": 7.047922533980342e-05, "loss": 0.043, "step": 39740 }, { "grad_norm": 0.5763852596282959, "learning_rate": 7.046414007512571e-05, "loss": 0.0585, "step": 39750 }, { "grad_norm": 0.5855180025100708, "learning_rate": 7.044905257252042e-05, "loss": 0.0591, "step": 39760 }, { "grad_norm": 0.6322004795074463, "learning_rate": 7.04339628336375e-05, "loss": 0.0495, "step": 39770 }, { "grad_norm": 0.6210305690765381, "learning_rate": 7.041887086012711e-05, "loss": 0.0421, "step": 39780 }, { "grad_norm": 0.6904467344284058, "learning_rate": 7.040377665363969e-05, "loss": 0.0485, "step": 39790 }, { "grad_norm": 0.7882780432701111, "learning_rate": 7.038868021582594e-05, "loss": 0.0798, "step": 39800 }, { "grad_norm": 0.6134046912193298, "learning_rate": 7.037358154833679e-05, "loss": 0.0528, "step": 39810 }, { "grad_norm": 0.6424406170845032, "learning_rate": 7.035848065282339e-05, "loss": 0.0425, "step": 39820 }, { "grad_norm": 0.5959106087684631, "learning_rate": 7.034337753093714e-05, "loss": 0.0529, "step": 39830 }, { "grad_norm": 0.6964280009269714, "learning_rate": 7.032827218432972e-05, "loss": 0.0582, "step": 39840 }, { "grad_norm": 0.798348069190979, "learning_rate": 7.031316461465302e-05, "loss": 0.0462, "step": 39850 }, { "grad_norm": 0.42979392409324646, "learning_rate": 7.029805482355915e-05, "loss": 0.0568, "step": 39860 }, { "grad_norm": 0.5795695781707764, "learning_rate": 7.028294281270055e-05, "loss": 0.0407, "step": 39870 }, { "grad_norm": 0.5195600986480713, "learning_rate": 7.026782858372982e-05, "loss": 0.0561, "step": 39880 }, { "grad_norm": 0.5019919276237488, "learning_rate": 7.025271213829982e-05, "loss": 0.0403, "step": 39890 }, { "grad_norm": 0.7275989651679993, "learning_rate": 7.023759347806366e-05, "loss": 0.052, "step": 39900 }, { "grad_norm": 0.6087260842323303, "learning_rate": 7.022247260467472e-05, "loss": 0.0556, "step": 39910 }, { "grad_norm": 0.5410177707672119, "learning_rate": 7.020734951978658e-05, "loss": 0.0375, "step": 39920 }, { "grad_norm": 0.2773137390613556, "learning_rate": 7.019222422505307e-05, "loss": 0.0484, "step": 39930 }, { "grad_norm": 0.5124337077140808, "learning_rate": 7.017709672212827e-05, "loss": 0.0402, "step": 39940 }, { "grad_norm": 0.4466988146305084, "learning_rate": 7.016196701266652e-05, "loss": 0.0454, "step": 39950 }, { "grad_norm": 0.602121114730835, "learning_rate": 7.014683509832238e-05, "loss": 0.0567, "step": 39960 }, { "grad_norm": 0.4546777307987213, "learning_rate": 7.013170098075063e-05, "loss": 0.0466, "step": 39970 }, { "grad_norm": 0.4978305995464325, "learning_rate": 7.011656466160632e-05, "loss": 0.0335, "step": 39980 }, { "grad_norm": 0.42459744215011597, "learning_rate": 7.010142614254475e-05, "loss": 0.0382, "step": 39990 }, { "grad_norm": 0.5190584063529968, "learning_rate": 7.008628542522147e-05, "loss": 0.0454, "step": 40000 }, { "grad_norm": 0.42553141713142395, "learning_rate": 7.007114251129218e-05, "loss": 0.0505, "step": 40010 }, { "grad_norm": 0.4659156799316406, "learning_rate": 7.005599740241293e-05, "loss": 0.0367, "step": 40020 }, { "grad_norm": 0.3709252178668976, "learning_rate": 7.004085010023996e-05, "loss": 0.038, "step": 40030 }, { "grad_norm": 0.5079686641693115, "learning_rate": 7.002570060642976e-05, "loss": 0.0467, "step": 40040 }, { "grad_norm": 0.5626863837242126, "learning_rate": 7.001054892263903e-05, "loss": 0.043, "step": 40050 }, { "grad_norm": 0.8861824870109558, "learning_rate": 6.999539505052477e-05, "loss": 0.0294, "step": 40060 }, { "grad_norm": 0.5293782353401184, "learning_rate": 6.998023899174418e-05, "loss": 0.0748, "step": 40070 }, { "grad_norm": 0.5251789093017578, "learning_rate": 6.996508074795467e-05, "loss": 0.039, "step": 40080 }, { "grad_norm": 0.5167202949523926, "learning_rate": 6.994992032081396e-05, "loss": 0.0383, "step": 40090 }, { "grad_norm": 0.8794233798980713, "learning_rate": 6.993475771197995e-05, "loss": 0.0482, "step": 40100 }, { "grad_norm": 0.6024871468544006, "learning_rate": 6.991959292311082e-05, "loss": 0.0366, "step": 40110 }, { "grad_norm": 0.6172850131988525, "learning_rate": 6.990442595586495e-05, "loss": 0.0391, "step": 40120 }, { "grad_norm": 0.34370437264442444, "learning_rate": 6.988925681190098e-05, "loss": 0.0388, "step": 40130 }, { "grad_norm": 0.6326910853385925, "learning_rate": 6.987408549287778e-05, "loss": 0.065, "step": 40140 }, { "grad_norm": 0.4835527539253235, "learning_rate": 6.985891200045449e-05, "loss": 0.0459, "step": 40150 }, { "grad_norm": 0.5197232365608215, "learning_rate": 6.984373633629045e-05, "loss": 0.0432, "step": 40160 }, { "grad_norm": 0.6397375464439392, "learning_rate": 6.98285585020452e-05, "loss": 0.0583, "step": 40170 }, { "grad_norm": 0.3254182040691376, "learning_rate": 6.981337849937864e-05, "loss": 0.0414, "step": 40180 }, { "grad_norm": 0.4519194960594177, "learning_rate": 6.979819632995078e-05, "loss": 0.0375, "step": 40190 }, { "grad_norm": 0.5335282683372498, "learning_rate": 6.978301199542193e-05, "loss": 0.0454, "step": 40200 }, { "grad_norm": 0.5468130707740784, "learning_rate": 6.976782549745263e-05, "loss": 0.0517, "step": 40210 }, { "grad_norm": 0.6543139219284058, "learning_rate": 6.975263683770367e-05, "loss": 0.0373, "step": 40220 }, { "grad_norm": 0.41169601678848267, "learning_rate": 6.973744601783604e-05, "loss": 0.0467, "step": 40230 }, { "grad_norm": 0.654420018196106, "learning_rate": 6.972225303951097e-05, "loss": 0.0469, "step": 40240 }, { "grad_norm": 0.6104034185409546, "learning_rate": 6.970705790438998e-05, "loss": 0.0497, "step": 40250 }, { "grad_norm": 0.5091560482978821, "learning_rate": 6.969186061413477e-05, "loss": 0.0413, "step": 40260 }, { "grad_norm": 0.8235982656478882, "learning_rate": 6.967666117040727e-05, "loss": 0.0483, "step": 40270 }, { "grad_norm": 0.5555811524391174, "learning_rate": 6.966145957486972e-05, "loss": 0.0405, "step": 40280 }, { "grad_norm": 0.44131478667259216, "learning_rate": 6.964625582918449e-05, "loss": 0.0571, "step": 40290 }, { "grad_norm": 0.4009896218776703, "learning_rate": 6.963104993501425e-05, "loss": 0.0446, "step": 40300 }, { "grad_norm": 0.594877302646637, "learning_rate": 6.961584189402192e-05, "loss": 0.0495, "step": 40310 }, { "grad_norm": 0.462342768907547, "learning_rate": 6.96006317078706e-05, "loss": 0.0468, "step": 40320 }, { "grad_norm": 0.5540574789047241, "learning_rate": 6.958541937822367e-05, "loss": 0.0651, "step": 40330 }, { "grad_norm": 0.69124436378479, "learning_rate": 6.957020490674473e-05, "loss": 0.049, "step": 40340 }, { "grad_norm": 0.3108363747596741, "learning_rate": 6.95549882950976e-05, "loss": 0.0463, "step": 40350 }, { "grad_norm": 0.32637977600097656, "learning_rate": 6.953976954494635e-05, "loss": 0.0461, "step": 40360 }, { "grad_norm": 0.5432504415512085, "learning_rate": 6.952454865795528e-05, "loss": 0.0389, "step": 40370 }, { "grad_norm": 0.48110371828079224, "learning_rate": 6.950932563578892e-05, "loss": 0.0339, "step": 40380 }, { "grad_norm": 0.6632357835769653, "learning_rate": 6.949410048011206e-05, "loss": 0.0443, "step": 40390 }, { "grad_norm": 0.8549562692642212, "learning_rate": 6.947887319258966e-05, "loss": 0.0457, "step": 40400 }, { "grad_norm": 0.6411885619163513, "learning_rate": 6.946364377488696e-05, "loss": 0.0474, "step": 40410 }, { "grad_norm": 0.5334661602973938, "learning_rate": 6.944841222866947e-05, "loss": 0.0438, "step": 40420 }, { "grad_norm": 0.47835272550582886, "learning_rate": 6.943317855560284e-05, "loss": 0.0531, "step": 40430 }, { "grad_norm": 0.6311628222465515, "learning_rate": 6.941794275735302e-05, "loss": 0.0348, "step": 40440 }, { "grad_norm": 0.39888864755630493, "learning_rate": 6.94027048355862e-05, "loss": 0.0365, "step": 40450 }, { "grad_norm": 0.886711597442627, "learning_rate": 6.938746479196871e-05, "loss": 0.0701, "step": 40460 }, { "grad_norm": 0.5415273904800415, "learning_rate": 6.937222262816724e-05, "loss": 0.0444, "step": 40470 }, { "grad_norm": 0.624238908290863, "learning_rate": 6.935697834584865e-05, "loss": 0.0443, "step": 40480 }, { "grad_norm": 0.3813030421733856, "learning_rate": 6.934173194667998e-05, "loss": 0.0477, "step": 40490 }, { "grad_norm": 0.45907407999038696, "learning_rate": 6.93264834323286e-05, "loss": 0.0414, "step": 40500 }, { "grad_norm": 1.4689618349075317, "learning_rate": 6.931123280446206e-05, "loss": 0.0446, "step": 40510 }, { "grad_norm": 0.6001729369163513, "learning_rate": 6.929598006474811e-05, "loss": 0.0361, "step": 40520 }, { "grad_norm": 0.4871310889720917, "learning_rate": 6.92807252148548e-05, "loss": 0.0338, "step": 40530 }, { "grad_norm": 0.4188902974128723, "learning_rate": 6.926546825645037e-05, "loss": 0.031, "step": 40540 }, { "grad_norm": 0.3810742199420929, "learning_rate": 6.92502091912033e-05, "loss": 0.0396, "step": 40550 }, { "grad_norm": 0.5839111804962158, "learning_rate": 6.923494802078229e-05, "loss": 0.0459, "step": 40560 }, { "grad_norm": 0.4609980285167694, "learning_rate": 6.921968474685628e-05, "loss": 0.0586, "step": 40570 }, { "grad_norm": 0.5019090175628662, "learning_rate": 6.920441937109445e-05, "loss": 0.0461, "step": 40580 }, { "grad_norm": 0.6022616624832153, "learning_rate": 6.91891518951662e-05, "loss": 0.0378, "step": 40590 }, { "grad_norm": 0.46397000551223755, "learning_rate": 6.917388232074114e-05, "loss": 0.043, "step": 40600 }, { "grad_norm": 0.6181228160858154, "learning_rate": 6.915861064948914e-05, "loss": 0.0545, "step": 40610 }, { "grad_norm": 0.6877040863037109, "learning_rate": 6.914333688308025e-05, "loss": 0.04, "step": 40620 }, { "grad_norm": 0.7453680634498596, "learning_rate": 6.912806102318487e-05, "loss": 0.0407, "step": 40630 }, { "grad_norm": 0.4757643938064575, "learning_rate": 6.911278307147347e-05, "loss": 0.047, "step": 40640 }, { "grad_norm": 1.1120423078536987, "learning_rate": 6.909750302961684e-05, "loss": 0.0445, "step": 40650 }, { "grad_norm": 0.5663095116615295, "learning_rate": 6.9082220899286e-05, "loss": 0.0365, "step": 40660 }, { "grad_norm": 0.4196541905403137, "learning_rate": 6.906693668215219e-05, "loss": 0.0429, "step": 40670 }, { "grad_norm": 0.33167171478271484, "learning_rate": 6.905165037988683e-05, "loss": 0.0528, "step": 40680 }, { "grad_norm": 0.7350439429283142, "learning_rate": 6.903636199416164e-05, "loss": 0.0523, "step": 40690 }, { "grad_norm": 0.7621098160743713, "learning_rate": 6.902107152664851e-05, "loss": 0.0566, "step": 40700 }, { "grad_norm": 0.42508450150489807, "learning_rate": 6.90057789790196e-05, "loss": 0.0552, "step": 40710 }, { "grad_norm": 0.3992989659309387, "learning_rate": 6.899048435294728e-05, "loss": 0.0451, "step": 40720 }, { "grad_norm": 0.4016650915145874, "learning_rate": 6.897518765010415e-05, "loss": 0.0423, "step": 40730 }, { "grad_norm": 0.6706708669662476, "learning_rate": 6.895988887216302e-05, "loss": 0.0478, "step": 40740 }, { "grad_norm": 0.5392720103263855, "learning_rate": 6.894458802079694e-05, "loss": 0.042, "step": 40750 }, { "grad_norm": 0.6942638158798218, "learning_rate": 6.892928509767922e-05, "loss": 0.0711, "step": 40760 }, { "grad_norm": 0.49227970838546753, "learning_rate": 6.891398010448333e-05, "loss": 0.046, "step": 40770 }, { "grad_norm": 0.4693014919757843, "learning_rate": 6.8898673042883e-05, "loss": 0.0464, "step": 40780 }, { "grad_norm": 0.5479590892791748, "learning_rate": 6.888336391455222e-05, "loss": 0.0401, "step": 40790 }, { "grad_norm": 0.5095242857933044, "learning_rate": 6.886805272116513e-05, "loss": 0.0379, "step": 40800 }, { "grad_norm": 0.3985646367073059, "learning_rate": 6.885273946439617e-05, "loss": 0.0378, "step": 40810 }, { "grad_norm": 0.4799807071685791, "learning_rate": 6.883742414591998e-05, "loss": 0.043, "step": 40820 }, { "grad_norm": 0.4290860593318939, "learning_rate": 6.88221067674114e-05, "loss": 0.0379, "step": 40830 }, { "grad_norm": 0.4219008684158325, "learning_rate": 6.88067873305455e-05, "loss": 0.0465, "step": 40840 }, { "grad_norm": 0.48238641023635864, "learning_rate": 6.879146583699765e-05, "loss": 0.0387, "step": 40850 }, { "grad_norm": 0.6856355667114258, "learning_rate": 6.877614228844334e-05, "loss": 0.0386, "step": 40860 }, { "grad_norm": 0.4832354187965393, "learning_rate": 6.876081668655832e-05, "loss": 0.045, "step": 40870 }, { "grad_norm": 0.38369303941726685, "learning_rate": 6.874548903301861e-05, "loss": 0.0458, "step": 40880 }, { "grad_norm": 0.41973280906677246, "learning_rate": 6.873015932950039e-05, "loss": 0.0403, "step": 40890 }, { "grad_norm": 0.36987313628196716, "learning_rate": 6.871482757768012e-05, "loss": 0.0434, "step": 40900 }, { "grad_norm": 0.8959362506866455, "learning_rate": 6.869949377923442e-05, "loss": 0.0826, "step": 40910 }, { "grad_norm": 0.41355398297309875, "learning_rate": 6.868415793584022e-05, "loss": 0.0454, "step": 40920 }, { "grad_norm": 0.6234316825866699, "learning_rate": 6.86688200491746e-05, "loss": 0.0456, "step": 40930 }, { "grad_norm": 0.41121864318847656, "learning_rate": 6.865348012091488e-05, "loss": 0.0581, "step": 40940 }, { "grad_norm": 0.6013553142547607, "learning_rate": 6.86381381527386e-05, "loss": 0.0439, "step": 40950 }, { "grad_norm": 0.6868756413459778, "learning_rate": 6.862279414632359e-05, "loss": 0.0448, "step": 40960 }, { "grad_norm": 0.699888288974762, "learning_rate": 6.860744810334778e-05, "loss": 0.0333, "step": 40970 }, { "grad_norm": 0.7671993374824524, "learning_rate": 6.859210002548943e-05, "loss": 0.034, "step": 40980 }, { "grad_norm": 0.40234309434890747, "learning_rate": 6.8576749914427e-05, "loss": 0.0314, "step": 40990 }, { "grad_norm": 0.40545234084129333, "learning_rate": 6.85613977718391e-05, "loss": 0.0446, "step": 41000 }, { "grad_norm": 0.48696109652519226, "learning_rate": 6.854604359940464e-05, "loss": 0.0483, "step": 41010 }, { "grad_norm": 0.68381667137146, "learning_rate": 6.853068739880276e-05, "loss": 0.0469, "step": 41020 }, { "grad_norm": 0.45369231700897217, "learning_rate": 6.851532917171276e-05, "loss": 0.0376, "step": 41030 }, { "grad_norm": 0.3266373574733734, "learning_rate": 6.849996891981418e-05, "loss": 0.0444, "step": 41040 }, { "grad_norm": 0.3982946574687958, "learning_rate": 6.848460664478684e-05, "loss": 0.0415, "step": 41050 }, { "grad_norm": 0.44625186920166016, "learning_rate": 6.846924234831069e-05, "loss": 0.0522, "step": 41060 }, { "grad_norm": 0.6359847784042358, "learning_rate": 6.845387603206597e-05, "loss": 0.038, "step": 41070 }, { "grad_norm": 0.4772109389305115, "learning_rate": 6.843850769773312e-05, "loss": 0.0392, "step": 41080 }, { "grad_norm": 0.6724666953086853, "learning_rate": 6.842313734699278e-05, "loss": 0.0406, "step": 41090 }, { "grad_norm": 0.5918323993682861, "learning_rate": 6.840776498152584e-05, "loss": 0.041, "step": 41100 }, { "grad_norm": 0.5652077794075012, "learning_rate": 6.83923906030134e-05, "loss": 0.0376, "step": 41110 }, { "grad_norm": 0.527122974395752, "learning_rate": 6.837701421313677e-05, "loss": 0.0406, "step": 41120 }, { "grad_norm": 0.6866210103034973, "learning_rate": 6.83616358135775e-05, "loss": 0.0374, "step": 41130 }, { "grad_norm": 0.47954705357551575, "learning_rate": 6.834625540601734e-05, "loss": 0.0423, "step": 41140 }, { "grad_norm": 0.7542096376419067, "learning_rate": 6.833087299213829e-05, "loss": 0.0444, "step": 41150 }, { "grad_norm": 0.7733188271522522, "learning_rate": 6.83154885736225e-05, "loss": 0.0323, "step": 41160 }, { "grad_norm": 0.749836802482605, "learning_rate": 6.830010215215244e-05, "loss": 0.0496, "step": 41170 }, { "grad_norm": 0.7936599254608154, "learning_rate": 6.828471372941072e-05, "loss": 0.0431, "step": 41180 }, { "grad_norm": 0.7887213826179504, "learning_rate": 6.826932330708018e-05, "loss": 0.0443, "step": 41190 }, { "grad_norm": 1.2122493982315063, "learning_rate": 6.825393088684393e-05, "loss": 0.0459, "step": 41200 }, { "grad_norm": 0.6925350427627563, "learning_rate": 6.823853647038524e-05, "loss": 0.0401, "step": 41210 }, { "grad_norm": 0.5942347645759583, "learning_rate": 6.822314005938762e-05, "loss": 0.0612, "step": 41220 }, { "grad_norm": 0.29337435960769653, "learning_rate": 6.820774165553479e-05, "loss": 0.0343, "step": 41230 }, { "grad_norm": 0.38077831268310547, "learning_rate": 6.819234126051072e-05, "loss": 0.0423, "step": 41240 }, { "grad_norm": 0.6297895312309265, "learning_rate": 6.817693887599956e-05, "loss": 0.0455, "step": 41250 }, { "grad_norm": 0.6067937016487122, "learning_rate": 6.816153450368568e-05, "loss": 0.0473, "step": 41260 }, { "grad_norm": 1.2064725160598755, "learning_rate": 6.814612814525372e-05, "loss": 0.0423, "step": 41270 }, { "grad_norm": 0.48257166147232056, "learning_rate": 6.813071980238846e-05, "loss": 0.04, "step": 41280 }, { "grad_norm": 0.4414083957672119, "learning_rate": 6.811530947677492e-05, "loss": 0.0378, "step": 41290 }, { "grad_norm": 0.40779638290405273, "learning_rate": 6.809989717009839e-05, "loss": 0.0475, "step": 41300 }, { "grad_norm": 0.6427592635154724, "learning_rate": 6.808448288404431e-05, "loss": 0.0326, "step": 41310 }, { "grad_norm": 1.0649971961975098, "learning_rate": 6.806906662029838e-05, "loss": 0.0614, "step": 41320 }, { "grad_norm": 0.44274333119392395, "learning_rate": 6.805364838054647e-05, "loss": 0.0344, "step": 41330 }, { "grad_norm": 0.8146663308143616, "learning_rate": 6.803822816647474e-05, "loss": 0.0388, "step": 41340 }, { "grad_norm": 0.7344682812690735, "learning_rate": 6.802280597976949e-05, "loss": 0.0426, "step": 41350 }, { "grad_norm": 0.39796000719070435, "learning_rate": 6.800738182211727e-05, "loss": 0.0398, "step": 41360 }, { "grad_norm": 0.5162355303764343, "learning_rate": 6.799195569520485e-05, "loss": 0.0293, "step": 41370 }, { "grad_norm": 0.6577728986740112, "learning_rate": 6.79765276007192e-05, "loss": 0.0475, "step": 41380 }, { "grad_norm": 0.7029166221618652, "learning_rate": 6.796109754034753e-05, "loss": 0.0538, "step": 41390 }, { "grad_norm": 0.40332165360450745, "learning_rate": 6.794566551577724e-05, "loss": 0.0507, "step": 41400 }, { "grad_norm": 0.594450056552887, "learning_rate": 6.793023152869594e-05, "loss": 0.0414, "step": 41410 }, { "grad_norm": 0.5895494818687439, "learning_rate": 6.791479558079147e-05, "loss": 0.0391, "step": 41420 }, { "grad_norm": 0.2536514103412628, "learning_rate": 6.789935767375191e-05, "loss": 0.0497, "step": 41430 }, { "grad_norm": 0.7098327875137329, "learning_rate": 6.78839178092655e-05, "loss": 0.0321, "step": 41440 }, { "grad_norm": 0.5596318244934082, "learning_rate": 6.786847598902072e-05, "loss": 0.0455, "step": 41450 }, { "grad_norm": 0.700771152973175, "learning_rate": 6.785303221470627e-05, "loss": 0.0583, "step": 41460 }, { "grad_norm": 0.35759133100509644, "learning_rate": 6.783758648801108e-05, "loss": 0.0329, "step": 41470 }, { "grad_norm": 0.37054577469825745, "learning_rate": 6.782213881062422e-05, "loss": 0.0315, "step": 41480 }, { "grad_norm": 0.3662903606891632, "learning_rate": 6.780668918423508e-05, "loss": 0.0477, "step": 41490 }, { "grad_norm": 0.7606661319732666, "learning_rate": 6.779123761053317e-05, "loss": 0.0521, "step": 41500 }, { "grad_norm": 0.542398989200592, "learning_rate": 6.777578409120826e-05, "loss": 0.0394, "step": 41510 }, { "grad_norm": 0.7923572063446045, "learning_rate": 6.776032862795033e-05, "loss": 0.0472, "step": 41520 }, { "grad_norm": 0.7001250982284546, "learning_rate": 6.774487122244956e-05, "loss": 0.0488, "step": 41530 }, { "grad_norm": 0.5246747136116028, "learning_rate": 6.772941187639637e-05, "loss": 0.0664, "step": 41540 }, { "grad_norm": 0.5705745220184326, "learning_rate": 6.771395059148134e-05, "loss": 0.0421, "step": 41550 }, { "grad_norm": 0.5907261371612549, "learning_rate": 6.76984873693953e-05, "loss": 0.0461, "step": 41560 }, { "grad_norm": 0.4098798334598541, "learning_rate": 6.768302221182931e-05, "loss": 0.0421, "step": 41570 }, { "grad_norm": 0.3607255220413208, "learning_rate": 6.766755512047457e-05, "loss": 0.0478, "step": 41580 }, { "grad_norm": 0.4573536217212677, "learning_rate": 6.765208609702259e-05, "loss": 0.0635, "step": 41590 }, { "grad_norm": 0.3651900291442871, "learning_rate": 6.763661514316499e-05, "loss": 0.0432, "step": 41600 }, { "grad_norm": 0.43347546458244324, "learning_rate": 6.76211422605937e-05, "loss": 0.0389, "step": 41610 }, { "grad_norm": 0.4539860188961029, "learning_rate": 6.760566745100076e-05, "loss": 0.0427, "step": 41620 }, { "grad_norm": 1.6178951263427734, "learning_rate": 6.75901907160785e-05, "loss": 0.0515, "step": 41630 }, { "grad_norm": 0.608616292476654, "learning_rate": 6.757471205751943e-05, "loss": 0.0536, "step": 41640 }, { "grad_norm": 0.5555578470230103, "learning_rate": 6.75592314770163e-05, "loss": 0.0411, "step": 41650 }, { "grad_norm": 0.5135046243667603, "learning_rate": 6.7543748976262e-05, "loss": 0.0446, "step": 41660 }, { "grad_norm": 0.5195563435554504, "learning_rate": 6.752826455694968e-05, "loss": 0.0569, "step": 41670 }, { "grad_norm": 0.9497635960578918, "learning_rate": 6.751277822077271e-05, "loss": 0.0475, "step": 41680 }, { "grad_norm": 0.4085886478424072, "learning_rate": 6.749728996942465e-05, "loss": 0.0536, "step": 41690 }, { "grad_norm": 0.36724281311035156, "learning_rate": 6.748179980459924e-05, "loss": 0.0426, "step": 41700 }, { "grad_norm": 0.3107476830482483, "learning_rate": 6.746630772799052e-05, "loss": 0.0571, "step": 41710 }, { "grad_norm": 0.4800924062728882, "learning_rate": 6.745081374129262e-05, "loss": 0.0407, "step": 41720 }, { "grad_norm": 0.4289737343788147, "learning_rate": 6.743531784619998e-05, "loss": 0.0451, "step": 41730 }, { "grad_norm": 0.4829464554786682, "learning_rate": 6.741982004440719e-05, "loss": 0.0393, "step": 41740 }, { "grad_norm": 0.4302888512611389, "learning_rate": 6.740432033760907e-05, "loss": 0.0484, "step": 41750 }, { "grad_norm": 0.7213699221611023, "learning_rate": 6.738881872750066e-05, "loss": 0.0409, "step": 41760 }, { "grad_norm": 0.5366184115409851, "learning_rate": 6.737331521577715e-05, "loss": 0.0402, "step": 41770 }, { "grad_norm": 0.7940341830253601, "learning_rate": 6.735780980413403e-05, "loss": 0.0456, "step": 41780 }, { "grad_norm": 0.8212419748306274, "learning_rate": 6.734230249426692e-05, "loss": 0.0474, "step": 41790 }, { "grad_norm": 0.6329543590545654, "learning_rate": 6.732679328787168e-05, "loss": 0.0545, "step": 41800 }, { "grad_norm": 0.689367949962616, "learning_rate": 6.731128218664438e-05, "loss": 0.0469, "step": 41810 }, { "grad_norm": 0.5718804001808167, "learning_rate": 6.729576919228129e-05, "loss": 0.048, "step": 41820 }, { "grad_norm": 0.4477439522743225, "learning_rate": 6.728025430647888e-05, "loss": 0.0489, "step": 41830 }, { "grad_norm": 0.4203040301799774, "learning_rate": 6.726473753093383e-05, "loss": 0.0342, "step": 41840 }, { "grad_norm": 0.5963268876075745, "learning_rate": 6.724921886734305e-05, "loss": 0.035, "step": 41850 }, { "grad_norm": 0.29994210600852966, "learning_rate": 6.723369831740363e-05, "loss": 0.0356, "step": 41860 }, { "grad_norm": 0.8181735277175903, "learning_rate": 6.721817588281288e-05, "loss": 0.0504, "step": 41870 }, { "grad_norm": 0.26966628432273865, "learning_rate": 6.720265156526828e-05, "loss": 0.0427, "step": 41880 }, { "grad_norm": 0.4193284809589386, "learning_rate": 6.718712536646758e-05, "loss": 0.0322, "step": 41890 }, { "grad_norm": 0.443450927734375, "learning_rate": 6.71715972881087e-05, "loss": 0.0382, "step": 41900 }, { "grad_norm": 0.6814678907394409, "learning_rate": 6.715606733188974e-05, "loss": 0.0411, "step": 41910 }, { "grad_norm": 0.43334317207336426, "learning_rate": 6.714053549950909e-05, "loss": 0.0367, "step": 41920 }, { "grad_norm": 0.7482848167419434, "learning_rate": 6.71250017926652e-05, "loss": 0.0454, "step": 41930 }, { "grad_norm": 0.5626275539398193, "learning_rate": 6.71094662130569e-05, "loss": 0.0415, "step": 41940 }, { "grad_norm": 0.6565224528312683, "learning_rate": 6.709392876238307e-05, "loss": 0.0428, "step": 41950 }, { "grad_norm": 0.4850108325481415, "learning_rate": 6.707838944234289e-05, "loss": 0.0449, "step": 41960 }, { "grad_norm": 0.48000824451446533, "learning_rate": 6.706284825463574e-05, "loss": 0.037, "step": 41970 }, { "grad_norm": 0.5530019998550415, "learning_rate": 6.704730520096114e-05, "loss": 0.0409, "step": 41980 }, { "grad_norm": 0.35539504885673523, "learning_rate": 6.703176028301888e-05, "loss": 0.0362, "step": 41990 }, { "grad_norm": 0.4020082950592041, "learning_rate": 6.701621350250892e-05, "loss": 0.0398, "step": 42000 }, { "grad_norm": 0.49360841512680054, "learning_rate": 6.700066486113144e-05, "loss": 0.0538, "step": 42010 }, { "grad_norm": 0.44502130150794983, "learning_rate": 6.69851143605868e-05, "loss": 0.0436, "step": 42020 }, { "grad_norm": 0.6370185017585754, "learning_rate": 6.696956200257561e-05, "loss": 0.046, "step": 42030 }, { "grad_norm": 0.388451486825943, "learning_rate": 6.695400778879863e-05, "loss": 0.0297, "step": 42040 }, { "grad_norm": 0.5328022241592407, "learning_rate": 6.693845172095683e-05, "loss": 0.0387, "step": 42050 }, { "grad_norm": 0.570099413394928, "learning_rate": 6.692289380075142e-05, "loss": 0.0505, "step": 42060 }, { "grad_norm": 0.4924107789993286, "learning_rate": 6.690733402988379e-05, "loss": 0.0377, "step": 42070 }, { "grad_norm": 0.3615560233592987, "learning_rate": 6.689177241005553e-05, "loss": 0.0366, "step": 42080 }, { "grad_norm": 0.4190600514411926, "learning_rate": 6.687620894296844e-05, "loss": 0.0413, "step": 42090 }, { "grad_norm": 0.47485679388046265, "learning_rate": 6.686064363032451e-05, "loss": 0.0372, "step": 42100 }, { "grad_norm": 0.3170727789402008, "learning_rate": 6.684507647382596e-05, "loss": 0.0524, "step": 42110 }, { "grad_norm": 0.44352179765701294, "learning_rate": 6.682950747517513e-05, "loss": 0.0535, "step": 42120 }, { "grad_norm": 0.6404318809509277, "learning_rate": 6.68139366360747e-05, "loss": 0.047, "step": 42130 }, { "grad_norm": 0.5964602828025818, "learning_rate": 6.679836395822744e-05, "loss": 0.0389, "step": 42140 }, { "grad_norm": 0.6249107122421265, "learning_rate": 6.678278944333633e-05, "loss": 0.0428, "step": 42150 }, { "grad_norm": 0.5915071964263916, "learning_rate": 6.676721309310462e-05, "loss": 0.0425, "step": 42160 }, { "grad_norm": 0.4044371545314789, "learning_rate": 6.67516349092357e-05, "loss": 0.048, "step": 42170 }, { "grad_norm": 0.5541946887969971, "learning_rate": 6.673605489343315e-05, "loss": 0.0455, "step": 42180 }, { "grad_norm": 0.5999897122383118, "learning_rate": 6.672047304740081e-05, "loss": 0.0473, "step": 42190 }, { "grad_norm": 0.46513795852661133, "learning_rate": 6.670488937284268e-05, "loss": 0.0513, "step": 42200 }, { "grad_norm": 0.562030017375946, "learning_rate": 6.668930387146294e-05, "loss": 0.0385, "step": 42210 }, { "grad_norm": 0.5644851326942444, "learning_rate": 6.667371654496605e-05, "loss": 0.0344, "step": 42220 }, { "grad_norm": 0.7369999885559082, "learning_rate": 6.665812739505659e-05, "loss": 0.0371, "step": 42230 }, { "grad_norm": 0.3929874002933502, "learning_rate": 6.664253642343935e-05, "loss": 0.0473, "step": 42240 }, { "grad_norm": 0.3062319755554199, "learning_rate": 6.662694363181935e-05, "loss": 0.0515, "step": 42250 }, { "grad_norm": 0.8789100050926208, "learning_rate": 6.66113490219018e-05, "loss": 0.0427, "step": 42260 }, { "grad_norm": 0.5712013244628906, "learning_rate": 6.659575259539207e-05, "loss": 0.0447, "step": 42270 }, { "grad_norm": 0.5186916589736938, "learning_rate": 6.65801543539958e-05, "loss": 0.0414, "step": 42280 }, { "grad_norm": 0.41842028498649597, "learning_rate": 6.65645542994188e-05, "loss": 0.0404, "step": 42290 }, { "grad_norm": 0.6476746797561646, "learning_rate": 6.654895243336702e-05, "loss": 0.0346, "step": 42300 }, { "grad_norm": 0.3226379454135895, "learning_rate": 6.65333487575467e-05, "loss": 0.0334, "step": 42310 }, { "grad_norm": 0.5633754730224609, "learning_rate": 6.651774327366422e-05, "loss": 0.0381, "step": 42320 }, { "grad_norm": 0.5305528044700623, "learning_rate": 6.650213598342613e-05, "loss": 0.0421, "step": 42330 }, { "grad_norm": 0.5662806034088135, "learning_rate": 6.648652688853931e-05, "loss": 0.0453, "step": 42340 }, { "grad_norm": 0.5596418380737305, "learning_rate": 6.647091599071066e-05, "loss": 0.0364, "step": 42350 }, { "grad_norm": 0.38123494386672974, "learning_rate": 6.645530329164743e-05, "loss": 0.0388, "step": 42360 }, { "grad_norm": 0.6262298226356506, "learning_rate": 6.643968879305694e-05, "loss": 0.0363, "step": 42370 }, { "grad_norm": 0.6030360460281372, "learning_rate": 6.642407249664683e-05, "loss": 0.0457, "step": 42380 }, { "grad_norm": 0.8406639099121094, "learning_rate": 6.640845440412483e-05, "loss": 0.0389, "step": 42390 }, { "grad_norm": 1.1989606618881226, "learning_rate": 6.639283451719893e-05, "loss": 0.0451, "step": 42400 }, { "grad_norm": 0.39051443338394165, "learning_rate": 6.637721283757729e-05, "loss": 0.0452, "step": 42410 }, { "grad_norm": 0.525603711605072, "learning_rate": 6.636158936696828e-05, "loss": 0.041, "step": 42420 }, { "grad_norm": 0.7297793030738831, "learning_rate": 6.634596410708047e-05, "loss": 0.0494, "step": 42430 }, { "grad_norm": 0.35154297947883606, "learning_rate": 6.633033705962257e-05, "loss": 0.0362, "step": 42440 }, { "grad_norm": 0.4759536385536194, "learning_rate": 6.631470822630359e-05, "loss": 0.051, "step": 42450 }, { "grad_norm": 0.584578812122345, "learning_rate": 6.629907760883263e-05, "loss": 0.0385, "step": 42460 }, { "grad_norm": 0.41934114694595337, "learning_rate": 6.628344520891905e-05, "loss": 0.0334, "step": 42470 }, { "grad_norm": 0.9778308272361755, "learning_rate": 6.62678110282724e-05, "loss": 0.0345, "step": 42480 }, { "grad_norm": 0.4756602346897125, "learning_rate": 6.625217506860237e-05, "loss": 0.037, "step": 42490 }, { "grad_norm": 0.5007375478744507, "learning_rate": 6.623653733161892e-05, "loss": 0.0377, "step": 42500 }, { "grad_norm": 0.47691142559051514, "learning_rate": 6.622089781903217e-05, "loss": 0.0498, "step": 42510 }, { "grad_norm": 0.4098455309867859, "learning_rate": 6.620525653255241e-05, "loss": 0.0477, "step": 42520 }, { "grad_norm": 0.39970651268959045, "learning_rate": 6.618961347389015e-05, "loss": 0.0445, "step": 42530 }, { "grad_norm": 0.3599296510219574, "learning_rate": 6.617396864475613e-05, "loss": 0.0459, "step": 42540 }, { "grad_norm": 0.660040020942688, "learning_rate": 6.61583220468612e-05, "loss": 0.0337, "step": 42550 }, { "grad_norm": 0.39633798599243164, "learning_rate": 6.614267368191645e-05, "loss": 0.0441, "step": 42560 }, { "grad_norm": 0.944604754447937, "learning_rate": 6.61270235516332e-05, "loss": 0.039, "step": 42570 }, { "grad_norm": 0.49399298429489136, "learning_rate": 6.611137165772287e-05, "loss": 0.0382, "step": 42580 }, { "grad_norm": 0.3906403183937073, "learning_rate": 6.60957180018972e-05, "loss": 0.0331, "step": 42590 }, { "grad_norm": 0.6167473793029785, "learning_rate": 6.608006258586797e-05, "loss": 0.0312, "step": 42600 }, { "grad_norm": 0.3475201725959778, "learning_rate": 6.606440541134731e-05, "loss": 0.0358, "step": 42610 }, { "grad_norm": 0.532035768032074, "learning_rate": 6.60487464800474e-05, "loss": 0.0367, "step": 42620 }, { "grad_norm": 0.35281211137771606, "learning_rate": 6.603308579368071e-05, "loss": 0.0345, "step": 42630 }, { "grad_norm": 0.46036025881767273, "learning_rate": 6.601742335395987e-05, "loss": 0.0408, "step": 42640 }, { "grad_norm": 0.6692774295806885, "learning_rate": 6.600175916259769e-05, "loss": 0.0448, "step": 42650 }, { "grad_norm": 0.5458018183708191, "learning_rate": 6.598609322130718e-05, "loss": 0.0641, "step": 42660 }, { "grad_norm": 0.5100794434547424, "learning_rate": 6.597042553180155e-05, "loss": 0.0412, "step": 42670 }, { "grad_norm": 0.5326998233795166, "learning_rate": 6.59547560957942e-05, "loss": 0.0346, "step": 42680 }, { "grad_norm": 0.35927116870880127, "learning_rate": 6.59390849149987e-05, "loss": 0.0505, "step": 42690 }, { "grad_norm": 0.3465171456336975, "learning_rate": 6.592341199112886e-05, "loss": 0.0383, "step": 42700 }, { "grad_norm": 1.069105863571167, "learning_rate": 6.590773732589861e-05, "loss": 0.0535, "step": 42710 }, { "grad_norm": 0.5164547562599182, "learning_rate": 6.589206092102213e-05, "loss": 0.0402, "step": 42720 }, { "grad_norm": 0.5252087116241455, "learning_rate": 6.587638277821375e-05, "loss": 0.0466, "step": 42730 }, { "grad_norm": 0.493525892496109, "learning_rate": 6.586070289918803e-05, "loss": 0.0288, "step": 42740 }, { "grad_norm": 0.5960071682929993, "learning_rate": 6.584502128565968e-05, "loss": 0.0412, "step": 42750 }, { "grad_norm": 0.4668791592121124, "learning_rate": 6.582933793934363e-05, "loss": 0.0413, "step": 42760 }, { "grad_norm": 0.5202819108963013, "learning_rate": 6.581365286195498e-05, "loss": 0.0352, "step": 42770 }, { "grad_norm": 0.5879470109939575, "learning_rate": 6.579796605520903e-05, "loss": 0.05, "step": 42780 }, { "grad_norm": 0.4599423408508301, "learning_rate": 6.578227752082127e-05, "loss": 0.0549, "step": 42790 }, { "grad_norm": 0.48058682680130005, "learning_rate": 6.576658726050735e-05, "loss": 0.0444, "step": 42800 }, { "grad_norm": 0.37817826867103577, "learning_rate": 6.575089527598316e-05, "loss": 0.0455, "step": 42810 }, { "grad_norm": 0.4614299237728119, "learning_rate": 6.573520156896476e-05, "loss": 0.0615, "step": 42820 }, { "grad_norm": 0.4605107009410858, "learning_rate": 6.571950614116835e-05, "loss": 0.0379, "step": 42830 }, { "grad_norm": 0.5078271627426147, "learning_rate": 6.570380899431039e-05, "loss": 0.0554, "step": 42840 }, { "grad_norm": 0.49722838401794434, "learning_rate": 6.568811013010749e-05, "loss": 0.0447, "step": 42850 }, { "grad_norm": 0.34518691897392273, "learning_rate": 6.567240955027645e-05, "loss": 0.0356, "step": 42860 }, { "grad_norm": 0.6144761443138123, "learning_rate": 6.565670725653427e-05, "loss": 0.0391, "step": 42870 }, { "grad_norm": 0.44064241647720337, "learning_rate": 6.564100325059812e-05, "loss": 0.0356, "step": 42880 }, { "grad_norm": 0.6878923177719116, "learning_rate": 6.562529753418538e-05, "loss": 0.0458, "step": 42890 }, { "grad_norm": 0.5259845852851868, "learning_rate": 6.56095901090136e-05, "loss": 0.0503, "step": 42900 }, { "grad_norm": 0.49163153767585754, "learning_rate": 6.55938809768005e-05, "loss": 0.0475, "step": 42910 }, { "grad_norm": 0.46708208322525024, "learning_rate": 6.557817013926402e-05, "loss": 0.0415, "step": 42920 }, { "grad_norm": 0.7728968858718872, "learning_rate": 6.556245759812227e-05, "loss": 0.0614, "step": 42930 }, { "grad_norm": 0.627383828163147, "learning_rate": 6.554674335509357e-05, "loss": 0.0492, "step": 42940 }, { "grad_norm": 0.4452661871910095, "learning_rate": 6.553102741189638e-05, "loss": 0.0339, "step": 42950 }, { "grad_norm": 0.8834837079048157, "learning_rate": 6.55153097702494e-05, "loss": 0.0451, "step": 42960 }, { "grad_norm": 0.595770537853241, "learning_rate": 6.549959043187146e-05, "loss": 0.0393, "step": 42970 }, { "grad_norm": 0.7848407626152039, "learning_rate": 6.54838693984816e-05, "loss": 0.0431, "step": 42980 }, { "grad_norm": 0.5040974617004395, "learning_rate": 6.546814667179909e-05, "loss": 0.0433, "step": 42990 }, { "grad_norm": 0.4771643280982971, "learning_rate": 6.545242225354328e-05, "loss": 0.0446, "step": 43000 }, { "grad_norm": 0.8934485912322998, "learning_rate": 6.543669614543383e-05, "loss": 0.0583, "step": 43010 }, { "grad_norm": 0.6215184330940247, "learning_rate": 6.542096834919049e-05, "loss": 0.0336, "step": 43020 }, { "grad_norm": 0.37819308042526245, "learning_rate": 6.540523886653324e-05, "loss": 0.0333, "step": 43030 }, { "grad_norm": 0.4553479254245758, "learning_rate": 6.538950769918222e-05, "loss": 0.0453, "step": 43040 }, { "grad_norm": 0.5210141539573669, "learning_rate": 6.537377484885779e-05, "loss": 0.0481, "step": 43050 }, { "grad_norm": 0.5370297431945801, "learning_rate": 6.535804031728045e-05, "loss": 0.0292, "step": 43060 }, { "grad_norm": 0.36858195066452026, "learning_rate": 6.534230410617092e-05, "loss": 0.0361, "step": 43070 }, { "grad_norm": 0.3827567994594574, "learning_rate": 6.532656621725007e-05, "loss": 0.0435, "step": 43080 }, { "grad_norm": 0.5544043779373169, "learning_rate": 6.531082665223898e-05, "loss": 0.0425, "step": 43090 }, { "grad_norm": 0.4980715811252594, "learning_rate": 6.529508541285889e-05, "loss": 0.0492, "step": 43100 }, { "grad_norm": 0.5786001682281494, "learning_rate": 6.527934250083128e-05, "loss": 0.0388, "step": 43110 }, { "grad_norm": 0.5146042108535767, "learning_rate": 6.526359791787775e-05, "loss": 0.0452, "step": 43120 }, { "grad_norm": 0.4626689851284027, "learning_rate": 6.524785166572009e-05, "loss": 0.0364, "step": 43130 }, { "grad_norm": 0.5561603307723999, "learning_rate": 6.523210374608029e-05, "loss": 0.0366, "step": 43140 }, { "grad_norm": 0.45071712136268616, "learning_rate": 6.521635416068054e-05, "loss": 0.036, "step": 43150 }, { "grad_norm": 0.38791424036026, "learning_rate": 6.520060291124317e-05, "loss": 0.0329, "step": 43160 }, { "grad_norm": 0.6015097498893738, "learning_rate": 6.518484999949072e-05, "loss": 0.0445, "step": 43170 }, { "grad_norm": 0.5520193576812744, "learning_rate": 6.516909542714591e-05, "loss": 0.0524, "step": 43180 }, { "grad_norm": 0.30831024050712585, "learning_rate": 6.515333919593161e-05, "loss": 0.0375, "step": 43190 }, { "grad_norm": 0.32627537846565247, "learning_rate": 6.513758130757094e-05, "loss": 0.0443, "step": 43200 }, { "grad_norm": 0.4423539638519287, "learning_rate": 6.512182176378713e-05, "loss": 0.0382, "step": 43210 }, { "grad_norm": 0.6923248171806335, "learning_rate": 6.510606056630362e-05, "loss": 0.064, "step": 43220 }, { "grad_norm": 0.3965635597705841, "learning_rate": 6.509029771684405e-05, "loss": 0.0424, "step": 43230 }, { "grad_norm": 0.42411720752716064, "learning_rate": 6.50745332171322e-05, "loss": 0.0334, "step": 43240 }, { "grad_norm": 0.4388883411884308, "learning_rate": 6.505876706889207e-05, "loss": 0.0463, "step": 43250 }, { "grad_norm": 0.48000404238700867, "learning_rate": 6.504299927384781e-05, "loss": 0.0338, "step": 43260 }, { "grad_norm": 0.6127128601074219, "learning_rate": 6.502722983372378e-05, "loss": 0.0421, "step": 43270 }, { "grad_norm": 0.48418766260147095, "learning_rate": 6.501145875024446e-05, "loss": 0.036, "step": 43280 }, { "grad_norm": 0.4264218509197235, "learning_rate": 6.499568602513459e-05, "loss": 0.0452, "step": 43290 }, { "grad_norm": 0.4809929132461548, "learning_rate": 6.497991166011903e-05, "loss": 0.0334, "step": 43300 }, { "grad_norm": 0.33077314496040344, "learning_rate": 6.49641356569229e-05, "loss": 0.0443, "step": 43310 }, { "grad_norm": 0.3185482621192932, "learning_rate": 6.494835801727135e-05, "loss": 0.0372, "step": 43320 }, { "grad_norm": 0.47616657614707947, "learning_rate": 6.493257874288986e-05, "loss": 0.0443, "step": 43330 }, { "grad_norm": 1.0101933479309082, "learning_rate": 6.491679783550399e-05, "loss": 0.0406, "step": 43340 }, { "grad_norm": 0.48547792434692383, "learning_rate": 6.490101529683957e-05, "loss": 0.0404, "step": 43350 }, { "grad_norm": 1.3732197284698486, "learning_rate": 6.488523112862248e-05, "loss": 0.0509, "step": 43360 }, { "grad_norm": 0.5586040616035461, "learning_rate": 6.486944533257891e-05, "loss": 0.0279, "step": 43370 }, { "grad_norm": 0.6557138562202454, "learning_rate": 6.485365791043517e-05, "loss": 0.0415, "step": 43380 }, { "grad_norm": 0.37977027893066406, "learning_rate": 6.483786886391772e-05, "loss": 0.0378, "step": 43390 }, { "grad_norm": 0.6779929399490356, "learning_rate": 6.482207819475323e-05, "loss": 0.0348, "step": 43400 }, { "grad_norm": 0.7332218289375305, "learning_rate": 6.480628590466857e-05, "loss": 0.0371, "step": 43410 }, { "grad_norm": 0.7402006387710571, "learning_rate": 6.479049199539074e-05, "loss": 0.0445, "step": 43420 }, { "grad_norm": 0.8176881074905396, "learning_rate": 6.477469646864693e-05, "loss": 0.0489, "step": 43430 }, { "grad_norm": 0.5723649859428406, "learning_rate": 6.475889932616454e-05, "loss": 0.0515, "step": 43440 }, { "grad_norm": 0.5872345566749573, "learning_rate": 6.474310056967111e-05, "loss": 0.036, "step": 43450 }, { "grad_norm": 0.646550714969635, "learning_rate": 6.472730020089437e-05, "loss": 0.0395, "step": 43460 }, { "grad_norm": 0.5592097640037537, "learning_rate": 6.471149822156222e-05, "loss": 0.0358, "step": 43470 }, { "grad_norm": 0.918241560459137, "learning_rate": 6.469569463340275e-05, "loss": 0.0405, "step": 43480 }, { "grad_norm": 0.4112730920314789, "learning_rate": 6.467988943814418e-05, "loss": 0.0366, "step": 43490 }, { "grad_norm": 0.5184336304664612, "learning_rate": 6.4664082637515e-05, "loss": 0.0393, "step": 43500 }, { "grad_norm": 0.7137280702590942, "learning_rate": 6.464827423324377e-05, "loss": 0.0333, "step": 43510 }, { "grad_norm": 0.7537974715232849, "learning_rate": 6.463246422705931e-05, "loss": 0.0641, "step": 43520 }, { "grad_norm": 0.5371395945549011, "learning_rate": 6.461665262069055e-05, "loss": 0.0518, "step": 43530 }, { "grad_norm": 0.5404565930366516, "learning_rate": 6.460083941586665e-05, "loss": 0.0374, "step": 43540 }, { "grad_norm": 0.5012213587760925, "learning_rate": 6.45850246143169e-05, "loss": 0.0509, "step": 43550 }, { "grad_norm": 0.7077435255050659, "learning_rate": 6.456920821777077e-05, "loss": 0.0743, "step": 43560 }, { "grad_norm": 0.32723042368888855, "learning_rate": 6.455339022795796e-05, "loss": 0.0333, "step": 43570 }, { "grad_norm": 0.42966124415397644, "learning_rate": 6.453757064660825e-05, "loss": 0.0373, "step": 43580 }, { "grad_norm": 0.31809064745903015, "learning_rate": 6.452174947545169e-05, "loss": 0.0355, "step": 43590 }, { "grad_norm": 0.6288097500801086, "learning_rate": 6.450592671621842e-05, "loss": 0.0655, "step": 43600 }, { "grad_norm": 0.5415629744529724, "learning_rate": 6.449010237063882e-05, "loss": 0.0372, "step": 43610 }, { "grad_norm": 0.5291100740432739, "learning_rate": 6.447427644044343e-05, "loss": 0.0438, "step": 43620 }, { "grad_norm": 0.6963885426521301, "learning_rate": 6.44584489273629e-05, "loss": 0.037, "step": 43630 }, { "grad_norm": 0.9202314019203186, "learning_rate": 6.444261983312817e-05, "loss": 0.0536, "step": 43640 }, { "grad_norm": 0.6561486721038818, "learning_rate": 6.442678915947023e-05, "loss": 0.0523, "step": 43650 }, { "grad_norm": 1.2635974884033203, "learning_rate": 6.441095690812032e-05, "loss": 0.0589, "step": 43660 }, { "grad_norm": 0.5254172682762146, "learning_rate": 6.439512308080982e-05, "loss": 0.0527, "step": 43670 }, { "grad_norm": 0.3906739056110382, "learning_rate": 6.43792876792703e-05, "loss": 0.0432, "step": 43680 }, { "grad_norm": 0.3994554579257965, "learning_rate": 6.43634507052335e-05, "loss": 0.0403, "step": 43690 }, { "grad_norm": 0.45236456394195557, "learning_rate": 6.434761216043135e-05, "loss": 0.0411, "step": 43700 }, { "grad_norm": 0.42376238107681274, "learning_rate": 6.433177204659589e-05, "loss": 0.0495, "step": 43710 }, { "grad_norm": 0.4362475275993347, "learning_rate": 6.431593036545938e-05, "loss": 0.0376, "step": 43720 }, { "grad_norm": 0.278316855430603, "learning_rate": 6.430008711875426e-05, "loss": 0.0395, "step": 43730 }, { "grad_norm": 0.6000173091888428, "learning_rate": 6.42842423082131e-05, "loss": 0.0392, "step": 43740 }, { "grad_norm": 0.540557324886322, "learning_rate": 6.426839593556868e-05, "loss": 0.0348, "step": 43750 }, { "grad_norm": 0.7583839297294617, "learning_rate": 6.425254800255392e-05, "loss": 0.0366, "step": 43760 }, { "grad_norm": 0.3931249678134918, "learning_rate": 6.423669851090196e-05, "loss": 0.0393, "step": 43770 }, { "grad_norm": 0.5056610107421875, "learning_rate": 6.422084746234604e-05, "loss": 0.0407, "step": 43780 }, { "grad_norm": 0.7000067830085754, "learning_rate": 6.420499485861965e-05, "loss": 0.0331, "step": 43790 }, { "grad_norm": 0.6762052178382874, "learning_rate": 6.418914070145636e-05, "loss": 0.0517, "step": 43800 }, { "grad_norm": 0.6063492298126221, "learning_rate": 6.417328499258996e-05, "loss": 0.0365, "step": 43810 }, { "grad_norm": 0.4058820307254791, "learning_rate": 6.415742773375446e-05, "loss": 0.0499, "step": 43820 }, { "grad_norm": 0.7079405188560486, "learning_rate": 6.414156892668393e-05, "loss": 0.0452, "step": 43830 }, { "grad_norm": 0.5530103445053101, "learning_rate": 6.412570857311267e-05, "loss": 0.0391, "step": 43840 }, { "grad_norm": 0.420765221118927, "learning_rate": 6.410984667477518e-05, "loss": 0.0461, "step": 43850 }, { "grad_norm": 0.5745368599891663, "learning_rate": 6.409398323340607e-05, "loss": 0.0306, "step": 43860 }, { "grad_norm": 0.5835544466972351, "learning_rate": 6.407811825074012e-05, "loss": 0.042, "step": 43870 }, { "grad_norm": 0.7078288793563843, "learning_rate": 6.406225172851234e-05, "loss": 0.0519, "step": 43880 }, { "grad_norm": 0.3632938265800476, "learning_rate": 6.404638366845786e-05, "loss": 0.0282, "step": 43890 }, { "grad_norm": 0.37309530377388, "learning_rate": 6.403051407231196e-05, "loss": 0.0416, "step": 43900 }, { "grad_norm": 0.5021750926971436, "learning_rate": 6.401464294181016e-05, "loss": 0.0375, "step": 43910 }, { "grad_norm": 0.4373999238014221, "learning_rate": 6.399877027868808e-05, "loss": 0.0321, "step": 43920 }, { "grad_norm": 0.5292659401893616, "learning_rate": 6.39828960846815e-05, "loss": 0.0448, "step": 43930 }, { "grad_norm": 0.39707666635513306, "learning_rate": 6.396702036152644e-05, "loss": 0.0369, "step": 43940 }, { "grad_norm": 0.5003921985626221, "learning_rate": 6.395114311095905e-05, "loss": 0.0262, "step": 43950 }, { "grad_norm": 0.8625183701515198, "learning_rate": 6.393526433471559e-05, "loss": 0.0518, "step": 43960 }, { "grad_norm": 0.8469435572624207, "learning_rate": 6.391938403453262e-05, "loss": 0.0303, "step": 43970 }, { "grad_norm": 0.5530310273170471, "learning_rate": 6.390350221214671e-05, "loss": 0.0515, "step": 43980 }, { "grad_norm": 0.6085091233253479, "learning_rate": 6.38876188692947e-05, "loss": 0.0455, "step": 43990 }, { "grad_norm": 0.47715553641319275, "learning_rate": 6.387173400771358e-05, "loss": 0.0426, "step": 44000 }, { "grad_norm": 0.6204472780227661, "learning_rate": 6.385584762914047e-05, "loss": 0.0469, "step": 44010 }, { "grad_norm": 1.7948346138000488, "learning_rate": 6.38399597353127e-05, "loss": 0.0387, "step": 44020 }, { "grad_norm": 0.4825550317764282, "learning_rate": 6.382407032796775e-05, "loss": 0.0498, "step": 44030 }, { "grad_norm": 0.5658043026924133, "learning_rate": 6.380817940884325e-05, "loss": 0.0404, "step": 44040 }, { "grad_norm": 0.48253077268600464, "learning_rate": 6.379228697967702e-05, "loss": 0.0612, "step": 44050 }, { "grad_norm": 0.430042564868927, "learning_rate": 6.3776393042207e-05, "loss": 0.0329, "step": 44060 }, { "grad_norm": 0.4645174741744995, "learning_rate": 6.376049759817137e-05, "loss": 0.0485, "step": 44070 }, { "grad_norm": 0.5250317454338074, "learning_rate": 6.37446006493084e-05, "loss": 0.0513, "step": 44080 }, { "grad_norm": 0.49667420983314514, "learning_rate": 6.372870219735656e-05, "loss": 0.0423, "step": 44090 }, { "grad_norm": 0.6319755911827087, "learning_rate": 6.371280224405452e-05, "loss": 0.0398, "step": 44100 }, { "grad_norm": 0.49237439036369324, "learning_rate": 6.369690079114101e-05, "loss": 0.032, "step": 44110 }, { "grad_norm": 0.6787002086639404, "learning_rate": 6.368099784035504e-05, "loss": 0.0382, "step": 44120 }, { "grad_norm": 0.3737016022205353, "learning_rate": 6.366509339343572e-05, "loss": 0.0396, "step": 44130 }, { "grad_norm": 0.34151989221572876, "learning_rate": 6.364918745212232e-05, "loss": 0.035, "step": 44140 }, { "grad_norm": 0.43510666489601135, "learning_rate": 6.363328001815431e-05, "loss": 0.0364, "step": 44150 }, { "grad_norm": 0.469826877117157, "learning_rate": 6.361737109327128e-05, "loss": 0.0338, "step": 44160 }, { "grad_norm": 0.5809003710746765, "learning_rate": 6.360146067921301e-05, "loss": 0.0509, "step": 44170 }, { "grad_norm": 0.4222644567489624, "learning_rate": 6.35855487777195e-05, "loss": 0.0465, "step": 44180 }, { "grad_norm": 0.4407500922679901, "learning_rate": 6.356963539053076e-05, "loss": 0.0323, "step": 44190 }, { "grad_norm": 0.43333935737609863, "learning_rate": 6.355372051938711e-05, "loss": 0.0348, "step": 44200 }, { "grad_norm": 0.732036292552948, "learning_rate": 6.353780416602894e-05, "loss": 0.0405, "step": 44210 }, { "grad_norm": 0.7895017266273499, "learning_rate": 6.352188633219689e-05, "loss": 0.0649, "step": 44220 }, { "grad_norm": 0.6632788181304932, "learning_rate": 6.350596701963166e-05, "loss": 0.0601, "step": 44230 }, { "grad_norm": 0.4980246424674988, "learning_rate": 6.349004623007419e-05, "loss": 0.044, "step": 44240 }, { "grad_norm": 0.657019853591919, "learning_rate": 6.347412396526555e-05, "loss": 0.0465, "step": 44250 }, { "grad_norm": 0.4366965591907501, "learning_rate": 6.345820022694696e-05, "loss": 0.0444, "step": 44260 }, { "grad_norm": 0.5992009043693542, "learning_rate": 6.344227501685984e-05, "loss": 0.0336, "step": 44270 }, { "grad_norm": 0.5255269408226013, "learning_rate": 6.342634833674572e-05, "loss": 0.0506, "step": 44280 }, { "grad_norm": 0.5223502516746521, "learning_rate": 6.341042018834635e-05, "loss": 0.0372, "step": 44290 }, { "grad_norm": 0.4201904237270355, "learning_rate": 6.339449057340359e-05, "loss": 0.0478, "step": 44300 }, { "grad_norm": 0.46850329637527466, "learning_rate": 6.337855949365945e-05, "loss": 0.0364, "step": 44310 }, { "grad_norm": 0.7432220578193665, "learning_rate": 6.336262695085619e-05, "loss": 0.035, "step": 44320 }, { "grad_norm": 0.535855233669281, "learning_rate": 6.334669294673612e-05, "loss": 0.0392, "step": 44330 }, { "grad_norm": 0.41116389632225037, "learning_rate": 6.333075748304179e-05, "loss": 0.0304, "step": 44340 }, { "grad_norm": 0.39463159441947937, "learning_rate": 6.331482056151585e-05, "loss": 0.0332, "step": 44350 }, { "grad_norm": 0.6140920519828796, "learning_rate": 6.329888218390117e-05, "loss": 0.0322, "step": 44360 }, { "grad_norm": 0.38431593775749207, "learning_rate": 6.328294235194072e-05, "loss": 0.0345, "step": 44370 }, { "grad_norm": 0.33497053384780884, "learning_rate": 6.326700106737765e-05, "loss": 0.0395, "step": 44380 }, { "grad_norm": 0.7762013077735901, "learning_rate": 6.325105833195531e-05, "loss": 0.0248, "step": 44390 }, { "grad_norm": 0.6603453755378723, "learning_rate": 6.323511414741715e-05, "loss": 0.0408, "step": 44400 }, { "grad_norm": 0.700968861579895, "learning_rate": 6.321916851550678e-05, "loss": 0.0451, "step": 44410 }, { "grad_norm": 0.6072859168052673, "learning_rate": 6.320322143796806e-05, "loss": 0.0375, "step": 44420 }, { "grad_norm": 0.3059663772583008, "learning_rate": 6.318727291654488e-05, "loss": 0.0388, "step": 44430 }, { "grad_norm": 0.34382864832878113, "learning_rate": 6.317132295298134e-05, "loss": 0.0378, "step": 44440 }, { "grad_norm": 0.44908133149147034, "learning_rate": 6.315537154902173e-05, "loss": 0.0299, "step": 44450 }, { "grad_norm": 0.6361789107322693, "learning_rate": 6.313941870641048e-05, "loss": 0.0339, "step": 44460 }, { "grad_norm": 0.40627291798591614, "learning_rate": 6.312346442689214e-05, "loss": 0.0452, "step": 44470 }, { "grad_norm": 0.884087324142456, "learning_rate": 6.310750871221147e-05, "loss": 0.0515, "step": 44480 }, { "grad_norm": 0.395519495010376, "learning_rate": 6.309155156411335e-05, "loss": 0.0228, "step": 44490 }, { "grad_norm": 0.5942389369010925, "learning_rate": 6.307559298434284e-05, "loss": 0.0404, "step": 44500 }, { "grad_norm": 0.5524536371231079, "learning_rate": 6.305963297464512e-05, "loss": 0.0328, "step": 44510 }, { "grad_norm": 0.5573508739471436, "learning_rate": 6.304367153676561e-05, "loss": 0.0472, "step": 44520 }, { "grad_norm": 0.7267699241638184, "learning_rate": 6.302770867244978e-05, "loss": 0.035, "step": 44530 }, { "grad_norm": 0.5200977921485901, "learning_rate": 6.301174438344328e-05, "loss": 0.0333, "step": 44540 }, { "grad_norm": 0.4391214847564697, "learning_rate": 6.299577867149202e-05, "loss": 0.0341, "step": 44550 }, { "grad_norm": 0.7035077810287476, "learning_rate": 6.297981153834192e-05, "loss": 0.0588, "step": 44560 }, { "grad_norm": 0.6335018277168274, "learning_rate": 6.296384298573916e-05, "loss": 0.0387, "step": 44570 }, { "grad_norm": 0.3254739046096802, "learning_rate": 6.294787301543001e-05, "loss": 0.0322, "step": 44580 }, { "grad_norm": 0.3814617097377777, "learning_rate": 6.293190162916095e-05, "loss": 0.04, "step": 44590 }, { "grad_norm": 0.40313902497291565, "learning_rate": 6.291592882867855e-05, "loss": 0.0319, "step": 44600 }, { "grad_norm": 0.45468762516975403, "learning_rate": 6.28999546157296e-05, "loss": 0.0425, "step": 44610 }, { "grad_norm": 0.5674170851707458, "learning_rate": 6.288397899206102e-05, "loss": 0.0323, "step": 44620 }, { "grad_norm": 0.503278911113739, "learning_rate": 6.286800195941984e-05, "loss": 0.0497, "step": 44630 }, { "grad_norm": 0.5241132378578186, "learning_rate": 6.285202351955334e-05, "loss": 0.0549, "step": 44640 }, { "grad_norm": 0.574643075466156, "learning_rate": 6.283604367420887e-05, "loss": 0.0306, "step": 44650 }, { "grad_norm": 0.47521528601646423, "learning_rate": 6.282006242513394e-05, "loss": 0.0383, "step": 44660 }, { "grad_norm": 0.5947495102882385, "learning_rate": 6.280407977407628e-05, "loss": 0.0548, "step": 44670 }, { "grad_norm": 0.28626319766044617, "learning_rate": 6.27880957227837e-05, "loss": 0.0404, "step": 44680 }, { "grad_norm": 0.8910302519798279, "learning_rate": 6.27721102730042e-05, "loss": 0.0466, "step": 44690 }, { "grad_norm": 0.6172066926956177, "learning_rate": 6.275612342648592e-05, "loss": 0.0479, "step": 44700 }, { "grad_norm": 0.6716508865356445, "learning_rate": 6.274013518497716e-05, "loss": 0.0353, "step": 44710 }, { "grad_norm": 0.4992518126964569, "learning_rate": 6.272414555022636e-05, "loss": 0.0376, "step": 44720 }, { "grad_norm": 1.4821265935897827, "learning_rate": 6.270815452398215e-05, "loss": 0.0473, "step": 44730 }, { "grad_norm": 0.5154579877853394, "learning_rate": 6.269216210799326e-05, "loss": 0.0605, "step": 44740 }, { "grad_norm": 0.341295063495636, "learning_rate": 6.26761683040086e-05, "loss": 0.0344, "step": 44750 }, { "grad_norm": 0.43455296754837036, "learning_rate": 6.266017311377723e-05, "loss": 0.0401, "step": 44760 }, { "grad_norm": 0.518577516078949, "learning_rate": 6.264417653904839e-05, "loss": 0.0331, "step": 44770 }, { "grad_norm": 0.5143943428993225, "learning_rate": 6.262817858157139e-05, "loss": 0.0371, "step": 44780 }, { "grad_norm": 0.543095588684082, "learning_rate": 6.261217924309576e-05, "loss": 0.046, "step": 44790 }, { "grad_norm": 0.6332974433898926, "learning_rate": 6.259617852537118e-05, "loss": 0.0468, "step": 44800 }, { "grad_norm": 0.7661228775978088, "learning_rate": 6.258017643014747e-05, "loss": 0.0655, "step": 44810 }, { "grad_norm": 0.46045365929603577, "learning_rate": 6.256417295917456e-05, "loss": 0.0509, "step": 44820 }, { "grad_norm": 0.8912178874015808, "learning_rate": 6.254816811420258e-05, "loss": 0.0438, "step": 44830 }, { "grad_norm": 0.9487501382827759, "learning_rate": 6.253216189698183e-05, "loss": 0.0512, "step": 44840 }, { "grad_norm": 0.4438384473323822, "learning_rate": 6.251615430926267e-05, "loss": 0.0355, "step": 44850 }, { "grad_norm": 0.5403282046318054, "learning_rate": 6.25001453527957e-05, "loss": 0.0402, "step": 44860 }, { "grad_norm": 0.39935338497161865, "learning_rate": 6.248413502933164e-05, "loss": 0.0396, "step": 44870 }, { "grad_norm": 0.5121784210205078, "learning_rate": 6.246812334062133e-05, "loss": 0.0423, "step": 44880 }, { "grad_norm": 0.4601176679134369, "learning_rate": 6.245211028841579e-05, "loss": 0.0371, "step": 44890 }, { "grad_norm": 0.4950484037399292, "learning_rate": 6.24360958744662e-05, "loss": 0.0388, "step": 44900 }, { "grad_norm": 0.46858954429626465, "learning_rate": 6.242008010052387e-05, "loss": 0.0339, "step": 44910 }, { "grad_norm": 0.47027233242988586, "learning_rate": 6.240406296834024e-05, "loss": 0.0422, "step": 44920 }, { "grad_norm": 0.7006424069404602, "learning_rate": 6.238804447966694e-05, "loss": 0.0366, "step": 44930 }, { "grad_norm": 0.4117923676967621, "learning_rate": 6.237202463625573e-05, "loss": 0.055, "step": 44940 }, { "grad_norm": 0.3877706527709961, "learning_rate": 6.235600343985848e-05, "loss": 0.037, "step": 44950 }, { "grad_norm": 0.5565269589424133, "learning_rate": 6.233998089222729e-05, "loss": 0.0436, "step": 44960 }, { "grad_norm": 0.32375863194465637, "learning_rate": 6.232395699511433e-05, "loss": 0.0419, "step": 44970 }, { "grad_norm": 0.6887921094894409, "learning_rate": 6.230793175027199e-05, "loss": 0.0381, "step": 44980 }, { "grad_norm": 0.5815576314926147, "learning_rate": 6.22919051594527e-05, "loss": 0.0428, "step": 44990 }, { "grad_norm": 0.575553297996521, "learning_rate": 6.227587722440917e-05, "loss": 0.0391, "step": 45000 }, { "grad_norm": 0.4797833561897278, "learning_rate": 6.225984794689414e-05, "loss": 0.0305, "step": 45010 }, { "grad_norm": 0.5105327367782593, "learning_rate": 6.22438173286606e-05, "loss": 0.0423, "step": 45020 }, { "grad_norm": 0.3197272717952728, "learning_rate": 6.222778537146159e-05, "loss": 0.0289, "step": 45030 }, { "grad_norm": 0.45791611075401306, "learning_rate": 6.221175207705037e-05, "loss": 0.0511, "step": 45040 }, { "grad_norm": 0.556300163269043, "learning_rate": 6.219571744718027e-05, "loss": 0.0712, "step": 45050 }, { "grad_norm": 0.5462717413902283, "learning_rate": 6.217968148360487e-05, "loss": 0.0459, "step": 45060 }, { "grad_norm": 0.42285457253456116, "learning_rate": 6.216364418807782e-05, "loss": 0.0321, "step": 45070 }, { "grad_norm": 0.5775290727615356, "learning_rate": 6.214760556235292e-05, "loss": 0.0308, "step": 45080 }, { "grad_norm": 0.45519617199897766, "learning_rate": 6.213156560818414e-05, "loss": 0.0433, "step": 45090 }, { "grad_norm": 0.7817742824554443, "learning_rate": 6.211552432732559e-05, "loss": 0.0451, "step": 45100 }, { "grad_norm": 0.4305228292942047, "learning_rate": 6.20994817215315e-05, "loss": 0.0334, "step": 45110 }, { "grad_norm": 0.6794783473014832, "learning_rate": 6.208343779255628e-05, "loss": 0.057, "step": 45120 }, { "grad_norm": 0.45609280467033386, "learning_rate": 6.206739254215449e-05, "loss": 0.0327, "step": 45130 }, { "grad_norm": 0.8795503973960876, "learning_rate": 6.205134597208077e-05, "loss": 0.038, "step": 45140 }, { "grad_norm": 0.4021459221839905, "learning_rate": 6.203529808409e-05, "loss": 0.0291, "step": 45150 }, { "grad_norm": 0.313688725233078, "learning_rate": 6.201924887993712e-05, "loss": 0.0297, "step": 45160 }, { "grad_norm": 0.5588819980621338, "learning_rate": 6.200319836137725e-05, "loss": 0.0404, "step": 45170 }, { "grad_norm": 0.39337828755378723, "learning_rate": 6.198714653016565e-05, "loss": 0.0315, "step": 45180 }, { "grad_norm": 0.40495070815086365, "learning_rate": 6.197109338805774e-05, "loss": 0.0303, "step": 45190 }, { "grad_norm": 0.7776375412940979, "learning_rate": 6.195503893680903e-05, "loss": 0.041, "step": 45200 }, { "grad_norm": 0.5941488146781921, "learning_rate": 6.193898317817524e-05, "loss": 0.0356, "step": 45210 }, { "grad_norm": 0.6553882956504822, "learning_rate": 6.192292611391221e-05, "loss": 0.0424, "step": 45220 }, { "grad_norm": 0.7319733500480652, "learning_rate": 6.190686774577591e-05, "loss": 0.0427, "step": 45230 }, { "grad_norm": 0.6818759441375732, "learning_rate": 6.189080807552245e-05, "loss": 0.0389, "step": 45240 }, { "grad_norm": 0.580646812915802, "learning_rate": 6.187474710490809e-05, "loss": 0.0335, "step": 45250 }, { "grad_norm": 0.5079045295715332, "learning_rate": 6.185868483568926e-05, "loss": 0.0441, "step": 45260 }, { "grad_norm": 0.6137645244598389, "learning_rate": 6.184262126962245e-05, "loss": 0.0493, "step": 45270 }, { "grad_norm": 0.5630025863647461, "learning_rate": 6.182655640846442e-05, "loss": 0.0442, "step": 45280 }, { "grad_norm": 0.3287871778011322, "learning_rate": 6.181049025397196e-05, "loss": 0.0339, "step": 45290 }, { "grad_norm": 0.5058879852294922, "learning_rate": 6.179442280790202e-05, "loss": 0.037, "step": 45300 }, { "grad_norm": 0.876714825630188, "learning_rate": 6.177835407201174e-05, "loss": 0.0326, "step": 45310 }, { "grad_norm": 0.688592255115509, "learning_rate": 6.176228404805839e-05, "loss": 0.0478, "step": 45320 }, { "grad_norm": 0.5278666615486145, "learning_rate": 6.174621273779932e-05, "loss": 0.0371, "step": 45330 }, { "grad_norm": 0.5694205164909363, "learning_rate": 6.17301401429921e-05, "loss": 0.0431, "step": 45340 }, { "grad_norm": 0.6761003732681274, "learning_rate": 6.171406626539436e-05, "loss": 0.0413, "step": 45350 }, { "grad_norm": 0.8174397349357605, "learning_rate": 6.169799110676398e-05, "loss": 0.0528, "step": 45360 }, { "grad_norm": 0.6986231207847595, "learning_rate": 6.168191466885885e-05, "loss": 0.0481, "step": 45370 }, { "grad_norm": 0.9935580492019653, "learning_rate": 6.166583695343711e-05, "loss": 0.0274, "step": 45380 }, { "grad_norm": 0.5118404626846313, "learning_rate": 6.164975796225698e-05, "loss": 0.0332, "step": 45390 }, { "grad_norm": 0.7913576364517212, "learning_rate": 6.163367769707683e-05, "loss": 0.0436, "step": 45400 }, { "grad_norm": 0.42860814929008484, "learning_rate": 6.161759615965519e-05, "loss": 0.0421, "step": 45410 }, { "grad_norm": 0.30694714188575745, "learning_rate": 6.16015133517507e-05, "loss": 0.0636, "step": 45420 }, { "grad_norm": 0.43803250789642334, "learning_rate": 6.158542927512214e-05, "loss": 0.0323, "step": 45430 }, { "grad_norm": 0.4599696695804596, "learning_rate": 6.156934393152846e-05, "loss": 0.0361, "step": 45440 }, { "grad_norm": 0.729347825050354, "learning_rate": 6.15532573227287e-05, "loss": 0.0405, "step": 45450 }, { "grad_norm": 0.43523454666137695, "learning_rate": 6.153716945048212e-05, "loss": 0.0323, "step": 45460 }, { "grad_norm": 0.4899122416973114, "learning_rate": 6.152108031654802e-05, "loss": 0.0388, "step": 45470 }, { "grad_norm": 0.555853545665741, "learning_rate": 6.15049899226859e-05, "loss": 0.0422, "step": 45480 }, { "grad_norm": 0.4887179434299469, "learning_rate": 6.148889827065537e-05, "loss": 0.0444, "step": 45490 }, { "grad_norm": 0.5466554164886475, "learning_rate": 6.147280536221622e-05, "loss": 0.0537, "step": 45500 }, { "grad_norm": 0.6493180990219116, "learning_rate": 6.145671119912832e-05, "loss": 0.0487, "step": 45510 }, { "grad_norm": 0.592431902885437, "learning_rate": 6.144061578315169e-05, "loss": 0.0426, "step": 45520 }, { "grad_norm": 0.4645792841911316, "learning_rate": 6.142451911604654e-05, "loss": 0.0378, "step": 45530 }, { "grad_norm": 0.4016173481941223, "learning_rate": 6.140842119957315e-05, "loss": 0.0398, "step": 45540 }, { "grad_norm": 0.5377594232559204, "learning_rate": 6.139232203549197e-05, "loss": 0.0344, "step": 45550 }, { "grad_norm": 0.6121097803115845, "learning_rate": 6.137622162556357e-05, "loss": 0.0408, "step": 45560 }, { "grad_norm": 0.4650006890296936, "learning_rate": 6.13601199715487e-05, "loss": 0.0302, "step": 45570 }, { "grad_norm": 0.505370020866394, "learning_rate": 6.134401707520816e-05, "loss": 0.0341, "step": 45580 }, { "grad_norm": 0.6292383670806885, "learning_rate": 6.132791293830298e-05, "loss": 0.0356, "step": 45590 }, { "grad_norm": 0.339362233877182, "learning_rate": 6.131180756259428e-05, "loss": 0.0307, "step": 45600 }, { "grad_norm": 0.5671970248222351, "learning_rate": 6.129570094984331e-05, "loss": 0.0404, "step": 45610 }, { "grad_norm": 1.2641322612762451, "learning_rate": 6.127959310181145e-05, "loss": 0.061, "step": 45620 }, { "grad_norm": 0.5820739269256592, "learning_rate": 6.126348402026026e-05, "loss": 0.0525, "step": 45630 }, { "grad_norm": 0.35724982619285583, "learning_rate": 6.12473737069514e-05, "loss": 0.0355, "step": 45640 }, { "grad_norm": 0.43037843704223633, "learning_rate": 6.123126216364665e-05, "loss": 0.0283, "step": 45650 }, { "grad_norm": 0.3537924587726593, "learning_rate": 6.121514939210797e-05, "loss": 0.0335, "step": 45660 }, { "grad_norm": 0.8109027743339539, "learning_rate": 6.119903539409741e-05, "loss": 0.0421, "step": 45670 }, { "grad_norm": 0.8114076256752014, "learning_rate": 6.118292017137716e-05, "loss": 0.0419, "step": 45680 }, { "grad_norm": 0.47734159231185913, "learning_rate": 6.116680372570959e-05, "loss": 0.0542, "step": 45690 }, { "grad_norm": 0.5411100387573242, "learning_rate": 6.115068605885713e-05, "loss": 0.0478, "step": 45700 }, { "grad_norm": 0.3976271450519562, "learning_rate": 6.113456717258243e-05, "loss": 0.0401, "step": 45710 }, { "grad_norm": 0.7009553909301758, "learning_rate": 6.11184470686482e-05, "loss": 0.0323, "step": 45720 }, { "grad_norm": 0.606822669506073, "learning_rate": 6.110232574881731e-05, "loss": 0.0303, "step": 45730 }, { "grad_norm": 0.7511770129203796, "learning_rate": 6.108620321485277e-05, "loss": 0.0364, "step": 45740 }, { "grad_norm": 0.593625545501709, "learning_rate": 6.107007946851773e-05, "loss": 0.0451, "step": 45750 }, { "grad_norm": 0.3704650104045868, "learning_rate": 6.105395451157542e-05, "loss": 0.051, "step": 45760 }, { "grad_norm": 0.5778931379318237, "learning_rate": 6.103782834578928e-05, "loss": 0.0433, "step": 45770 }, { "grad_norm": 0.40039992332458496, "learning_rate": 6.102170097292281e-05, "loss": 0.0326, "step": 45780 }, { "grad_norm": 0.5000478029251099, "learning_rate": 6.1005572394739716e-05, "loss": 0.0585, "step": 45790 }, { "grad_norm": 0.6582210063934326, "learning_rate": 6.0989442613003765e-05, "loss": 0.0431, "step": 45800 }, { "grad_norm": 0.674427330493927, "learning_rate": 6.097331162947888e-05, "loss": 0.0473, "step": 45810 }, { "grad_norm": 0.47627487778663635, "learning_rate": 6.095717944592914e-05, "loss": 0.0372, "step": 45820 }, { "grad_norm": 0.3391026556491852, "learning_rate": 6.094104606411873e-05, "loss": 0.0454, "step": 45830 }, { "grad_norm": 0.5090001821517944, "learning_rate": 6.0924911485811966e-05, "loss": 0.0443, "step": 45840 }, { "grad_norm": 0.4289797246456146, "learning_rate": 6.090877571277331e-05, "loss": 0.0437, "step": 45850 }, { "grad_norm": 0.578790545463562, "learning_rate": 6.089263874676736e-05, "loss": 0.0429, "step": 45860 }, { "grad_norm": 0.6239549517631531, "learning_rate": 6.0876500589558796e-05, "loss": 0.0453, "step": 45870 }, { "grad_norm": 0.5183854103088379, "learning_rate": 6.086036124291248e-05, "loss": 0.0317, "step": 45880 }, { "grad_norm": 0.3528290390968323, "learning_rate": 6.084422070859339e-05, "loss": 0.0403, "step": 45890 }, { "grad_norm": 0.48850902915000916, "learning_rate": 6.082807898836663e-05, "loss": 0.0598, "step": 45900 }, { "grad_norm": 0.7016517519950867, "learning_rate": 6.081193608399742e-05, "loss": 0.0401, "step": 45910 }, { "grad_norm": 0.4472510516643524, "learning_rate": 6.0795791997251164e-05, "loss": 0.0369, "step": 45920 }, { "grad_norm": 0.6063622236251831, "learning_rate": 6.0779646729893294e-05, "loss": 0.0413, "step": 45930 }, { "grad_norm": 0.4850773215293884, "learning_rate": 6.0763500283689476e-05, "loss": 0.033, "step": 45940 }, { "grad_norm": 0.3916773498058319, "learning_rate": 6.0747352660405455e-05, "loss": 0.0286, "step": 45950 }, { "grad_norm": 0.5814676880836487, "learning_rate": 6.073120386180709e-05, "loss": 0.0404, "step": 45960 }, { "grad_norm": 0.5111261606216431, "learning_rate": 6.0715053889660425e-05, "loss": 0.0335, "step": 45970 }, { "grad_norm": 0.5178468823432922, "learning_rate": 6.069890274573157e-05, "loss": 0.0403, "step": 45980 }, { "grad_norm": 0.4388749897480011, "learning_rate": 6.068275043178679e-05, "loss": 0.0302, "step": 45990 }, { "grad_norm": 0.6580687165260315, "learning_rate": 6.066659694959248e-05, "loss": 0.0354, "step": 46000 }, { "grad_norm": 0.43187397718429565, "learning_rate": 6.0650442300915176e-05, "loss": 0.0384, "step": 46010 }, { "grad_norm": 0.5052251815795898, "learning_rate": 6.063428648752152e-05, "loss": 0.0431, "step": 46020 }, { "grad_norm": 0.4855790138244629, "learning_rate": 6.0618129511178266e-05, "loss": 0.0407, "step": 46030 }, { "grad_norm": 0.3905932307243347, "learning_rate": 6.060197137365234e-05, "loss": 0.028, "step": 46040 }, { "grad_norm": 0.5362650156021118, "learning_rate": 6.058581207671077e-05, "loss": 0.0452, "step": 46050 }, { "grad_norm": 0.4570986032485962, "learning_rate": 6.056965162212072e-05, "loss": 0.0403, "step": 46060 }, { "grad_norm": 1.4559646844863892, "learning_rate": 6.055349001164943e-05, "loss": 0.0497, "step": 46070 }, { "grad_norm": 0.6159012913703918, "learning_rate": 6.0537327247064347e-05, "loss": 0.0481, "step": 46080 }, { "grad_norm": 0.29853761196136475, "learning_rate": 6.0521163330133014e-05, "loss": 0.0592, "step": 46090 }, { "grad_norm": 0.7411194443702698, "learning_rate": 6.050499826262306e-05, "loss": 0.0452, "step": 46100 }, { "grad_norm": 0.8844083547592163, "learning_rate": 6.0488832046302294e-05, "loss": 0.0531, "step": 46110 }, { "grad_norm": 0.618937075138092, "learning_rate": 6.0472664682938626e-05, "loss": 0.0366, "step": 46120 }, { "grad_norm": 0.5124964118003845, "learning_rate": 6.045649617430009e-05, "loss": 0.0307, "step": 46130 }, { "grad_norm": 0.5809561014175415, "learning_rate": 6.0440326522154866e-05, "loss": 0.0395, "step": 46140 }, { "grad_norm": 0.7254514694213867, "learning_rate": 6.0424155728271224e-05, "loss": 0.0432, "step": 46150 }, { "grad_norm": 0.4292505383491516, "learning_rate": 6.040798379441758e-05, "loss": 0.0287, "step": 46160 }, { "grad_norm": 0.5720456838607788, "learning_rate": 6.0391810722362485e-05, "loss": 0.0411, "step": 46170 }, { "grad_norm": 0.3410074710845947, "learning_rate": 6.037563651387458e-05, "loss": 0.0286, "step": 46180 }, { "grad_norm": 0.4878728985786438, "learning_rate": 6.0359461170722666e-05, "loss": 0.0534, "step": 46190 }, { "grad_norm": 0.49770936369895935, "learning_rate": 6.034328469467566e-05, "loss": 0.0368, "step": 46200 }, { "grad_norm": 0.7295461297035217, "learning_rate": 6.0327107087502596e-05, "loss": 0.0664, "step": 46210 }, { "grad_norm": 0.42270177602767944, "learning_rate": 6.031092835097262e-05, "loss": 0.0317, "step": 46220 }, { "grad_norm": 0.4473024308681488, "learning_rate": 6.0294748486855024e-05, "loss": 0.0412, "step": 46230 }, { "grad_norm": 0.5114226937294006, "learning_rate": 6.0278567496919216e-05, "loss": 0.039, "step": 46240 }, { "grad_norm": 0.6414662599563599, "learning_rate": 6.026238538293472e-05, "loss": 0.0403, "step": 46250 }, { "grad_norm": 0.47857123613357544, "learning_rate": 6.024620214667118e-05, "loss": 0.0318, "step": 46260 }, { "grad_norm": 0.6363354325294495, "learning_rate": 6.0230017789898384e-05, "loss": 0.0305, "step": 46270 }, { "grad_norm": 0.4637919068336487, "learning_rate": 6.021383231438622e-05, "loss": 0.0345, "step": 46280 }, { "grad_norm": 0.8478460907936096, "learning_rate": 6.0197645721904704e-05, "loss": 0.0673, "step": 46290 }, { "grad_norm": 0.5177402496337891, "learning_rate": 6.0181458014224e-05, "loss": 0.0302, "step": 46300 }, { "grad_norm": 0.451823890209198, "learning_rate": 6.016526919311435e-05, "loss": 0.0356, "step": 46310 }, { "grad_norm": 0.6658415198326111, "learning_rate": 6.014907926034613e-05, "loss": 0.0357, "step": 46320 }, { "grad_norm": 0.32162734866142273, "learning_rate": 6.0132888217689875e-05, "loss": 0.0275, "step": 46330 }, { "grad_norm": 0.3204077184200287, "learning_rate": 6.01166960669162e-05, "loss": 0.0441, "step": 46340 }, { "grad_norm": 0.44344180822372437, "learning_rate": 6.0100502809795845e-05, "loss": 0.0271, "step": 46350 }, { "grad_norm": 0.5713291764259338, "learning_rate": 6.00843084480997e-05, "loss": 0.0317, "step": 46360 }, { "grad_norm": 0.3682006597518921, "learning_rate": 6.0068112983598736e-05, "loss": 0.0477, "step": 46370 }, { "grad_norm": 1.4604438543319702, "learning_rate": 6.005191641806407e-05, "loss": 0.0453, "step": 46380 }, { "grad_norm": 0.3193345069885254, "learning_rate": 6.003571875326694e-05, "loss": 0.0313, "step": 46390 }, { "grad_norm": 0.40302589535713196, "learning_rate": 6.00195199909787e-05, "loss": 0.041, "step": 46400 }, { "grad_norm": 0.39835211634635925, "learning_rate": 6.000332013297082e-05, "loss": 0.0374, "step": 46410 }, { "grad_norm": 0.4394910931587219, "learning_rate": 5.998711918101487e-05, "loss": 0.0374, "step": 46420 }, { "grad_norm": 0.5165179371833801, "learning_rate": 5.997091713688261e-05, "loss": 0.0349, "step": 46430 }, { "grad_norm": 0.696989893913269, "learning_rate": 5.9954714002345836e-05, "loss": 0.0429, "step": 46440 }, { "grad_norm": 1.0708011388778687, "learning_rate": 5.993850977917649e-05, "loss": 0.0438, "step": 46450 }, { "grad_norm": 0.43379727005958557, "learning_rate": 5.992230446914667e-05, "loss": 0.044, "step": 46460 }, { "grad_norm": 0.5035563111305237, "learning_rate": 5.990609807402855e-05, "loss": 0.0443, "step": 46470 }, { "grad_norm": 0.7139365077018738, "learning_rate": 5.988989059559443e-05, "loss": 0.0388, "step": 46480 }, { "grad_norm": 0.49217939376831055, "learning_rate": 5.987368203561675e-05, "loss": 0.0461, "step": 46490 }, { "grad_norm": 0.6869942545890808, "learning_rate": 5.9857472395868055e-05, "loss": 0.0295, "step": 46500 }, { "grad_norm": 0.9041708707809448, "learning_rate": 5.9841261678120983e-05, "loss": 0.0316, "step": 46510 }, { "grad_norm": 0.4318874776363373, "learning_rate": 5.982504988414834e-05, "loss": 0.0308, "step": 46520 }, { "grad_norm": 0.5745075941085815, "learning_rate": 5.9808837015723015e-05, "loss": 0.0314, "step": 46530 }, { "grad_norm": 0.9014844298362732, "learning_rate": 5.9792623074618016e-05, "loss": 0.0431, "step": 46540 }, { "grad_norm": 0.6289086937904358, "learning_rate": 5.977640806260648e-05, "loss": 0.0354, "step": 46550 }, { "grad_norm": 0.7405497431755066, "learning_rate": 5.976019198146166e-05, "loss": 0.0384, "step": 46560 }, { "grad_norm": 0.78238445520401, "learning_rate": 5.974397483295692e-05, "loss": 0.0556, "step": 46570 }, { "grad_norm": 0.7057822942733765, "learning_rate": 5.972775661886575e-05, "loss": 0.0348, "step": 46580 }, { "grad_norm": 0.38331401348114014, "learning_rate": 5.971153734096173e-05, "loss": 0.0363, "step": 46590 }, { "grad_norm": 0.5274115204811096, "learning_rate": 5.969531700101859e-05, "loss": 0.046, "step": 46600 }, { "grad_norm": 0.4827587604522705, "learning_rate": 5.9679095600810155e-05, "loss": 0.0475, "step": 46610 }, { "grad_norm": 0.4659257233142853, "learning_rate": 5.9662873142110384e-05, "loss": 0.0453, "step": 46620 }, { "grad_norm": 0.4485527276992798, "learning_rate": 5.964664962669333e-05, "loss": 0.0329, "step": 46630 }, { "grad_norm": 0.7033471465110779, "learning_rate": 5.9630425056333186e-05, "loss": 0.0362, "step": 46640 }, { "grad_norm": 0.5655105113983154, "learning_rate": 5.961419943280422e-05, "loss": 0.0343, "step": 46650 }, { "grad_norm": 0.42667949199676514, "learning_rate": 5.959797275788087e-05, "loss": 0.048, "step": 46660 }, { "grad_norm": 0.5615009069442749, "learning_rate": 5.958174503333765e-05, "loss": 0.0435, "step": 46670 }, { "grad_norm": 0.33480679988861084, "learning_rate": 5.9565516260949195e-05, "loss": 0.0433, "step": 46680 }, { "grad_norm": 0.7375448942184448, "learning_rate": 5.9549286442490273e-05, "loss": 0.0318, "step": 46690 }, { "grad_norm": 0.30541718006134033, "learning_rate": 5.953305557973572e-05, "loss": 0.0468, "step": 46700 }, { "grad_norm": 0.3404403328895569, "learning_rate": 5.9516823674460564e-05, "loss": 0.0296, "step": 46710 }, { "grad_norm": 0.5769661068916321, "learning_rate": 5.9500590728439875e-05, "loss": 0.0431, "step": 46720 }, { "grad_norm": 0.5672650337219238, "learning_rate": 5.948435674344886e-05, "loss": 0.047, "step": 46730 }, { "grad_norm": 0.43507125973701477, "learning_rate": 5.946812172126285e-05, "loss": 0.0314, "step": 46740 }, { "grad_norm": 0.38744980096817017, "learning_rate": 5.945188566365729e-05, "loss": 0.0415, "step": 46750 }, { "grad_norm": 0.5270951986312866, "learning_rate": 5.943564857240773e-05, "loss": 0.0555, "step": 46760 }, { "grad_norm": 0.6422924399375916, "learning_rate": 5.941941044928981e-05, "loss": 0.0319, "step": 46770 }, { "grad_norm": 0.5851060748100281, "learning_rate": 5.940317129607935e-05, "loss": 0.0591, "step": 46780 }, { "grad_norm": 0.8330283761024475, "learning_rate": 5.9386931114552204e-05, "loss": 0.0381, "step": 46790 }, { "grad_norm": 0.2838006615638733, "learning_rate": 5.937068990648438e-05, "loss": 0.0532, "step": 46800 }, { "grad_norm": 0.6139370203018188, "learning_rate": 5.935444767365199e-05, "loss": 0.0418, "step": 46810 }, { "grad_norm": 0.44011542201042175, "learning_rate": 5.933820441783129e-05, "loss": 0.035, "step": 46820 }, { "grad_norm": 0.3986303210258484, "learning_rate": 5.932196014079857e-05, "loss": 0.0313, "step": 46830 }, { "grad_norm": 0.23853115737438202, "learning_rate": 5.930571484433032e-05, "loss": 0.0298, "step": 46840 }, { "grad_norm": 0.6165457963943481, "learning_rate": 5.928946853020309e-05, "loss": 0.0378, "step": 46850 }, { "grad_norm": 0.4404699504375458, "learning_rate": 5.927322120019352e-05, "loss": 0.0332, "step": 46860 }, { "grad_norm": 0.3191690742969513, "learning_rate": 5.925697285607844e-05, "loss": 0.0293, "step": 46870 }, { "grad_norm": 0.5101690888404846, "learning_rate": 5.924072349963472e-05, "loss": 0.0306, "step": 46880 }, { "grad_norm": 0.6836748123168945, "learning_rate": 5.922447313263938e-05, "loss": 0.0361, "step": 46890 }, { "grad_norm": 0.487621933221817, "learning_rate": 5.920822175686952e-05, "loss": 0.0526, "step": 46900 }, { "grad_norm": 0.6266036629676819, "learning_rate": 5.919196937410239e-05, "loss": 0.0332, "step": 46910 }, { "grad_norm": 0.6454750299453735, "learning_rate": 5.917571598611529e-05, "loss": 0.0393, "step": 46920 }, { "grad_norm": 1.173021674156189, "learning_rate": 5.9159461594685686e-05, "loss": 0.0436, "step": 46930 }, { "grad_norm": 0.7111793756484985, "learning_rate": 5.914320620159114e-05, "loss": 0.0379, "step": 46940 }, { "grad_norm": 0.548673689365387, "learning_rate": 5.912694980860931e-05, "loss": 0.0431, "step": 46950 }, { "grad_norm": 0.35136866569519043, "learning_rate": 5.9110692417517964e-05, "loss": 0.03, "step": 46960 }, { "grad_norm": 0.5937889814376831, "learning_rate": 5.9094434030095e-05, "loss": 0.0413, "step": 46970 }, { "grad_norm": 0.6154346466064453, "learning_rate": 5.907817464811841e-05, "loss": 0.0451, "step": 46980 }, { "grad_norm": 0.5365429520606995, "learning_rate": 5.906191427336627e-05, "loss": 0.0353, "step": 46990 }, { "grad_norm": 0.4466266632080078, "learning_rate": 5.9045652907616835e-05, "loss": 0.0715, "step": 47000 }, { "grad_norm": 0.34026235342025757, "learning_rate": 5.902939055264838e-05, "loss": 0.0406, "step": 47010 }, { "grad_norm": 0.3602653443813324, "learning_rate": 5.901312721023935e-05, "loss": 0.0359, "step": 47020 }, { "grad_norm": 0.6471030116081238, "learning_rate": 5.8996862882168294e-05, "loss": 0.0726, "step": 47030 }, { "grad_norm": 0.8775890469551086, "learning_rate": 5.8980597570213826e-05, "loss": 0.0649, "step": 47040 }, { "grad_norm": 0.6931150555610657, "learning_rate": 5.896433127615471e-05, "loss": 0.04, "step": 47050 }, { "grad_norm": 1.3234245777130127, "learning_rate": 5.894806400176981e-05, "loss": 0.0358, "step": 47060 }, { "grad_norm": 0.5407373309135437, "learning_rate": 5.893179574883808e-05, "loss": 0.0459, "step": 47070 }, { "grad_norm": 0.454738587141037, "learning_rate": 5.8915526519138585e-05, "loss": 0.0406, "step": 47080 }, { "grad_norm": 0.5812658667564392, "learning_rate": 5.889925631445053e-05, "loss": 0.0355, "step": 47090 }, { "grad_norm": 0.6418092250823975, "learning_rate": 5.888298513655318e-05, "loss": 0.0345, "step": 47100 }, { "grad_norm": 0.36239656805992126, "learning_rate": 5.886671298722592e-05, "loss": 0.0343, "step": 47110 }, { "grad_norm": 0.43646514415740967, "learning_rate": 5.885043986824828e-05, "loss": 0.0365, "step": 47120 }, { "grad_norm": 0.497951477766037, "learning_rate": 5.883416578139982e-05, "loss": 0.0392, "step": 47130 }, { "grad_norm": 0.4810960590839386, "learning_rate": 5.881789072846029e-05, "loss": 0.0313, "step": 47140 }, { "grad_norm": 0.5305639505386353, "learning_rate": 5.8801614711209474e-05, "loss": 0.0304, "step": 47150 }, { "grad_norm": 0.6393002271652222, "learning_rate": 5.878533773142732e-05, "loss": 0.0392, "step": 47160 }, { "grad_norm": 0.3152593672275543, "learning_rate": 5.8769059790893834e-05, "loss": 0.0353, "step": 47170 }, { "grad_norm": 1.259609341621399, "learning_rate": 5.8752780891389145e-05, "loss": 0.0545, "step": 47180 }, { "grad_norm": 0.32524019479751587, "learning_rate": 5.873650103469351e-05, "loss": 0.0354, "step": 47190 }, { "grad_norm": 0.44509902596473694, "learning_rate": 5.872022022258725e-05, "loss": 0.0341, "step": 47200 }, { "grad_norm": 0.6280832886695862, "learning_rate": 5.8703938456850814e-05, "loss": 0.0492, "step": 47210 }, { "grad_norm": 0.3168761432170868, "learning_rate": 5.868765573926477e-05, "loss": 0.0289, "step": 47220 }, { "grad_norm": 0.39110425114631653, "learning_rate": 5.867137207160974e-05, "loss": 0.0315, "step": 47230 }, { "grad_norm": 0.47348299622535706, "learning_rate": 5.86550874556665e-05, "loss": 0.0308, "step": 47240 }, { "grad_norm": 0.7762520909309387, "learning_rate": 5.863880189321592e-05, "loss": 0.0283, "step": 47250 }, { "grad_norm": 0.36883190274238586, "learning_rate": 5.862251538603894e-05, "loss": 0.0348, "step": 47260 }, { "grad_norm": 0.3376723825931549, "learning_rate": 5.8606227935916656e-05, "loss": 0.0338, "step": 47270 }, { "grad_norm": 0.4752686023712158, "learning_rate": 5.858993954463021e-05, "loss": 0.0548, "step": 47280 }, { "grad_norm": 0.5922755002975464, "learning_rate": 5.8573650213960896e-05, "loss": 0.028, "step": 47290 }, { "grad_norm": 0.5161023139953613, "learning_rate": 5.8557359945690084e-05, "loss": 0.0413, "step": 47300 }, { "grad_norm": 0.8777897357940674, "learning_rate": 5.8541068741599246e-05, "loss": 0.0386, "step": 47310 }, { "grad_norm": 0.5122294425964355, "learning_rate": 5.8524776603469985e-05, "loss": 0.0345, "step": 47320 }, { "grad_norm": 0.4807484447956085, "learning_rate": 5.850848353308397e-05, "loss": 0.0286, "step": 47330 }, { "grad_norm": 0.3327494263648987, "learning_rate": 5.849218953222297e-05, "loss": 0.0393, "step": 47340 }, { "grad_norm": 0.554995596408844, "learning_rate": 5.847589460266891e-05, "loss": 0.0461, "step": 47350 }, { "grad_norm": 0.4586021602153778, "learning_rate": 5.845959874620375e-05, "loss": 0.0404, "step": 47360 }, { "grad_norm": 0.4453204572200775, "learning_rate": 5.84433019646096e-05, "loss": 0.0353, "step": 47370 }, { "grad_norm": 0.4367438554763794, "learning_rate": 5.842700425966863e-05, "loss": 0.0275, "step": 47380 }, { "grad_norm": 0.3615090548992157, "learning_rate": 5.841070563316315e-05, "loss": 0.0434, "step": 47390 }, { "grad_norm": 0.44654330611228943, "learning_rate": 5.839440608687554e-05, "loss": 0.0344, "step": 47400 }, { "grad_norm": 0.41758185625076294, "learning_rate": 5.837810562258831e-05, "loss": 0.0275, "step": 47410 }, { "grad_norm": 0.44572675228118896, "learning_rate": 5.836180424208405e-05, "loss": 0.0357, "step": 47420 }, { "grad_norm": 0.5052928328514099, "learning_rate": 5.834550194714543e-05, "loss": 0.0354, "step": 47430 }, { "grad_norm": 0.733173668384552, "learning_rate": 5.832919873955528e-05, "loss": 0.0314, "step": 47440 }, { "grad_norm": 0.4120323956012726, "learning_rate": 5.831289462109648e-05, "loss": 0.0378, "step": 47450 }, { "grad_norm": 0.7593982815742493, "learning_rate": 5.829658959355202e-05, "loss": 0.0535, "step": 47460 }, { "grad_norm": 0.6753876805305481, "learning_rate": 5.8280283658704995e-05, "loss": 0.0313, "step": 47470 }, { "grad_norm": 0.7134599685668945, "learning_rate": 5.826397681833859e-05, "loss": 0.0355, "step": 47480 }, { "grad_norm": 0.7200716733932495, "learning_rate": 5.824766907423612e-05, "loss": 0.0454, "step": 47490 }, { "grad_norm": 0.43732112646102905, "learning_rate": 5.8231360428180945e-05, "loss": 0.0356, "step": 47500 }, { "grad_norm": 0.3907637596130371, "learning_rate": 5.821505088195658e-05, "loss": 0.0371, "step": 47510 }, { "grad_norm": 0.5073890089988708, "learning_rate": 5.819874043734661e-05, "loss": 0.0418, "step": 47520 }, { "grad_norm": 1.0128214359283447, "learning_rate": 5.8182429096134695e-05, "loss": 0.0352, "step": 47530 }, { "grad_norm": 0.4936371445655823, "learning_rate": 5.816611686010465e-05, "loss": 0.0331, "step": 47540 }, { "grad_norm": 0.5994740128517151, "learning_rate": 5.814980373104033e-05, "loss": 0.0366, "step": 47550 }, { "grad_norm": 0.48228809237480164, "learning_rate": 5.813348971072572e-05, "loss": 0.0444, "step": 47560 }, { "grad_norm": 0.46252313256263733, "learning_rate": 5.811717480094492e-05, "loss": 0.034, "step": 47570 }, { "grad_norm": 0.8946103453636169, "learning_rate": 5.810085900348209e-05, "loss": 0.0307, "step": 47580 }, { "grad_norm": 0.5054081082344055, "learning_rate": 5.8084542320121483e-05, "loss": 0.0411, "step": 47590 }, { "grad_norm": 0.4133581519126892, "learning_rate": 5.8068224752647497e-05, "loss": 0.0352, "step": 47600 }, { "grad_norm": 0.5700311660766602, "learning_rate": 5.805190630284457e-05, "loss": 0.0326, "step": 47610 }, { "grad_norm": 0.4083300232887268, "learning_rate": 5.803558697249729e-05, "loss": 0.041, "step": 47620 }, { "grad_norm": 0.4083648920059204, "learning_rate": 5.8019266763390276e-05, "loss": 0.0285, "step": 47630 }, { "grad_norm": 0.4414835274219513, "learning_rate": 5.800294567730833e-05, "loss": 0.0363, "step": 47640 }, { "grad_norm": 0.4444383680820465, "learning_rate": 5.798662371603628e-05, "loss": 0.0272, "step": 47650 }, { "grad_norm": 0.556306779384613, "learning_rate": 5.797030088135904e-05, "loss": 0.0353, "step": 47660 }, { "grad_norm": 0.923940122127533, "learning_rate": 5.7953977175061705e-05, "loss": 0.0331, "step": 47670 }, { "grad_norm": 0.5738679766654968, "learning_rate": 5.793765259892938e-05, "loss": 0.0399, "step": 47680 }, { "grad_norm": 0.45419201254844666, "learning_rate": 5.792132715474729e-05, "loss": 0.0284, "step": 47690 }, { "grad_norm": 0.8528549671173096, "learning_rate": 5.790500084430078e-05, "loss": 0.023, "step": 47700 }, { "grad_norm": 0.6205198168754578, "learning_rate": 5.7888673669375274e-05, "loss": 0.0302, "step": 47710 }, { "grad_norm": 0.814326822757721, "learning_rate": 5.787234563175625e-05, "loss": 0.0351, "step": 47720 }, { "grad_norm": 0.858762264251709, "learning_rate": 5.7856016733229355e-05, "loss": 0.0392, "step": 47730 }, { "grad_norm": 0.5093608498573303, "learning_rate": 5.7839686975580297e-05, "loss": 0.0325, "step": 47740 }, { "grad_norm": 0.5719572901725769, "learning_rate": 5.782335636059484e-05, "loss": 0.0311, "step": 47750 }, { "grad_norm": 0.5367909073829651, "learning_rate": 5.780702489005889e-05, "loss": 0.0244, "step": 47760 }, { "grad_norm": 0.29245108366012573, "learning_rate": 5.779069256575845e-05, "loss": 0.0249, "step": 47770 }, { "grad_norm": 0.3579544126987457, "learning_rate": 5.7774359389479574e-05, "loss": 0.0529, "step": 47780 }, { "grad_norm": 0.5003787279129028, "learning_rate": 5.775802536300845e-05, "loss": 0.0291, "step": 47790 }, { "grad_norm": 0.7387018203735352, "learning_rate": 5.774169048813134e-05, "loss": 0.0321, "step": 47800 }, { "grad_norm": 0.5356366038322449, "learning_rate": 5.77253547666346e-05, "loss": 0.0328, "step": 47810 }, { "grad_norm": 0.4286780059337616, "learning_rate": 5.770901820030465e-05, "loss": 0.0326, "step": 47820 }, { "grad_norm": 0.408565878868103, "learning_rate": 5.769268079092809e-05, "loss": 0.0388, "step": 47830 }, { "grad_norm": 0.35753288865089417, "learning_rate": 5.767634254029151e-05, "loss": 0.0534, "step": 47840 }, { "grad_norm": 0.47993624210357666, "learning_rate": 5.7660003450181655e-05, "loss": 0.0344, "step": 47850 }, { "grad_norm": 0.3925694525241852, "learning_rate": 5.764366352238534e-05, "loss": 0.0294, "step": 47860 }, { "grad_norm": 0.23366039991378784, "learning_rate": 5.7627322758689474e-05, "loss": 0.0257, "step": 47870 }, { "grad_norm": 0.4321248233318329, "learning_rate": 5.761098116088105e-05, "loss": 0.0307, "step": 47880 }, { "grad_norm": 0.6739298105239868, "learning_rate": 5.759463873074717e-05, "loss": 0.0321, "step": 47890 }, { "grad_norm": 0.630996823310852, "learning_rate": 5.757829547007504e-05, "loss": 0.0317, "step": 47900 }, { "grad_norm": 0.5200311541557312, "learning_rate": 5.756195138065189e-05, "loss": 0.0496, "step": 47910 }, { "grad_norm": 0.42908206582069397, "learning_rate": 5.754560646426511e-05, "loss": 0.0502, "step": 47920 }, { "grad_norm": 0.5745263695716858, "learning_rate": 5.752926072270216e-05, "loss": 0.0491, "step": 47930 }, { "grad_norm": 0.40788379311561584, "learning_rate": 5.7512914157750563e-05, "loss": 0.041, "step": 47940 }, { "grad_norm": 0.6373555064201355, "learning_rate": 5.749656677119798e-05, "loss": 0.0412, "step": 47950 }, { "grad_norm": 0.4869467616081238, "learning_rate": 5.7480218564832125e-05, "loss": 0.043, "step": 47960 }, { "grad_norm": 0.5826699137687683, "learning_rate": 5.746386954044082e-05, "loss": 0.0491, "step": 47970 }, { "grad_norm": 0.4235965609550476, "learning_rate": 5.744751969981195e-05, "loss": 0.0432, "step": 47980 }, { "grad_norm": 0.34114357829093933, "learning_rate": 5.7431169044733526e-05, "loss": 0.0288, "step": 47990 }, { "grad_norm": 0.5618627667427063, "learning_rate": 5.741481757699364e-05, "loss": 0.0383, "step": 48000 }, { "grad_norm": 0.48600801825523376, "learning_rate": 5.7398465298380434e-05, "loss": 0.0289, "step": 48010 }, { "grad_norm": 0.47837138175964355, "learning_rate": 5.7382112210682193e-05, "loss": 0.0291, "step": 48020 }, { "grad_norm": 0.4416406750679016, "learning_rate": 5.7365758315687266e-05, "loss": 0.0421, "step": 48030 }, { "grad_norm": 0.30381709337234497, "learning_rate": 5.734940361518407e-05, "loss": 0.0257, "step": 48040 }, { "grad_norm": 0.46924465894699097, "learning_rate": 5.733304811096116e-05, "loss": 0.0304, "step": 48050 }, { "grad_norm": 0.5631760954856873, "learning_rate": 5.731669180480713e-05, "loss": 0.0447, "step": 48060 }, { "grad_norm": 0.5226015448570251, "learning_rate": 5.730033469851067e-05, "loss": 0.0281, "step": 48070 }, { "grad_norm": 0.5638256072998047, "learning_rate": 5.7283976793860607e-05, "loss": 0.0401, "step": 48080 }, { "grad_norm": 0.7112981081008911, "learning_rate": 5.726761809264577e-05, "loss": 0.0411, "step": 48090 }, { "grad_norm": 0.333517462015152, "learning_rate": 5.7251258596655155e-05, "loss": 0.0312, "step": 48100 }, { "grad_norm": 0.5448212623596191, "learning_rate": 5.72348983076778e-05, "loss": 0.0323, "step": 48110 }, { "grad_norm": 0.41244518756866455, "learning_rate": 5.7218537227502854e-05, "loss": 0.0434, "step": 48120 }, { "grad_norm": 0.38040176033973694, "learning_rate": 5.720217535791951e-05, "loss": 0.0272, "step": 48130 }, { "grad_norm": 0.5197267532348633, "learning_rate": 5.718581270071711e-05, "loss": 0.0334, "step": 48140 }, { "grad_norm": 0.2913355529308319, "learning_rate": 5.716944925768505e-05, "loss": 0.0403, "step": 48150 }, { "grad_norm": 0.7915660738945007, "learning_rate": 5.7153085030612786e-05, "loss": 0.0322, "step": 48160 }, { "grad_norm": 0.31522274017333984, "learning_rate": 5.713672002128989e-05, "loss": 0.03, "step": 48170 }, { "grad_norm": 0.2528863847255707, "learning_rate": 5.712035423150602e-05, "loss": 0.0325, "step": 48180 }, { "grad_norm": 0.5808204412460327, "learning_rate": 5.710398766305094e-05, "loss": 0.0431, "step": 48190 }, { "grad_norm": 0.5277796387672424, "learning_rate": 5.708762031771442e-05, "loss": 0.0357, "step": 48200 }, { "grad_norm": 0.645632266998291, "learning_rate": 5.707125219728642e-05, "loss": 0.0374, "step": 48210 }, { "grad_norm": 0.33531078696250916, "learning_rate": 5.7054883303556905e-05, "loss": 0.0273, "step": 48220 }, { "grad_norm": 0.5638363361358643, "learning_rate": 5.703851363831595e-05, "loss": 0.0299, "step": 48230 }, { "grad_norm": 0.30921247601509094, "learning_rate": 5.702214320335374e-05, "loss": 0.0309, "step": 48240 }, { "grad_norm": 0.33940941095352173, "learning_rate": 5.7005772000460514e-05, "loss": 0.0556, "step": 48250 }, { "grad_norm": 0.49991750717163086, "learning_rate": 5.698940003142656e-05, "loss": 0.0437, "step": 48260 }, { "grad_norm": 0.41236281394958496, "learning_rate": 5.697302729804236e-05, "loss": 0.0332, "step": 48270 }, { "grad_norm": 0.5482643246650696, "learning_rate": 5.695665380209837e-05, "loss": 0.0477, "step": 48280 }, { "grad_norm": 0.705350935459137, "learning_rate": 5.6940279545385165e-05, "loss": 0.0382, "step": 48290 }, { "grad_norm": 0.36718064546585083, "learning_rate": 5.692390452969344e-05, "loss": 0.0288, "step": 48300 }, { "grad_norm": 0.47818607091903687, "learning_rate": 5.690752875681392e-05, "loss": 0.0322, "step": 48310 }, { "grad_norm": 0.28407904505729675, "learning_rate": 5.6891152228537435e-05, "loss": 0.0304, "step": 48320 }, { "grad_norm": 0.2969317138195038, "learning_rate": 5.687477494665492e-05, "loss": 0.0349, "step": 48330 }, { "grad_norm": 0.5806385278701782, "learning_rate": 5.685839691295733e-05, "loss": 0.0365, "step": 48340 }, { "grad_norm": 0.6068897247314453, "learning_rate": 5.6842018129235786e-05, "loss": 0.0369, "step": 48350 }, { "grad_norm": 0.5753882527351379, "learning_rate": 5.6825638597281404e-05, "loss": 0.0289, "step": 48360 }, { "grad_norm": 0.3788544535636902, "learning_rate": 5.680925831888546e-05, "loss": 0.0282, "step": 48370 }, { "grad_norm": 0.29867106676101685, "learning_rate": 5.6792877295839274e-05, "loss": 0.0289, "step": 48380 }, { "grad_norm": 0.3578714430332184, "learning_rate": 5.6776495529934224e-05, "loss": 0.026, "step": 48390 }, { "grad_norm": 0.3415161371231079, "learning_rate": 5.6760113022961824e-05, "loss": 0.0361, "step": 48400 }, { "grad_norm": 0.4522944390773773, "learning_rate": 5.6743729776713617e-05, "loss": 0.0236, "step": 48410 }, { "grad_norm": 0.7215755581855774, "learning_rate": 5.672734579298126e-05, "loss": 0.0451, "step": 48420 }, { "grad_norm": 0.44044339656829834, "learning_rate": 5.671096107355649e-05, "loss": 0.0317, "step": 48430 }, { "grad_norm": 0.544217586517334, "learning_rate": 5.66945756202311e-05, "loss": 0.0455, "step": 48440 }, { "grad_norm": 0.4743748605251312, "learning_rate": 5.667818943479699e-05, "loss": 0.0352, "step": 48450 }, { "grad_norm": 0.2908952534198761, "learning_rate": 5.666180251904612e-05, "loss": 0.0392, "step": 48460 }, { "grad_norm": 0.44139018654823303, "learning_rate": 5.6645414874770555e-05, "loss": 0.0316, "step": 48470 }, { "grad_norm": 0.42737582325935364, "learning_rate": 5.66290265037624e-05, "loss": 0.0324, "step": 48480 }, { "grad_norm": 0.3454590141773224, "learning_rate": 5.661263740781386e-05, "loss": 0.0303, "step": 48490 }, { "grad_norm": 0.6273218989372253, "learning_rate": 5.6596247588717254e-05, "loss": 0.0359, "step": 48500 }, { "grad_norm": 0.4315100312232971, "learning_rate": 5.6579857048264926e-05, "loss": 0.0384, "step": 48510 }, { "grad_norm": 0.5694676041603088, "learning_rate": 5.6563465788249314e-05, "loss": 0.0326, "step": 48520 }, { "grad_norm": 0.6257622838020325, "learning_rate": 5.6547073810462956e-05, "loss": 0.0354, "step": 48530 }, { "grad_norm": 0.5935635566711426, "learning_rate": 5.653068111669846e-05, "loss": 0.0353, "step": 48540 }, { "grad_norm": 0.5780514478683472, "learning_rate": 5.651428770874848e-05, "loss": 0.0319, "step": 48550 }, { "grad_norm": 0.6215280890464783, "learning_rate": 5.64978935884058e-05, "loss": 0.0383, "step": 48560 }, { "grad_norm": 0.3824189603328705, "learning_rate": 5.6481498757463244e-05, "loss": 0.0294, "step": 48570 }, { "grad_norm": 0.6199069023132324, "learning_rate": 5.646510321771373e-05, "loss": 0.0445, "step": 48580 }, { "grad_norm": 0.5136793851852417, "learning_rate": 5.644870697095024e-05, "loss": 0.0529, "step": 48590 }, { "grad_norm": 0.6329749822616577, "learning_rate": 5.643231001896586e-05, "loss": 0.0347, "step": 48600 }, { "grad_norm": 0.33754974603652954, "learning_rate": 5.6415912363553726e-05, "loss": 0.0346, "step": 48610 }, { "grad_norm": 0.5246559977531433, "learning_rate": 5.639951400650706e-05, "loss": 0.0525, "step": 48620 }, { "grad_norm": 0.2830900549888611, "learning_rate": 5.6383114949619165e-05, "loss": 0.0312, "step": 48630 }, { "grad_norm": 0.5600969791412354, "learning_rate": 5.636671519468342e-05, "loss": 0.0368, "step": 48640 }, { "grad_norm": 0.547409176826477, "learning_rate": 5.635031474349327e-05, "loss": 0.0302, "step": 48650 }, { "grad_norm": 0.657559871673584, "learning_rate": 5.6333913597842246e-05, "loss": 0.0357, "step": 48660 }, { "grad_norm": 0.6704307198524475, "learning_rate": 5.6317511759523955e-05, "loss": 0.0402, "step": 48670 }, { "grad_norm": 0.5459592938423157, "learning_rate": 5.630110923033207e-05, "loss": 0.0403, "step": 48680 }, { "grad_norm": 0.573108971118927, "learning_rate": 5.628470601206036e-05, "loss": 0.0421, "step": 48690 }, { "grad_norm": 0.40911492705345154, "learning_rate": 5.626830210650266e-05, "loss": 0.0372, "step": 48700 }, { "grad_norm": 0.4408343434333801, "learning_rate": 5.625189751545285e-05, "loss": 0.0346, "step": 48710 }, { "grad_norm": 0.5498644709587097, "learning_rate": 5.6235492240704936e-05, "loss": 0.0214, "step": 48720 }, { "grad_norm": 0.5491772890090942, "learning_rate": 5.621908628405296e-05, "loss": 0.0452, "step": 48730 }, { "grad_norm": 0.4891643524169922, "learning_rate": 5.620267964729106e-05, "loss": 0.0281, "step": 48740 }, { "grad_norm": 0.4599913954734802, "learning_rate": 5.618627233221344e-05, "loss": 0.033, "step": 48750 }, { "grad_norm": 0.42992883920669556, "learning_rate": 5.616986434061438e-05, "loss": 0.0372, "step": 48760 }, { "grad_norm": 0.460012286901474, "learning_rate": 5.615345567428822e-05, "loss": 0.0327, "step": 48770 }, { "grad_norm": 0.3428276479244232, "learning_rate": 5.613704633502941e-05, "loss": 0.0347, "step": 48780 }, { "grad_norm": 1.0259588956832886, "learning_rate": 5.612063632463245e-05, "loss": 0.0366, "step": 48790 }, { "grad_norm": 0.5720424056053162, "learning_rate": 5.610422564489188e-05, "loss": 0.0504, "step": 48800 }, { "grad_norm": 0.5508673191070557, "learning_rate": 5.608781429760239e-05, "loss": 0.0319, "step": 48810 }, { "grad_norm": 0.32144975662231445, "learning_rate": 5.607140228455866e-05, "loss": 0.0334, "step": 48820 }, { "grad_norm": 0.3878611922264099, "learning_rate": 5.605498960755553e-05, "loss": 0.0257, "step": 48830 }, { "grad_norm": 0.393741637468338, "learning_rate": 5.603857626838782e-05, "loss": 0.0355, "step": 48840 }, { "grad_norm": 0.617824912071228, "learning_rate": 5.60221622688505e-05, "loss": 0.043, "step": 48850 }, { "grad_norm": 0.5165732502937317, "learning_rate": 5.6005747610738565e-05, "loss": 0.0396, "step": 48860 }, { "grad_norm": 0.5100282430648804, "learning_rate": 5.59893322958471e-05, "loss": 0.0297, "step": 48870 }, { "grad_norm": 0.4480036497116089, "learning_rate": 5.5972916325971256e-05, "loss": 0.0275, "step": 48880 }, { "grad_norm": 0.48258987069129944, "learning_rate": 5.595649970290628e-05, "loss": 0.0321, "step": 48890 }, { "grad_norm": 0.7150671482086182, "learning_rate": 5.5940082428447426e-05, "loss": 0.0455, "step": 48900 }, { "grad_norm": 0.40982264280319214, "learning_rate": 5.592366450439012e-05, "loss": 0.0441, "step": 48910 }, { "grad_norm": 0.3692775368690491, "learning_rate": 5.590724593252975e-05, "loss": 0.0414, "step": 48920 }, { "grad_norm": 0.4675212800502777, "learning_rate": 5.589082671466184e-05, "loss": 0.0377, "step": 48930 }, { "grad_norm": 0.4629203677177429, "learning_rate": 5.587440685258199e-05, "loss": 0.0433, "step": 48940 }, { "grad_norm": 0.423069030046463, "learning_rate": 5.585798634808583e-05, "loss": 0.0284, "step": 48950 }, { "grad_norm": 0.5247565507888794, "learning_rate": 5.584156520296909e-05, "loss": 0.0466, "step": 48960 }, { "grad_norm": 0.29022741317749023, "learning_rate": 5.582514341902757e-05, "loss": 0.0302, "step": 48970 }, { "grad_norm": 0.5186642408370972, "learning_rate": 5.580872099805713e-05, "loss": 0.0371, "step": 48980 }, { "grad_norm": 0.6099215149879456, "learning_rate": 5.5792297941853674e-05, "loss": 0.0382, "step": 48990 }, { "grad_norm": 0.3812929689884186, "learning_rate": 5.5775874252213247e-05, "loss": 0.0364, "step": 49000 }, { "grad_norm": 0.4171590507030487, "learning_rate": 5.575944993093189e-05, "loss": 0.0324, "step": 49010 }, { "grad_norm": 0.465679407119751, "learning_rate": 5.574302497980574e-05, "loss": 0.0303, "step": 49020 }, { "grad_norm": 0.8032529354095459, "learning_rate": 5.5726599400631005e-05, "loss": 0.0454, "step": 49030 }, { "grad_norm": 0.52031409740448, "learning_rate": 5.571017319520399e-05, "loss": 0.0275, "step": 49040 }, { "grad_norm": 0.9319069385528564, "learning_rate": 5.5693746365320984e-05, "loss": 0.0254, "step": 49050 }, { "grad_norm": 0.45984166860580444, "learning_rate": 5.567731891277848e-05, "loss": 0.0331, "step": 49060 }, { "grad_norm": 0.4077000319957733, "learning_rate": 5.566089083937288e-05, "loss": 0.0359, "step": 49070 }, { "grad_norm": 0.5794187188148499, "learning_rate": 5.564446214690079e-05, "loss": 0.0333, "step": 49080 }, { "grad_norm": 0.6058686375617981, "learning_rate": 5.56280328371588e-05, "loss": 0.0363, "step": 49090 }, { "grad_norm": 0.5772956013679504, "learning_rate": 5.5611602911943596e-05, "loss": 0.0292, "step": 49100 }, { "grad_norm": 0.5587334036827087, "learning_rate": 5.559517237305194e-05, "loss": 0.039, "step": 49110 }, { "grad_norm": 0.31100910902023315, "learning_rate": 5.557874122228064e-05, "loss": 0.0312, "step": 49120 }, { "grad_norm": 0.332252562046051, "learning_rate": 5.55623094614266e-05, "loss": 0.0308, "step": 49130 }, { "grad_norm": 0.5151214599609375, "learning_rate": 5.5545877092286743e-05, "loss": 0.0452, "step": 49140 }, { "grad_norm": 0.4246450364589691, "learning_rate": 5.552944411665811e-05, "loss": 0.0293, "step": 49150 }, { "grad_norm": 0.8447694182395935, "learning_rate": 5.5513010536337795e-05, "loss": 0.0427, "step": 49160 }, { "grad_norm": 0.49907752871513367, "learning_rate": 5.5496576353122934e-05, "loss": 0.033, "step": 49170 }, { "grad_norm": 0.8641324639320374, "learning_rate": 5.548014156881074e-05, "loss": 0.0343, "step": 49180 }, { "grad_norm": 0.41454893350601196, "learning_rate": 5.5463706185198494e-05, "loss": 0.0221, "step": 49190 }, { "grad_norm": 0.4017502963542938, "learning_rate": 5.544727020408358e-05, "loss": 0.031, "step": 49200 }, { "grad_norm": 0.42667660117149353, "learning_rate": 5.5430833627263366e-05, "loss": 0.0398, "step": 49210 }, { "grad_norm": 0.7806093096733093, "learning_rate": 5.5414396456535354e-05, "loss": 0.04, "step": 49220 }, { "grad_norm": 0.5345382690429688, "learning_rate": 5.539795869369709e-05, "loss": 0.0312, "step": 49230 }, { "grad_norm": 0.46442195773124695, "learning_rate": 5.538152034054618e-05, "loss": 0.026, "step": 49240 }, { "grad_norm": 0.3317699134349823, "learning_rate": 5.536508139888028e-05, "loss": 0.0386, "step": 49250 }, { "grad_norm": 1.077884316444397, "learning_rate": 5.534864187049716e-05, "loss": 0.0374, "step": 49260 }, { "grad_norm": 0.6941362619400024, "learning_rate": 5.5332201757194615e-05, "loss": 0.0408, "step": 49270 }, { "grad_norm": 0.5163547992706299, "learning_rate": 5.531576106077049e-05, "loss": 0.025, "step": 49280 }, { "grad_norm": 0.3448512554168701, "learning_rate": 5.529931978302272e-05, "loss": 0.0301, "step": 49290 }, { "grad_norm": 0.5010901093482971, "learning_rate": 5.528287792574932e-05, "loss": 0.0319, "step": 49300 }, { "grad_norm": 0.34834927320480347, "learning_rate": 5.5266435490748324e-05, "loss": 0.0385, "step": 49310 }, { "grad_norm": 0.5355767607688904, "learning_rate": 5.524999247981787e-05, "loss": 0.0375, "step": 49320 }, { "grad_norm": 0.6203731894493103, "learning_rate": 5.523354889475613e-05, "loss": 0.04, "step": 49330 }, { "grad_norm": 0.4171624779701233, "learning_rate": 5.521710473736134e-05, "loss": 0.0467, "step": 49340 }, { "grad_norm": 0.792027473449707, "learning_rate": 5.5200660009431826e-05, "loss": 0.0271, "step": 49350 }, { "grad_norm": 0.4319539964199066, "learning_rate": 5.518421471276596e-05, "loss": 0.035, "step": 49360 }, { "grad_norm": 0.3688984811306, "learning_rate": 5.5167768849162174e-05, "loss": 0.0385, "step": 49370 }, { "grad_norm": 0.4782676696777344, "learning_rate": 5.515132242041893e-05, "loss": 0.0393, "step": 49380 }, { "grad_norm": 0.5002743601799011, "learning_rate": 5.513487542833483e-05, "loss": 0.037, "step": 49390 }, { "grad_norm": 0.5110666155815125, "learning_rate": 5.5118427874708466e-05, "loss": 0.0463, "step": 49400 }, { "grad_norm": 0.49927711486816406, "learning_rate": 5.510197976133853e-05, "loss": 0.0401, "step": 49410 }, { "grad_norm": 0.5889466404914856, "learning_rate": 5.508553109002376e-05, "loss": 0.0377, "step": 49420 }, { "grad_norm": 0.43014058470726013, "learning_rate": 5.5069081862562957e-05, "loss": 0.0413, "step": 49430 }, { "grad_norm": 0.5648043155670166, "learning_rate": 5.5052632080754965e-05, "loss": 0.048, "step": 49440 }, { "grad_norm": 0.5354810953140259, "learning_rate": 5.503618174639874e-05, "loss": 0.0337, "step": 49450 }, { "grad_norm": 0.6240143179893494, "learning_rate": 5.501973086129325e-05, "loss": 0.0388, "step": 49460 }, { "grad_norm": 0.7109888792037964, "learning_rate": 5.500327942723753e-05, "loss": 0.038, "step": 49470 }, { "grad_norm": 0.6509689092636108, "learning_rate": 5.498682744603071e-05, "loss": 0.0376, "step": 49480 }, { "grad_norm": 0.434044748544693, "learning_rate": 5.497037491947194e-05, "loss": 0.0503, "step": 49490 }, { "grad_norm": 0.8085863590240479, "learning_rate": 5.4953921849360424e-05, "loss": 0.0417, "step": 49500 }, { "grad_norm": 0.6679142117500305, "learning_rate": 5.493746823749547e-05, "loss": 0.0418, "step": 49510 }, { "grad_norm": 0.8086990118026733, "learning_rate": 5.49210140856764e-05, "loss": 0.0491, "step": 49520 }, { "grad_norm": 0.655702531337738, "learning_rate": 5.4904559395702647e-05, "loss": 0.0463, "step": 49530 }, { "grad_norm": 0.2551746964454651, "learning_rate": 5.488810416937364e-05, "loss": 0.0368, "step": 49540 }, { "grad_norm": 0.5412835478782654, "learning_rate": 5.4871648408488926e-05, "loss": 0.0384, "step": 49550 }, { "grad_norm": 0.41066527366638184, "learning_rate": 5.485519211484807e-05, "loss": 0.0365, "step": 49560 }, { "grad_norm": 1.307033896446228, "learning_rate": 5.4838735290250676e-05, "loss": 0.0392, "step": 49570 }, { "grad_norm": 0.7894662022590637, "learning_rate": 5.482227793649649e-05, "loss": 0.0447, "step": 49580 }, { "grad_norm": 0.628361165523529, "learning_rate": 5.480582005538524e-05, "loss": 0.0296, "step": 49590 }, { "grad_norm": 0.4906902313232422, "learning_rate": 5.478936164871671e-05, "loss": 0.0409, "step": 49600 }, { "grad_norm": 0.22746960818767548, "learning_rate": 5.477290271829081e-05, "loss": 0.0294, "step": 49610 }, { "grad_norm": 0.39473825693130493, "learning_rate": 5.475644326590744e-05, "loss": 0.0445, "step": 49620 }, { "grad_norm": 0.612957775592804, "learning_rate": 5.473998329336658e-05, "loss": 0.0304, "step": 49630 }, { "grad_norm": 0.7559658885002136, "learning_rate": 5.4723522802468286e-05, "loss": 0.0346, "step": 49640 }, { "grad_norm": 0.7923377156257629, "learning_rate": 5.470706179501264e-05, "loss": 0.0537, "step": 49650 }, { "grad_norm": 0.3510439991950989, "learning_rate": 5.4690600272799776e-05, "loss": 0.0299, "step": 49660 }, { "grad_norm": 0.5137256383895874, "learning_rate": 5.467413823762993e-05, "loss": 0.0349, "step": 49670 }, { "grad_norm": 0.5734778642654419, "learning_rate": 5.465767569130335e-05, "loss": 0.0315, "step": 49680 }, { "grad_norm": 0.6012734770774841, "learning_rate": 5.464121263562036e-05, "loss": 0.0327, "step": 49690 }, { "grad_norm": 0.3997322618961334, "learning_rate": 5.4624749072381343e-05, "loss": 0.0335, "step": 49700 }, { "grad_norm": 0.5050919055938721, "learning_rate": 5.460828500338672e-05, "loss": 0.0391, "step": 49710 }, { "grad_norm": 0.5869150161743164, "learning_rate": 5.459182043043698e-05, "loss": 0.0348, "step": 49720 }, { "grad_norm": 0.46812430024147034, "learning_rate": 5.457535535533265e-05, "loss": 0.0393, "step": 49730 }, { "grad_norm": 0.43068188428878784, "learning_rate": 5.4558889779874334e-05, "loss": 0.0512, "step": 49740 }, { "grad_norm": 0.3916361927986145, "learning_rate": 5.454242370586269e-05, "loss": 0.039, "step": 49750 }, { "grad_norm": 0.36946648359298706, "learning_rate": 5.452595713509843e-05, "loss": 0.0453, "step": 49760 }, { "grad_norm": 0.6103852987289429, "learning_rate": 5.4509490069382274e-05, "loss": 0.0426, "step": 49770 }, { "grad_norm": 0.4682578146457672, "learning_rate": 5.449302251051509e-05, "loss": 0.0362, "step": 49780 }, { "grad_norm": 0.4178363084793091, "learning_rate": 5.44765544602977e-05, "loss": 0.0395, "step": 49790 }, { "grad_norm": 0.46991610527038574, "learning_rate": 5.446008592053107e-05, "loss": 0.0357, "step": 49800 }, { "grad_norm": 0.274906188249588, "learning_rate": 5.4443616893016135e-05, "loss": 0.028, "step": 49810 }, { "grad_norm": 0.5725603103637695, "learning_rate": 5.4427147379553934e-05, "loss": 0.0337, "step": 49820 }, { "grad_norm": 0.3993805944919586, "learning_rate": 5.441067738194556e-05, "loss": 0.0327, "step": 49830 }, { "grad_norm": 0.48392418026924133, "learning_rate": 5.439420690199214e-05, "loss": 0.0356, "step": 49840 }, { "grad_norm": 0.5730889439582825, "learning_rate": 5.437773594149484e-05, "loss": 0.033, "step": 49850 }, { "grad_norm": 0.5745247006416321, "learning_rate": 5.436126450225495e-05, "loss": 0.0397, "step": 49860 }, { "grad_norm": 0.5760297179222107, "learning_rate": 5.434479258607371e-05, "loss": 0.0472, "step": 49870 }, { "grad_norm": 0.504349946975708, "learning_rate": 5.43283201947525e-05, "loss": 0.0378, "step": 49880 }, { "grad_norm": 0.18816567957401276, "learning_rate": 5.431184733009268e-05, "loss": 0.026, "step": 49890 }, { "grad_norm": 0.41989120841026306, "learning_rate": 5.4295373993895736e-05, "loss": 0.0492, "step": 49900 }, { "grad_norm": 0.49728867411613464, "learning_rate": 5.4278900187963157e-05, "loss": 0.033, "step": 49910 }, { "grad_norm": 0.6026683449745178, "learning_rate": 5.426242591409646e-05, "loss": 0.04, "step": 49920 }, { "grad_norm": 0.6354518532752991, "learning_rate": 5.42459511740973e-05, "loss": 0.0382, "step": 49930 }, { "grad_norm": 0.5452233552932739, "learning_rate": 5.422947596976729e-05, "loss": 0.0375, "step": 49940 }, { "grad_norm": 0.6903913617134094, "learning_rate": 5.4213000302908134e-05, "loss": 0.0452, "step": 49950 }, { "grad_norm": 0.5373862981796265, "learning_rate": 5.419652417532162e-05, "loss": 0.0321, "step": 49960 }, { "grad_norm": 0.47756871581077576, "learning_rate": 5.4180047588809534e-05, "loss": 0.0356, "step": 49970 }, { "grad_norm": 0.3785707950592041, "learning_rate": 5.4163570545173704e-05, "loss": 0.0523, "step": 49980 }, { "grad_norm": 0.5166054368019104, "learning_rate": 5.414709304621608e-05, "loss": 0.0346, "step": 49990 }, { "grad_norm": 0.6954185962677002, "learning_rate": 5.413061509373858e-05, "loss": 0.0436, "step": 50000 }, { "grad_norm": 0.5440680384635925, "learning_rate": 5.411413668954324e-05, "loss": 0.0432, "step": 50010 }, { "grad_norm": 0.43846243619918823, "learning_rate": 5.4097657835432083e-05, "loss": 0.0338, "step": 50020 }, { "grad_norm": 0.3396415412425995, "learning_rate": 5.408117853320723e-05, "loss": 0.0299, "step": 50030 }, { "grad_norm": 0.43188032507896423, "learning_rate": 5.406469878467084e-05, "loss": 0.0434, "step": 50040 }, { "grad_norm": 0.4953576922416687, "learning_rate": 5.404821859162509e-05, "loss": 0.0321, "step": 50050 }, { "grad_norm": 0.5606303811073303, "learning_rate": 5.403173795587225e-05, "loss": 0.056, "step": 50060 }, { "grad_norm": 0.6203931570053101, "learning_rate": 5.4015256879214606e-05, "loss": 0.0331, "step": 50070 }, { "grad_norm": 0.5324113965034485, "learning_rate": 5.39987753634545e-05, "loss": 0.0313, "step": 50080 }, { "grad_norm": 0.456556499004364, "learning_rate": 5.3982293410394335e-05, "loss": 0.0411, "step": 50090 }, { "grad_norm": 0.43262648582458496, "learning_rate": 5.396581102183655e-05, "loss": 0.0297, "step": 50100 }, { "grad_norm": 1.0241138935089111, "learning_rate": 5.3949328199583634e-05, "loss": 0.0431, "step": 50110 }, { "grad_norm": 0.5513630509376526, "learning_rate": 5.393284494543813e-05, "loss": 0.034, "step": 50120 }, { "grad_norm": 0.6187970638275146, "learning_rate": 5.391636126120262e-05, "loss": 0.0351, "step": 50130 }, { "grad_norm": 0.37620481848716736, "learning_rate": 5.389987714867971e-05, "loss": 0.0232, "step": 50140 }, { "grad_norm": 0.6690865755081177, "learning_rate": 5.3883392609672114e-05, "loss": 0.0399, "step": 50150 }, { "grad_norm": 0.7363924980163574, "learning_rate": 5.3866907645982545e-05, "loss": 0.051, "step": 50160 }, { "grad_norm": 0.603299081325531, "learning_rate": 5.385042225941375e-05, "loss": 0.0294, "step": 50170 }, { "grad_norm": 0.8049163222312927, "learning_rate": 5.3833936451768585e-05, "loss": 0.0475, "step": 50180 }, { "grad_norm": 0.4578104615211487, "learning_rate": 5.3817450224849884e-05, "loss": 0.0333, "step": 50190 }, { "grad_norm": 0.31485268473625183, "learning_rate": 5.380096358046056e-05, "loss": 0.0399, "step": 50200 }, { "grad_norm": 0.47431132197380066, "learning_rate": 5.378447652040359e-05, "loss": 0.0361, "step": 50210 }, { "grad_norm": 0.5150892734527588, "learning_rate": 5.376798904648195e-05, "loss": 0.0318, "step": 50220 }, { "grad_norm": 0.5348941683769226, "learning_rate": 5.375150116049869e-05, "loss": 0.0303, "step": 50230 }, { "grad_norm": 0.6946293115615845, "learning_rate": 5.373501286425691e-05, "loss": 0.0355, "step": 50240 }, { "grad_norm": 0.32771527767181396, "learning_rate": 5.3718524159559726e-05, "loss": 0.0594, "step": 50250 }, { "grad_norm": 0.21866728365421295, "learning_rate": 5.370203504821034e-05, "loss": 0.0348, "step": 50260 }, { "grad_norm": 0.5933965444564819, "learning_rate": 5.368554553201196e-05, "loss": 0.0364, "step": 50270 }, { "grad_norm": 0.49158307909965515, "learning_rate": 5.3669055612767874e-05, "loss": 0.0377, "step": 50280 }, { "grad_norm": 0.46341365575790405, "learning_rate": 5.3652565292281376e-05, "loss": 0.0255, "step": 50290 }, { "grad_norm": 0.6613458395004272, "learning_rate": 5.3636074572355824e-05, "loss": 0.0319, "step": 50300 }, { "grad_norm": 0.353593111038208, "learning_rate": 5.361958345479463e-05, "loss": 0.0352, "step": 50310 }, { "grad_norm": 0.30574995279312134, "learning_rate": 5.360309194140123e-05, "loss": 0.0249, "step": 50320 }, { "grad_norm": 0.6439571976661682, "learning_rate": 5.3586600033979106e-05, "loss": 0.0386, "step": 50330 }, { "grad_norm": 0.5634507536888123, "learning_rate": 5.357010773433181e-05, "loss": 0.045, "step": 50340 }, { "grad_norm": 0.4217360317707062, "learning_rate": 5.3553615044262895e-05, "loss": 0.0346, "step": 50350 }, { "grad_norm": 0.3120757043361664, "learning_rate": 5.3537121965575964e-05, "loss": 0.0313, "step": 50360 }, { "grad_norm": 0.44473373889923096, "learning_rate": 5.352062850007471e-05, "loss": 0.041, "step": 50370 }, { "grad_norm": 0.3310357630252838, "learning_rate": 5.350413464956282e-05, "loss": 0.0278, "step": 50380 }, { "grad_norm": 0.9471684098243713, "learning_rate": 5.348764041584403e-05, "loss": 0.0328, "step": 50390 }, { "grad_norm": 0.47243833541870117, "learning_rate": 5.347114580072212e-05, "loss": 0.0316, "step": 50400 }, { "grad_norm": 0.3600432276725769, "learning_rate": 5.345465080600093e-05, "loss": 0.0395, "step": 50410 }, { "grad_norm": 0.47226572036743164, "learning_rate": 5.3438155433484326e-05, "loss": 0.0483, "step": 50420 }, { "grad_norm": 0.3750462234020233, "learning_rate": 5.3421659684976197e-05, "loss": 0.0322, "step": 50430 }, { "grad_norm": 0.36642658710479736, "learning_rate": 5.340516356228052e-05, "loss": 0.0298, "step": 50440 }, { "grad_norm": 0.6763122081756592, "learning_rate": 5.338866706720128e-05, "loss": 0.0473, "step": 50450 }, { "grad_norm": 0.5267098546028137, "learning_rate": 5.337217020154249e-05, "loss": 0.0472, "step": 50460 }, { "grad_norm": 0.4600614905357361, "learning_rate": 5.335567296710825e-05, "loss": 0.0281, "step": 50470 }, { "grad_norm": 0.38772791624069214, "learning_rate": 5.333917536570265e-05, "loss": 0.0373, "step": 50480 }, { "grad_norm": 0.47143980860710144, "learning_rate": 5.332267739912986e-05, "loss": 0.0365, "step": 50490 }, { "grad_norm": 0.4885306656360626, "learning_rate": 5.330617906919405e-05, "loss": 0.0253, "step": 50500 }, { "grad_norm": 0.3192148804664612, "learning_rate": 5.328968037769949e-05, "loss": 0.0338, "step": 50510 }, { "grad_norm": 0.37678614258766174, "learning_rate": 5.327318132645042e-05, "loss": 0.0298, "step": 50520 }, { "grad_norm": 0.6548759341239929, "learning_rate": 5.325668191725116e-05, "loss": 0.0265, "step": 50530 }, { "grad_norm": 0.7972645163536072, "learning_rate": 5.324018215190606e-05, "loss": 0.0429, "step": 50540 }, { "grad_norm": 0.41989919543266296, "learning_rate": 5.3223682032219515e-05, "loss": 0.0347, "step": 50550 }, { "grad_norm": 0.32455673813819885, "learning_rate": 5.320718155999595e-05, "loss": 0.0317, "step": 50560 }, { "grad_norm": 0.5135205984115601, "learning_rate": 5.3190680737039835e-05, "loss": 0.0295, "step": 50570 }, { "grad_norm": 0.2962495684623718, "learning_rate": 5.317417956515567e-05, "loss": 0.0325, "step": 50580 }, { "grad_norm": 0.469789057970047, "learning_rate": 5.315767804614798e-05, "loss": 0.0441, "step": 50590 }, { "grad_norm": 0.5065226554870605, "learning_rate": 5.3141176181821395e-05, "loss": 0.0292, "step": 50600 }, { "grad_norm": 0.3570669889450073, "learning_rate": 5.31246739739805e-05, "loss": 0.0265, "step": 50610 }, { "grad_norm": 0.803942859172821, "learning_rate": 5.310817142442994e-05, "loss": 0.0372, "step": 50620 }, { "grad_norm": 0.6215888857841492, "learning_rate": 5.309166853497445e-05, "loss": 0.0285, "step": 50630 }, { "grad_norm": 0.5279948711395264, "learning_rate": 5.307516530741873e-05, "loss": 0.0334, "step": 50640 }, { "grad_norm": 0.2945486605167389, "learning_rate": 5.305866174356754e-05, "loss": 0.0524, "step": 50650 }, { "grad_norm": 0.5057479739189148, "learning_rate": 5.304215784522571e-05, "loss": 0.0365, "step": 50660 }, { "grad_norm": 0.5444679856300354, "learning_rate": 5.302565361419808e-05, "loss": 0.0347, "step": 50670 }, { "grad_norm": 0.3083181381225586, "learning_rate": 5.3009149052289507e-05, "loss": 0.0309, "step": 50680 }, { "grad_norm": 0.26472413539886475, "learning_rate": 5.299264416130493e-05, "loss": 0.0319, "step": 50690 }, { "grad_norm": 0.2788454592227936, "learning_rate": 5.297613894304928e-05, "loss": 0.0448, "step": 50700 }, { "grad_norm": 0.508951723575592, "learning_rate": 5.2959633399327534e-05, "loss": 0.0349, "step": 50710 }, { "grad_norm": 0.5093986988067627, "learning_rate": 5.294312753194476e-05, "loss": 0.0325, "step": 50720 }, { "grad_norm": 0.42637550830841064, "learning_rate": 5.292662134270596e-05, "loss": 0.0393, "step": 50730 }, { "grad_norm": 0.6151197552680969, "learning_rate": 5.291011483341626e-05, "loss": 0.0344, "step": 50740 }, { "grad_norm": 0.5928164124488831, "learning_rate": 5.2893608005880767e-05, "loss": 0.0327, "step": 50750 }, { "grad_norm": 0.8972004652023315, "learning_rate": 5.287710086190467e-05, "loss": 0.0388, "step": 50760 }, { "grad_norm": 0.6737184524536133, "learning_rate": 5.286059340329314e-05, "loss": 0.0389, "step": 50770 }, { "grad_norm": 0.46213436126708984, "learning_rate": 5.284408563185141e-05, "loss": 0.0342, "step": 50780 }, { "grad_norm": 0.5012590289115906, "learning_rate": 5.282757754938476e-05, "loss": 0.032, "step": 50790 }, { "grad_norm": 0.6699250936508179, "learning_rate": 5.281106915769849e-05, "loss": 0.0256, "step": 50800 }, { "grad_norm": 0.5134585499763489, "learning_rate": 5.2794560458597897e-05, "loss": 0.036, "step": 50810 }, { "grad_norm": 0.7187100052833557, "learning_rate": 5.27780514538884e-05, "loss": 0.0328, "step": 50820 }, { "grad_norm": 0.38014471530914307, "learning_rate": 5.2761542145375365e-05, "loss": 0.0293, "step": 50830 }, { "grad_norm": 0.5545541644096375, "learning_rate": 5.274503253486421e-05, "loss": 0.0423, "step": 50840 }, { "grad_norm": 0.4839503765106201, "learning_rate": 5.272852262416046e-05, "loss": 0.0411, "step": 50850 }, { "grad_norm": 0.5535187125205994, "learning_rate": 5.2712012415069555e-05, "loss": 0.0491, "step": 50860 }, { "grad_norm": 0.5829978585243225, "learning_rate": 5.269550190939705e-05, "loss": 0.0446, "step": 50870 }, { "grad_norm": 0.40988484025001526, "learning_rate": 5.267899110894852e-05, "loss": 0.0344, "step": 50880 }, { "grad_norm": 0.4203360080718994, "learning_rate": 5.266248001552955e-05, "loss": 0.0548, "step": 50890 }, { "grad_norm": 0.48410922288894653, "learning_rate": 5.264596863094575e-05, "loss": 0.0362, "step": 50900 }, { "grad_norm": 0.4024907648563385, "learning_rate": 5.262945695700282e-05, "loss": 0.0297, "step": 50910 }, { "grad_norm": 0.41709911823272705, "learning_rate": 5.261294499550643e-05, "loss": 0.0317, "step": 50920 }, { "grad_norm": 0.47282347083091736, "learning_rate": 5.25964327482623e-05, "loss": 0.0445, "step": 50930 }, { "grad_norm": 0.6311399936676025, "learning_rate": 5.257992021707617e-05, "loss": 0.0353, "step": 50940 }, { "grad_norm": 0.49442344903945923, "learning_rate": 5.256340740375387e-05, "loss": 0.0308, "step": 50950 }, { "grad_norm": 0.6401200294494629, "learning_rate": 5.254689431010117e-05, "loss": 0.0428, "step": 50960 }, { "grad_norm": 0.44210779666900635, "learning_rate": 5.253038093792395e-05, "loss": 0.0305, "step": 50970 }, { "grad_norm": 0.4113073945045471, "learning_rate": 5.251386728902806e-05, "loss": 0.0269, "step": 50980 }, { "grad_norm": 0.4905489981174469, "learning_rate": 5.2497353365219446e-05, "loss": 0.0281, "step": 50990 }, { "grad_norm": 0.41499459743499756, "learning_rate": 5.2480839168304e-05, "loss": 0.0393, "step": 51000 }, { "grad_norm": 0.4225313365459442, "learning_rate": 5.246432470008773e-05, "loss": 0.0371, "step": 51010 }, { "grad_norm": 0.33881276845932007, "learning_rate": 5.2447809962376616e-05, "loss": 0.0332, "step": 51020 }, { "grad_norm": 0.6068467497825623, "learning_rate": 5.243129495697666e-05, "loss": 0.0298, "step": 51030 }, { "grad_norm": 0.2932214140892029, "learning_rate": 5.241477968569397e-05, "loss": 0.0296, "step": 51040 }, { "grad_norm": 0.6843432784080505, "learning_rate": 5.23982641503346e-05, "loss": 0.0415, "step": 51050 }, { "grad_norm": 0.4840449094772339, "learning_rate": 5.238174835270465e-05, "loss": 0.0366, "step": 51060 }, { "grad_norm": 0.27128803730010986, "learning_rate": 5.23652322946103e-05, "loss": 0.0248, "step": 51070 }, { "grad_norm": 0.6312944293022156, "learning_rate": 5.234871597785771e-05, "loss": 0.0426, "step": 51080 }, { "grad_norm": 0.3182927668094635, "learning_rate": 5.233219940425306e-05, "loss": 0.0319, "step": 51090 }, { "grad_norm": 0.5282608866691589, "learning_rate": 5.2315682575602576e-05, "loss": 0.0356, "step": 51100 }, { "grad_norm": 0.4802302122116089, "learning_rate": 5.229916549371254e-05, "loss": 0.032, "step": 51110 }, { "grad_norm": 0.500589907169342, "learning_rate": 5.228264816038924e-05, "loss": 0.0428, "step": 51120 }, { "grad_norm": 0.46542149782180786, "learning_rate": 5.2266130577438935e-05, "loss": 0.0309, "step": 51130 }, { "grad_norm": 1.2305532693862915, "learning_rate": 5.224961274666802e-05, "loss": 0.0371, "step": 51140 }, { "grad_norm": 0.5218974351882935, "learning_rate": 5.223309466988283e-05, "loss": 0.0316, "step": 51150 }, { "grad_norm": 0.2965162694454193, "learning_rate": 5.221657634888976e-05, "loss": 0.0378, "step": 51160 }, { "grad_norm": 0.5597740411758423, "learning_rate": 5.2200057785495235e-05, "loss": 0.03, "step": 51170 }, { "grad_norm": 0.5321778655052185, "learning_rate": 5.2183538981505706e-05, "loss": 0.0332, "step": 51180 }, { "grad_norm": 0.4956342279911041, "learning_rate": 5.216701993872762e-05, "loss": 0.0372, "step": 51190 }, { "grad_norm": 1.2473526000976562, "learning_rate": 5.215050065896752e-05, "loss": 0.0371, "step": 51200 }, { "grad_norm": 0.49856260418891907, "learning_rate": 5.2133981144031875e-05, "loss": 0.0485, "step": 51210 }, { "grad_norm": 0.36447834968566895, "learning_rate": 5.211746139572727e-05, "loss": 0.039, "step": 51220 }, { "grad_norm": 0.5446431040763855, "learning_rate": 5.210094141586026e-05, "loss": 0.0468, "step": 51230 }, { "grad_norm": 0.5457251071929932, "learning_rate": 5.2084421206237476e-05, "loss": 0.0335, "step": 51240 }, { "grad_norm": 0.420281320810318, "learning_rate": 5.20679007686655e-05, "loss": 0.0377, "step": 51250 }, { "grad_norm": 0.5806899070739746, "learning_rate": 5.205138010495102e-05, "loss": 0.0534, "step": 51260 }, { "grad_norm": 0.39269447326660156, "learning_rate": 5.20348592169007e-05, "loss": 0.0337, "step": 51270 }, { "grad_norm": 0.2940201163291931, "learning_rate": 5.201833810632123e-05, "loss": 0.0267, "step": 51280 }, { "grad_norm": 0.4283692538738251, "learning_rate": 5.200181677501933e-05, "loss": 0.0283, "step": 51290 }, { "grad_norm": 0.3765804171562195, "learning_rate": 5.1985295224801777e-05, "loss": 0.0436, "step": 51300 }, { "grad_norm": 0.3926578760147095, "learning_rate": 5.196877345747531e-05, "loss": 0.0245, "step": 51310 }, { "grad_norm": 0.7706429362297058, "learning_rate": 5.195225147484674e-05, "loss": 0.0304, "step": 51320 }, { "grad_norm": 0.613819420337677, "learning_rate": 5.193572927872288e-05, "loss": 0.0439, "step": 51330 }, { "grad_norm": 0.32752418518066406, "learning_rate": 5.191920687091059e-05, "loss": 0.0364, "step": 51340 }, { "grad_norm": 0.45447272062301636, "learning_rate": 5.1902684253216695e-05, "loss": 0.0293, "step": 51350 }, { "grad_norm": 0.8090012073516846, "learning_rate": 5.1886161427448134e-05, "loss": 0.0343, "step": 51360 }, { "grad_norm": 0.31521427631378174, "learning_rate": 5.186963839541179e-05, "loss": 0.0237, "step": 51370 }, { "grad_norm": 0.3692028522491455, "learning_rate": 5.185311515891459e-05, "loss": 0.038, "step": 51380 }, { "grad_norm": 0.6139435172080994, "learning_rate": 5.1836591719763516e-05, "loss": 0.0334, "step": 51390 }, { "grad_norm": 0.42027226090431213, "learning_rate": 5.182006807976554e-05, "loss": 0.0306, "step": 51400 }, { "grad_norm": 0.629979133605957, "learning_rate": 5.180354424072763e-05, "loss": 0.0423, "step": 51410 }, { "grad_norm": 0.5978937149047852, "learning_rate": 5.178702020445686e-05, "loss": 0.0351, "step": 51420 }, { "grad_norm": 0.4558480381965637, "learning_rate": 5.177049597276024e-05, "loss": 0.0358, "step": 51430 }, { "grad_norm": 0.43703413009643555, "learning_rate": 5.175397154744482e-05, "loss": 0.0253, "step": 51440 }, { "grad_norm": 0.39343398809432983, "learning_rate": 5.1737446930317736e-05, "loss": 0.0486, "step": 51450 }, { "grad_norm": 0.6907020211219788, "learning_rate": 5.172092212318604e-05, "loss": 0.0437, "step": 51460 }, { "grad_norm": 0.8468671441078186, "learning_rate": 5.170439712785692e-05, "loss": 0.0353, "step": 51470 }, { "grad_norm": 0.6646105647087097, "learning_rate": 5.1687871946137467e-05, "loss": 0.0316, "step": 51480 }, { "grad_norm": 0.41396260261535645, "learning_rate": 5.16713465798349e-05, "loss": 0.031, "step": 51490 }, { "grad_norm": 0.452501118183136, "learning_rate": 5.165482103075637e-05, "loss": 0.0328, "step": 51500 }, { "grad_norm": 0.3218111991882324, "learning_rate": 5.163829530070909e-05, "loss": 0.0371, "step": 51510 }, { "grad_norm": 0.45889633893966675, "learning_rate": 5.162176939150032e-05, "loss": 0.0316, "step": 51520 }, { "grad_norm": 0.4702692925930023, "learning_rate": 5.160524330493728e-05, "loss": 0.0309, "step": 51530 }, { "grad_norm": 0.5556879043579102, "learning_rate": 5.158871704282724e-05, "loss": 0.029, "step": 51540 }, { "grad_norm": 0.5419033765792847, "learning_rate": 5.15721906069775e-05, "loss": 0.0271, "step": 51550 }, { "grad_norm": 0.6358541250228882, "learning_rate": 5.155566399919536e-05, "loss": 0.0265, "step": 51560 }, { "grad_norm": 0.5900490283966064, "learning_rate": 5.153913722128813e-05, "loss": 0.0305, "step": 51570 }, { "grad_norm": 0.7923049926757812, "learning_rate": 5.1522610275063176e-05, "loss": 0.0376, "step": 51580 }, { "grad_norm": 0.722632110118866, "learning_rate": 5.150608316232786e-05, "loss": 0.0267, "step": 51590 }, { "grad_norm": 0.3246087431907654, "learning_rate": 5.148955588488955e-05, "loss": 0.0307, "step": 51600 }, { "grad_norm": 0.3244600296020508, "learning_rate": 5.147302844455564e-05, "loss": 0.0324, "step": 51610 }, { "grad_norm": 0.45003408193588257, "learning_rate": 5.145650084313356e-05, "loss": 0.0325, "step": 51620 }, { "grad_norm": 0.5713023543357849, "learning_rate": 5.1439973082430734e-05, "loss": 0.0304, "step": 51630 }, { "grad_norm": 0.6699866056442261, "learning_rate": 5.142344516425461e-05, "loss": 0.0267, "step": 51640 }, { "grad_norm": 0.5241699814796448, "learning_rate": 5.1406917090412673e-05, "loss": 0.0368, "step": 51650 }, { "grad_norm": 1.3264615535736084, "learning_rate": 5.1390388862712404e-05, "loss": 0.0296, "step": 51660 }, { "grad_norm": 0.6242450475692749, "learning_rate": 5.137386048296129e-05, "loss": 0.0462, "step": 51670 }, { "grad_norm": 0.6842703223228455, "learning_rate": 5.135733195296686e-05, "loss": 0.0279, "step": 51680 }, { "grad_norm": 0.5630202293395996, "learning_rate": 5.134080327453664e-05, "loss": 0.0349, "step": 51690 }, { "grad_norm": 0.3624936044216156, "learning_rate": 5.1324274449478204e-05, "loss": 0.0397, "step": 51700 }, { "grad_norm": 0.6141699552536011, "learning_rate": 5.130774547959909e-05, "loss": 0.0354, "step": 51710 }, { "grad_norm": 0.6019532680511475, "learning_rate": 5.129121636670692e-05, "loss": 0.0325, "step": 51720 }, { "grad_norm": 0.48119962215423584, "learning_rate": 5.1274687112609246e-05, "loss": 0.0427, "step": 51730 }, { "grad_norm": 0.48043760657310486, "learning_rate": 5.125815771911372e-05, "loss": 0.042, "step": 51740 }, { "grad_norm": 0.3379809856414795, "learning_rate": 5.1241628188027956e-05, "loss": 0.0433, "step": 51750 }, { "grad_norm": 0.5148066878318787, "learning_rate": 5.1225098521159585e-05, "loss": 0.0353, "step": 51760 }, { "grad_norm": 0.5273358821868896, "learning_rate": 5.12085687203163e-05, "loss": 0.0362, "step": 51770 }, { "grad_norm": 0.3393116295337677, "learning_rate": 5.119203878730575e-05, "loss": 0.0209, "step": 51780 }, { "grad_norm": 0.31848475337028503, "learning_rate": 5.117550872393564e-05, "loss": 0.0252, "step": 51790 }, { "grad_norm": 0.3814146816730499, "learning_rate": 5.1158978532013645e-05, "loss": 0.031, "step": 51800 }, { "grad_norm": 0.4178673028945923, "learning_rate": 5.114244821334751e-05, "loss": 0.0344, "step": 51810 }, { "grad_norm": 0.3909122347831726, "learning_rate": 5.112591776974497e-05, "loss": 0.037, "step": 51820 }, { "grad_norm": 0.5029231905937195, "learning_rate": 5.110938720301374e-05, "loss": 0.0289, "step": 51830 }, { "grad_norm": 0.5089048743247986, "learning_rate": 5.1092856514961595e-05, "loss": 0.0357, "step": 51840 }, { "grad_norm": 0.48800426721572876, "learning_rate": 5.107632570739631e-05, "loss": 0.0358, "step": 51850 }, { "grad_norm": 0.4619179666042328, "learning_rate": 5.1059794782125645e-05, "loss": 0.0379, "step": 51860 }, { "grad_norm": 0.3748224079608917, "learning_rate": 5.104326374095744e-05, "loss": 0.0295, "step": 51870 }, { "grad_norm": 0.5012024641036987, "learning_rate": 5.102673258569947e-05, "loss": 0.0332, "step": 51880 }, { "grad_norm": 0.6317352652549744, "learning_rate": 5.101020131815956e-05, "loss": 0.0327, "step": 51890 }, { "grad_norm": 0.31846436858177185, "learning_rate": 5.099366994014556e-05, "loss": 0.0304, "step": 51900 }, { "grad_norm": 0.5346493124961853, "learning_rate": 5.097713845346532e-05, "loss": 0.0378, "step": 51910 }, { "grad_norm": 0.4511321187019348, "learning_rate": 5.096060685992668e-05, "loss": 0.0353, "step": 51920 }, { "grad_norm": 0.5927000641822815, "learning_rate": 5.0944075161337514e-05, "loss": 0.0386, "step": 51930 }, { "grad_norm": 0.44122642278671265, "learning_rate": 5.09275433595057e-05, "loss": 0.0369, "step": 51940 }, { "grad_norm": 0.4608015716075897, "learning_rate": 5.0911011456239157e-05, "loss": 0.0227, "step": 51950 }, { "grad_norm": 0.2918066084384918, "learning_rate": 5.089447945334574e-05, "loss": 0.0338, "step": 51960 }, { "grad_norm": 0.6275266408920288, "learning_rate": 5.087794735263343e-05, "loss": 0.0442, "step": 51970 }, { "grad_norm": 0.43509602546691895, "learning_rate": 5.086141515591011e-05, "loss": 0.0282, "step": 51980 }, { "grad_norm": 0.4709314703941345, "learning_rate": 5.084488286498371e-05, "loss": 0.0371, "step": 51990 }, { "grad_norm": 0.7816606760025024, "learning_rate": 5.08283504816622e-05, "loss": 0.042, "step": 52000 }, { "grad_norm": 0.43648526072502136, "learning_rate": 5.081181800775353e-05, "loss": 0.0313, "step": 52010 }, { "grad_norm": 0.5525174140930176, "learning_rate": 5.079528544506565e-05, "loss": 0.0472, "step": 52020 }, { "grad_norm": 0.6976786255836487, "learning_rate": 5.077875279540657e-05, "loss": 0.0424, "step": 52030 }, { "grad_norm": 0.5790350437164307, "learning_rate": 5.076222006058424e-05, "loss": 0.0346, "step": 52040 }, { "grad_norm": 0.3427538275718689, "learning_rate": 5.074568724240667e-05, "loss": 0.0265, "step": 52050 }, { "grad_norm": 0.4165568947792053, "learning_rate": 5.072915434268187e-05, "loss": 0.0355, "step": 52060 }, { "grad_norm": 0.5301558375358582, "learning_rate": 5.071262136321785e-05, "loss": 0.0381, "step": 52070 }, { "grad_norm": 0.5312427282333374, "learning_rate": 5.069608830582262e-05, "loss": 0.0383, "step": 52080 }, { "grad_norm": 0.4540577828884125, "learning_rate": 5.067955517230423e-05, "loss": 0.0333, "step": 52090 }, { "grad_norm": 0.5940041542053223, "learning_rate": 5.06630219644707e-05, "loss": 0.044, "step": 52100 }, { "grad_norm": 0.37693050503730774, "learning_rate": 5.064648868413008e-05, "loss": 0.0348, "step": 52110 }, { "grad_norm": 1.0711945295333862, "learning_rate": 5.062995533309044e-05, "loss": 0.0326, "step": 52120 }, { "grad_norm": 0.6641495823860168, "learning_rate": 5.061342191315983e-05, "loss": 0.0483, "step": 52130 }, { "grad_norm": 0.43550971150398254, "learning_rate": 5.0596888426146325e-05, "loss": 0.0451, "step": 52140 }, { "grad_norm": 0.4411577880382538, "learning_rate": 5.058035487385798e-05, "loss": 0.0403, "step": 52150 }, { "grad_norm": 0.6162967085838318, "learning_rate": 5.0563821258102915e-05, "loss": 0.0761, "step": 52160 }, { "grad_norm": 0.49575212597846985, "learning_rate": 5.05472875806892e-05, "loss": 0.0364, "step": 52170 }, { "grad_norm": 0.27399376034736633, "learning_rate": 5.053075384342494e-05, "loss": 0.0472, "step": 52180 }, { "grad_norm": 0.3881027400493622, "learning_rate": 5.0514220048118234e-05, "loss": 0.0293, "step": 52190 }, { "grad_norm": 0.32618260383605957, "learning_rate": 5.049768619657721e-05, "loss": 0.04, "step": 52200 }, { "grad_norm": 0.41990533471107483, "learning_rate": 5.048115229060996e-05, "loss": 0.0279, "step": 52210 }, { "grad_norm": 0.4486388862133026, "learning_rate": 5.046461833202462e-05, "loss": 0.0276, "step": 52220 }, { "grad_norm": 0.48815450072288513, "learning_rate": 5.044808432262932e-05, "loss": 0.0314, "step": 52230 }, { "grad_norm": 0.37220582365989685, "learning_rate": 5.043155026423219e-05, "loss": 0.039, "step": 52240 }, { "grad_norm": 0.311348021030426, "learning_rate": 5.041501615864137e-05, "loss": 0.0323, "step": 52250 }, { "grad_norm": 0.4410335123538971, "learning_rate": 5.039848200766502e-05, "loss": 0.0248, "step": 52260 }, { "grad_norm": 0.4658949375152588, "learning_rate": 5.038194781311125e-05, "loss": 0.0362, "step": 52270 }, { "grad_norm": 0.47291243076324463, "learning_rate": 5.0365413576788264e-05, "loss": 0.0321, "step": 52280 }, { "grad_norm": 0.5425276160240173, "learning_rate": 5.0348879300504194e-05, "loss": 0.0464, "step": 52290 }, { "grad_norm": 0.4626060724258423, "learning_rate": 5.033234498606721e-05, "loss": 0.0269, "step": 52300 }, { "grad_norm": 0.545005202293396, "learning_rate": 5.0315810635285476e-05, "loss": 0.0292, "step": 52310 }, { "grad_norm": 0.4735749065876007, "learning_rate": 5.0299276249967164e-05, "loss": 0.0246, "step": 52320 }, { "grad_norm": 0.4033064842224121, "learning_rate": 5.0282741831920454e-05, "loss": 0.0263, "step": 52330 }, { "grad_norm": 0.6180170178413391, "learning_rate": 5.026620738295351e-05, "loss": 0.034, "step": 52340 }, { "grad_norm": 0.3470180332660675, "learning_rate": 5.024967290487455e-05, "loss": 0.0305, "step": 52350 }, { "grad_norm": 0.47019124031066895, "learning_rate": 5.023313839949172e-05, "loss": 0.0331, "step": 52360 }, { "grad_norm": 0.5143560767173767, "learning_rate": 5.0216603868613224e-05, "loss": 0.035, "step": 52370 }, { "grad_norm": 0.32493284344673157, "learning_rate": 5.020006931404726e-05, "loss": 0.031, "step": 52380 }, { "grad_norm": 0.416568785905838, "learning_rate": 5.0183534737602014e-05, "loss": 0.0338, "step": 52390 }, { "grad_norm": 0.3232403099536896, "learning_rate": 5.0167000141085686e-05, "loss": 0.0333, "step": 52400 }, { "grad_norm": 0.7791218757629395, "learning_rate": 5.015046552630647e-05, "loss": 0.0267, "step": 52410 }, { "grad_norm": 0.5984809994697571, "learning_rate": 5.013393089507258e-05, "loss": 0.037, "step": 52420 }, { "grad_norm": 0.5752379298210144, "learning_rate": 5.011739624919219e-05, "loss": 0.0435, "step": 52430 }, { "grad_norm": 0.5320286154747009, "learning_rate": 5.0100861590473545e-05, "loss": 0.0388, "step": 52440 }, { "grad_norm": 0.566378653049469, "learning_rate": 5.008432692072482e-05, "loss": 0.0465, "step": 52450 }, { "grad_norm": 0.6605237126350403, "learning_rate": 5.006779224175423e-05, "loss": 0.0269, "step": 52460 }, { "grad_norm": 0.593847393989563, "learning_rate": 5.005125755536998e-05, "loss": 0.047, "step": 52470 }, { "grad_norm": 0.6780304908752441, "learning_rate": 5.003472286338029e-05, "loss": 0.0364, "step": 52480 }, { "grad_norm": 0.22948729991912842, "learning_rate": 5.001818816759335e-05, "loss": 0.0233, "step": 52490 }, { "grad_norm": 0.6728953719139099, "learning_rate": 5.000165346981738e-05, "loss": 0.0566, "step": 52500 }, { "grad_norm": 0.6249719858169556, "learning_rate": 4.9985118771860595e-05, "loss": 0.0463, "step": 52510 }, { "grad_norm": 0.5262007713317871, "learning_rate": 4.996858407553119e-05, "loss": 0.0524, "step": 52520 }, { "grad_norm": 0.7346469759941101, "learning_rate": 4.99520493826374e-05, "loss": 0.0534, "step": 52530 }, { "grad_norm": 0.35633236169815063, "learning_rate": 4.9935514694987384e-05, "loss": 0.0273, "step": 52540 }, { "grad_norm": 0.6342657208442688, "learning_rate": 4.991898001438941e-05, "loss": 0.0631, "step": 52550 }, { "grad_norm": 0.433599054813385, "learning_rate": 4.990244534265166e-05, "loss": 0.0327, "step": 52560 }, { "grad_norm": 0.8467435836791992, "learning_rate": 4.988591068158234e-05, "loss": 0.0522, "step": 52570 }, { "grad_norm": 0.37215456366539, "learning_rate": 4.986937603298965e-05, "loss": 0.0561, "step": 52580 }, { "grad_norm": 0.5543404817581177, "learning_rate": 4.985284139868181e-05, "loss": 0.0438, "step": 52590 }, { "grad_norm": 0.6034116148948669, "learning_rate": 4.9836306780467e-05, "loss": 0.0449, "step": 52600 }, { "grad_norm": 0.4375680685043335, "learning_rate": 4.981977218015343e-05, "loss": 0.0357, "step": 52610 }, { "grad_norm": 0.4713754653930664, "learning_rate": 4.980323759954932e-05, "loss": 0.0536, "step": 52620 }, { "grad_norm": 0.5100083947181702, "learning_rate": 4.978670304046284e-05, "loss": 0.0345, "step": 52630 }, { "grad_norm": 0.713539719581604, "learning_rate": 4.9770168504702195e-05, "loss": 0.0402, "step": 52640 }, { "grad_norm": 0.6782763600349426, "learning_rate": 4.975363399407559e-05, "loss": 0.0484, "step": 52650 }, { "grad_norm": 0.6228240132331848, "learning_rate": 4.9737099510391194e-05, "loss": 0.0326, "step": 52660 }, { "grad_norm": 0.6250206232070923, "learning_rate": 4.972056505545719e-05, "loss": 0.0387, "step": 52670 }, { "grad_norm": 0.3012140691280365, "learning_rate": 4.9704030631081775e-05, "loss": 0.0386, "step": 52680 }, { "grad_norm": 0.4517218768596649, "learning_rate": 4.968749623907313e-05, "loss": 0.0304, "step": 52690 }, { "grad_norm": 0.33448442816734314, "learning_rate": 4.967096188123943e-05, "loss": 0.0278, "step": 52700 }, { "grad_norm": 0.2634274363517761, "learning_rate": 4.965442755938884e-05, "loss": 0.0299, "step": 52710 }, { "grad_norm": 0.3478478491306305, "learning_rate": 4.963789327532954e-05, "loss": 0.0401, "step": 52720 }, { "grad_norm": 0.35660743713378906, "learning_rate": 4.962135903086968e-05, "loss": 0.0227, "step": 52730 }, { "grad_norm": 0.34876805543899536, "learning_rate": 4.960482482781741e-05, "loss": 0.0312, "step": 52740 }, { "grad_norm": 0.6525951027870178, "learning_rate": 4.958829066798092e-05, "loss": 0.0364, "step": 52750 }, { "grad_norm": 0.7319929003715515, "learning_rate": 4.957175655316835e-05, "loss": 0.0271, "step": 52760 }, { "grad_norm": 0.6901270747184753, "learning_rate": 4.9555222485187834e-05, "loss": 0.0423, "step": 52770 }, { "grad_norm": 0.6259992718696594, "learning_rate": 4.953868846584752e-05, "loss": 0.034, "step": 52780 }, { "grad_norm": 0.5035828351974487, "learning_rate": 4.9522154496955544e-05, "loss": 0.0331, "step": 52790 }, { "grad_norm": 0.4892840087413788, "learning_rate": 4.950562058032001e-05, "loss": 0.0263, "step": 52800 }, { "grad_norm": 0.5221060514450073, "learning_rate": 4.948908671774909e-05, "loss": 0.0327, "step": 52810 }, { "grad_norm": 0.38518592715263367, "learning_rate": 4.947255291105087e-05, "loss": 0.0268, "step": 52820 }, { "grad_norm": 0.2642107903957367, "learning_rate": 4.945601916203348e-05, "loss": 0.0191, "step": 52830 }, { "grad_norm": 0.3245396912097931, "learning_rate": 4.9439485472505016e-05, "loss": 0.0334, "step": 52840 }, { "grad_norm": 0.4102567136287689, "learning_rate": 4.942295184427356e-05, "loss": 0.0354, "step": 52850 }, { "grad_norm": 0.5198910236358643, "learning_rate": 4.9406418279147234e-05, "loss": 0.0311, "step": 52860 }, { "grad_norm": 0.35041362047195435, "learning_rate": 4.9389884778934116e-05, "loss": 0.0379, "step": 52870 }, { "grad_norm": 0.4181135892868042, "learning_rate": 4.937335134544228e-05, "loss": 0.0248, "step": 52880 }, { "grad_norm": 0.5102434754371643, "learning_rate": 4.935681798047979e-05, "loss": 0.0479, "step": 52890 }, { "grad_norm": 0.3940427899360657, "learning_rate": 4.934028468585473e-05, "loss": 0.0233, "step": 52900 }, { "grad_norm": 0.37114688754081726, "learning_rate": 4.932375146337514e-05, "loss": 0.0302, "step": 52910 }, { "grad_norm": 0.5510526299476624, "learning_rate": 4.930721831484908e-05, "loss": 0.0443, "step": 52920 }, { "grad_norm": 0.46958616375923157, "learning_rate": 4.929068524208456e-05, "loss": 0.0365, "step": 52930 }, { "grad_norm": 0.6083042621612549, "learning_rate": 4.927415224688965e-05, "loss": 0.033, "step": 52940 }, { "grad_norm": 0.30038946866989136, "learning_rate": 4.9257619331072375e-05, "loss": 0.023, "step": 52950 }, { "grad_norm": 0.45151886343955994, "learning_rate": 4.924108649644072e-05, "loss": 0.0325, "step": 52960 }, { "grad_norm": 0.5277974605560303, "learning_rate": 4.922455374480271e-05, "loss": 0.0328, "step": 52970 }, { "grad_norm": 0.34347105026245117, "learning_rate": 4.920802107796634e-05, "loss": 0.0328, "step": 52980 }, { "grad_norm": 0.37708941102027893, "learning_rate": 4.9191488497739564e-05, "loss": 0.0388, "step": 52990 }, { "grad_norm": 0.5151813626289368, "learning_rate": 4.917495600593042e-05, "loss": 0.0368, "step": 53000 }, { "grad_norm": 0.38570651412010193, "learning_rate": 4.915842360434684e-05, "loss": 0.0555, "step": 53010 }, { "grad_norm": 0.6257079839706421, "learning_rate": 4.914189129479679e-05, "loss": 0.0308, "step": 53020 }, { "grad_norm": 0.24730969965457916, "learning_rate": 4.912535907908822e-05, "loss": 0.0292, "step": 53030 }, { "grad_norm": 0.5121248364448547, "learning_rate": 4.910882695902907e-05, "loss": 0.0437, "step": 53040 }, { "grad_norm": 0.2766617238521576, "learning_rate": 4.9092294936427234e-05, "loss": 0.0341, "step": 53050 }, { "grad_norm": 0.42447033524513245, "learning_rate": 4.9075763013090685e-05, "loss": 0.04, "step": 53060 }, { "grad_norm": 0.3607490360736847, "learning_rate": 4.9059231190827305e-05, "loss": 0.0388, "step": 53070 }, { "grad_norm": 0.5345454216003418, "learning_rate": 4.904269947144499e-05, "loss": 0.0363, "step": 53080 }, { "grad_norm": 0.528853714466095, "learning_rate": 4.902616785675161e-05, "loss": 0.032, "step": 53090 }, { "grad_norm": 0.630715548992157, "learning_rate": 4.900963634855505e-05, "loss": 0.0351, "step": 53100 }, { "grad_norm": 0.6120406985282898, "learning_rate": 4.8993104948663173e-05, "loss": 0.0421, "step": 53110 }, { "grad_norm": 1.0366698503494263, "learning_rate": 4.8976573658883823e-05, "loss": 0.0413, "step": 53120 }, { "grad_norm": 0.4998018741607666, "learning_rate": 4.896004248102484e-05, "loss": 0.0332, "step": 53130 }, { "grad_norm": 0.41688933968544006, "learning_rate": 4.8943511416894045e-05, "loss": 0.0252, "step": 53140 }, { "grad_norm": 0.43431583046913147, "learning_rate": 4.892698046829926e-05, "loss": 0.032, "step": 53150 }, { "grad_norm": 0.4211648106575012, "learning_rate": 4.8910449637048275e-05, "loss": 0.029, "step": 53160 }, { "grad_norm": 0.6614252924919128, "learning_rate": 4.8893918924948884e-05, "loss": 0.0347, "step": 53170 }, { "grad_norm": 0.5990672707557678, "learning_rate": 4.8877388333808835e-05, "loss": 0.0287, "step": 53180 }, { "grad_norm": 0.37627124786376953, "learning_rate": 4.8860857865435925e-05, "loss": 0.0324, "step": 53190 }, { "grad_norm": 0.9666886925697327, "learning_rate": 4.884432752163789e-05, "loss": 0.0337, "step": 53200 }, { "grad_norm": 0.3994467258453369, "learning_rate": 4.8827797304222466e-05, "loss": 0.0251, "step": 53210 }, { "grad_norm": 0.27115514874458313, "learning_rate": 4.8811267214997366e-05, "loss": 0.0319, "step": 53220 }, { "grad_norm": 0.34428974986076355, "learning_rate": 4.879473725577029e-05, "loss": 0.0294, "step": 53230 }, { "grad_norm": 0.32977601885795593, "learning_rate": 4.877820742834891e-05, "loss": 0.0452, "step": 53240 }, { "grad_norm": 0.2793072462081909, "learning_rate": 4.8761677734540956e-05, "loss": 0.0431, "step": 53250 }, { "grad_norm": 0.5745869278907776, "learning_rate": 4.874514817615407e-05, "loss": 0.0264, "step": 53260 }, { "grad_norm": 0.44278717041015625, "learning_rate": 4.872861875499588e-05, "loss": 0.0279, "step": 53270 }, { "grad_norm": 0.4116177558898926, "learning_rate": 4.871208947287404e-05, "loss": 0.0245, "step": 53280 }, { "grad_norm": 0.5040559768676758, "learning_rate": 4.869556033159615e-05, "loss": 0.0262, "step": 53290 }, { "grad_norm": 0.30094674229621887, "learning_rate": 4.8679031332969835e-05, "loss": 0.0213, "step": 53300 }, { "grad_norm": 0.42743006348609924, "learning_rate": 4.8662502478802635e-05, "loss": 0.033, "step": 53310 }, { "grad_norm": 0.43925940990448, "learning_rate": 4.8645973770902176e-05, "loss": 0.0355, "step": 53320 }, { "grad_norm": 0.23542699217796326, "learning_rate": 4.862944521107599e-05, "loss": 0.026, "step": 53330 }, { "grad_norm": 0.6917351484298706, "learning_rate": 4.861291680113162e-05, "loss": 0.0254, "step": 53340 }, { "grad_norm": 0.40775609016418457, "learning_rate": 4.8596388542876563e-05, "loss": 0.0356, "step": 53350 }, { "grad_norm": 0.42677736282348633, "learning_rate": 4.857986043811836e-05, "loss": 0.031, "step": 53360 }, { "grad_norm": 0.5736457705497742, "learning_rate": 4.856333248866448e-05, "loss": 0.0308, "step": 53370 }, { "grad_norm": 0.7806362509727478, "learning_rate": 4.85468046963224e-05, "loss": 0.0358, "step": 53380 }, { "grad_norm": 0.474711537361145, "learning_rate": 4.853027706289957e-05, "loss": 0.0441, "step": 53390 }, { "grad_norm": 0.7633859515190125, "learning_rate": 4.851374959020344e-05, "loss": 0.0463, "step": 53400 }, { "grad_norm": 0.42871391773223877, "learning_rate": 4.849722228004141e-05, "loss": 0.0353, "step": 53410 }, { "grad_norm": 0.36887428164482117, "learning_rate": 4.848069513422091e-05, "loss": 0.0365, "step": 53420 }, { "grad_norm": 0.2698582708835602, "learning_rate": 4.8464168154549294e-05, "loss": 0.0348, "step": 53430 }, { "grad_norm": 0.4149673283100128, "learning_rate": 4.844764134283392e-05, "loss": 0.0314, "step": 53440 }, { "grad_norm": 0.4317920207977295, "learning_rate": 4.843111470088218e-05, "loss": 0.0301, "step": 53450 }, { "grad_norm": 0.41115760803222656, "learning_rate": 4.841458823050137e-05, "loss": 0.0239, "step": 53460 }, { "grad_norm": 0.49034827947616577, "learning_rate": 4.8398061933498816e-05, "loss": 0.0368, "step": 53470 }, { "grad_norm": 0.403840571641922, "learning_rate": 4.83815358116818e-05, "loss": 0.0283, "step": 53480 }, { "grad_norm": 0.48838427662849426, "learning_rate": 4.8365009866857584e-05, "loss": 0.0269, "step": 53490 }, { "grad_norm": 0.46319252252578735, "learning_rate": 4.834848410083342e-05, "loss": 0.0431, "step": 53500 }, { "grad_norm": 0.6565258502960205, "learning_rate": 4.833195851541657e-05, "loss": 0.0288, "step": 53510 }, { "grad_norm": 0.36424344778060913, "learning_rate": 4.831543311241423e-05, "loss": 0.0335, "step": 53520 }, { "grad_norm": 0.5051030516624451, "learning_rate": 4.829890789363359e-05, "loss": 0.0287, "step": 53530 }, { "grad_norm": 0.4036886394023895, "learning_rate": 4.828238286088182e-05, "loss": 0.0269, "step": 53540 }, { "grad_norm": 0.5589853525161743, "learning_rate": 4.826585801596609e-05, "loss": 0.0345, "step": 53550 }, { "grad_norm": 0.7147602438926697, "learning_rate": 4.8249333360693486e-05, "loss": 0.0308, "step": 53560 }, { "grad_norm": 0.7478481531143188, "learning_rate": 4.8232808896871185e-05, "loss": 0.0422, "step": 53570 }, { "grad_norm": 0.22371327877044678, "learning_rate": 4.821628462630624e-05, "loss": 0.0219, "step": 53580 }, { "grad_norm": 0.3776855170726776, "learning_rate": 4.8199760550805735e-05, "loss": 0.0335, "step": 53590 }, { "grad_norm": 1.303812026977539, "learning_rate": 4.818323667217669e-05, "loss": 0.0357, "step": 53600 }, { "grad_norm": 0.39915502071380615, "learning_rate": 4.8166712992226165e-05, "loss": 0.0309, "step": 53610 }, { "grad_norm": 0.39612290263175964, "learning_rate": 4.8150189512761154e-05, "loss": 0.0308, "step": 53620 }, { "grad_norm": 0.42400258779525757, "learning_rate": 4.813366623558862e-05, "loss": 0.0364, "step": 53630 }, { "grad_norm": 0.4030560851097107, "learning_rate": 4.811714316251554e-05, "loss": 0.034, "step": 53640 }, { "grad_norm": 0.5010875463485718, "learning_rate": 4.810062029534886e-05, "loss": 0.0229, "step": 53650 }, { "grad_norm": 0.5241807699203491, "learning_rate": 4.8084097635895505e-05, "loss": 0.0386, "step": 53660 }, { "grad_norm": 0.4566687047481537, "learning_rate": 4.8067575185962335e-05, "loss": 0.0302, "step": 53670 }, { "grad_norm": 0.5652801394462585, "learning_rate": 4.805105294735625e-05, "loss": 0.0432, "step": 53680 }, { "grad_norm": 0.2545115351676941, "learning_rate": 4.803453092188406e-05, "loss": 0.0307, "step": 53690 }, { "grad_norm": 0.7347168326377869, "learning_rate": 4.801800911135263e-05, "loss": 0.0371, "step": 53700 }, { "grad_norm": 0.2697025239467621, "learning_rate": 4.8001487517568744e-05, "loss": 0.0364, "step": 53710 }, { "grad_norm": 0.2992554008960724, "learning_rate": 4.798496614233918e-05, "loss": 0.0316, "step": 53720 }, { "grad_norm": 0.26931124925613403, "learning_rate": 4.796844498747069e-05, "loss": 0.0283, "step": 53730 }, { "grad_norm": 0.4026371240615845, "learning_rate": 4.795192405477e-05, "loss": 0.0259, "step": 53740 }, { "grad_norm": 0.45361563563346863, "learning_rate": 4.79354033460438e-05, "loss": 0.0294, "step": 53750 }, { "grad_norm": 0.47981610894203186, "learning_rate": 4.79188828630988e-05, "loss": 0.0418, "step": 53760 }, { "grad_norm": 0.3103068172931671, "learning_rate": 4.790236260774165e-05, "loss": 0.0317, "step": 53770 }, { "grad_norm": 0.38389357924461365, "learning_rate": 4.788584258177896e-05, "loss": 0.0278, "step": 53780 }, { "grad_norm": 0.46879518032073975, "learning_rate": 4.7869322787017356e-05, "loss": 0.0326, "step": 53790 }, { "grad_norm": 0.5191639065742493, "learning_rate": 4.78528032252634e-05, "loss": 0.0307, "step": 53800 }, { "grad_norm": 0.5070299506187439, "learning_rate": 4.783628389832366e-05, "loss": 0.0456, "step": 53810 }, { "grad_norm": 0.28753888607025146, "learning_rate": 4.7819764808004644e-05, "loss": 0.0388, "step": 53820 }, { "grad_norm": 0.6620809435844421, "learning_rate": 4.780324595611289e-05, "loss": 0.033, "step": 53830 }, { "grad_norm": 0.5189014077186584, "learning_rate": 4.7786727344454854e-05, "loss": 0.0372, "step": 53840 }, { "grad_norm": 0.400642454624176, "learning_rate": 4.777020897483697e-05, "loss": 0.0344, "step": 53850 }, { "grad_norm": 0.6429850459098816, "learning_rate": 4.775369084906569e-05, "loss": 0.0317, "step": 53860 }, { "grad_norm": 0.6631628274917603, "learning_rate": 4.773717296894742e-05, "loss": 0.0488, "step": 53870 }, { "grad_norm": 0.7786402702331543, "learning_rate": 4.772065533628848e-05, "loss": 0.0535, "step": 53880 }, { "grad_norm": 0.4650348126888275, "learning_rate": 4.7704137952895265e-05, "loss": 0.0379, "step": 53890 }, { "grad_norm": 0.43824639916419983, "learning_rate": 4.768762082057407e-05, "loss": 0.0313, "step": 53900 }, { "grad_norm": 0.673926055431366, "learning_rate": 4.7671103941131186e-05, "loss": 0.0295, "step": 53910 }, { "grad_norm": 0.33418595790863037, "learning_rate": 4.765458731637287e-05, "loss": 0.0349, "step": 53920 }, { "grad_norm": 0.4565182626247406, "learning_rate": 4.763807094810537e-05, "loss": 0.0501, "step": 53930 }, { "grad_norm": 0.3462187349796295, "learning_rate": 4.762155483813487e-05, "loss": 0.0361, "step": 53940 }, { "grad_norm": 0.701815664768219, "learning_rate": 4.760503898826754e-05, "loss": 0.031, "step": 53950 }, { "grad_norm": 0.49150919914245605, "learning_rate": 4.758852340030957e-05, "loss": 0.0215, "step": 53960 }, { "grad_norm": 0.8293952345848083, "learning_rate": 4.757200807606705e-05, "loss": 0.0421, "step": 53970 }, { "grad_norm": 0.628908634185791, "learning_rate": 4.755549301734607e-05, "loss": 0.0402, "step": 53980 }, { "grad_norm": 0.7534639835357666, "learning_rate": 4.753897822595271e-05, "loss": 0.0285, "step": 53990 }, { "grad_norm": 0.6237165331840515, "learning_rate": 4.7522463703692994e-05, "loss": 0.0289, "step": 54000 }, { "grad_norm": 0.3210429251194, "learning_rate": 4.75059494523729e-05, "loss": 0.0281, "step": 54010 }, { "grad_norm": 0.4722844064235687, "learning_rate": 4.748943547379843e-05, "loss": 0.0226, "step": 54020 }, { "grad_norm": 0.6799517869949341, "learning_rate": 4.7472921769775535e-05, "loss": 0.0324, "step": 54030 }, { "grad_norm": 0.5310957431793213, "learning_rate": 4.745640834211011e-05, "loss": 0.0349, "step": 54040 }, { "grad_norm": 0.43615391850471497, "learning_rate": 4.7439895192608054e-05, "loss": 0.0223, "step": 54050 }, { "grad_norm": 0.4918540120124817, "learning_rate": 4.7423382323075186e-05, "loss": 0.0253, "step": 54060 }, { "grad_norm": 0.5499719977378845, "learning_rate": 4.740686973531737e-05, "loss": 0.0368, "step": 54070 }, { "grad_norm": 0.26072442531585693, "learning_rate": 4.7390357431140375e-05, "loss": 0.0457, "step": 54080 }, { "grad_norm": 0.2469758689403534, "learning_rate": 4.737384541234998e-05, "loss": 0.0216, "step": 54090 }, { "grad_norm": 0.5148982405662537, "learning_rate": 4.735733368075189e-05, "loss": 0.0358, "step": 54100 }, { "grad_norm": 0.5353986024856567, "learning_rate": 4.734082223815182e-05, "loss": 0.0337, "step": 54110 }, { "grad_norm": 0.3125280737876892, "learning_rate": 4.7324311086355436e-05, "loss": 0.0224, "step": 54120 }, { "grad_norm": 0.29935839772224426, "learning_rate": 4.730780022716836e-05, "loss": 0.0313, "step": 54130 }, { "grad_norm": 0.558957576751709, "learning_rate": 4.729128966239619e-05, "loss": 0.032, "step": 54140 }, { "grad_norm": 0.5646008253097534, "learning_rate": 4.7274779393844536e-05, "loss": 0.0272, "step": 54150 }, { "grad_norm": 0.4434354603290558, "learning_rate": 4.725826942331891e-05, "loss": 0.0272, "step": 54160 }, { "grad_norm": 0.6837260127067566, "learning_rate": 4.724175975262481e-05, "loss": 0.0367, "step": 54170 }, { "grad_norm": 0.2539790868759155, "learning_rate": 4.722525038356772e-05, "loss": 0.0261, "step": 54180 }, { "grad_norm": 0.5105404853820801, "learning_rate": 4.7208741317953075e-05, "loss": 0.0311, "step": 54190 }, { "grad_norm": 0.41080254316329956, "learning_rate": 4.7192232557586264e-05, "loss": 0.0227, "step": 54200 }, { "grad_norm": 0.5968176126480103, "learning_rate": 4.717572410427269e-05, "loss": 0.0258, "step": 54210 }, { "grad_norm": 0.632993221282959, "learning_rate": 4.71592159598177e-05, "loss": 0.0361, "step": 54220 }, { "grad_norm": 0.43754151463508606, "learning_rate": 4.714270812602657e-05, "loss": 0.033, "step": 54230 }, { "grad_norm": 0.37339547276496887, "learning_rate": 4.712620060470458e-05, "loss": 0.0362, "step": 54240 }, { "grad_norm": 0.4144604206085205, "learning_rate": 4.7109693397656986e-05, "loss": 0.0291, "step": 54250 }, { "grad_norm": 0.52293860912323, "learning_rate": 4.709318650668894e-05, "loss": 0.0344, "step": 54260 }, { "grad_norm": 0.6565057635307312, "learning_rate": 4.7076679933605676e-05, "loss": 0.0405, "step": 54270 }, { "grad_norm": 0.4770655333995819, "learning_rate": 4.70601736802123e-05, "loss": 0.0326, "step": 54280 }, { "grad_norm": 0.455731064081192, "learning_rate": 4.704366774831391e-05, "loss": 0.027, "step": 54290 }, { "grad_norm": 0.34233197569847107, "learning_rate": 4.702716213971557e-05, "loss": 0.0236, "step": 54300 }, { "grad_norm": 0.529655933380127, "learning_rate": 4.70106568562223e-05, "loss": 0.0291, "step": 54310 }, { "grad_norm": 1.734764575958252, "learning_rate": 4.6994151899639105e-05, "loss": 0.0359, "step": 54320 }, { "grad_norm": 0.5971866250038147, "learning_rate": 4.697764727177093e-05, "loss": 0.0345, "step": 54330 }, { "grad_norm": 0.6168492436408997, "learning_rate": 4.696114297442272e-05, "loss": 0.0308, "step": 54340 }, { "grad_norm": 0.49029433727264404, "learning_rate": 4.694463900939933e-05, "loss": 0.0318, "step": 54350 }, { "grad_norm": 0.6721840500831604, "learning_rate": 4.692813537850564e-05, "loss": 0.047, "step": 54360 }, { "grad_norm": 0.727932870388031, "learning_rate": 4.6911632083546445e-05, "loss": 0.0409, "step": 54370 }, { "grad_norm": 0.819068431854248, "learning_rate": 4.6895129126326525e-05, "loss": 0.0361, "step": 54380 }, { "grad_norm": 0.7240658402442932, "learning_rate": 4.687862650865059e-05, "loss": 0.0457, "step": 54390 }, { "grad_norm": 0.39427322149276733, "learning_rate": 4.686212423232339e-05, "loss": 0.0437, "step": 54400 }, { "grad_norm": 0.4496326744556427, "learning_rate": 4.6845622299149565e-05, "loss": 0.0356, "step": 54410 }, { "grad_norm": 0.43093565106391907, "learning_rate": 4.682912071093374e-05, "loss": 0.0254, "step": 54420 }, { "grad_norm": 0.301481693983078, "learning_rate": 4.681261946948052e-05, "loss": 0.0255, "step": 54430 }, { "grad_norm": 0.7523221969604492, "learning_rate": 4.679611857659443e-05, "loss": 0.0325, "step": 54440 }, { "grad_norm": 0.5204411149024963, "learning_rate": 4.6779618034079975e-05, "loss": 0.0281, "step": 54450 }, { "grad_norm": 0.5845060348510742, "learning_rate": 4.6763117843741675e-05, "loss": 0.0281, "step": 54460 }, { "grad_norm": 0.5645264983177185, "learning_rate": 4.674661800738394e-05, "loss": 0.0275, "step": 54470 }, { "grad_norm": 0.3217347264289856, "learning_rate": 4.673011852681117e-05, "loss": 0.0309, "step": 54480 }, { "grad_norm": 0.6871273517608643, "learning_rate": 4.671361940382772e-05, "loss": 0.0651, "step": 54490 }, { "grad_norm": 0.7212827801704407, "learning_rate": 4.669712064023791e-05, "loss": 0.0338, "step": 54500 }, { "grad_norm": 0.4060833752155304, "learning_rate": 4.668062223784602e-05, "loss": 0.023, "step": 54510 }, { "grad_norm": 0.6155139803886414, "learning_rate": 4.6664124198456274e-05, "loss": 0.0274, "step": 54520 }, { "grad_norm": 0.4018639028072357, "learning_rate": 4.6647626523872905e-05, "loss": 0.0239, "step": 54530 }, { "grad_norm": 0.23201990127563477, "learning_rate": 4.6631129215900066e-05, "loss": 0.0253, "step": 54540 }, { "grad_norm": 0.18408025801181793, "learning_rate": 4.661463227634187e-05, "loss": 0.0237, "step": 54550 }, { "grad_norm": 0.6620102524757385, "learning_rate": 4.659813570700238e-05, "loss": 0.0373, "step": 54560 }, { "grad_norm": 0.2456628531217575, "learning_rate": 4.658163950968568e-05, "loss": 0.027, "step": 54570 }, { "grad_norm": 0.5423017144203186, "learning_rate": 4.6565143686195726e-05, "loss": 0.0229, "step": 54580 }, { "grad_norm": 0.3303506672382355, "learning_rate": 4.654864823833649e-05, "loss": 0.0253, "step": 54590 }, { "grad_norm": 0.4381656050682068, "learning_rate": 4.65321531679119e-05, "loss": 0.03, "step": 54600 }, { "grad_norm": 0.5438292622566223, "learning_rate": 4.6515658476725834e-05, "loss": 0.0385, "step": 54610 }, { "grad_norm": 0.5479652285575867, "learning_rate": 4.649916416658212e-05, "loss": 0.0381, "step": 54620 }, { "grad_norm": 0.29149237275123596, "learning_rate": 4.648267023928455e-05, "loss": 0.033, "step": 54630 }, { "grad_norm": 0.5014554262161255, "learning_rate": 4.646617669663687e-05, "loss": 0.0417, "step": 54640 }, { "grad_norm": 0.4071817398071289, "learning_rate": 4.644968354044277e-05, "loss": 0.0254, "step": 54650 }, { "grad_norm": 0.35969987511634827, "learning_rate": 4.643319077250596e-05, "loss": 0.0263, "step": 54660 }, { "grad_norm": 0.3914814591407776, "learning_rate": 4.6416698394630055e-05, "loss": 0.0258, "step": 54670 }, { "grad_norm": 0.5296677350997925, "learning_rate": 4.640020640861862e-05, "loss": 0.026, "step": 54680 }, { "grad_norm": 0.6010584235191345, "learning_rate": 4.6383714816275195e-05, "loss": 0.0392, "step": 54690 }, { "grad_norm": 0.43746015429496765, "learning_rate": 4.636722361940329e-05, "loss": 0.0299, "step": 54700 }, { "grad_norm": 0.41235825419425964, "learning_rate": 4.635073281980632e-05, "loss": 0.0379, "step": 54710 }, { "grad_norm": 0.5155618190765381, "learning_rate": 4.633424241928775e-05, "loss": 0.0285, "step": 54720 }, { "grad_norm": 0.5708706974983215, "learning_rate": 4.6317752419650914e-05, "loss": 0.0326, "step": 54730 }, { "grad_norm": 0.4343256950378418, "learning_rate": 4.630126282269914e-05, "loss": 0.0248, "step": 54740 }, { "grad_norm": 0.45706042647361755, "learning_rate": 4.628477363023569e-05, "loss": 0.0309, "step": 54750 }, { "grad_norm": 0.47956958413124084, "learning_rate": 4.626828484406383e-05, "loss": 0.0401, "step": 54760 }, { "grad_norm": 0.47034314274787903, "learning_rate": 4.625179646598669e-05, "loss": 0.0342, "step": 54770 }, { "grad_norm": 0.41612696647644043, "learning_rate": 4.6235308497807487e-05, "loss": 0.0227, "step": 54780 }, { "grad_norm": 0.26538145542144775, "learning_rate": 4.6218820941329276e-05, "loss": 0.0264, "step": 54790 }, { "grad_norm": 0.4885602593421936, "learning_rate": 4.620233379835513e-05, "loss": 0.0234, "step": 54800 }, { "grad_norm": 0.5580989122390747, "learning_rate": 4.618584707068804e-05, "loss": 0.0417, "step": 54810 }, { "grad_norm": 0.3862396776676178, "learning_rate": 4.616936076013099e-05, "loss": 0.0326, "step": 54820 }, { "grad_norm": 0.32268381118774414, "learning_rate": 4.615287486848688e-05, "loss": 0.024, "step": 54830 }, { "grad_norm": 0.3621905446052551, "learning_rate": 4.6136389397558584e-05, "loss": 0.0329, "step": 54840 }, { "grad_norm": 0.6265318393707275, "learning_rate": 4.6119904349148935e-05, "loss": 0.0313, "step": 54850 }, { "grad_norm": 0.5026485323905945, "learning_rate": 4.6103419725060726e-05, "loss": 0.031, "step": 54860 }, { "grad_norm": 0.5680094361305237, "learning_rate": 4.608693552709667e-05, "loss": 0.0345, "step": 54870 }, { "grad_norm": 0.7981262803077698, "learning_rate": 4.607045175705947e-05, "loss": 0.0364, "step": 54880 }, { "grad_norm": 0.43815377354621887, "learning_rate": 4.605396841675176e-05, "loss": 0.0312, "step": 54890 }, { "grad_norm": 0.367766410112381, "learning_rate": 4.6037485507976094e-05, "loss": 0.0304, "step": 54900 }, { "grad_norm": 0.33387795090675354, "learning_rate": 4.602100303253509e-05, "loss": 0.0266, "step": 54910 }, { "grad_norm": 0.35395076870918274, "learning_rate": 4.600452099223122e-05, "loss": 0.0302, "step": 54920 }, { "grad_norm": 0.5412142276763916, "learning_rate": 4.598803938886691e-05, "loss": 0.0329, "step": 54930 }, { "grad_norm": 0.4756264090538025, "learning_rate": 4.597155822424459e-05, "loss": 0.0491, "step": 54940 }, { "grad_norm": 0.6406821608543396, "learning_rate": 4.59550775001666e-05, "loss": 0.0344, "step": 54950 }, { "grad_norm": 0.521117627620697, "learning_rate": 4.5938597218435234e-05, "loss": 0.0297, "step": 54960 }, { "grad_norm": 0.5295270085334778, "learning_rate": 4.592211738085279e-05, "loss": 0.0329, "step": 54970 }, { "grad_norm": 0.5285102128982544, "learning_rate": 4.590563798922146e-05, "loss": 0.0412, "step": 54980 }, { "grad_norm": 0.26951801776885986, "learning_rate": 4.5889159045343404e-05, "loss": 0.0289, "step": 54990 }, { "grad_norm": 0.45374637842178345, "learning_rate": 4.5872680551020734e-05, "loss": 0.0507, "step": 55000 }, { "grad_norm": 0.36069291830062866, "learning_rate": 4.585620250805552e-05, "loss": 0.0349, "step": 55010 }, { "grad_norm": 0.7530139684677124, "learning_rate": 4.583972491824975e-05, "loss": 0.0365, "step": 55020 }, { "grad_norm": 0.358264684677124, "learning_rate": 4.582324778340541e-05, "loss": 0.0278, "step": 55030 }, { "grad_norm": 0.3540635406970978, "learning_rate": 4.5806771105324424e-05, "loss": 0.0494, "step": 55040 }, { "grad_norm": 0.42629683017730713, "learning_rate": 4.579029488580865e-05, "loss": 0.0368, "step": 55050 }, { "grad_norm": 0.5879757404327393, "learning_rate": 4.5773819126659876e-05, "loss": 0.0368, "step": 55060 }, { "grad_norm": 0.3961437940597534, "learning_rate": 4.575734382967991e-05, "loss": 0.0415, "step": 55070 }, { "grad_norm": 0.4083178639411926, "learning_rate": 4.574086899667044e-05, "loss": 0.022, "step": 55080 }, { "grad_norm": 0.5232065320014954, "learning_rate": 4.5724394629433106e-05, "loss": 0.0358, "step": 55090 }, { "grad_norm": 0.7170877456665039, "learning_rate": 4.570792072976958e-05, "loss": 0.0416, "step": 55100 }, { "grad_norm": 0.37021446228027344, "learning_rate": 4.569144729948138e-05, "loss": 0.027, "step": 55110 }, { "grad_norm": 0.42416149377822876, "learning_rate": 4.5674974340370033e-05, "loss": 0.0239, "step": 55120 }, { "grad_norm": 0.5200952291488647, "learning_rate": 4.565850185423699e-05, "loss": 0.0325, "step": 55130 }, { "grad_norm": 0.6199479103088379, "learning_rate": 4.564202984288365e-05, "loss": 0.0285, "step": 55140 }, { "grad_norm": 0.5867946147918701, "learning_rate": 4.5625558308111356e-05, "loss": 0.0247, "step": 55150 }, { "grad_norm": 0.5579968690872192, "learning_rate": 4.560908725172145e-05, "loss": 0.0256, "step": 55160 }, { "grad_norm": 0.8409668207168579, "learning_rate": 4.559261667551516e-05, "loss": 0.0387, "step": 55170 }, { "grad_norm": 1.2289109230041504, "learning_rate": 4.5576146581293685e-05, "loss": 0.0303, "step": 55180 }, { "grad_norm": 0.6207820773124695, "learning_rate": 4.5559676970858164e-05, "loss": 0.0428, "step": 55190 }, { "grad_norm": 0.5449680685997009, "learning_rate": 4.5543207846009705e-05, "loss": 0.0373, "step": 55200 }, { "grad_norm": 0.5441644787788391, "learning_rate": 4.552673920854933e-05, "loss": 0.0373, "step": 55210 }, { "grad_norm": 0.5060887336730957, "learning_rate": 4.5510271060278006e-05, "loss": 0.0404, "step": 55220 }, { "grad_norm": 0.5286913514137268, "learning_rate": 4.5493803402996704e-05, "loss": 0.0331, "step": 55230 }, { "grad_norm": 0.5822490453720093, "learning_rate": 4.5477336238506295e-05, "loss": 0.0278, "step": 55240 }, { "grad_norm": 0.4981083273887634, "learning_rate": 4.546086956860759e-05, "loss": 0.026, "step": 55250 }, { "grad_norm": 0.5654203295707703, "learning_rate": 4.5444403395101374e-05, "loss": 0.0258, "step": 55260 }, { "grad_norm": 0.41052281856536865, "learning_rate": 4.542793771978834e-05, "loss": 0.0285, "step": 55270 }, { "grad_norm": 0.5407291650772095, "learning_rate": 4.541147254446916e-05, "loss": 0.0271, "step": 55280 }, { "grad_norm": 0.5432749390602112, "learning_rate": 4.539500787094445e-05, "loss": 0.0341, "step": 55290 }, { "grad_norm": 0.5430466532707214, "learning_rate": 4.537854370101475e-05, "loss": 0.0477, "step": 55300 }, { "grad_norm": 0.4803856909275055, "learning_rate": 4.5362080036480566e-05, "loss": 0.0237, "step": 55310 }, { "grad_norm": 0.3935672640800476, "learning_rate": 4.534561687914235e-05, "loss": 0.0537, "step": 55320 }, { "grad_norm": 0.7216311693191528, "learning_rate": 4.5329154230800466e-05, "loss": 0.0449, "step": 55330 }, { "grad_norm": 0.3347742557525635, "learning_rate": 4.531269209325525e-05, "loss": 0.033, "step": 55340 }, { "grad_norm": 0.33175918459892273, "learning_rate": 4.529623046830696e-05, "loss": 0.0308, "step": 55350 }, { "grad_norm": 0.3734349012374878, "learning_rate": 4.5279769357755866e-05, "loss": 0.0352, "step": 55360 }, { "grad_norm": 0.5823997259140015, "learning_rate": 4.5263308763402084e-05, "loss": 0.0267, "step": 55370 }, { "grad_norm": 0.2693466544151306, "learning_rate": 4.524684868704574e-05, "loss": 0.0298, "step": 55380 }, { "grad_norm": 0.6429613828659058, "learning_rate": 4.523038913048688e-05, "loss": 0.0394, "step": 55390 }, { "grad_norm": 0.2696307301521301, "learning_rate": 4.521393009552549e-05, "loss": 0.0343, "step": 55400 }, { "grad_norm": 0.4827648997306824, "learning_rate": 4.5197471583961484e-05, "loss": 0.0359, "step": 55410 }, { "grad_norm": 0.4299778640270233, "learning_rate": 4.51810135975948e-05, "loss": 0.0265, "step": 55420 }, { "grad_norm": 0.21535329520702362, "learning_rate": 4.516455613822521e-05, "loss": 0.0237, "step": 55430 }, { "grad_norm": 0.32725685834884644, "learning_rate": 4.514809920765249e-05, "loss": 0.0304, "step": 55440 }, { "grad_norm": 0.3932625353336334, "learning_rate": 4.5131642807676356e-05, "loss": 0.0307, "step": 55450 }, { "grad_norm": 0.2589971721172333, "learning_rate": 4.511518694009644e-05, "loss": 0.0278, "step": 55460 }, { "grad_norm": 0.4089033305644989, "learning_rate": 4.509873160671231e-05, "loss": 0.0257, "step": 55470 }, { "grad_norm": 0.31404560804367065, "learning_rate": 4.508227680932356e-05, "loss": 0.0287, "step": 55480 }, { "grad_norm": 0.5982537269592285, "learning_rate": 4.5065822549729605e-05, "loss": 0.0339, "step": 55490 }, { "grad_norm": 0.5146327614784241, "learning_rate": 4.504936882972989e-05, "loss": 0.0284, "step": 55500 }, { "grad_norm": 0.4942052364349365, "learning_rate": 4.503291565112374e-05, "loss": 0.0263, "step": 55510 }, { "grad_norm": 0.5133792757987976, "learning_rate": 4.501646301571048e-05, "loss": 0.0307, "step": 55520 }, { "grad_norm": 0.7793782353401184, "learning_rate": 4.500001092528933e-05, "loss": 0.0438, "step": 55530 }, { "grad_norm": 0.5173954963684082, "learning_rate": 4.4983559381659455e-05, "loss": 0.0372, "step": 55540 }, { "grad_norm": 0.3433986008167267, "learning_rate": 4.496710838661999e-05, "loss": 0.0334, "step": 55550 }, { "grad_norm": 0.9412435293197632, "learning_rate": 4.495065794196999e-05, "loss": 0.0329, "step": 55560 }, { "grad_norm": 0.6514267325401306, "learning_rate": 4.4934208049508455e-05, "loss": 0.0374, "step": 55570 }, { "grad_norm": 0.49582162499427795, "learning_rate": 4.4917758711034303e-05, "loss": 0.0411, "step": 55580 }, { "grad_norm": 1.484492301940918, "learning_rate": 4.490130992834642e-05, "loss": 0.0347, "step": 55590 }, { "grad_norm": 0.563549816608429, "learning_rate": 4.488486170324359e-05, "loss": 0.0321, "step": 55600 }, { "grad_norm": 0.38008201122283936, "learning_rate": 4.486841403752463e-05, "loss": 0.0287, "step": 55610 }, { "grad_norm": 0.5721786618232727, "learning_rate": 4.4851966932988174e-05, "loss": 0.0616, "step": 55620 }, { "grad_norm": 0.33316782116889954, "learning_rate": 4.4835520391432894e-05, "loss": 0.0266, "step": 55630 }, { "grad_norm": 0.2924897372722626, "learning_rate": 4.481907441465733e-05, "loss": 0.029, "step": 55640 }, { "grad_norm": 0.4484623670578003, "learning_rate": 4.480262900446e-05, "loss": 0.0292, "step": 55650 }, { "grad_norm": 0.37958917021751404, "learning_rate": 4.478618416263933e-05, "loss": 0.027, "step": 55660 }, { "grad_norm": 0.4124063551425934, "learning_rate": 4.476973989099373e-05, "loss": 0.0343, "step": 55670 }, { "grad_norm": 0.673806369304657, "learning_rate": 4.475329619132153e-05, "loss": 0.0363, "step": 55680 }, { "grad_norm": 0.5105194449424744, "learning_rate": 4.4736853065420956e-05, "loss": 0.0306, "step": 55690 }, { "grad_norm": 0.3177066743373871, "learning_rate": 4.472041051509023e-05, "loss": 0.031, "step": 55700 }, { "grad_norm": 0.24673965573310852, "learning_rate": 4.470396854212746e-05, "loss": 0.0243, "step": 55710 }, { "grad_norm": 0.6229809522628784, "learning_rate": 4.468752714833072e-05, "loss": 0.0322, "step": 55720 }, { "grad_norm": 0.5253458619117737, "learning_rate": 4.467108633549803e-05, "loss": 0.0274, "step": 55730 }, { "grad_norm": 0.3329172730445862, "learning_rate": 4.4654646105427335e-05, "loss": 0.0301, "step": 55740 }, { "grad_norm": 0.4723927080631256, "learning_rate": 4.463820645991651e-05, "loss": 0.0352, "step": 55750 }, { "grad_norm": 1.018502950668335, "learning_rate": 4.462176740076334e-05, "loss": 0.0616, "step": 55760 }, { "grad_norm": 0.549685537815094, "learning_rate": 4.4605328929765624e-05, "loss": 0.0441, "step": 55770 }, { "grad_norm": 0.3780834376811981, "learning_rate": 4.458889104872102e-05, "loss": 0.0309, "step": 55780 }, { "grad_norm": 0.3569270372390747, "learning_rate": 4.457245375942715e-05, "loss": 0.0288, "step": 55790 }, { "grad_norm": 0.4997591972351074, "learning_rate": 4.455601706368157e-05, "loss": 0.037, "step": 55800 }, { "grad_norm": 0.4949020445346832, "learning_rate": 4.453958096328179e-05, "loss": 0.0343, "step": 55810 }, { "grad_norm": 0.5192452073097229, "learning_rate": 4.452314546002523e-05, "loss": 0.0247, "step": 55820 }, { "grad_norm": 0.37547221779823303, "learning_rate": 4.4506710555709246e-05, "loss": 0.0275, "step": 55830 }, { "grad_norm": 0.24368147552013397, "learning_rate": 4.449027625213114e-05, "loss": 0.0315, "step": 55840 }, { "grad_norm": 0.4562852680683136, "learning_rate": 4.4473842551088135e-05, "loss": 0.0322, "step": 55850 }, { "grad_norm": 1.720686674118042, "learning_rate": 4.445740945437738e-05, "loss": 0.0361, "step": 55860 }, { "grad_norm": 0.5834934115409851, "learning_rate": 4.444097696379602e-05, "loss": 0.0379, "step": 55870 }, { "grad_norm": 0.3912636339664459, "learning_rate": 4.442454508114106e-05, "loss": 0.0297, "step": 55880 }, { "grad_norm": 0.4348730444908142, "learning_rate": 4.4408113808209465e-05, "loss": 0.0318, "step": 55890 }, { "grad_norm": 0.3179060220718384, "learning_rate": 4.439168314679813e-05, "loss": 0.0316, "step": 55900 }, { "grad_norm": 0.3745516538619995, "learning_rate": 4.43752530987039e-05, "loss": 0.023, "step": 55910 }, { "grad_norm": 0.3107320964336395, "learning_rate": 4.4358823665723515e-05, "loss": 0.0523, "step": 55920 }, { "grad_norm": 0.47763699293136597, "learning_rate": 4.434239484965371e-05, "loss": 0.0244, "step": 55930 }, { "grad_norm": 0.442209392786026, "learning_rate": 4.4325966652291103e-05, "loss": 0.0275, "step": 55940 }, { "grad_norm": 0.40592408180236816, "learning_rate": 4.430953907543225e-05, "loss": 0.0261, "step": 55950 }, { "grad_norm": 0.3545991778373718, "learning_rate": 4.4293112120873645e-05, "loss": 0.0255, "step": 55960 }, { "grad_norm": 0.46535682678222656, "learning_rate": 4.427668579041172e-05, "loss": 0.0324, "step": 55970 }, { "grad_norm": 0.41499465703964233, "learning_rate": 4.4260260085842816e-05, "loss": 0.0299, "step": 55980 }, { "grad_norm": 0.5912322998046875, "learning_rate": 4.4243835008963256e-05, "loss": 0.0406, "step": 55990 }, { "grad_norm": 0.4073815941810608, "learning_rate": 4.422741056156925e-05, "loss": 0.0282, "step": 56000 }, { "grad_norm": 0.36484289169311523, "learning_rate": 4.421098674545693e-05, "loss": 0.0294, "step": 56010 }, { "grad_norm": 0.7235339879989624, "learning_rate": 4.419456356242241e-05, "loss": 0.0293, "step": 56020 }, { "grad_norm": 0.5844442248344421, "learning_rate": 4.4178141014261685e-05, "loss": 0.0379, "step": 56030 }, { "grad_norm": 0.5732998847961426, "learning_rate": 4.416171910277071e-05, "loss": 0.0389, "step": 56040 }, { "grad_norm": 0.5763537287712097, "learning_rate": 4.414529782974533e-05, "loss": 0.0367, "step": 56050 }, { "grad_norm": 0.5144556164741516, "learning_rate": 4.41288771969814e-05, "loss": 0.0268, "step": 56060 }, { "grad_norm": 0.37471213936805725, "learning_rate": 4.411245720627464e-05, "loss": 0.0326, "step": 56070 }, { "grad_norm": 0.2553597092628479, "learning_rate": 4.40960378594207e-05, "loss": 0.0373, "step": 56080 }, { "grad_norm": 0.3880774974822998, "learning_rate": 4.407961915821519e-05, "loss": 0.0298, "step": 56090 }, { "grad_norm": 0.4008074998855591, "learning_rate": 4.406320110445362e-05, "loss": 0.0327, "step": 56100 }, { "grad_norm": 0.45662617683410645, "learning_rate": 4.404678369993144e-05, "loss": 0.0277, "step": 56110 }, { "grad_norm": 0.5034385919570923, "learning_rate": 4.403036694644406e-05, "loss": 0.0303, "step": 56120 }, { "grad_norm": 0.5200672745704651, "learning_rate": 4.4013950845786764e-05, "loss": 0.0233, "step": 56130 }, { "grad_norm": 0.4750920236110687, "learning_rate": 4.399753539975482e-05, "loss": 0.0353, "step": 56140 }, { "grad_norm": 0.3312378227710724, "learning_rate": 4.398112061014337e-05, "loss": 0.0232, "step": 56150 }, { "grad_norm": 0.36318671703338623, "learning_rate": 4.396470647874753e-05, "loss": 0.0258, "step": 56160 }, { "grad_norm": 0.4172528088092804, "learning_rate": 4.394829300736229e-05, "loss": 0.0254, "step": 56170 }, { "grad_norm": 0.5623169541358948, "learning_rate": 4.393188019778265e-05, "loss": 0.037, "step": 56180 }, { "grad_norm": 0.25633808970451355, "learning_rate": 4.391546805180347e-05, "loss": 0.0348, "step": 56190 }, { "grad_norm": 0.8081329464912415, "learning_rate": 4.389905657121955e-05, "loss": 0.0287, "step": 56200 }, { "grad_norm": 0.47072210907936096, "learning_rate": 4.3882645757825635e-05, "loss": 0.0395, "step": 56210 }, { "grad_norm": 0.38355615735054016, "learning_rate": 4.386623561341637e-05, "loss": 0.0275, "step": 56220 }, { "grad_norm": 0.47099852561950684, "learning_rate": 4.384982613978637e-05, "loss": 0.035, "step": 56230 }, { "grad_norm": 0.7935210466384888, "learning_rate": 4.383341733873012e-05, "loss": 0.0263, "step": 56240 }, { "grad_norm": 0.4221004545688629, "learning_rate": 4.381700921204209e-05, "loss": 0.0278, "step": 56250 }, { "grad_norm": 0.37240684032440186, "learning_rate": 4.380060176151663e-05, "loss": 0.0232, "step": 56260 }, { "grad_norm": 0.869385302066803, "learning_rate": 4.378419498894805e-05, "loss": 0.0289, "step": 56270 }, { "grad_norm": 0.4789959192276001, "learning_rate": 4.376778889613056e-05, "loss": 0.0291, "step": 56280 }, { "grad_norm": 0.6104966402053833, "learning_rate": 4.375138348485831e-05, "loss": 0.0285, "step": 56290 }, { "grad_norm": 0.4368179440498352, "learning_rate": 4.3734978756925346e-05, "loss": 0.0438, "step": 56300 }, { "grad_norm": 0.5050607919692993, "learning_rate": 4.3718574714125706e-05, "loss": 0.0281, "step": 56310 }, { "grad_norm": 0.9411633610725403, "learning_rate": 4.370217135825329e-05, "loss": 0.042, "step": 56320 }, { "grad_norm": 1.7791153192520142, "learning_rate": 4.368576869110194e-05, "loss": 0.0256, "step": 56330 }, { "grad_norm": 0.3346923291683197, "learning_rate": 4.366936671446544e-05, "loss": 0.033, "step": 56340 }, { "grad_norm": 0.6098805665969849, "learning_rate": 4.3652965430137474e-05, "loss": 0.0444, "step": 56350 }, { "grad_norm": 0.3466915190219879, "learning_rate": 4.3636564839911646e-05, "loss": 0.0211, "step": 56360 }, { "grad_norm": 0.3869677186012268, "learning_rate": 4.3620164945581545e-05, "loss": 0.0412, "step": 56370 }, { "grad_norm": 0.27137041091918945, "learning_rate": 4.360376574894061e-05, "loss": 0.0404, "step": 56380 }, { "grad_norm": 0.38270246982574463, "learning_rate": 4.358736725178224e-05, "loss": 0.03, "step": 56390 }, { "grad_norm": 0.6562985777854919, "learning_rate": 4.357096945589974e-05, "loss": 0.0364, "step": 56400 }, { "grad_norm": 0.4580192565917969, "learning_rate": 4.355457236308636e-05, "loss": 0.0346, "step": 56410 }, { "grad_norm": 0.31432223320007324, "learning_rate": 4.353817597513526e-05, "loss": 0.0225, "step": 56420 }, { "grad_norm": 0.5346508622169495, "learning_rate": 4.352178029383948e-05, "loss": 0.031, "step": 56430 }, { "grad_norm": 0.17443659901618958, "learning_rate": 4.3505385320992105e-05, "loss": 0.0303, "step": 56440 }, { "grad_norm": 0.6354203820228577, "learning_rate": 4.348899105838602e-05, "loss": 0.0231, "step": 56450 }, { "grad_norm": 0.23365025222301483, "learning_rate": 4.3472597507814087e-05, "loss": 0.0348, "step": 56460 }, { "grad_norm": 0.3415464162826538, "learning_rate": 4.3456204671069066e-05, "loss": 0.0324, "step": 56470 }, { "grad_norm": 0.5596753358840942, "learning_rate": 4.343981254994367e-05, "loss": 0.025, "step": 56480 }, { "grad_norm": 0.37893861532211304, "learning_rate": 4.34234211462305e-05, "loss": 0.0332, "step": 56490 }, { "grad_norm": 0.45366352796554565, "learning_rate": 4.340703046172213e-05, "loss": 0.0433, "step": 56500 }, { "grad_norm": 0.4006165862083435, "learning_rate": 4.339064049821097e-05, "loss": 0.0291, "step": 56510 }, { "grad_norm": 0.44609344005584717, "learning_rate": 4.3374251257489446e-05, "loss": 0.0397, "step": 56520 }, { "grad_norm": 0.339187353849411, "learning_rate": 4.335786274134984e-05, "loss": 0.0265, "step": 56530 }, { "grad_norm": 0.34003111720085144, "learning_rate": 4.334147495158438e-05, "loss": 0.0273, "step": 56540 }, { "grad_norm": 0.3283267021179199, "learning_rate": 4.33250878899852e-05, "loss": 0.0341, "step": 56550 }, { "grad_norm": 0.6175225377082825, "learning_rate": 4.330870155834437e-05, "loss": 0.0257, "step": 56560 }, { "grad_norm": 0.3921773433685303, "learning_rate": 4.3292315958453884e-05, "loss": 0.0334, "step": 56570 }, { "grad_norm": 0.3122110068798065, "learning_rate": 4.3275931092105646e-05, "loss": 0.0196, "step": 56580 }, { "grad_norm": 0.5045640468597412, "learning_rate": 4.325954696109147e-05, "loss": 0.0354, "step": 56590 }, { "grad_norm": 0.3655394911766052, "learning_rate": 4.32431635672031e-05, "loss": 0.0251, "step": 56600 }, { "grad_norm": 0.36625936627388, "learning_rate": 4.3226780912232214e-05, "loss": 0.0352, "step": 56610 }, { "grad_norm": 0.3581089973449707, "learning_rate": 4.321039899797036e-05, "loss": 0.0298, "step": 56620 }, { "grad_norm": 0.47798511385917664, "learning_rate": 4.319401782620908e-05, "loss": 0.0282, "step": 56630 }, { "grad_norm": 0.3659927546977997, "learning_rate": 4.317763739873978e-05, "loss": 0.0224, "step": 56640 }, { "grad_norm": 0.5244653224945068, "learning_rate": 4.31612577173538e-05, "loss": 0.0265, "step": 56650 }, { "grad_norm": 0.5219129920005798, "learning_rate": 4.314487878384239e-05, "loss": 0.0358, "step": 56660 }, { "grad_norm": 0.5383313894271851, "learning_rate": 4.3128500599996736e-05, "loss": 0.0325, "step": 56670 }, { "grad_norm": 0.38365989923477173, "learning_rate": 4.31121231676079e-05, "loss": 0.0256, "step": 56680 }, { "grad_norm": 0.4778825342655182, "learning_rate": 4.309574648846694e-05, "loss": 0.0328, "step": 56690 }, { "grad_norm": 0.38704586029052734, "learning_rate": 4.3079370564364755e-05, "loss": 0.0247, "step": 56700 }, { "grad_norm": 0.3470279276371002, "learning_rate": 4.306299539709221e-05, "loss": 0.0272, "step": 56710 }, { "grad_norm": 0.2560126483440399, "learning_rate": 4.3046620988440046e-05, "loss": 0.0239, "step": 56720 }, { "grad_norm": 0.3498140573501587, "learning_rate": 4.303024734019897e-05, "loss": 0.0291, "step": 56730 }, { "grad_norm": 0.5082994699478149, "learning_rate": 4.3013874454159565e-05, "loss": 0.0317, "step": 56740 }, { "grad_norm": 0.4747190773487091, "learning_rate": 4.299750233211233e-05, "loss": 0.0276, "step": 56750 }, { "grad_norm": 0.6473188996315002, "learning_rate": 4.2981130975847715e-05, "loss": 0.0335, "step": 56760 }, { "grad_norm": 0.40461820363998413, "learning_rate": 4.2964760387156076e-05, "loss": 0.0233, "step": 56770 }, { "grad_norm": 0.4187439978122711, "learning_rate": 4.294839056782767e-05, "loss": 0.0424, "step": 56780 }, { "grad_norm": 0.33573538064956665, "learning_rate": 4.293202151965266e-05, "loss": 0.0226, "step": 56790 }, { "grad_norm": 0.6145443916320801, "learning_rate": 4.291565324442115e-05, "loss": 0.0326, "step": 56800 }, { "grad_norm": 0.44724351167678833, "learning_rate": 4.2899285743923135e-05, "loss": 0.0285, "step": 56810 }, { "grad_norm": 0.5973689556121826, "learning_rate": 4.288291901994857e-05, "loss": 0.039, "step": 56820 }, { "grad_norm": 0.30843302607536316, "learning_rate": 4.286655307428729e-05, "loss": 0.044, "step": 56830 }, { "grad_norm": 0.4437859058380127, "learning_rate": 4.285018790872903e-05, "loss": 0.0295, "step": 56840 }, { "grad_norm": 0.36344730854034424, "learning_rate": 4.283382352506347e-05, "loss": 0.0388, "step": 56850 }, { "grad_norm": 0.7183606624603271, "learning_rate": 4.2817459925080205e-05, "loss": 0.0344, "step": 56860 }, { "grad_norm": 0.5695903301239014, "learning_rate": 4.28010971105687e-05, "loss": 0.0324, "step": 56870 }, { "grad_norm": 0.3991449773311615, "learning_rate": 4.278473508331841e-05, "loss": 0.0372, "step": 56880 }, { "grad_norm": 0.2758060097694397, "learning_rate": 4.276837384511864e-05, "loss": 0.0277, "step": 56890 }, { "grad_norm": 0.28567394614219666, "learning_rate": 4.275201339775864e-05, "loss": 0.0322, "step": 56900 }, { "grad_norm": 0.30435609817504883, "learning_rate": 4.273565374302756e-05, "loss": 0.028, "step": 56910 }, { "grad_norm": 0.5606180429458618, "learning_rate": 4.2719294882714454e-05, "loss": 0.0374, "step": 56920 }, { "grad_norm": 0.33340346813201904, "learning_rate": 4.270293681860831e-05, "loss": 0.0296, "step": 56930 }, { "grad_norm": 0.4342455267906189, "learning_rate": 4.268657955249802e-05, "loss": 0.0271, "step": 56940 }, { "grad_norm": 0.4993112087249756, "learning_rate": 4.267022308617241e-05, "loss": 0.0393, "step": 56950 }, { "grad_norm": 0.5546606779098511, "learning_rate": 4.265386742142018e-05, "loss": 0.041, "step": 56960 }, { "grad_norm": 0.4140022397041321, "learning_rate": 4.263751256002996e-05, "loss": 0.0286, "step": 56970 }, { "grad_norm": 0.4658932685852051, "learning_rate": 4.26211585037903e-05, "loss": 0.0327, "step": 56980 }, { "grad_norm": 0.3161769211292267, "learning_rate": 4.2604805254489656e-05, "loss": 0.0324, "step": 56990 }, { "grad_norm": 0.42331504821777344, "learning_rate": 4.258845281391638e-05, "loss": 0.0312, "step": 57000 }, { "grad_norm": 0.33444201946258545, "learning_rate": 4.257210118385876e-05, "loss": 0.0243, "step": 57010 }, { "grad_norm": 0.47096797823905945, "learning_rate": 4.2555750366104994e-05, "loss": 0.0301, "step": 57020 }, { "grad_norm": 0.3225822448730469, "learning_rate": 4.2539400362443184e-05, "loss": 0.037, "step": 57030 }, { "grad_norm": 0.37952157855033875, "learning_rate": 4.252305117466133e-05, "loss": 0.0336, "step": 57040 }, { "grad_norm": 0.4017888009548187, "learning_rate": 4.2506702804547356e-05, "loss": 0.0347, "step": 57050 }, { "grad_norm": 0.40146759152412415, "learning_rate": 4.249035525388907e-05, "loss": 0.0385, "step": 57060 }, { "grad_norm": 0.40909892320632935, "learning_rate": 4.247400852447427e-05, "loss": 0.0301, "step": 57070 }, { "grad_norm": 0.2687731683254242, "learning_rate": 4.245766261809059e-05, "loss": 0.0275, "step": 57080 }, { "grad_norm": 0.47112584114074707, "learning_rate": 4.244131753652558e-05, "loss": 0.0374, "step": 57090 }, { "grad_norm": 0.8086903691291809, "learning_rate": 4.242497328156672e-05, "loss": 0.032, "step": 57100 }, { "grad_norm": 0.39481887221336365, "learning_rate": 4.240862985500139e-05, "loss": 0.037, "step": 57110 }, { "grad_norm": 0.4963906407356262, "learning_rate": 4.2392287258616894e-05, "loss": 0.0307, "step": 57120 }, { "grad_norm": 0.3557317852973938, "learning_rate": 4.2375945494200396e-05, "loss": 0.055, "step": 57130 }, { "grad_norm": 0.5290271043777466, "learning_rate": 4.2359604563539054e-05, "loss": 0.0341, "step": 57140 }, { "grad_norm": 1.0663243532180786, "learning_rate": 4.2343264468419885e-05, "loss": 0.0486, "step": 57150 }, { "grad_norm": 0.26051607728004456, "learning_rate": 4.232692521062979e-05, "loss": 0.0303, "step": 57160 }, { "grad_norm": 0.26992514729499817, "learning_rate": 4.231058679195562e-05, "loss": 0.0301, "step": 57170 }, { "grad_norm": 0.3699473738670349, "learning_rate": 4.229424921418411e-05, "loss": 0.0368, "step": 57180 }, { "grad_norm": 0.3391129672527313, "learning_rate": 4.2277912479101916e-05, "loss": 0.0376, "step": 57190 }, { "grad_norm": 0.32758858799934387, "learning_rate": 4.226157658849562e-05, "loss": 0.0287, "step": 57200 }, { "grad_norm": 0.39072608947753906, "learning_rate": 4.2245241544151664e-05, "loss": 0.0509, "step": 57210 }, { "grad_norm": 0.3287012577056885, "learning_rate": 4.2228907347856425e-05, "loss": 0.0401, "step": 57220 }, { "grad_norm": 0.5121421217918396, "learning_rate": 4.2212574001396214e-05, "loss": 0.0296, "step": 57230 }, { "grad_norm": 0.42435961961746216, "learning_rate": 4.21962415065572e-05, "loss": 0.0305, "step": 57240 }, { "grad_norm": 0.5917856097221375, "learning_rate": 4.217990986512548e-05, "loss": 0.0301, "step": 57250 }, { "grad_norm": 0.47142601013183594, "learning_rate": 4.216357907888704e-05, "loss": 0.0218, "step": 57260 }, { "grad_norm": 0.27367842197418213, "learning_rate": 4.2147249149627824e-05, "loss": 0.0286, "step": 57270 }, { "grad_norm": 1.0009666681289673, "learning_rate": 4.213092007913364e-05, "loss": 0.0314, "step": 57280 }, { "grad_norm": 0.4717470407485962, "learning_rate": 4.2114591869190205e-05, "loss": 0.0301, "step": 57290 }, { "grad_norm": 0.4153082072734833, "learning_rate": 4.209826452158315e-05, "loss": 0.0269, "step": 57300 }, { "grad_norm": 0.4974222481250763, "learning_rate": 4.2081938038098e-05, "loss": 0.0293, "step": 57310 }, { "grad_norm": 0.582513689994812, "learning_rate": 4.2065612420520185e-05, "loss": 0.0415, "step": 57320 }, { "grad_norm": 0.4373595416545868, "learning_rate": 4.204928767063509e-05, "loss": 0.0206, "step": 57330 }, { "grad_norm": 0.3401395082473755, "learning_rate": 4.2032963790227944e-05, "loss": 0.0315, "step": 57340 }, { "grad_norm": 0.5435667037963867, "learning_rate": 4.2016640781083887e-05, "loss": 0.0425, "step": 57350 }, { "grad_norm": 0.9168472290039062, "learning_rate": 4.2000318644988004e-05, "loss": 0.0345, "step": 57360 }, { "grad_norm": 0.35546573996543884, "learning_rate": 4.198399738372524e-05, "loss": 0.0313, "step": 57370 }, { "grad_norm": 0.3664940595626831, "learning_rate": 4.196767699908045e-05, "loss": 0.0265, "step": 57380 }, { "grad_norm": 0.3943156898021698, "learning_rate": 4.195135749283845e-05, "loss": 0.0453, "step": 57390 }, { "grad_norm": 0.7073175311088562, "learning_rate": 4.1935038866783894e-05, "loss": 0.0225, "step": 57400 }, { "grad_norm": 0.42814692854881287, "learning_rate": 4.1918721122701355e-05, "loss": 0.0265, "step": 57410 }, { "grad_norm": 0.48171466588974, "learning_rate": 4.190240426237533e-05, "loss": 0.0359, "step": 57420 }, { "grad_norm": 0.5016931891441345, "learning_rate": 4.188608828759019e-05, "loss": 0.0202, "step": 57430 }, { "grad_norm": 0.31329819560050964, "learning_rate": 4.1869773200130246e-05, "loss": 0.0471, "step": 57440 }, { "grad_norm": 0.6783474087715149, "learning_rate": 4.185345900177967e-05, "loss": 0.0283, "step": 57450 }, { "grad_norm": 0.4365290105342865, "learning_rate": 4.183714569432258e-05, "loss": 0.0588, "step": 57460 }, { "grad_norm": 0.4371166527271271, "learning_rate": 4.182083327954295e-05, "loss": 0.0308, "step": 57470 }, { "grad_norm": 0.4112355709075928, "learning_rate": 4.1804521759224714e-05, "loss": 0.0259, "step": 57480 }, { "grad_norm": 0.3764275014400482, "learning_rate": 4.178821113515165e-05, "loss": 0.0258, "step": 57490 }, { "grad_norm": 0.4631001949310303, "learning_rate": 4.1771901409107476e-05, "loss": 0.0285, "step": 57500 }, { "grad_norm": 0.4823700785636902, "learning_rate": 4.175559258287577e-05, "loss": 0.042, "step": 57510 }, { "grad_norm": 0.366572767496109, "learning_rate": 4.173928465824008e-05, "loss": 0.0291, "step": 57520 }, { "grad_norm": 0.4603232443332672, "learning_rate": 4.17229776369838e-05, "loss": 0.0263, "step": 57530 }, { "grad_norm": 0.602341890335083, "learning_rate": 4.170667152089025e-05, "loss": 0.0324, "step": 57540 }, { "grad_norm": 0.5789105296134949, "learning_rate": 4.1690366311742626e-05, "loss": 0.0373, "step": 57550 }, { "grad_norm": 0.40108031034469604, "learning_rate": 4.1674062011324055e-05, "loss": 0.0337, "step": 57560 }, { "grad_norm": 0.5127894878387451, "learning_rate": 4.165775862141752e-05, "loss": 0.0374, "step": 57570 }, { "grad_norm": 0.33041274547576904, "learning_rate": 4.164145614380598e-05, "loss": 0.0352, "step": 57580 }, { "grad_norm": 0.348892480134964, "learning_rate": 4.162515458027223e-05, "loss": 0.0319, "step": 57590 }, { "grad_norm": 0.2947714328765869, "learning_rate": 4.160885393259899e-05, "loss": 0.023, "step": 57600 }, { "grad_norm": 0.4511187970638275, "learning_rate": 4.159255420256886e-05, "loss": 0.0271, "step": 57610 }, { "grad_norm": 0.3951359689235687, "learning_rate": 4.157625539196436e-05, "loss": 0.0257, "step": 57620 }, { "grad_norm": 0.6521416306495667, "learning_rate": 4.155995750256791e-05, "loss": 0.0328, "step": 57630 }, { "grad_norm": 0.4872557818889618, "learning_rate": 4.15436605361618e-05, "loss": 0.0338, "step": 57640 }, { "grad_norm": 0.36815404891967773, "learning_rate": 4.152736449452827e-05, "loss": 0.0265, "step": 57650 }, { "grad_norm": 0.2105143815279007, "learning_rate": 4.151106937944943e-05, "loss": 0.0285, "step": 57660 }, { "grad_norm": 0.46440720558166504, "learning_rate": 4.149477519270727e-05, "loss": 0.0252, "step": 57670 }, { "grad_norm": 0.6148982048034668, "learning_rate": 4.147848193608371e-05, "loss": 0.0279, "step": 57680 }, { "grad_norm": 0.5944820642471313, "learning_rate": 4.146218961136056e-05, "loss": 0.0222, "step": 57690 }, { "grad_norm": 0.5203352570533752, "learning_rate": 4.14458982203195e-05, "loss": 0.0393, "step": 57700 }, { "grad_norm": 0.31650128960609436, "learning_rate": 4.1429607764742165e-05, "loss": 0.0323, "step": 57710 }, { "grad_norm": 0.2655096650123596, "learning_rate": 4.141331824641003e-05, "loss": 0.0254, "step": 57720 }, { "grad_norm": 0.5847546458244324, "learning_rate": 4.139702966710452e-05, "loss": 0.0347, "step": 57730 }, { "grad_norm": 0.4418269693851471, "learning_rate": 4.13807420286069e-05, "loss": 0.0297, "step": 57740 }, { "grad_norm": 0.25481465458869934, "learning_rate": 4.136445533269837e-05, "loss": 0.0341, "step": 57750 }, { "grad_norm": 0.5012801289558411, "learning_rate": 4.134816958116003e-05, "loss": 0.0328, "step": 57760 }, { "grad_norm": 0.5743717551231384, "learning_rate": 4.133188477577283e-05, "loss": 0.0268, "step": 57770 }, { "grad_norm": 0.42587295174598694, "learning_rate": 4.131560091831771e-05, "loss": 0.0229, "step": 57780 }, { "grad_norm": 0.5329195857048035, "learning_rate": 4.1299318010575405e-05, "loss": 0.0269, "step": 57790 }, { "grad_norm": 0.42626118659973145, "learning_rate": 4.128303605432661e-05, "loss": 0.0191, "step": 57800 }, { "grad_norm": 0.3587830662727356, "learning_rate": 4.126675505135188e-05, "loss": 0.0256, "step": 57810 }, { "grad_norm": 0.5519593954086304, "learning_rate": 4.125047500343169e-05, "loss": 0.0326, "step": 57820 }, { "grad_norm": 0.43282291293144226, "learning_rate": 4.1234195912346376e-05, "loss": 0.0253, "step": 57830 }, { "grad_norm": 0.38659918308258057, "learning_rate": 4.1217917779876235e-05, "loss": 0.0259, "step": 57840 }, { "grad_norm": 0.47642195224761963, "learning_rate": 4.120164060780141e-05, "loss": 0.0263, "step": 57850 }, { "grad_norm": 0.49917104840278625, "learning_rate": 4.1185364397901935e-05, "loss": 0.0279, "step": 57860 }, { "grad_norm": 0.46067866683006287, "learning_rate": 4.116908915195776e-05, "loss": 0.0214, "step": 57870 }, { "grad_norm": 0.6510335803031921, "learning_rate": 4.115281487174872e-05, "loss": 0.0295, "step": 57880 }, { "grad_norm": 0.42733293771743774, "learning_rate": 4.113654155905452e-05, "loss": 0.034, "step": 57890 }, { "grad_norm": 0.23344428837299347, "learning_rate": 4.112026921565483e-05, "loss": 0.0303, "step": 57900 }, { "grad_norm": 0.3442455232143402, "learning_rate": 4.1103997843329147e-05, "loss": 0.0323, "step": 57910 }, { "grad_norm": 0.3337825536727905, "learning_rate": 4.108772744385689e-05, "loss": 0.0238, "step": 57920 }, { "grad_norm": 0.3619404733181, "learning_rate": 4.1071458019017346e-05, "loss": 0.0341, "step": 57930 }, { "grad_norm": 1.0170345306396484, "learning_rate": 4.1055189570589734e-05, "loss": 0.0277, "step": 57940 }, { "grad_norm": 0.3976888656616211, "learning_rate": 4.103892210035315e-05, "loss": 0.0225, "step": 57950 }, { "grad_norm": 0.31477090716362, "learning_rate": 4.102265561008656e-05, "loss": 0.0203, "step": 57960 }, { "grad_norm": 0.39169833064079285, "learning_rate": 4.100639010156885e-05, "loss": 0.0214, "step": 57970 }, { "grad_norm": 0.5306956768035889, "learning_rate": 4.099012557657882e-05, "loss": 0.0274, "step": 57980 }, { "grad_norm": 0.6153080463409424, "learning_rate": 4.09738620368951e-05, "loss": 0.0319, "step": 57990 }, { "grad_norm": 0.3953703045845032, "learning_rate": 4.095759948429626e-05, "loss": 0.0206, "step": 58000 }, { "grad_norm": 0.4187508225440979, "learning_rate": 4.094133792056075e-05, "loss": 0.0308, "step": 58010 }, { "grad_norm": 0.4560694992542267, "learning_rate": 4.092507734746688e-05, "loss": 0.0225, "step": 58020 }, { "grad_norm": 0.3750382363796234, "learning_rate": 4.090881776679293e-05, "loss": 0.0274, "step": 58030 }, { "grad_norm": 0.25347042083740234, "learning_rate": 4.0892559180316994e-05, "loss": 0.0279, "step": 58040 }, { "grad_norm": 0.3576776385307312, "learning_rate": 4.08763015898171e-05, "loss": 0.0274, "step": 58050 }, { "grad_norm": 0.18532970547676086, "learning_rate": 4.0860044997071145e-05, "loss": 0.0286, "step": 58060 }, { "grad_norm": 0.48726770281791687, "learning_rate": 4.084378940385693e-05, "loss": 0.0343, "step": 58070 }, { "grad_norm": 0.3693099617958069, "learning_rate": 4.082753481195211e-05, "loss": 0.0441, "step": 58080 }, { "grad_norm": 0.633404552936554, "learning_rate": 4.0811281223134324e-05, "loss": 0.035, "step": 58090 }, { "grad_norm": 0.33302128314971924, "learning_rate": 4.0795028639181e-05, "loss": 0.0356, "step": 58100 }, { "grad_norm": 0.39011919498443604, "learning_rate": 4.077877706186951e-05, "loss": 0.0295, "step": 58110 }, { "grad_norm": 0.7639243006706238, "learning_rate": 4.076252649297709e-05, "loss": 0.0451, "step": 58120 }, { "grad_norm": 0.6558593511581421, "learning_rate": 4.07462769342809e-05, "loss": 0.0328, "step": 58130 }, { "grad_norm": 1.0758248567581177, "learning_rate": 4.073002838755793e-05, "loss": 0.0321, "step": 58140 }, { "grad_norm": 0.5848369002342224, "learning_rate": 4.0713780854585115e-05, "loss": 0.0323, "step": 58150 }, { "grad_norm": 0.378814697265625, "learning_rate": 4.069753433713927e-05, "loss": 0.0328, "step": 58160 }, { "grad_norm": 0.7294735312461853, "learning_rate": 4.068128883699709e-05, "loss": 0.0317, "step": 58170 }, { "grad_norm": 0.27353528141975403, "learning_rate": 4.0665044355935155e-05, "loss": 0.0297, "step": 58180 }, { "grad_norm": 0.5227489471435547, "learning_rate": 4.064880089572993e-05, "loss": 0.025, "step": 58190 }, { "grad_norm": 0.4464496374130249, "learning_rate": 4.063255845815779e-05, "loss": 0.029, "step": 58200 }, { "grad_norm": 0.3881493806838989, "learning_rate": 4.0616317044994945e-05, "loss": 0.0294, "step": 58210 }, { "grad_norm": 0.9173793792724609, "learning_rate": 4.0600076658017585e-05, "loss": 0.0304, "step": 58220 }, { "grad_norm": 0.4833660423755646, "learning_rate": 4.0583837299001706e-05, "loss": 0.0233, "step": 58230 }, { "grad_norm": 0.24831244349479675, "learning_rate": 4.056759896972322e-05, "loss": 0.0338, "step": 58240 }, { "grad_norm": 0.42982953786849976, "learning_rate": 4.055136167195793e-05, "loss": 0.0305, "step": 58250 }, { "grad_norm": 0.4278218448162079, "learning_rate": 4.053512540748152e-05, "loss": 0.0221, "step": 58260 }, { "grad_norm": 0.3939071595668793, "learning_rate": 4.051889017806955e-05, "loss": 0.03, "step": 58270 }, { "grad_norm": 0.4026348888874054, "learning_rate": 4.0502655985497516e-05, "loss": 0.0335, "step": 58280 }, { "grad_norm": 0.4113391935825348, "learning_rate": 4.048642283154074e-05, "loss": 0.0305, "step": 58290 }, { "grad_norm": 0.20752669870853424, "learning_rate": 4.0470190717974464e-05, "loss": 0.0308, "step": 58300 }, { "grad_norm": 0.2750304341316223, "learning_rate": 4.045395964657381e-05, "loss": 0.0405, "step": 58310 }, { "grad_norm": 0.7597982883453369, "learning_rate": 4.043772961911376e-05, "loss": 0.0294, "step": 58320 }, { "grad_norm": 0.5207206606864929, "learning_rate": 4.042150063736924e-05, "loss": 0.0335, "step": 58330 }, { "grad_norm": 0.4976537525653839, "learning_rate": 4.0405272703114976e-05, "loss": 0.0441, "step": 58340 }, { "grad_norm": 0.5088873505592346, "learning_rate": 4.038904581812569e-05, "loss": 0.0355, "step": 58350 }, { "grad_norm": 0.3753615915775299, "learning_rate": 4.03728199841759e-05, "loss": 0.0283, "step": 58360 }, { "grad_norm": 0.2705828845500946, "learning_rate": 4.0356595203040044e-05, "loss": 0.0192, "step": 58370 }, { "grad_norm": 0.4649571478366852, "learning_rate": 4.0340371476492426e-05, "loss": 0.0331, "step": 58380 }, { "grad_norm": 0.3758942782878876, "learning_rate": 4.032414880630727e-05, "loss": 0.0286, "step": 58390 }, { "grad_norm": 0.5745285153388977, "learning_rate": 4.0307927194258646e-05, "loss": 0.0288, "step": 58400 }, { "grad_norm": 0.5657630562782288, "learning_rate": 4.029170664212054e-05, "loss": 0.0255, "step": 58410 }, { "grad_norm": 0.5585871338844299, "learning_rate": 4.0275487151666794e-05, "loss": 0.033, "step": 58420 }, { "grad_norm": 0.265087753534317, "learning_rate": 4.025926872467116e-05, "loss": 0.039, "step": 58430 }, { "grad_norm": 0.2725234925746918, "learning_rate": 4.024305136290725e-05, "loss": 0.0328, "step": 58440 }, { "grad_norm": 0.279704213142395, "learning_rate": 4.022683506814858e-05, "loss": 0.0231, "step": 58450 }, { "grad_norm": 0.26416563987731934, "learning_rate": 4.021061984216853e-05, "loss": 0.0288, "step": 58460 }, { "grad_norm": 0.8102436065673828, "learning_rate": 4.019440568674035e-05, "loss": 0.0514, "step": 58470 }, { "grad_norm": 0.32222476601600647, "learning_rate": 4.017819260363724e-05, "loss": 0.0254, "step": 58480 }, { "grad_norm": 0.8794733881950378, "learning_rate": 4.016198059463223e-05, "loss": 0.0241, "step": 58490 }, { "grad_norm": 0.645253598690033, "learning_rate": 4.0145769661498233e-05, "loss": 0.0237, "step": 58500 }, { "grad_norm": 0.3734169900417328, "learning_rate": 4.012955980600804e-05, "loss": 0.0302, "step": 58510 }, { "grad_norm": 0.5732899308204651, "learning_rate": 4.0113351029934355e-05, "loss": 0.0311, "step": 58520 }, { "grad_norm": 0.37438058853149414, "learning_rate": 4.0097143335049714e-05, "loss": 0.0256, "step": 58530 }, { "grad_norm": 0.3604678809642792, "learning_rate": 4.0080936723126606e-05, "loss": 0.0285, "step": 58540 }, { "grad_norm": 0.6722224950790405, "learning_rate": 4.006473119593735e-05, "loss": 0.0279, "step": 58550 }, { "grad_norm": 0.5164196491241455, "learning_rate": 4.004852675525415e-05, "loss": 0.0316, "step": 58560 }, { "grad_norm": 0.5090042948722839, "learning_rate": 4.00323234028491e-05, "loss": 0.0415, "step": 58570 }, { "grad_norm": 0.5020487904548645, "learning_rate": 4.0016121140494186e-05, "loss": 0.0254, "step": 58580 }, { "grad_norm": 0.39730408787727356, "learning_rate": 3.999991996996122e-05, "loss": 0.0322, "step": 58590 }, { "grad_norm": 0.2605755627155304, "learning_rate": 3.9983719893022e-05, "loss": 0.026, "step": 58600 }, { "grad_norm": 0.2709169387817383, "learning_rate": 3.996752091144812e-05, "loss": 0.0336, "step": 58610 }, { "grad_norm": 1.6168025732040405, "learning_rate": 3.9951323027011055e-05, "loss": 0.0374, "step": 58620 }, { "grad_norm": 0.4529217481613159, "learning_rate": 3.993512624148219e-05, "loss": 0.0338, "step": 58630 }, { "grad_norm": 0.2365112453699112, "learning_rate": 3.99189305566328e-05, "loss": 0.0292, "step": 58640 }, { "grad_norm": 0.24628128111362457, "learning_rate": 3.990273597423401e-05, "loss": 0.0188, "step": 58650 }, { "grad_norm": 0.3763720989227295, "learning_rate": 3.988654249605681e-05, "loss": 0.0314, "step": 58660 }, { "grad_norm": 0.2858726978302002, "learning_rate": 3.987035012387213e-05, "loss": 0.028, "step": 58670 }, { "grad_norm": 0.6236218810081482, "learning_rate": 3.9854158859450735e-05, "loss": 0.0261, "step": 58680 }, { "grad_norm": 0.3497120440006256, "learning_rate": 3.983796870456328e-05, "loss": 0.0272, "step": 58690 }, { "grad_norm": 0.5562028288841248, "learning_rate": 3.982177966098028e-05, "loss": 0.0318, "step": 58700 }, { "grad_norm": 0.44417670369148254, "learning_rate": 3.980559173047216e-05, "loss": 0.0255, "step": 58710 }, { "grad_norm": 0.475508451461792, "learning_rate": 3.9789404914809176e-05, "loss": 0.0289, "step": 58720 }, { "grad_norm": 0.4471508264541626, "learning_rate": 3.9773219215761544e-05, "loss": 0.0295, "step": 58730 }, { "grad_norm": 0.37892434000968933, "learning_rate": 3.975703463509928e-05, "loss": 0.0288, "step": 58740 }, { "grad_norm": 0.27553820610046387, "learning_rate": 3.974085117459231e-05, "loss": 0.0293, "step": 58750 }, { "grad_norm": 0.34191232919692993, "learning_rate": 3.972466883601043e-05, "loss": 0.0272, "step": 58760 }, { "grad_norm": 0.7957545518875122, "learning_rate": 3.9708487621123316e-05, "loss": 0.0278, "step": 58770 }, { "grad_norm": 0.43815216422080994, "learning_rate": 3.969230753170051e-05, "loss": 0.0218, "step": 58780 }, { "grad_norm": 0.5223745703697205, "learning_rate": 3.967612856951146e-05, "loss": 0.0335, "step": 58790 }, { "grad_norm": 0.38601383566856384, "learning_rate": 3.9659950736325476e-05, "loss": 0.019, "step": 58800 }, { "grad_norm": 0.26911860704421997, "learning_rate": 3.964377403391174e-05, "loss": 0.0225, "step": 58810 }, { "grad_norm": 0.5120274424552917, "learning_rate": 3.9627598464039303e-05, "loss": 0.0356, "step": 58820 }, { "grad_norm": 0.2902222275733948, "learning_rate": 3.9611424028477096e-05, "loss": 0.0224, "step": 58830 }, { "grad_norm": 0.3927972912788391, "learning_rate": 3.959525072899394e-05, "loss": 0.0282, "step": 58840 }, { "grad_norm": 0.32733169198036194, "learning_rate": 3.9579078567358506e-05, "loss": 0.0217, "step": 58850 }, { "grad_norm": 0.4030626714229584, "learning_rate": 3.956290754533939e-05, "loss": 0.0333, "step": 58860 }, { "grad_norm": 0.3977598547935486, "learning_rate": 3.954673766470501e-05, "loss": 0.0204, "step": 58870 }, { "grad_norm": 0.7640965580940247, "learning_rate": 3.9530568927223674e-05, "loss": 0.0404, "step": 58880 }, { "grad_norm": 0.5336241722106934, "learning_rate": 3.95144013346636e-05, "loss": 0.0401, "step": 58890 }, { "grad_norm": 0.3648128807544708, "learning_rate": 3.949823488879282e-05, "loss": 0.0243, "step": 58900 }, { "grad_norm": 0.2561493217945099, "learning_rate": 3.9482069591379276e-05, "loss": 0.0271, "step": 58910 }, { "grad_norm": 0.44906026124954224, "learning_rate": 3.9465905444190784e-05, "loss": 0.0182, "step": 58920 }, { "grad_norm": 0.8319763541221619, "learning_rate": 3.944974244899505e-05, "loss": 0.0247, "step": 58930 }, { "grad_norm": 0.41404861211776733, "learning_rate": 3.9433580607559616e-05, "loss": 0.028, "step": 58940 }, { "grad_norm": 0.5670837759971619, "learning_rate": 3.941741992165193e-05, "loss": 0.0377, "step": 58950 }, { "grad_norm": 0.4913533627986908, "learning_rate": 3.940126039303928e-05, "loss": 0.027, "step": 58960 }, { "grad_norm": 0.40894049406051636, "learning_rate": 3.938510202348886e-05, "loss": 0.0266, "step": 58970 }, { "grad_norm": 0.2762669622898102, "learning_rate": 3.93689448147677e-05, "loss": 0.029, "step": 58980 }, { "grad_norm": 0.765204131603241, "learning_rate": 3.935278876864278e-05, "loss": 0.0362, "step": 58990 }, { "grad_norm": 0.43664371967315674, "learning_rate": 3.933663388688087e-05, "loss": 0.0262, "step": 59000 }, { "grad_norm": 0.43817317485809326, "learning_rate": 3.9320480171248645e-05, "loss": 0.0246, "step": 59010 }, { "grad_norm": 0.45956850051879883, "learning_rate": 3.930432762351265e-05, "loss": 0.0251, "step": 59020 }, { "grad_norm": 0.33764737844467163, "learning_rate": 3.928817624543931e-05, "loss": 0.0219, "step": 59030 }, { "grad_norm": 0.24722476303577423, "learning_rate": 3.9272026038794885e-05, "loss": 0.0289, "step": 59040 }, { "grad_norm": 0.28938204050064087, "learning_rate": 3.9255877005345586e-05, "loss": 0.0242, "step": 59050 }, { "grad_norm": 0.40187621116638184, "learning_rate": 3.923972914685743e-05, "loss": 0.0308, "step": 59060 }, { "grad_norm": 0.483320415019989, "learning_rate": 3.922358246509631e-05, "loss": 0.0307, "step": 59070 }, { "grad_norm": 0.48877570033073425, "learning_rate": 3.920743696182801e-05, "loss": 0.0504, "step": 59080 }, { "grad_norm": 0.40741968154907227, "learning_rate": 3.9191292638818164e-05, "loss": 0.0276, "step": 59090 }, { "grad_norm": 0.3937186896800995, "learning_rate": 3.9175149497832316e-05, "loss": 0.0307, "step": 59100 }, { "grad_norm": 0.42215245962142944, "learning_rate": 3.915900754063584e-05, "loss": 0.0263, "step": 59110 }, { "grad_norm": 0.3185754418373108, "learning_rate": 3.9142866768994014e-05, "loss": 0.0231, "step": 59120 }, { "grad_norm": 0.5713803768157959, "learning_rate": 3.912672718467193e-05, "loss": 0.0271, "step": 59130 }, { "grad_norm": 0.43947601318359375, "learning_rate": 3.911058878943463e-05, "loss": 0.0241, "step": 59140 }, { "grad_norm": 0.4029775857925415, "learning_rate": 3.9094451585046974e-05, "loss": 0.0379, "step": 59150 }, { "grad_norm": 0.28621554374694824, "learning_rate": 3.90783155732737e-05, "loss": 0.032, "step": 59160 }, { "grad_norm": 0.2512027621269226, "learning_rate": 3.9062180755879373e-05, "loss": 0.0275, "step": 59170 }, { "grad_norm": 0.3722873032093048, "learning_rate": 3.904604713462855e-05, "loss": 0.0309, "step": 59180 }, { "grad_norm": 0.22355149686336517, "learning_rate": 3.902991471128554e-05, "loss": 0.0224, "step": 59190 }, { "grad_norm": 0.6496069431304932, "learning_rate": 3.901378348761456e-05, "loss": 0.0223, "step": 59200 }, { "grad_norm": 0.7902256846427917, "learning_rate": 3.899765346537971e-05, "loss": 0.0329, "step": 59210 }, { "grad_norm": 0.8146674036979675, "learning_rate": 3.898152464634493e-05, "loss": 0.0211, "step": 59220 }, { "grad_norm": 0.22272832691669464, "learning_rate": 3.896539703227402e-05, "loss": 0.031, "step": 59230 }, { "grad_norm": 3.0403482913970947, "learning_rate": 3.894927062493073e-05, "loss": 0.0281, "step": 59240 }, { "grad_norm": 0.3155088424682617, "learning_rate": 3.893314542607859e-05, "loss": 0.0305, "step": 59250 }, { "grad_norm": 0.580988347530365, "learning_rate": 3.891702143748103e-05, "loss": 0.0273, "step": 59260 }, { "grad_norm": 0.6640663146972656, "learning_rate": 3.890089866090133e-05, "loss": 0.0321, "step": 59270 }, { "grad_norm": 0.5139829516410828, "learning_rate": 3.888477709810267e-05, "loss": 0.0294, "step": 59280 }, { "grad_norm": 0.5541282296180725, "learning_rate": 3.886865675084806e-05, "loss": 0.0332, "step": 59290 }, { "grad_norm": 0.38627976179122925, "learning_rate": 3.885253762090043e-05, "loss": 0.0297, "step": 59300 }, { "grad_norm": 0.40797802805900574, "learning_rate": 3.8836419710022534e-05, "loss": 0.0303, "step": 59310 }, { "grad_norm": 0.30893898010253906, "learning_rate": 3.882030301997698e-05, "loss": 0.0412, "step": 59320 }, { "grad_norm": 0.46471545100212097, "learning_rate": 3.880418755252629e-05, "loss": 0.0363, "step": 59330 }, { "grad_norm": 0.3650834262371063, "learning_rate": 3.87880733094328e-05, "loss": 0.025, "step": 59340 }, { "grad_norm": 0.3818226754665375, "learning_rate": 3.877196029245877e-05, "loss": 0.0314, "step": 59350 }, { "grad_norm": 0.5012707114219666, "learning_rate": 3.875584850336627e-05, "loss": 0.0331, "step": 59360 }, { "grad_norm": 0.44456759095191956, "learning_rate": 3.873973794391728e-05, "loss": 0.0287, "step": 59370 }, { "grad_norm": 0.8192908763885498, "learning_rate": 3.872362861587361e-05, "loss": 0.0373, "step": 59380 }, { "grad_norm": 0.6398888826370239, "learning_rate": 3.870752052099698e-05, "loss": 0.0277, "step": 59390 }, { "grad_norm": 0.6308113932609558, "learning_rate": 3.8691413661048916e-05, "loss": 0.0356, "step": 59400 }, { "grad_norm": 0.48758581280708313, "learning_rate": 3.867530803779085e-05, "loss": 0.0261, "step": 59410 }, { "grad_norm": 0.5085527300834656, "learning_rate": 3.8659203652984054e-05, "loss": 0.0472, "step": 59420 }, { "grad_norm": 0.5525022745132446, "learning_rate": 3.864310050838972e-05, "loss": 0.032, "step": 59430 }, { "grad_norm": 0.65977543592453, "learning_rate": 3.8626998605768835e-05, "loss": 0.0268, "step": 59440 }, { "grad_norm": 0.5238434076309204, "learning_rate": 3.8610897946882294e-05, "loss": 0.0325, "step": 59450 }, { "grad_norm": 0.5930488705635071, "learning_rate": 3.859479853349082e-05, "loss": 0.0529, "step": 59460 }, { "grad_norm": 0.2642799913883209, "learning_rate": 3.857870036735503e-05, "loss": 0.0224, "step": 59470 }, { "grad_norm": 0.36563438177108765, "learning_rate": 3.856260345023538e-05, "loss": 0.0335, "step": 59480 }, { "grad_norm": 0.6221231818199158, "learning_rate": 3.854650778389224e-05, "loss": 0.0261, "step": 59490 }, { "grad_norm": 0.3469806909561157, "learning_rate": 3.8530413370085786e-05, "loss": 0.0332, "step": 59500 }, { "grad_norm": 0.39816251397132874, "learning_rate": 3.851432021057608e-05, "loss": 0.0272, "step": 59510 }, { "grad_norm": 0.36649709939956665, "learning_rate": 3.849822830712306e-05, "loss": 0.0351, "step": 59520 }, { "grad_norm": 0.41052740812301636, "learning_rate": 3.848213766148649e-05, "loss": 0.034, "step": 59530 }, { "grad_norm": 0.44317352771759033, "learning_rate": 3.846604827542602e-05, "loss": 0.0221, "step": 59540 }, { "grad_norm": 0.48150762915611267, "learning_rate": 3.844996015070117e-05, "loss": 0.0214, "step": 59550 }, { "grad_norm": 0.24065251648426056, "learning_rate": 3.843387328907132e-05, "loss": 0.0228, "step": 59560 }, { "grad_norm": 0.41100242733955383, "learning_rate": 3.8417787692295696e-05, "loss": 0.0265, "step": 59570 }, { "grad_norm": 0.33723825216293335, "learning_rate": 3.84017033621334e-05, "loss": 0.0182, "step": 59580 }, { "grad_norm": 1.1356415748596191, "learning_rate": 3.838562030034338e-05, "loss": 0.0388, "step": 59590 }, { "grad_norm": 0.34565016627311707, "learning_rate": 3.836953850868447e-05, "loss": 0.0239, "step": 59600 }, { "grad_norm": 0.6678682565689087, "learning_rate": 3.835345798891533e-05, "loss": 0.0276, "step": 59610 }, { "grad_norm": 0.6476311087608337, "learning_rate": 3.833737874279452e-05, "loss": 0.0285, "step": 59620 }, { "grad_norm": 0.5533714294433594, "learning_rate": 3.8321300772080433e-05, "loss": 0.0216, "step": 59630 }, { "grad_norm": 0.25823384523391724, "learning_rate": 3.8305224078531345e-05, "loss": 0.0294, "step": 59640 }, { "grad_norm": 0.5854607224464417, "learning_rate": 3.828914866390537e-05, "loss": 0.0418, "step": 59650 }, { "grad_norm": 0.5987792611122131, "learning_rate": 3.827307452996048e-05, "loss": 0.0297, "step": 59660 }, { "grad_norm": 0.44111138582229614, "learning_rate": 3.825700167845454e-05, "loss": 0.0298, "step": 59670 }, { "grad_norm": 0.24293799698352814, "learning_rate": 3.8240930111145214e-05, "loss": 0.025, "step": 59680 }, { "grad_norm": 0.18458911776542664, "learning_rate": 3.8224859829790116e-05, "loss": 0.0356, "step": 59690 }, { "grad_norm": 0.31560516357421875, "learning_rate": 3.820879083614664e-05, "loss": 0.0234, "step": 59700 }, { "grad_norm": 0.7221130132675171, "learning_rate": 3.819272313197208e-05, "loss": 0.0312, "step": 59710 }, { "grad_norm": 0.3484918177127838, "learning_rate": 3.817665671902355e-05, "loss": 0.0285, "step": 59720 }, { "grad_norm": 0.4458434283733368, "learning_rate": 3.816059159905808e-05, "loss": 0.0281, "step": 59730 }, { "grad_norm": 0.4024519622325897, "learning_rate": 3.8144527773832476e-05, "loss": 0.0213, "step": 59740 }, { "grad_norm": 0.5779972672462463, "learning_rate": 3.812846524510352e-05, "loss": 0.0327, "step": 59750 }, { "grad_norm": 0.34962546825408936, "learning_rate": 3.811240401462775e-05, "loss": 0.0294, "step": 59760 }, { "grad_norm": 0.2802976667881012, "learning_rate": 3.809634408416162e-05, "loss": 0.0197, "step": 59770 }, { "grad_norm": 0.3968326449394226, "learning_rate": 3.8080285455461394e-05, "loss": 0.026, "step": 59780 }, { "grad_norm": 0.3566715717315674, "learning_rate": 3.806422813028323e-05, "loss": 0.0482, "step": 59790 }, { "grad_norm": 0.6349256038665771, "learning_rate": 3.8048172110383114e-05, "loss": 0.0355, "step": 59800 }, { "grad_norm": 0.4604831337928772, "learning_rate": 3.8032117397516944e-05, "loss": 0.0289, "step": 59810 }, { "grad_norm": 0.49808257818222046, "learning_rate": 3.801606399344042e-05, "loss": 0.0317, "step": 59820 }, { "grad_norm": 0.5388044714927673, "learning_rate": 3.800001189990913e-05, "loss": 0.0326, "step": 59830 }, { "grad_norm": 0.5055914521217346, "learning_rate": 3.798396111867847e-05, "loss": 0.0233, "step": 59840 }, { "grad_norm": 0.6240856051445007, "learning_rate": 3.796791165150378e-05, "loss": 0.0293, "step": 59850 }, { "grad_norm": 0.31615135073661804, "learning_rate": 3.795186350014018e-05, "loss": 0.0422, "step": 59860 }, { "grad_norm": 0.45041611790657043, "learning_rate": 3.793581666634266e-05, "loss": 0.0319, "step": 59870 }, { "grad_norm": 0.4436202049255371, "learning_rate": 3.791977115186609e-05, "loss": 0.0324, "step": 59880 }, { "grad_norm": 0.5257066488265991, "learning_rate": 3.79037269584652e-05, "loss": 0.03, "step": 59890 }, { "grad_norm": 0.48171934485435486, "learning_rate": 3.7887684087894545e-05, "loss": 0.0244, "step": 59900 }, { "grad_norm": 0.6690279841423035, "learning_rate": 3.787164254190854e-05, "loss": 0.034, "step": 59910 }, { "grad_norm": 0.5759469866752625, "learning_rate": 3.7855602322261473e-05, "loss": 0.0276, "step": 59920 }, { "grad_norm": 0.3452184796333313, "learning_rate": 3.783956343070746e-05, "loss": 0.0234, "step": 59930 }, { "grad_norm": 0.3418416380882263, "learning_rate": 3.7823525869000526e-05, "loss": 0.0228, "step": 59940 }, { "grad_norm": 0.4455966353416443, "learning_rate": 3.78074896388945e-05, "loss": 0.0238, "step": 59950 }, { "grad_norm": 0.6245043277740479, "learning_rate": 3.7791454742143075e-05, "loss": 0.0353, "step": 59960 }, { "grad_norm": 0.6504950523376465, "learning_rate": 3.7775421180499795e-05, "loss": 0.0269, "step": 59970 }, { "grad_norm": 0.2251807302236557, "learning_rate": 3.7759388955718086e-05, "loss": 0.0278, "step": 59980 }, { "grad_norm": 0.6382627487182617, "learning_rate": 3.774335806955117e-05, "loss": 0.028, "step": 59990 }, { "grad_norm": 0.3897418677806854, "learning_rate": 3.7727328523752215e-05, "loss": 0.0292, "step": 60000 }, { "grad_norm": 0.6091395020484924, "learning_rate": 3.771130032007415e-05, "loss": 0.0236, "step": 60010 }, { "grad_norm": 0.4591229259967804, "learning_rate": 3.769527346026982e-05, "loss": 0.0286, "step": 60020 }, { "grad_norm": 0.5630825161933899, "learning_rate": 3.767924794609188e-05, "loss": 0.0229, "step": 60030 }, { "grad_norm": 0.606180727481842, "learning_rate": 3.7663223779292854e-05, "loss": 0.0279, "step": 60040 }, { "grad_norm": 0.22947733104228973, "learning_rate": 3.764720096162512e-05, "loss": 0.0271, "step": 60050 }, { "grad_norm": 0.24944956600666046, "learning_rate": 3.763117949484092e-05, "loss": 0.0181, "step": 60060 }, { "grad_norm": 0.27558574080467224, "learning_rate": 3.7615159380692344e-05, "loss": 0.029, "step": 60070 }, { "grad_norm": 0.46192479133605957, "learning_rate": 3.759914062093132e-05, "loss": 0.0289, "step": 60080 }, { "grad_norm": 0.675787091255188, "learning_rate": 3.7583123217309615e-05, "loss": 0.0457, "step": 60090 }, { "grad_norm": 0.18804357945919037, "learning_rate": 3.7567107171578904e-05, "loss": 0.0309, "step": 60100 }, { "grad_norm": 0.4969225823879242, "learning_rate": 3.755109248549066e-05, "loss": 0.0347, "step": 60110 }, { "grad_norm": 0.2800215184688568, "learning_rate": 3.75350791607962e-05, "loss": 0.0215, "step": 60120 }, { "grad_norm": 0.32485517859458923, "learning_rate": 3.751906719924676e-05, "loss": 0.0244, "step": 60130 }, { "grad_norm": 0.36662229895591736, "learning_rate": 3.750305660259337e-05, "loss": 0.0223, "step": 60140 }, { "grad_norm": 0.613726019859314, "learning_rate": 3.748704737258693e-05, "loss": 0.0391, "step": 60150 }, { "grad_norm": 0.3646060526371002, "learning_rate": 3.747103951097816e-05, "loss": 0.0182, "step": 60160 }, { "grad_norm": 0.2098173052072525, "learning_rate": 3.745503301951768e-05, "loss": 0.0239, "step": 60170 }, { "grad_norm": 0.2720406949520111, "learning_rate": 3.743902789995592e-05, "loss": 0.0331, "step": 60180 }, { "grad_norm": 0.6648995876312256, "learning_rate": 3.7423024154043183e-05, "loss": 0.0272, "step": 60190 }, { "grad_norm": 0.33547288179397583, "learning_rate": 3.740702178352963e-05, "loss": 0.0358, "step": 60200 }, { "grad_norm": 0.4456021785736084, "learning_rate": 3.739102079016523e-05, "loss": 0.0331, "step": 60210 }, { "grad_norm": 0.7052063345909119, "learning_rate": 3.7375021175699846e-05, "loss": 0.0335, "step": 60220 }, { "grad_norm": 0.9864174127578735, "learning_rate": 3.735902294188317e-05, "loss": 0.0239, "step": 60230 }, { "grad_norm": 0.4017731249332428, "learning_rate": 3.734302609046473e-05, "loss": 0.0291, "step": 60240 }, { "grad_norm": 0.6352142691612244, "learning_rate": 3.732703062319391e-05, "loss": 0.0261, "step": 60250 }, { "grad_norm": 0.31644296646118164, "learning_rate": 3.731103654181999e-05, "loss": 0.0253, "step": 60260 }, { "grad_norm": 0.6095880270004272, "learning_rate": 3.729504384809203e-05, "loss": 0.0325, "step": 60270 }, { "grad_norm": 0.3087562024593353, "learning_rate": 3.727905254375898e-05, "loss": 0.023, "step": 60280 }, { "grad_norm": 0.4012301564216614, "learning_rate": 3.7263062630569614e-05, "loss": 0.0299, "step": 60290 }, { "grad_norm": 0.35922786593437195, "learning_rate": 3.724707411027256e-05, "loss": 0.0335, "step": 60300 }, { "grad_norm": 0.251553475856781, "learning_rate": 3.723108698461631e-05, "loss": 0.0245, "step": 60310 }, { "grad_norm": 0.3287111818790436, "learning_rate": 3.72151012553492e-05, "loss": 0.0207, "step": 60320 }, { "grad_norm": 0.37171778082847595, "learning_rate": 3.719911692421939e-05, "loss": 0.0235, "step": 60330 }, { "grad_norm": 0.38168880343437195, "learning_rate": 3.71831339929749e-05, "loss": 0.0368, "step": 60340 }, { "grad_norm": 0.5163967609405518, "learning_rate": 3.716715246336362e-05, "loss": 0.0267, "step": 60350 }, { "grad_norm": 0.2991746664047241, "learning_rate": 3.715117233713324e-05, "loss": 0.0252, "step": 60360 }, { "grad_norm": 0.36772146821022034, "learning_rate": 3.713519361603135e-05, "loss": 0.0243, "step": 60370 }, { "grad_norm": 0.6929294466972351, "learning_rate": 3.711921630180532e-05, "loss": 0.025, "step": 60380 }, { "grad_norm": 0.27445724606513977, "learning_rate": 3.710324039620245e-05, "loss": 0.0251, "step": 60390 }, { "grad_norm": 0.7771074175834656, "learning_rate": 3.708726590096982e-05, "loss": 0.0314, "step": 60400 }, { "grad_norm": 0.42221084237098694, "learning_rate": 3.707129281785437e-05, "loss": 0.0236, "step": 60410 }, { "grad_norm": 0.4852900803089142, "learning_rate": 3.705532114860291e-05, "loss": 0.0277, "step": 60420 }, { "grad_norm": 0.27435213327407837, "learning_rate": 3.703935089496207e-05, "loss": 0.0185, "step": 60430 }, { "grad_norm": 0.38745179772377014, "learning_rate": 3.702338205867829e-05, "loss": 0.0236, "step": 60440 }, { "grad_norm": 0.53750079870224, "learning_rate": 3.7007414641497976e-05, "loss": 0.02, "step": 60450 }, { "grad_norm": 0.7414544224739075, "learning_rate": 3.699144864516726e-05, "loss": 0.0286, "step": 60460 }, { "grad_norm": 0.3617421090602875, "learning_rate": 3.6975484071432153e-05, "loss": 0.0305, "step": 60470 }, { "grad_norm": 0.4704335331916809, "learning_rate": 3.695952092203854e-05, "loss": 0.0369, "step": 60480 }, { "grad_norm": 0.5095056891441345, "learning_rate": 3.69435591987321e-05, "loss": 0.0249, "step": 60490 }, { "grad_norm": 0.3521387279033661, "learning_rate": 3.6927598903258374e-05, "loss": 0.0229, "step": 60500 }, { "grad_norm": 0.33572864532470703, "learning_rate": 3.691164003736281e-05, "loss": 0.0265, "step": 60510 }, { "grad_norm": 0.42575737833976746, "learning_rate": 3.689568260279061e-05, "loss": 0.0194, "step": 60520 }, { "grad_norm": 0.4777062237262726, "learning_rate": 3.687972660128686e-05, "loss": 0.0295, "step": 60530 }, { "grad_norm": 0.5977767109870911, "learning_rate": 3.686377203459648e-05, "loss": 0.0254, "step": 60540 }, { "grad_norm": 0.46117696166038513, "learning_rate": 3.6847818904464226e-05, "loss": 0.0268, "step": 60550 }, { "grad_norm": 0.414815217256546, "learning_rate": 3.683186721263474e-05, "loss": 0.0491, "step": 60560 }, { "grad_norm": 0.6370900273323059, "learning_rate": 3.6815916960852435e-05, "loss": 0.0259, "step": 60570 }, { "grad_norm": 0.6057232022285461, "learning_rate": 3.679996815086165e-05, "loss": 0.0309, "step": 60580 }, { "grad_norm": 0.593208909034729, "learning_rate": 3.678402078440649e-05, "loss": 0.0248, "step": 60590 }, { "grad_norm": 0.4672481417655945, "learning_rate": 3.676807486323096e-05, "loss": 0.0265, "step": 60600 }, { "grad_norm": 0.4087575376033783, "learning_rate": 3.6752130389078864e-05, "loss": 0.0403, "step": 60610 }, { "grad_norm": 0.6483577489852905, "learning_rate": 3.673618736369388e-05, "loss": 0.0268, "step": 60620 }, { "grad_norm": 0.5151599645614624, "learning_rate": 3.6720245788819474e-05, "loss": 0.025, "step": 60630 }, { "grad_norm": 0.47432979941368103, "learning_rate": 3.6704305666199044e-05, "loss": 0.0315, "step": 60640 }, { "grad_norm": 0.4186747968196869, "learning_rate": 3.668836699757576e-05, "loss": 0.0243, "step": 60650 }, { "grad_norm": 0.21924002468585968, "learning_rate": 3.667242978469265e-05, "loss": 0.0431, "step": 60660 }, { "grad_norm": 0.39181336760520935, "learning_rate": 3.6656494029292575e-05, "loss": 0.0294, "step": 60670 }, { "grad_norm": 0.3729739487171173, "learning_rate": 3.664055973311825e-05, "loss": 0.0336, "step": 60680 }, { "grad_norm": 0.5926548838615417, "learning_rate": 3.662462689791221e-05, "loss": 0.0264, "step": 60690 }, { "grad_norm": 0.3822999596595764, "learning_rate": 3.660869552541689e-05, "loss": 0.0229, "step": 60700 }, { "grad_norm": 0.6049081683158875, "learning_rate": 3.659276561737448e-05, "loss": 0.0364, "step": 60710 }, { "grad_norm": 0.3175472021102905, "learning_rate": 3.657683717552708e-05, "loss": 0.0308, "step": 60720 }, { "grad_norm": 0.34447383880615234, "learning_rate": 3.6560910201616574e-05, "loss": 0.0261, "step": 60730 }, { "grad_norm": 0.36460691690444946, "learning_rate": 3.654498469738472e-05, "loss": 0.0268, "step": 60740 }, { "grad_norm": 0.3204174339771271, "learning_rate": 3.6529060664573126e-05, "loss": 0.0255, "step": 60750 }, { "grad_norm": 0.48384615778923035, "learning_rate": 3.6513138104923176e-05, "loss": 0.0325, "step": 60760 }, { "grad_norm": 0.5812432169914246, "learning_rate": 3.6497217020176176e-05, "loss": 0.0345, "step": 60770 }, { "grad_norm": 0.2881401777267456, "learning_rate": 3.648129741207323e-05, "loss": 0.0301, "step": 60780 }, { "grad_norm": 0.373718798160553, "learning_rate": 3.646537928235527e-05, "loss": 0.0238, "step": 60790 }, { "grad_norm": 0.5520724058151245, "learning_rate": 3.6449462632763075e-05, "loss": 0.024, "step": 60800 }, { "grad_norm": 0.6878964304924011, "learning_rate": 3.643354746503729e-05, "loss": 0.0238, "step": 60810 }, { "grad_norm": 0.3061474561691284, "learning_rate": 3.6417633780918345e-05, "loss": 0.023, "step": 60820 }, { "grad_norm": 0.38632693886756897, "learning_rate": 3.640172158214655e-05, "loss": 0.0177, "step": 60830 }, { "grad_norm": 0.20487810671329498, "learning_rate": 3.6385810870462035e-05, "loss": 0.0265, "step": 60840 }, { "grad_norm": 0.5678831338882446, "learning_rate": 3.636990164760479e-05, "loss": 0.0336, "step": 60850 }, { "grad_norm": 0.3435019552707672, "learning_rate": 3.6353993915314596e-05, "loss": 0.0376, "step": 60860 }, { "grad_norm": 0.3486764132976532, "learning_rate": 3.6338087675331124e-05, "loss": 0.0176, "step": 60870 }, { "grad_norm": 0.49862509965896606, "learning_rate": 3.632218292939383e-05, "loss": 0.0226, "step": 60880 }, { "grad_norm": 0.42205810546875, "learning_rate": 3.6306279679242033e-05, "loss": 0.0241, "step": 60890 }, { "grad_norm": 0.5294206142425537, "learning_rate": 3.629037792661492e-05, "loss": 0.0214, "step": 60900 }, { "grad_norm": 0.3425261080265045, "learning_rate": 3.627447767325147e-05, "loss": 0.0352, "step": 60910 }, { "grad_norm": 0.3886183500289917, "learning_rate": 3.62585789208905e-05, "loss": 0.0232, "step": 60920 }, { "grad_norm": 0.32209348678588867, "learning_rate": 3.624268167127068e-05, "loss": 0.0234, "step": 60930 }, { "grad_norm": 0.39374634623527527, "learning_rate": 3.622678592613051e-05, "loss": 0.0295, "step": 60940 }, { "grad_norm": 0.3208768367767334, "learning_rate": 3.6210891687208304e-05, "loss": 0.03, "step": 60950 }, { "grad_norm": 0.1659345030784607, "learning_rate": 3.619499895624228e-05, "loss": 0.0241, "step": 60960 }, { "grad_norm": 0.3808448016643524, "learning_rate": 3.61791077349704e-05, "loss": 0.024, "step": 60970 }, { "grad_norm": 0.5953986644744873, "learning_rate": 3.616321802513053e-05, "loss": 0.0301, "step": 60980 }, { "grad_norm": 0.26027610898017883, "learning_rate": 3.6147329828460325e-05, "loss": 0.0317, "step": 60990 }, { "grad_norm": 0.3994484543800354, "learning_rate": 3.6131443146697306e-05, "loss": 0.0285, "step": 61000 }, { "grad_norm": 0.612472653388977, "learning_rate": 3.611555798157879e-05, "loss": 0.0277, "step": 61010 }, { "grad_norm": 0.4938116669654846, "learning_rate": 3.6099674334842e-05, "loss": 0.0292, "step": 61020 }, { "grad_norm": 0.34233129024505615, "learning_rate": 3.6083792208223934e-05, "loss": 0.0249, "step": 61030 }, { "grad_norm": 0.5370622873306274, "learning_rate": 3.606791160346142e-05, "loss": 0.0274, "step": 61040 }, { "grad_norm": 0.5219064950942993, "learning_rate": 3.6052032522291134e-05, "loss": 0.0395, "step": 61050 }, { "grad_norm": 0.29923513531684875, "learning_rate": 3.603615496644962e-05, "loss": 0.023, "step": 61060 }, { "grad_norm": 0.458146333694458, "learning_rate": 3.60202789376732e-05, "loss": 0.0279, "step": 61070 }, { "grad_norm": 0.5116509795188904, "learning_rate": 3.6004404437698036e-05, "loss": 0.0286, "step": 61080 }, { "grad_norm": 0.2511070966720581, "learning_rate": 3.5988531468260184e-05, "loss": 0.0199, "step": 61090 }, { "grad_norm": 0.3321871757507324, "learning_rate": 3.597266003109546e-05, "loss": 0.0309, "step": 61100 }, { "grad_norm": 0.39937514066696167, "learning_rate": 3.5956790127939557e-05, "loss": 0.0222, "step": 61110 }, { "grad_norm": 0.6492334008216858, "learning_rate": 3.594092176052796e-05, "loss": 0.0326, "step": 61120 }, { "grad_norm": 0.7212228775024414, "learning_rate": 3.592505493059603e-05, "loss": 0.0332, "step": 61130 }, { "grad_norm": 0.5328919291496277, "learning_rate": 3.5909189639878905e-05, "loss": 0.0374, "step": 61140 }, { "grad_norm": 0.8120834231376648, "learning_rate": 3.5893325890111626e-05, "loss": 0.0438, "step": 61150 }, { "grad_norm": 0.4912027418613434, "learning_rate": 3.5877463683029034e-05, "loss": 0.0226, "step": 61160 }, { "grad_norm": 0.27638742327690125, "learning_rate": 3.586160302036578e-05, "loss": 0.0269, "step": 61170 }, { "grad_norm": 0.46674054861068726, "learning_rate": 3.584574390385636e-05, "loss": 0.0332, "step": 61180 }, { "grad_norm": 0.5264849066734314, "learning_rate": 3.582988633523511e-05, "loss": 0.0294, "step": 61190 }, { "grad_norm": 0.5205361247062683, "learning_rate": 3.581403031623616e-05, "loss": 0.0287, "step": 61200 }, { "grad_norm": 0.2856141924858093, "learning_rate": 3.579817584859354e-05, "loss": 0.0293, "step": 61210 }, { "grad_norm": 0.5572342872619629, "learning_rate": 3.578232293404107e-05, "loss": 0.0218, "step": 61220 }, { "grad_norm": 0.7262758612632751, "learning_rate": 3.5766471574312385e-05, "loss": 0.0402, "step": 61230 }, { "grad_norm": 0.5062760710716248, "learning_rate": 3.5750621771140964e-05, "loss": 0.0248, "step": 61240 }, { "grad_norm": 0.5806033611297607, "learning_rate": 3.5734773526260115e-05, "loss": 0.031, "step": 61250 }, { "grad_norm": 0.3221057653427124, "learning_rate": 3.5718926841402996e-05, "loss": 0.0211, "step": 61260 }, { "grad_norm": 0.3988267481327057, "learning_rate": 3.5703081718302555e-05, "loss": 0.0228, "step": 61270 }, { "grad_norm": 0.6657896041870117, "learning_rate": 3.568723815869162e-05, "loss": 0.0313, "step": 61280 }, { "grad_norm": 0.3908320367336273, "learning_rate": 3.5671396164302777e-05, "loss": 0.0277, "step": 61290 }, { "grad_norm": 0.41986730694770813, "learning_rate": 3.565555573686853e-05, "loss": 0.0276, "step": 61300 }, { "grad_norm": 0.3507155478000641, "learning_rate": 3.563971687812113e-05, "loss": 0.0295, "step": 61310 }, { "grad_norm": 0.4506720006465912, "learning_rate": 3.5623879589792706e-05, "loss": 0.0191, "step": 61320 }, { "grad_norm": 0.43256640434265137, "learning_rate": 3.5608043873615185e-05, "loss": 0.0267, "step": 61330 }, { "grad_norm": 0.4065607488155365, "learning_rate": 3.559220973132035e-05, "loss": 0.0272, "step": 61340 }, { "grad_norm": 0.3525390028953552, "learning_rate": 3.557637716463981e-05, "loss": 0.0242, "step": 61350 }, { "grad_norm": 0.501265823841095, "learning_rate": 3.556054617530497e-05, "loss": 0.0277, "step": 61360 }, { "grad_norm": 0.3037167489528656, "learning_rate": 3.5544716765047106e-05, "loss": 0.0218, "step": 61370 }, { "grad_norm": 0.3472358286380768, "learning_rate": 3.5528888935597274e-05, "loss": 0.0304, "step": 61380 }, { "grad_norm": 0.5135502219200134, "learning_rate": 3.551306268868636e-05, "loss": 0.0259, "step": 61390 }, { "grad_norm": 0.3245714604854584, "learning_rate": 3.5497238026045166e-05, "loss": 0.0296, "step": 61400 }, { "grad_norm": 0.4447912573814392, "learning_rate": 3.548141494940421e-05, "loss": 0.027, "step": 61410 }, { "grad_norm": 0.41653668880462646, "learning_rate": 3.546559346049388e-05, "loss": 0.0221, "step": 61420 }, { "grad_norm": 0.3330276608467102, "learning_rate": 3.544977356104441e-05, "loss": 0.0197, "step": 61430 }, { "grad_norm": 0.4521268904209137, "learning_rate": 3.543395525278582e-05, "loss": 0.0245, "step": 61440 }, { "grad_norm": 0.354360431432724, "learning_rate": 3.5418138537447985e-05, "loss": 0.0255, "step": 61450 }, { "grad_norm": 0.45690426230430603, "learning_rate": 3.5402323416760574e-05, "loss": 0.0433, "step": 61460 }, { "grad_norm": 0.3494199812412262, "learning_rate": 3.538650989245316e-05, "loss": 0.0225, "step": 61470 }, { "grad_norm": 0.4436492621898651, "learning_rate": 3.537069796625504e-05, "loss": 0.0277, "step": 61480 }, { "grad_norm": 0.7600253820419312, "learning_rate": 3.53548876398954e-05, "loss": 0.0234, "step": 61490 }, { "grad_norm": 0.32537880539894104, "learning_rate": 3.5339078915103214e-05, "loss": 0.0245, "step": 61500 }, { "grad_norm": 0.6166929602622986, "learning_rate": 3.532327179360733e-05, "loss": 0.028, "step": 61510 }, { "grad_norm": 0.38697564601898193, "learning_rate": 3.530746627713636e-05, "loss": 0.0272, "step": 61520 }, { "grad_norm": 0.40802720189094543, "learning_rate": 3.5291662367418805e-05, "loss": 0.0245, "step": 61530 }, { "grad_norm": 0.3873223066329956, "learning_rate": 3.527586006618293e-05, "loss": 0.0237, "step": 61540 }, { "grad_norm": 0.658054530620575, "learning_rate": 3.526005937515687e-05, "loss": 0.0207, "step": 61550 }, { "grad_norm": 0.23832052946090698, "learning_rate": 3.524426029606856e-05, "loss": 0.0275, "step": 61560 }, { "grad_norm": 0.3209041357040405, "learning_rate": 3.5228462830645746e-05, "loss": 0.0203, "step": 61570 }, { "grad_norm": 0.4283449053764343, "learning_rate": 3.5212666980616036e-05, "loss": 0.0257, "step": 61580 }, { "grad_norm": 0.4947495460510254, "learning_rate": 3.519687274770681e-05, "loss": 0.0174, "step": 61590 }, { "grad_norm": 0.5535489320755005, "learning_rate": 3.518108013364535e-05, "loss": 0.036, "step": 61600 }, { "grad_norm": 0.4850211441516876, "learning_rate": 3.516528914015868e-05, "loss": 0.0296, "step": 61610 }, { "grad_norm": 0.7144868969917297, "learning_rate": 3.5149499768973694e-05, "loss": 0.0299, "step": 61620 }, { "grad_norm": 0.2588171660900116, "learning_rate": 3.513371202181708e-05, "loss": 0.0287, "step": 61630 }, { "grad_norm": 0.41572320461273193, "learning_rate": 3.511792590041537e-05, "loss": 0.0263, "step": 61640 }, { "grad_norm": 0.37611910700798035, "learning_rate": 3.510214140649488e-05, "loss": 0.0219, "step": 61650 }, { "grad_norm": 0.24207644164562225, "learning_rate": 3.508635854178184e-05, "loss": 0.0329, "step": 61660 }, { "grad_norm": 0.4779569208621979, "learning_rate": 3.50705773080022e-05, "loss": 0.0291, "step": 61670 }, { "grad_norm": 0.6818452477455139, "learning_rate": 3.505479770688178e-05, "loss": 0.0258, "step": 61680 }, { "grad_norm": 0.3643985092639923, "learning_rate": 3.503901974014622e-05, "loss": 0.0163, "step": 61690 }, { "grad_norm": 0.440468966960907, "learning_rate": 3.502324340952095e-05, "loss": 0.0234, "step": 61700 }, { "grad_norm": 0.45300349593162537, "learning_rate": 3.500746871673126e-05, "loss": 0.0336, "step": 61710 }, { "grad_norm": 0.3757050335407257, "learning_rate": 3.499169566350227e-05, "loss": 0.0306, "step": 61720 }, { "grad_norm": 0.3467967212200165, "learning_rate": 3.4975924251558876e-05, "loss": 0.025, "step": 61730 }, { "grad_norm": 0.5343717932701111, "learning_rate": 3.496015448262582e-05, "loss": 0.0258, "step": 61740 }, { "grad_norm": 0.3676266074180603, "learning_rate": 3.494438635842764e-05, "loss": 0.0285, "step": 61750 }, { "grad_norm": 0.3747934401035309, "learning_rate": 3.492861988068874e-05, "loss": 0.0234, "step": 61760 }, { "grad_norm": 0.5224794149398804, "learning_rate": 3.491285505113332e-05, "loss": 0.0285, "step": 61770 }, { "grad_norm": 0.41955360770225525, "learning_rate": 3.4897091871485376e-05, "loss": 0.031, "step": 61780 }, { "grad_norm": 0.4037942886352539, "learning_rate": 3.488133034346875e-05, "loss": 0.0425, "step": 61790 }, { "grad_norm": 0.291866272687912, "learning_rate": 3.486557046880713e-05, "loss": 0.0238, "step": 61800 }, { "grad_norm": 0.4886951148509979, "learning_rate": 3.484981224922395e-05, "loss": 0.0292, "step": 61810 }, { "grad_norm": 0.5388455986976624, "learning_rate": 3.4834055686442544e-05, "loss": 0.0312, "step": 61820 }, { "grad_norm": 0.45102354884147644, "learning_rate": 3.4818300782186e-05, "loss": 0.0409, "step": 61830 }, { "grad_norm": 0.20349450409412384, "learning_rate": 3.4802547538177236e-05, "loss": 0.0293, "step": 61840 }, { "grad_norm": 0.33168527483940125, "learning_rate": 3.478679595613905e-05, "loss": 0.0231, "step": 61850 }, { "grad_norm": 0.4270045757293701, "learning_rate": 3.4771046037793975e-05, "loss": 0.0298, "step": 61860 }, { "grad_norm": 0.518171489238739, "learning_rate": 3.4755297784864416e-05, "loss": 0.0223, "step": 61870 }, { "grad_norm": 0.8567748069763184, "learning_rate": 3.473955119907257e-05, "loss": 0.0366, "step": 61880 }, { "grad_norm": 0.6054223775863647, "learning_rate": 3.472380628214046e-05, "loss": 0.0316, "step": 61890 }, { "grad_norm": 0.35525786876678467, "learning_rate": 3.4708063035789906e-05, "loss": 0.0177, "step": 61900 }, { "grad_norm": 0.34814196825027466, "learning_rate": 3.4692321461742614e-05, "loss": 0.0258, "step": 61910 }, { "grad_norm": 0.5205919146537781, "learning_rate": 3.4676581561720023e-05, "loss": 0.0273, "step": 61920 }, { "grad_norm": 0.3290151059627533, "learning_rate": 3.466084333744344e-05, "loss": 0.0236, "step": 61930 }, { "grad_norm": 0.32885465025901794, "learning_rate": 3.4645106790633964e-05, "loss": 0.0317, "step": 61940 }, { "grad_norm": 0.36823734641075134, "learning_rate": 3.4629371923012526e-05, "loss": 0.0213, "step": 61950 }, { "grad_norm": 0.39870139956474304, "learning_rate": 3.4613638736299856e-05, "loss": 0.0236, "step": 61960 }, { "grad_norm": 0.3150848150253296, "learning_rate": 3.459790723221652e-05, "loss": 0.0198, "step": 61970 }, { "grad_norm": 0.5109792351722717, "learning_rate": 3.45821774124829e-05, "loss": 0.0297, "step": 61980 }, { "grad_norm": 0.4752870500087738, "learning_rate": 3.456644927881918e-05, "loss": 0.0265, "step": 61990 }, { "grad_norm": 0.37296050786972046, "learning_rate": 3.4550722832945354e-05, "loss": 0.0265, "step": 62000 }, { "grad_norm": 0.43574708700180054, "learning_rate": 3.453499807658126e-05, "loss": 0.0272, "step": 62010 }, { "grad_norm": 0.2670745253562927, "learning_rate": 3.451927501144653e-05, "loss": 0.0235, "step": 62020 }, { "grad_norm": 0.4518741965293884, "learning_rate": 3.450355363926059e-05, "loss": 0.0256, "step": 62030 }, { "grad_norm": 0.5220279693603516, "learning_rate": 3.448783396174273e-05, "loss": 0.0176, "step": 62040 }, { "grad_norm": 0.6242671012878418, "learning_rate": 3.4472115980612034e-05, "loss": 0.0259, "step": 62050 }, { "grad_norm": 0.48607876896858215, "learning_rate": 3.4456399697587396e-05, "loss": 0.0243, "step": 62060 }, { "grad_norm": 0.29445093870162964, "learning_rate": 3.4440685114387506e-05, "loss": 0.0242, "step": 62070 }, { "grad_norm": 0.543441116809845, "learning_rate": 3.442497223273091e-05, "loss": 0.0289, "step": 62080 }, { "grad_norm": 0.325928658246994, "learning_rate": 3.4409261054335926e-05, "loss": 0.0235, "step": 62090 }, { "grad_norm": 0.28584951162338257, "learning_rate": 3.439355158092069e-05, "loss": 0.0223, "step": 62100 }, { "grad_norm": 0.26754099130630493, "learning_rate": 3.437784381420322e-05, "loss": 0.0223, "step": 62110 }, { "grad_norm": 0.2793971598148346, "learning_rate": 3.436213775590126e-05, "loss": 0.0136, "step": 62120 }, { "grad_norm": 0.30312997102737427, "learning_rate": 3.4346433407732405e-05, "loss": 0.0263, "step": 62130 }, { "grad_norm": 0.3881233334541321, "learning_rate": 3.433073077141405e-05, "loss": 0.0221, "step": 62140 }, { "grad_norm": 0.3995305001735687, "learning_rate": 3.431502984866343e-05, "loss": 0.0215, "step": 62150 }, { "grad_norm": 0.3983164131641388, "learning_rate": 3.429933064119753e-05, "loss": 0.0272, "step": 62160 }, { "grad_norm": 0.505955696105957, "learning_rate": 3.4283633150733255e-05, "loss": 0.0342, "step": 62170 }, { "grad_norm": 0.23032642900943756, "learning_rate": 3.4267937378987225e-05, "loss": 0.0189, "step": 62180 }, { "grad_norm": 0.5396579504013062, "learning_rate": 3.425224332767592e-05, "loss": 0.0241, "step": 62190 }, { "grad_norm": 0.4568687975406647, "learning_rate": 3.4236550998515595e-05, "loss": 0.0362, "step": 62200 }, { "grad_norm": 0.23069727420806885, "learning_rate": 3.422086039322235e-05, "loss": 0.0266, "step": 62210 }, { "grad_norm": 0.5939576029777527, "learning_rate": 3.420517151351209e-05, "loss": 0.0287, "step": 62220 }, { "grad_norm": 0.3753196895122528, "learning_rate": 3.418948436110053e-05, "loss": 0.0247, "step": 62230 }, { "grad_norm": 0.444815069437027, "learning_rate": 3.417379893770319e-05, "loss": 0.0317, "step": 62240 }, { "grad_norm": 0.2535163462162018, "learning_rate": 3.41581152450354e-05, "loss": 0.0234, "step": 62250 }, { "grad_norm": 0.40602797269821167, "learning_rate": 3.414243328481231e-05, "loss": 0.0274, "step": 62260 }, { "grad_norm": 0.2245919108390808, "learning_rate": 3.412675305874888e-05, "loss": 0.0244, "step": 62270 }, { "grad_norm": 0.403800904750824, "learning_rate": 3.4111074568559876e-05, "loss": 0.0229, "step": 62280 }, { "grad_norm": 0.6321827173233032, "learning_rate": 3.4095397815959846e-05, "loss": 0.0258, "step": 62290 }, { "grad_norm": 0.3005104959011078, "learning_rate": 3.407972280266322e-05, "loss": 0.0291, "step": 62300 }, { "grad_norm": 0.7123910188674927, "learning_rate": 3.4064049530384166e-05, "loss": 0.0264, "step": 62310 }, { "grad_norm": 0.28279411792755127, "learning_rate": 3.40483780008367e-05, "loss": 0.0179, "step": 62320 }, { "grad_norm": 0.35155460238456726, "learning_rate": 3.4032708215734635e-05, "loss": 0.0239, "step": 62330 }, { "grad_norm": 0.46755513548851013, "learning_rate": 3.4017040176791584e-05, "loss": 0.0327, "step": 62340 }, { "grad_norm": 0.3043578267097473, "learning_rate": 3.4001373885720976e-05, "loss": 0.0395, "step": 62350 }, { "grad_norm": 0.3628482222557068, "learning_rate": 3.398570934423608e-05, "loss": 0.036, "step": 62360 }, { "grad_norm": 0.22069038450717926, "learning_rate": 3.397004655404994e-05, "loss": 0.0283, "step": 62370 }, { "grad_norm": 0.35910582542419434, "learning_rate": 3.395438551687541e-05, "loss": 0.0329, "step": 62380 }, { "grad_norm": 0.584378719329834, "learning_rate": 3.393872623442515e-05, "loss": 0.0228, "step": 62390 }, { "grad_norm": 0.43244466185569763, "learning_rate": 3.392306870841164e-05, "loss": 0.0245, "step": 62400 }, { "grad_norm": 0.46658948063850403, "learning_rate": 3.390741294054715e-05, "loss": 0.038, "step": 62410 }, { "grad_norm": 0.5073156952857971, "learning_rate": 3.38917589325438e-05, "loss": 0.0309, "step": 62420 }, { "grad_norm": 0.49538326263427734, "learning_rate": 3.387610668611349e-05, "loss": 0.0256, "step": 62430 }, { "grad_norm": 0.3711690604686737, "learning_rate": 3.38604562029679e-05, "loss": 0.0268, "step": 62440 }, { "grad_norm": 0.8846796154975891, "learning_rate": 3.384480748481855e-05, "loss": 0.0346, "step": 62450 }, { "grad_norm": 0.3913891017436981, "learning_rate": 3.382916053337676e-05, "loss": 0.0279, "step": 62460 }, { "grad_norm": 0.5635326504707336, "learning_rate": 3.381351535035366e-05, "loss": 0.0353, "step": 62470 }, { "grad_norm": 0.4702206254005432, "learning_rate": 3.379787193746019e-05, "loss": 0.0228, "step": 62480 }, { "grad_norm": 0.4610627591609955, "learning_rate": 3.378223029640709e-05, "loss": 0.0265, "step": 62490 }, { "grad_norm": 0.4217020273208618, "learning_rate": 3.376659042890489e-05, "loss": 0.0268, "step": 62500 }, { "grad_norm": 0.3761882483959198, "learning_rate": 3.375095233666397e-05, "loss": 0.0329, "step": 62510 }, { "grad_norm": 0.7081243395805359, "learning_rate": 3.373531602139447e-05, "loss": 0.067, "step": 62520 }, { "grad_norm": 0.47047457098960876, "learning_rate": 3.371968148480635e-05, "loss": 0.0188, "step": 62530 }, { "grad_norm": 0.3190332055091858, "learning_rate": 3.370404872860937e-05, "loss": 0.0221, "step": 62540 }, { "grad_norm": 0.36893028020858765, "learning_rate": 3.3688417754513145e-05, "loss": 0.0225, "step": 62550 }, { "grad_norm": 0.3940223157405853, "learning_rate": 3.367278856422703e-05, "loss": 0.0436, "step": 62560 }, { "grad_norm": 0.34323933720588684, "learning_rate": 3.365716115946021e-05, "loss": 0.0227, "step": 62570 }, { "grad_norm": 0.41021591424942017, "learning_rate": 3.3641535541921666e-05, "loss": 0.0332, "step": 62580 }, { "grad_norm": 0.34644976258277893, "learning_rate": 3.36259117133202e-05, "loss": 0.0294, "step": 62590 }, { "grad_norm": 0.44227147102355957, "learning_rate": 3.36102896753644e-05, "loss": 0.0418, "step": 62600 }, { "grad_norm": 0.35851749777793884, "learning_rate": 3.35946694297627e-05, "loss": 0.0457, "step": 62610 }, { "grad_norm": 0.3095875680446625, "learning_rate": 3.357905097822328e-05, "loss": 0.0326, "step": 62620 }, { "grad_norm": 0.2603227198123932, "learning_rate": 3.356343432245416e-05, "loss": 0.0237, "step": 62630 }, { "grad_norm": 0.37236881256103516, "learning_rate": 3.354781946416315e-05, "loss": 0.0164, "step": 62640 }, { "grad_norm": 0.4358791708946228, "learning_rate": 3.3532206405057864e-05, "loss": 0.0223, "step": 62650 }, { "grad_norm": 0.5214610695838928, "learning_rate": 3.351659514684573e-05, "loss": 0.026, "step": 62660 }, { "grad_norm": 0.34650859236717224, "learning_rate": 3.350098569123395e-05, "loss": 0.0214, "step": 62670 }, { "grad_norm": 0.6114858984947205, "learning_rate": 3.348537803992958e-05, "loss": 0.025, "step": 62680 }, { "grad_norm": 0.3651021718978882, "learning_rate": 3.346977219463945e-05, "loss": 0.0204, "step": 62690 }, { "grad_norm": 0.4083826541900635, "learning_rate": 3.345416815707017e-05, "loss": 0.0365, "step": 62700 }, { "grad_norm": 0.2423582822084427, "learning_rate": 3.3438565928928185e-05, "loss": 0.0247, "step": 62710 }, { "grad_norm": 0.21967938542366028, "learning_rate": 3.3422965511919736e-05, "loss": 0.0339, "step": 62720 }, { "grad_norm": 0.5235101580619812, "learning_rate": 3.3407366907750846e-05, "loss": 0.0216, "step": 62730 }, { "grad_norm": 0.7593891024589539, "learning_rate": 3.339177011812737e-05, "loss": 0.0371, "step": 62740 }, { "grad_norm": 0.26788899302482605, "learning_rate": 3.3376175144754935e-05, "loss": 0.0288, "step": 62750 }, { "grad_norm": 0.5666612386703491, "learning_rate": 3.3360581989339004e-05, "loss": 0.0263, "step": 62760 }, { "grad_norm": 0.5195663571357727, "learning_rate": 3.3344990653584804e-05, "loss": 0.0269, "step": 62770 }, { "grad_norm": 0.23695825040340424, "learning_rate": 3.332940113919739e-05, "loss": 0.027, "step": 62780 }, { "grad_norm": 0.3248499035835266, "learning_rate": 3.331381344788161e-05, "loss": 0.0197, "step": 62790 }, { "grad_norm": 0.34019196033477783, "learning_rate": 3.3298227581342076e-05, "loss": 0.0265, "step": 62800 }, { "grad_norm": 0.411165714263916, "learning_rate": 3.328264354128329e-05, "loss": 0.022, "step": 62810 }, { "grad_norm": 0.4678449034690857, "learning_rate": 3.3267061329409465e-05, "loss": 0.0234, "step": 62820 }, { "grad_norm": 0.2824901342391968, "learning_rate": 3.325148094742466e-05, "loss": 0.0168, "step": 62830 }, { "grad_norm": 0.19942955672740936, "learning_rate": 3.323590239703271e-05, "loss": 0.0279, "step": 62840 }, { "grad_norm": 0.298807293176651, "learning_rate": 3.322032567993728e-05, "loss": 0.0312, "step": 62850 }, { "grad_norm": 0.32895004749298096, "learning_rate": 3.3204750797841774e-05, "loss": 0.026, "step": 62860 }, { "grad_norm": 0.42032691836357117, "learning_rate": 3.31891777524495e-05, "loss": 0.0291, "step": 62870 }, { "grad_norm": 0.4112412929534912, "learning_rate": 3.317360654546347e-05, "loss": 0.0233, "step": 62880 }, { "grad_norm": 0.3350737392902374, "learning_rate": 3.3158037178586534e-05, "loss": 0.0266, "step": 62890 }, { "grad_norm": 0.6683763265609741, "learning_rate": 3.3142469653521316e-05, "loss": 0.0239, "step": 62900 }, { "grad_norm": 0.9233478903770447, "learning_rate": 3.3126903971970284e-05, "loss": 0.0291, "step": 62910 }, { "grad_norm": 0.4270220696926117, "learning_rate": 3.311134013563564e-05, "loss": 0.0206, "step": 62920 }, { "grad_norm": 0.2520073652267456, "learning_rate": 3.309577814621947e-05, "loss": 0.024, "step": 62930 }, { "grad_norm": 0.36350560188293457, "learning_rate": 3.308021800542359e-05, "loss": 0.0281, "step": 62940 }, { "grad_norm": 0.6165485382080078, "learning_rate": 3.306465971494963e-05, "loss": 0.0297, "step": 62950 }, { "grad_norm": 0.4380011558532715, "learning_rate": 3.304910327649901e-05, "loss": 0.0346, "step": 62960 }, { "grad_norm": 0.6411231756210327, "learning_rate": 3.303354869177297e-05, "loss": 0.0314, "step": 62970 }, { "grad_norm": 0.2150651067495346, "learning_rate": 3.301799596247255e-05, "loss": 0.0237, "step": 62980 }, { "grad_norm": 0.45358189940452576, "learning_rate": 3.300244509029854e-05, "loss": 0.0359, "step": 62990 }, { "grad_norm": 0.635371744632721, "learning_rate": 3.2986896076951585e-05, "loss": 0.0225, "step": 63000 }, { "grad_norm": 0.4402652978897095, "learning_rate": 3.297134892413211e-05, "loss": 0.0256, "step": 63010 }, { "grad_norm": 0.40341463685035706, "learning_rate": 3.295580363354031e-05, "loss": 0.0254, "step": 63020 }, { "grad_norm": 0.5696441531181335, "learning_rate": 3.2940260206876194e-05, "loss": 0.029, "step": 63030 }, { "grad_norm": 0.36614951491355896, "learning_rate": 3.292471864583958e-05, "loss": 0.0296, "step": 63040 }, { "grad_norm": 0.39498335123062134, "learning_rate": 3.290917895213004e-05, "loss": 0.034, "step": 63050 }, { "grad_norm": 0.4300706684589386, "learning_rate": 3.289364112744701e-05, "loss": 0.0377, "step": 63060 }, { "grad_norm": 0.3464142978191376, "learning_rate": 3.287810517348967e-05, "loss": 0.0232, "step": 63070 }, { "grad_norm": 0.20825868844985962, "learning_rate": 3.2862571091957014e-05, "loss": 0.0184, "step": 63080 }, { "grad_norm": 0.6022765636444092, "learning_rate": 3.284703888454781e-05, "loss": 0.0274, "step": 63090 }, { "grad_norm": 0.42671462893486023, "learning_rate": 3.2831508552960646e-05, "loss": 0.0259, "step": 63100 }, { "grad_norm": 0.40072840452194214, "learning_rate": 3.281598009889387e-05, "loss": 0.0339, "step": 63110 }, { "grad_norm": 0.34709811210632324, "learning_rate": 3.28004535240457e-05, "loss": 0.0248, "step": 63120 }, { "grad_norm": 0.30661582946777344, "learning_rate": 3.278492883011407e-05, "loss": 0.0338, "step": 63130 }, { "grad_norm": 0.3192675709724426, "learning_rate": 3.276940601879675e-05, "loss": 0.0284, "step": 63140 }, { "grad_norm": 0.34413787722587585, "learning_rate": 3.275388509179127e-05, "loss": 0.0234, "step": 63150 }, { "grad_norm": 0.40446552634239197, "learning_rate": 3.2738366050794985e-05, "loss": 0.0229, "step": 63160 }, { "grad_norm": 0.3614238202571869, "learning_rate": 3.2722848897505034e-05, "loss": 0.0216, "step": 63170 }, { "grad_norm": 0.4361209273338318, "learning_rate": 3.2707333633618356e-05, "loss": 0.028, "step": 63180 }, { "grad_norm": 0.29430335760116577, "learning_rate": 3.269182026083167e-05, "loss": 0.0181, "step": 63190 }, { "grad_norm": 0.43701601028442383, "learning_rate": 3.2676308780841516e-05, "loss": 0.026, "step": 63200 }, { "grad_norm": 0.2820853292942047, "learning_rate": 3.266079919534416e-05, "loss": 0.0264, "step": 63210 }, { "grad_norm": 0.4776044487953186, "learning_rate": 3.264529150603575e-05, "loss": 0.0384, "step": 63220 }, { "grad_norm": 0.6838534474372864, "learning_rate": 3.262978571461217e-05, "loss": 0.0305, "step": 63230 }, { "grad_norm": 0.5201625227928162, "learning_rate": 3.26142818227691e-05, "loss": 0.0281, "step": 63240 }, { "grad_norm": 0.2290903478860855, "learning_rate": 3.259877983220203e-05, "loss": 0.0212, "step": 63250 }, { "grad_norm": 0.9154102206230164, "learning_rate": 3.2583279744606246e-05, "loss": 0.0252, "step": 63260 }, { "grad_norm": 0.5204060077667236, "learning_rate": 3.25677815616768e-05, "loss": 0.0266, "step": 63270 }, { "grad_norm": 0.5317113995552063, "learning_rate": 3.255228528510856e-05, "loss": 0.0276, "step": 63280 }, { "grad_norm": 0.3400419056415558, "learning_rate": 3.253679091659617e-05, "loss": 0.0227, "step": 63290 }, { "grad_norm": 0.2772385776042938, "learning_rate": 3.252129845783404e-05, "loss": 0.0244, "step": 63300 }, { "grad_norm": 0.47263333201408386, "learning_rate": 3.2505807910516456e-05, "loss": 0.0248, "step": 63310 }, { "grad_norm": 0.45838719606399536, "learning_rate": 3.249031927633742e-05, "loss": 0.0209, "step": 63320 }, { "grad_norm": 0.5171664953231812, "learning_rate": 3.2474832556990737e-05, "loss": 0.0265, "step": 63330 }, { "grad_norm": 0.44562625885009766, "learning_rate": 3.245934775417002e-05, "loss": 0.0269, "step": 63340 }, { "grad_norm": 0.38976648449897766, "learning_rate": 3.244386486956866e-05, "loss": 0.0261, "step": 63350 }, { "grad_norm": 0.2767089903354645, "learning_rate": 3.2428383904879846e-05, "loss": 0.024, "step": 63360 }, { "grad_norm": 0.34035956859588623, "learning_rate": 3.241290486179652e-05, "loss": 0.0184, "step": 63370 }, { "grad_norm": 0.4072038531303406, "learning_rate": 3.239742774201151e-05, "loss": 0.0244, "step": 63380 }, { "grad_norm": 0.2871328294277191, "learning_rate": 3.238195254721733e-05, "loss": 0.0278, "step": 63390 }, { "grad_norm": 0.29980164766311646, "learning_rate": 3.236647927910633e-05, "loss": 0.02, "step": 63400 }, { "grad_norm": 0.5215563774108887, "learning_rate": 3.2351007939370646e-05, "loss": 0.0334, "step": 63410 }, { "grad_norm": 0.29645484685897827, "learning_rate": 3.2335538529702184e-05, "loss": 0.018, "step": 63420 }, { "grad_norm": 0.33213284611701965, "learning_rate": 3.232007105179268e-05, "loss": 0.0293, "step": 63430 }, { "grad_norm": 0.5462307929992676, "learning_rate": 3.230460550733363e-05, "loss": 0.0414, "step": 63440 }, { "grad_norm": 0.4188348054885864, "learning_rate": 3.2289141898016315e-05, "loss": 0.0232, "step": 63450 }, { "grad_norm": 0.6050399541854858, "learning_rate": 3.227368022553181e-05, "loss": 0.0238, "step": 63460 }, { "grad_norm": 0.4533573389053345, "learning_rate": 3.225822049157099e-05, "loss": 0.0373, "step": 63470 }, { "grad_norm": 0.4511139690876007, "learning_rate": 3.2242762697824494e-05, "loss": 0.0237, "step": 63480 }, { "grad_norm": 0.4977497458457947, "learning_rate": 3.222730684598279e-05, "loss": 0.0251, "step": 63490 }, { "grad_norm": 0.49430051445961, "learning_rate": 3.221185293773605e-05, "loss": 0.0269, "step": 63500 }, { "grad_norm": 0.2167532742023468, "learning_rate": 3.2196400974774354e-05, "loss": 0.0239, "step": 63510 }, { "grad_norm": 0.40359944105148315, "learning_rate": 3.2180950958787485e-05, "loss": 0.024, "step": 63520 }, { "grad_norm": 0.5883824229240417, "learning_rate": 3.2165502891465024e-05, "loss": 0.0408, "step": 63530 }, { "grad_norm": 0.4324372410774231, "learning_rate": 3.215005677449636e-05, "loss": 0.0234, "step": 63540 }, { "grad_norm": 0.485112726688385, "learning_rate": 3.2134612609570644e-05, "loss": 0.0367, "step": 63550 }, { "grad_norm": 0.43235141038894653, "learning_rate": 3.211917039837682e-05, "loss": 0.0252, "step": 63560 }, { "grad_norm": 0.3139662742614746, "learning_rate": 3.210373014260365e-05, "loss": 0.0202, "step": 63570 }, { "grad_norm": 0.3586916923522949, "learning_rate": 3.208829184393965e-05, "loss": 0.0468, "step": 63580 }, { "grad_norm": 0.28643712401390076, "learning_rate": 3.207285550407312e-05, "loss": 0.0295, "step": 63590 }, { "grad_norm": 0.37276118993759155, "learning_rate": 3.205742112469217e-05, "loss": 0.0256, "step": 63600 }, { "grad_norm": 0.29594817757606506, "learning_rate": 3.204198870748465e-05, "loss": 0.0546, "step": 63610 }, { "grad_norm": 0.21625708043575287, "learning_rate": 3.202655825413824e-05, "loss": 0.0227, "step": 63620 }, { "grad_norm": 0.5064502358436584, "learning_rate": 3.2011129766340416e-05, "loss": 0.0254, "step": 63630 }, { "grad_norm": 0.2793135344982147, "learning_rate": 3.1995703245778384e-05, "loss": 0.0256, "step": 63640 }, { "grad_norm": 0.4342672824859619, "learning_rate": 3.1980278694139186e-05, "loss": 0.0254, "step": 63650 }, { "grad_norm": 0.3681219518184662, "learning_rate": 3.196485611310962e-05, "loss": 0.038, "step": 63660 }, { "grad_norm": 0.5073325037956238, "learning_rate": 3.1949435504376257e-05, "loss": 0.0434, "step": 63670 }, { "grad_norm": 0.36944693326950073, "learning_rate": 3.1934016869625505e-05, "loss": 0.0251, "step": 63680 }, { "grad_norm": 0.530417263507843, "learning_rate": 3.191860021054349e-05, "loss": 0.0215, "step": 63690 }, { "grad_norm": 0.24586312472820282, "learning_rate": 3.190318552881618e-05, "loss": 0.0371, "step": 63700 }, { "grad_norm": 0.49514254927635193, "learning_rate": 3.1887772826129285e-05, "loss": 0.0169, "step": 63710 }, { "grad_norm": 0.26681968569755554, "learning_rate": 3.1872362104168335e-05, "loss": 0.0214, "step": 63720 }, { "grad_norm": 0.5103306770324707, "learning_rate": 3.185695336461861e-05, "loss": 0.0267, "step": 63730 }, { "grad_norm": 0.3266035318374634, "learning_rate": 3.184154660916519e-05, "loss": 0.0242, "step": 63740 }, { "grad_norm": 0.2678591012954712, "learning_rate": 3.1826141839492906e-05, "loss": 0.0253, "step": 63750 }, { "grad_norm": 0.3537883162498474, "learning_rate": 3.181073905728644e-05, "loss": 0.024, "step": 63760 }, { "grad_norm": 0.2556129992008209, "learning_rate": 3.1795338264230214e-05, "loss": 0.0274, "step": 63770 }, { "grad_norm": 0.6419453620910645, "learning_rate": 3.1779939462008426e-05, "loss": 0.0454, "step": 63780 }, { "grad_norm": 0.22078055143356323, "learning_rate": 3.1764542652305066e-05, "loss": 0.0223, "step": 63790 }, { "grad_norm": 0.43421128392219543, "learning_rate": 3.17491478368039e-05, "loss": 0.0332, "step": 63800 }, { "grad_norm": 0.4725186824798584, "learning_rate": 3.173375501718847e-05, "loss": 0.0298, "step": 63810 }, { "grad_norm": 0.47273966670036316, "learning_rate": 3.171836419514215e-05, "loss": 0.0419, "step": 63820 }, { "grad_norm": 0.4053094685077667, "learning_rate": 3.170297537234803e-05, "loss": 0.0245, "step": 63830 }, { "grad_norm": 0.42897334694862366, "learning_rate": 3.168758855048901e-05, "loss": 0.0436, "step": 63840 }, { "grad_norm": 2.580519676208496, "learning_rate": 3.1672203731247785e-05, "loss": 0.0257, "step": 63850 }, { "grad_norm": 0.39621809124946594, "learning_rate": 3.16568209163068e-05, "loss": 0.0213, "step": 63860 }, { "grad_norm": 0.5187084078788757, "learning_rate": 3.1641440107348295e-05, "loss": 0.0198, "step": 63870 }, { "grad_norm": 0.31264281272888184, "learning_rate": 3.162606130605429e-05, "loss": 0.0238, "step": 63880 }, { "grad_norm": 0.541378378868103, "learning_rate": 3.16106845141066e-05, "loss": 0.0263, "step": 63890 }, { "grad_norm": 0.4809972047805786, "learning_rate": 3.159530973318682e-05, "loss": 0.0316, "step": 63900 }, { "grad_norm": 0.39335450530052185, "learning_rate": 3.1579936964976286e-05, "loss": 0.021, "step": 63910 }, { "grad_norm": 0.2899989187717438, "learning_rate": 3.156456621115615e-05, "loss": 0.0379, "step": 63920 }, { "grad_norm": 0.30116942524909973, "learning_rate": 3.154919747340734e-05, "loss": 0.0229, "step": 63930 }, { "grad_norm": 0.3352719247341156, "learning_rate": 3.153383075341054e-05, "loss": 0.0309, "step": 63940 }, { "grad_norm": 0.4610649049282074, "learning_rate": 3.1518466052846264e-05, "loss": 0.0253, "step": 63950 }, { "grad_norm": 0.47977712750434875, "learning_rate": 3.150310337339474e-05, "loss": 0.0286, "step": 63960 }, { "grad_norm": 0.6913794875144958, "learning_rate": 3.148774271673603e-05, "loss": 0.0296, "step": 63970 }, { "grad_norm": 0.3214974105358124, "learning_rate": 3.147238408454995e-05, "loss": 0.0369, "step": 63980 }, { "grad_norm": 0.7176304459571838, "learning_rate": 3.1457027478516086e-05, "loss": 0.024, "step": 63990 }, { "grad_norm": 0.4903135895729065, "learning_rate": 3.144167290031382e-05, "loss": 0.0313, "step": 64000 }, { "grad_norm": 0.3827296793460846, "learning_rate": 3.142632035162227e-05, "loss": 0.0303, "step": 64010 }, { "grad_norm": 0.3839775323867798, "learning_rate": 3.141096983412044e-05, "loss": 0.0245, "step": 64020 }, { "grad_norm": 0.40626439452171326, "learning_rate": 3.139562134948699e-05, "loss": 0.0312, "step": 64030 }, { "grad_norm": 0.5061067938804626, "learning_rate": 3.1380274899400424e-05, "loss": 0.0256, "step": 64040 }, { "grad_norm": 0.32799357175827026, "learning_rate": 3.1364930485539e-05, "loss": 0.0192, "step": 64050 }, { "grad_norm": 0.8128136396408081, "learning_rate": 3.134958810958076e-05, "loss": 0.0322, "step": 64060 }, { "grad_norm": 0.3718286454677582, "learning_rate": 3.13342477732035e-05, "loss": 0.0224, "step": 64070 }, { "grad_norm": 0.5962384343147278, "learning_rate": 3.131890947808487e-05, "loss": 0.0349, "step": 64080 }, { "grad_norm": 0.3171875476837158, "learning_rate": 3.13035732259022e-05, "loss": 0.0258, "step": 64090 }, { "grad_norm": 0.5455929636955261, "learning_rate": 3.1288239018332656e-05, "loss": 0.0238, "step": 64100 }, { "grad_norm": 0.3235395848751068, "learning_rate": 3.1272906857053164e-05, "loss": 0.0294, "step": 64110 }, { "grad_norm": 0.485723078250885, "learning_rate": 3.125757674374042e-05, "loss": 0.0287, "step": 64120 }, { "grad_norm": 0.45902860164642334, "learning_rate": 3.124224868007088e-05, "loss": 0.0302, "step": 64130 }, { "grad_norm": 0.3439926207065582, "learning_rate": 3.1226922667720854e-05, "loss": 0.0271, "step": 64140 }, { "grad_norm": 0.34299686551094055, "learning_rate": 3.121159870836633e-05, "loss": 0.017, "step": 64150 }, { "grad_norm": 0.3994394838809967, "learning_rate": 3.119627680368312e-05, "loss": 0.0187, "step": 64160 }, { "grad_norm": 0.49471110105514526, "learning_rate": 3.1180956955346816e-05, "loss": 0.0348, "step": 64170 }, { "grad_norm": 0.2738630473613739, "learning_rate": 3.116563916503276e-05, "loss": 0.0272, "step": 64180 }, { "grad_norm": 0.34582972526550293, "learning_rate": 3.1150323434416095e-05, "loss": 0.0309, "step": 64190 }, { "grad_norm": 0.5056164860725403, "learning_rate": 3.1135009765171696e-05, "loss": 0.0226, "step": 64200 }, { "grad_norm": 0.337516188621521, "learning_rate": 3.111969815897429e-05, "loss": 0.0341, "step": 64210 }, { "grad_norm": 0.45935845375061035, "learning_rate": 3.1104388617498295e-05, "loss": 0.0256, "step": 64220 }, { "grad_norm": 0.43900251388549805, "learning_rate": 3.1089081142417966e-05, "loss": 0.0206, "step": 64230 }, { "grad_norm": 0.3964497148990631, "learning_rate": 3.1073775735407285e-05, "loss": 0.0233, "step": 64240 }, { "grad_norm": 0.3506081700325012, "learning_rate": 3.1058472398140035e-05, "loss": 0.0227, "step": 64250 }, { "grad_norm": 0.7073218822479248, "learning_rate": 3.1043171132289736e-05, "loss": 0.0317, "step": 64260 }, { "grad_norm": 0.7724215984344482, "learning_rate": 3.1027871939529774e-05, "loss": 0.0316, "step": 64270 }, { "grad_norm": 0.3084127902984619, "learning_rate": 3.1012574821533214e-05, "loss": 0.0168, "step": 64280 }, { "grad_norm": 0.16543367505073547, "learning_rate": 3.099727977997291e-05, "loss": 0.018, "step": 64290 }, { "grad_norm": 0.5476890206336975, "learning_rate": 3.098198681652154e-05, "loss": 0.0252, "step": 64300 }, { "grad_norm": 0.8037078976631165, "learning_rate": 3.0966695932851475e-05, "loss": 0.0238, "step": 64310 }, { "grad_norm": 0.41731494665145874, "learning_rate": 3.0951407130634926e-05, "loss": 0.0274, "step": 64320 }, { "grad_norm": 1.7852917909622192, "learning_rate": 3.0936120411543864e-05, "loss": 0.0322, "step": 64330 }, { "grad_norm": 0.3502584397792816, "learning_rate": 3.0920835777250015e-05, "loss": 0.0289, "step": 64340 }, { "grad_norm": 0.4569863975048065, "learning_rate": 3.090555322942488e-05, "loss": 0.023, "step": 64350 }, { "grad_norm": 0.764544665813446, "learning_rate": 3.089027276973973e-05, "loss": 0.0249, "step": 64360 }, { "grad_norm": 0.38599565625190735, "learning_rate": 3.0874994399865615e-05, "loss": 0.0191, "step": 64370 }, { "grad_norm": 0.31230083107948303, "learning_rate": 3.085971812147337e-05, "loss": 0.0185, "step": 64380 }, { "grad_norm": 0.36743319034576416, "learning_rate": 3.0844443936233557e-05, "loss": 0.0195, "step": 64390 }, { "grad_norm": 0.3862672448158264, "learning_rate": 3.082917184581657e-05, "loss": 0.0273, "step": 64400 }, { "grad_norm": 0.5685907602310181, "learning_rate": 3.0813901851892513e-05, "loss": 0.0278, "step": 64410 }, { "grad_norm": 0.4048730731010437, "learning_rate": 3.0798633956131316e-05, "loss": 0.0352, "step": 64420 }, { "grad_norm": 0.3497196137905121, "learning_rate": 3.078336816020264e-05, "loss": 0.0198, "step": 64430 }, { "grad_norm": 0.4255886971950531, "learning_rate": 3.0768104465775935e-05, "loss": 0.0299, "step": 64440 }, { "grad_norm": 0.38129058480262756, "learning_rate": 3.0752842874520396e-05, "loss": 0.0187, "step": 64450 }, { "grad_norm": 0.3658240735530853, "learning_rate": 3.073758338810503e-05, "loss": 0.0184, "step": 64460 }, { "grad_norm": 0.4730420410633087, "learning_rate": 3.072232600819859e-05, "loss": 0.0303, "step": 64470 }, { "grad_norm": 0.5485438108444214, "learning_rate": 3.070707073646959e-05, "loss": 0.0226, "step": 64480 }, { "grad_norm": 0.2471257746219635, "learning_rate": 3.069181757458633e-05, "loss": 0.0281, "step": 64490 }, { "grad_norm": 0.780146062374115, "learning_rate": 3.067656652421687e-05, "loss": 0.0316, "step": 64500 }, { "grad_norm": 0.26690441370010376, "learning_rate": 3.0661317587029025e-05, "loss": 0.028, "step": 64510 }, { "grad_norm": 0.47469690442085266, "learning_rate": 3.064607076469043e-05, "loss": 0.0262, "step": 64520 }, { "grad_norm": 0.4620678424835205, "learning_rate": 3.063082605886845e-05, "loss": 0.0298, "step": 64530 }, { "grad_norm": 0.2850009500980377, "learning_rate": 3.061558347123019e-05, "loss": 0.0194, "step": 64540 }, { "grad_norm": 0.25346118211746216, "learning_rate": 3.060034300344259e-05, "loss": 0.0194, "step": 64550 }, { "grad_norm": 0.45698195695877075, "learning_rate": 3.058510465717231e-05, "loss": 0.03, "step": 64560 }, { "grad_norm": 0.24862177670001984, "learning_rate": 3.0569868434085804e-05, "loss": 0.0214, "step": 64570 }, { "grad_norm": 0.33414483070373535, "learning_rate": 3.055463433584924e-05, "loss": 0.0269, "step": 64580 }, { "grad_norm": 0.36962127685546875, "learning_rate": 3.0539402364128656e-05, "loss": 0.0208, "step": 64590 }, { "grad_norm": 0.4355773627758026, "learning_rate": 3.052417252058977e-05, "loss": 0.019, "step": 64600 }, { "grad_norm": 0.3226814568042755, "learning_rate": 3.0508944806898092e-05, "loss": 0.0321, "step": 64610 }, { "grad_norm": 0.4916728138923645, "learning_rate": 3.04937192247189e-05, "loss": 0.0285, "step": 64620 }, { "grad_norm": 0.33166101574897766, "learning_rate": 3.047849577571725e-05, "loss": 0.0221, "step": 64630 }, { "grad_norm": 0.47438013553619385, "learning_rate": 3.046327446155795e-05, "loss": 0.0217, "step": 64640 }, { "grad_norm": 0.21724414825439453, "learning_rate": 3.0448055283905587e-05, "loss": 0.0308, "step": 64650 }, { "grad_norm": 0.6217232942581177, "learning_rate": 3.0432838244424494e-05, "loss": 0.0207, "step": 64660 }, { "grad_norm": 0.48127850890159607, "learning_rate": 3.04176233447788e-05, "loss": 0.029, "step": 64670 }, { "grad_norm": 0.7339156270027161, "learning_rate": 3.040241058663238e-05, "loss": 0.0218, "step": 64680 }, { "grad_norm": 0.5074663162231445, "learning_rate": 3.038719997164887e-05, "loss": 0.0197, "step": 64690 }, { "grad_norm": 0.4557279646396637, "learning_rate": 3.0371991501491683e-05, "loss": 0.0209, "step": 64700 }, { "grad_norm": 0.3904956877231598, "learning_rate": 3.0356785177823975e-05, "loss": 0.0243, "step": 64710 }, { "grad_norm": 0.4469676613807678, "learning_rate": 3.0341581002308726e-05, "loss": 0.0225, "step": 64720 }, { "grad_norm": 0.29899728298187256, "learning_rate": 3.0326378976608627e-05, "loss": 0.0244, "step": 64730 }, { "grad_norm": 0.4030550718307495, "learning_rate": 3.0311179102386146e-05, "loss": 0.0214, "step": 64740 }, { "grad_norm": 0.2779548466205597, "learning_rate": 3.0295981381303513e-05, "loss": 0.0233, "step": 64750 }, { "grad_norm": 0.3003369867801666, "learning_rate": 3.0280785815022727e-05, "loss": 0.0222, "step": 64760 }, { "grad_norm": 0.4922226071357727, "learning_rate": 3.0265592405205544e-05, "loss": 0.0165, "step": 64770 }, { "grad_norm": 0.5133985280990601, "learning_rate": 3.0250401153513517e-05, "loss": 0.0166, "step": 64780 }, { "grad_norm": 0.39762064814567566, "learning_rate": 3.0235212061607936e-05, "loss": 0.025, "step": 64790 }, { "grad_norm": 0.9468638896942139, "learning_rate": 3.0220025131149842e-05, "loss": 0.0365, "step": 64800 }, { "grad_norm": 0.7828836441040039, "learning_rate": 3.020484036380007e-05, "loss": 0.0249, "step": 64810 }, { "grad_norm": 0.3601033389568329, "learning_rate": 3.0189657761219188e-05, "loss": 0.0285, "step": 64820 }, { "grad_norm": 0.5524725317955017, "learning_rate": 3.017447732506753e-05, "loss": 0.017, "step": 64830 }, { "grad_norm": 0.6284120678901672, "learning_rate": 3.015929905700524e-05, "loss": 0.0332, "step": 64840 }, { "grad_norm": 0.36702993512153625, "learning_rate": 3.014412295869218e-05, "loss": 0.0334, "step": 64850 }, { "grad_norm": 0.8709197044372559, "learning_rate": 3.0128949031787977e-05, "loss": 0.0243, "step": 64860 }, { "grad_norm": 0.5426575541496277, "learning_rate": 3.0113777277952022e-05, "loss": 0.0315, "step": 64870 }, { "grad_norm": 0.27310478687286377, "learning_rate": 3.0098607698843496e-05, "loss": 0.0224, "step": 64880 }, { "grad_norm": 0.430075466632843, "learning_rate": 3.0083440296121312e-05, "loss": 0.0283, "step": 64890 }, { "grad_norm": 0.46477141976356506, "learning_rate": 3.0068275071444142e-05, "loss": 0.0262, "step": 64900 }, { "grad_norm": 0.3623978793621063, "learning_rate": 3.0053112026470442e-05, "loss": 0.0265, "step": 64910 }, { "grad_norm": 0.5795591473579407, "learning_rate": 3.0037951162858434e-05, "loss": 0.0229, "step": 64920 }, { "grad_norm": 0.3183465301990509, "learning_rate": 3.002279248226607e-05, "loss": 0.0291, "step": 64930 }, { "grad_norm": 0.2835332453250885, "learning_rate": 3.000763598635109e-05, "loss": 0.0319, "step": 64940 }, { "grad_norm": 0.43388301134109497, "learning_rate": 2.999248167677098e-05, "loss": 0.0288, "step": 64950 }, { "grad_norm": 0.38243499398231506, "learning_rate": 2.9977329555182976e-05, "loss": 0.0258, "step": 64960 }, { "grad_norm": 0.39632275700569153, "learning_rate": 2.9962179623244124e-05, "loss": 0.02, "step": 64970 }, { "grad_norm": 0.5510632395744324, "learning_rate": 2.9947031882611175e-05, "loss": 0.0241, "step": 64980 }, { "grad_norm": 0.37136125564575195, "learning_rate": 2.9931886334940685e-05, "loss": 0.0178, "step": 64990 }, { "grad_norm": 0.49916917085647583, "learning_rate": 2.9916742981888923e-05, "loss": 0.0268, "step": 65000 }, { "grad_norm": 0.35916849970817566, "learning_rate": 2.990160182511196e-05, "loss": 0.018, "step": 65010 }, { "grad_norm": 0.45161399245262146, "learning_rate": 2.9886462866265574e-05, "loss": 0.0243, "step": 65020 }, { "grad_norm": 0.26938536763191223, "learning_rate": 2.9871326107005392e-05, "loss": 0.0202, "step": 65030 }, { "grad_norm": 0.4129895865917206, "learning_rate": 2.9856191548986735e-05, "loss": 0.0192, "step": 65040 }, { "grad_norm": 0.37963056564331055, "learning_rate": 2.9841059193864672e-05, "loss": 0.0181, "step": 65050 }, { "grad_norm": 0.18844106793403625, "learning_rate": 2.982592904329407e-05, "loss": 0.0166, "step": 65060 }, { "grad_norm": 0.37893691658973694, "learning_rate": 2.981080109892954e-05, "loss": 0.0252, "step": 65070 }, { "grad_norm": 1.1058945655822754, "learning_rate": 2.9795675362425434e-05, "loss": 0.0257, "step": 65080 }, { "grad_norm": 0.6537178158760071, "learning_rate": 2.97805518354359e-05, "loss": 0.0371, "step": 65090 }, { "grad_norm": 0.4904627203941345, "learning_rate": 2.9765430519614825e-05, "loss": 0.0271, "step": 65100 }, { "grad_norm": 0.48568394780158997, "learning_rate": 2.975031141661584e-05, "loss": 0.0257, "step": 65110 }, { "grad_norm": 0.28436997532844543, "learning_rate": 2.9735194528092347e-05, "loss": 0.0241, "step": 65120 }, { "grad_norm": 0.3147582709789276, "learning_rate": 2.9720079855697525e-05, "loss": 0.0254, "step": 65130 }, { "grad_norm": 0.39449596405029297, "learning_rate": 2.970496740108427e-05, "loss": 0.0308, "step": 65140 }, { "grad_norm": 0.6335060000419617, "learning_rate": 2.968985716590527e-05, "loss": 0.0221, "step": 65150 }, { "grad_norm": 0.4303458631038666, "learning_rate": 2.9674749151812953e-05, "loss": 0.027, "step": 65160 }, { "grad_norm": 0.5219939947128296, "learning_rate": 2.965964336045951e-05, "loss": 0.0232, "step": 65170 }, { "grad_norm": 0.8100031614303589, "learning_rate": 2.9644539793496894e-05, "loss": 0.0257, "step": 65180 }, { "grad_norm": 0.3599408268928528, "learning_rate": 2.9629438452576807e-05, "loss": 0.0185, "step": 65190 }, { "grad_norm": 0.45007091760635376, "learning_rate": 2.9614339339350698e-05, "loss": 0.0229, "step": 65200 }, { "grad_norm": 0.2742367088794708, "learning_rate": 2.959924245546979e-05, "loss": 0.0224, "step": 65210 }, { "grad_norm": 0.4803943336009979, "learning_rate": 2.9584147802585038e-05, "loss": 0.0247, "step": 65220 }, { "grad_norm": 0.6865038871765137, "learning_rate": 2.956905538234721e-05, "loss": 0.0333, "step": 65230 }, { "grad_norm": 0.35200056433677673, "learning_rate": 2.9553965196406764e-05, "loss": 0.0253, "step": 65240 }, { "grad_norm": 0.4323226809501648, "learning_rate": 2.9538877246413943e-05, "loss": 0.0296, "step": 65250 }, { "grad_norm": 0.4411715865135193, "learning_rate": 2.952379153401874e-05, "loss": 0.017, "step": 65260 }, { "grad_norm": 0.28192558884620667, "learning_rate": 2.9508708060870914e-05, "loss": 0.0256, "step": 65270 }, { "grad_norm": 0.2867196202278137, "learning_rate": 2.9493626828619948e-05, "loss": 0.0213, "step": 65280 }, { "grad_norm": 0.8140136003494263, "learning_rate": 2.9478547838915137e-05, "loss": 0.0262, "step": 65290 }, { "grad_norm": 0.30189573764801025, "learning_rate": 2.946347109340548e-05, "loss": 0.0216, "step": 65300 }, { "grad_norm": 0.437460720539093, "learning_rate": 2.944839659373976e-05, "loss": 0.0227, "step": 65310 }, { "grad_norm": 0.419077068567276, "learning_rate": 2.943332434156647e-05, "loss": 0.0244, "step": 65320 }, { "grad_norm": 0.15681275725364685, "learning_rate": 2.9418254338533912e-05, "loss": 0.0266, "step": 65330 }, { "grad_norm": 1.4918791055679321, "learning_rate": 2.940318658629011e-05, "loss": 0.0283, "step": 65340 }, { "grad_norm": 0.5292578935623169, "learning_rate": 2.938812108648287e-05, "loss": 0.0225, "step": 65350 }, { "grad_norm": 0.9622846841812134, "learning_rate": 2.937305784075972e-05, "loss": 0.0243, "step": 65360 }, { "grad_norm": 0.4253517985343933, "learning_rate": 2.9357996850767937e-05, "loss": 0.0294, "step": 65370 }, { "grad_norm": 0.657929539680481, "learning_rate": 2.9342938118154595e-05, "loss": 0.0209, "step": 65380 }, { "grad_norm": 0.7413333654403687, "learning_rate": 2.9327881644566485e-05, "loss": 0.0249, "step": 65390 }, { "grad_norm": 0.26871728897094727, "learning_rate": 2.9312827431650157e-05, "loss": 0.0233, "step": 65400 }, { "grad_norm": 0.6660961508750916, "learning_rate": 2.92977754810519e-05, "loss": 0.0239, "step": 65410 }, { "grad_norm": 0.7479501962661743, "learning_rate": 2.928272579441781e-05, "loss": 0.0369, "step": 65420 }, { "grad_norm": 0.6615031957626343, "learning_rate": 2.9267678373393676e-05, "loss": 0.0382, "step": 65430 }, { "grad_norm": 0.3082142472267151, "learning_rate": 2.925263321962507e-05, "loss": 0.02, "step": 65440 }, { "grad_norm": 0.23938138782978058, "learning_rate": 2.9237590334757292e-05, "loss": 0.024, "step": 65450 }, { "grad_norm": 0.4278719425201416, "learning_rate": 2.922254972043543e-05, "loss": 0.0187, "step": 65460 }, { "grad_norm": 0.35779136419296265, "learning_rate": 2.9207511378304265e-05, "loss": 0.0202, "step": 65470 }, { "grad_norm": 0.4159371852874756, "learning_rate": 2.9192475310008428e-05, "loss": 0.027, "step": 65480 }, { "grad_norm": 0.2815870940685272, "learning_rate": 2.917744151719219e-05, "loss": 0.0214, "step": 65490 }, { "grad_norm": 0.5095306634902954, "learning_rate": 2.916241000149964e-05, "loss": 0.0277, "step": 65500 }, { "grad_norm": 0.6528527736663818, "learning_rate": 2.9147380764574622e-05, "loss": 0.0239, "step": 65510 }, { "grad_norm": 0.20992203056812286, "learning_rate": 2.9132353808060674e-05, "loss": 0.0194, "step": 65520 }, { "grad_norm": 0.26008015871047974, "learning_rate": 2.9117329133601134e-05, "loss": 0.0178, "step": 65530 }, { "grad_norm": 0.2486124038696289, "learning_rate": 2.9102306742839093e-05, "loss": 0.0256, "step": 65540 }, { "grad_norm": 0.4105188548564911, "learning_rate": 2.9087286637417375e-05, "loss": 0.0353, "step": 65550 }, { "grad_norm": 0.8091117143630981, "learning_rate": 2.9072268818978537e-05, "loss": 0.0274, "step": 65560 }, { "grad_norm": 0.3582206964492798, "learning_rate": 2.905725328916493e-05, "loss": 0.0212, "step": 65570 }, { "grad_norm": 0.24634146690368652, "learning_rate": 2.9042240049618608e-05, "loss": 0.0182, "step": 65580 }, { "grad_norm": 0.3857291638851166, "learning_rate": 2.9027229101981413e-05, "loss": 0.0314, "step": 65590 }, { "grad_norm": 0.31489136815071106, "learning_rate": 2.901222044789488e-05, "loss": 0.0242, "step": 65600 }, { "grad_norm": 0.3007563054561615, "learning_rate": 2.89972140890004e-05, "loss": 0.0193, "step": 65610 }, { "grad_norm": 0.3923858106136322, "learning_rate": 2.8982210026938988e-05, "loss": 0.0227, "step": 65620 }, { "grad_norm": 0.39339691400527954, "learning_rate": 2.896720826335151e-05, "loss": 0.026, "step": 65630 }, { "grad_norm": 0.400256872177124, "learning_rate": 2.8952208799878493e-05, "loss": 0.0203, "step": 65640 }, { "grad_norm": 0.2963939607143402, "learning_rate": 2.89372116381603e-05, "loss": 0.0191, "step": 65650 }, { "grad_norm": 0.9973160624504089, "learning_rate": 2.8922216779836932e-05, "loss": 0.0455, "step": 65660 }, { "grad_norm": 0.4086117744445801, "learning_rate": 2.8907224226548285e-05, "loss": 0.0197, "step": 65670 }, { "grad_norm": 0.3709625005722046, "learning_rate": 2.8892233979933868e-05, "loss": 0.0281, "step": 65680 }, { "grad_norm": 0.4179450571537018, "learning_rate": 2.8877246041633026e-05, "loss": 0.0354, "step": 65690 }, { "grad_norm": 0.35969263315200806, "learning_rate": 2.886226041328478e-05, "loss": 0.0248, "step": 65700 }, { "grad_norm": 0.27418461441993713, "learning_rate": 2.8847277096527957e-05, "loss": 0.0227, "step": 65710 }, { "grad_norm": 0.37528830766677856, "learning_rate": 2.8832296093001105e-05, "loss": 0.0352, "step": 65720 }, { "grad_norm": 0.5146607160568237, "learning_rate": 2.8817317404342547e-05, "loss": 0.0267, "step": 65730 }, { "grad_norm": 0.3276301324367523, "learning_rate": 2.880234103219029e-05, "loss": 0.0239, "step": 65740 }, { "grad_norm": 0.6191139817237854, "learning_rate": 2.878736697818215e-05, "loss": 0.0316, "step": 65750 }, { "grad_norm": 0.7479645609855652, "learning_rate": 2.877239524395568e-05, "loss": 0.0219, "step": 65760 }, { "grad_norm": 0.3509233593940735, "learning_rate": 2.8757425831148134e-05, "loss": 0.0255, "step": 65770 }, { "grad_norm": 0.34914833307266235, "learning_rate": 2.874245874139658e-05, "loss": 0.022, "step": 65780 }, { "grad_norm": 0.5716358423233032, "learning_rate": 2.8727493976337737e-05, "loss": 0.0347, "step": 65790 }, { "grad_norm": 0.1572932004928589, "learning_rate": 2.8712531537608218e-05, "loss": 0.0261, "step": 65800 }, { "grad_norm": 0.3338248133659363, "learning_rate": 2.8697571426844217e-05, "loss": 0.0269, "step": 65810 }, { "grad_norm": 0.41040244698524475, "learning_rate": 2.86826136456818e-05, "loss": 0.0263, "step": 65820 }, { "grad_norm": 0.5943193435668945, "learning_rate": 2.8667658195756685e-05, "loss": 0.0244, "step": 65830 }, { "grad_norm": 0.4884960949420929, "learning_rate": 2.8652705078704412e-05, "loss": 0.0308, "step": 65840 }, { "grad_norm": 0.4060570001602173, "learning_rate": 2.863775429616018e-05, "loss": 0.0276, "step": 65850 }, { "grad_norm": 0.31022903323173523, "learning_rate": 2.862280584975906e-05, "loss": 0.0196, "step": 65860 }, { "grad_norm": 0.6277173757553101, "learning_rate": 2.8607859741135735e-05, "loss": 0.0249, "step": 65870 }, { "grad_norm": 0.2604329288005829, "learning_rate": 2.8592915971924723e-05, "loss": 0.0219, "step": 65880 }, { "grad_norm": 0.37351223826408386, "learning_rate": 2.8577974543760212e-05, "loss": 0.0251, "step": 65890 }, { "grad_norm": 0.5722813606262207, "learning_rate": 2.8563035458276222e-05, "loss": 0.0252, "step": 65900 }, { "grad_norm": 0.46634751558303833, "learning_rate": 2.8548098717106424e-05, "loss": 0.0232, "step": 65910 }, { "grad_norm": 0.4761523902416229, "learning_rate": 2.85331643218843e-05, "loss": 0.0244, "step": 65920 }, { "grad_norm": 0.5731577277183533, "learning_rate": 2.8518232274243046e-05, "loss": 0.0276, "step": 65930 }, { "grad_norm": 0.3583638370037079, "learning_rate": 2.8503302575815638e-05, "loss": 0.0228, "step": 65940 }, { "grad_norm": 0.43000394105911255, "learning_rate": 2.8488375228234716e-05, "loss": 0.0177, "step": 65950 }, { "grad_norm": 0.31503796577453613, "learning_rate": 2.8473450233132738e-05, "loss": 0.027, "step": 65960 }, { "grad_norm": 0.262462854385376, "learning_rate": 2.8458527592141898e-05, "loss": 0.0204, "step": 65970 }, { "grad_norm": 0.34796056151390076, "learning_rate": 2.8443607306894073e-05, "loss": 0.0206, "step": 65980 }, { "grad_norm": 0.5178652405738831, "learning_rate": 2.8428689379020946e-05, "loss": 0.028, "step": 65990 }, { "grad_norm": 0.443951278924942, "learning_rate": 2.841377381015391e-05, "loss": 0.0273, "step": 66000 }, { "grad_norm": 0.26625296473503113, "learning_rate": 2.8398860601924144e-05, "loss": 0.0187, "step": 66010 }, { "grad_norm": 0.34301358461380005, "learning_rate": 2.838394975596248e-05, "loss": 0.0275, "step": 66020 }, { "grad_norm": 0.20589211583137512, "learning_rate": 2.83690412738996e-05, "loss": 0.0162, "step": 66030 }, { "grad_norm": 0.3486347198486328, "learning_rate": 2.835413515736579e-05, "loss": 0.0217, "step": 66040 }, { "grad_norm": 0.24874424934387207, "learning_rate": 2.8339231407991267e-05, "loss": 0.0275, "step": 66050 }, { "grad_norm": 0.6053945422172546, "learning_rate": 2.832433002740581e-05, "loss": 0.0171, "step": 66060 }, { "grad_norm": 0.5208368301391602, "learning_rate": 2.8309431017239047e-05, "loss": 0.0208, "step": 66070 }, { "grad_norm": 0.4673827290534973, "learning_rate": 2.829453437912028e-05, "loss": 0.0203, "step": 66080 }, { "grad_norm": 0.4826878011226654, "learning_rate": 2.8279640114678617e-05, "loss": 0.0317, "step": 66090 }, { "grad_norm": 0.5465542674064636, "learning_rate": 2.8264748225542836e-05, "loss": 0.0295, "step": 66100 }, { "grad_norm": 0.7217364311218262, "learning_rate": 2.8249858713341516e-05, "loss": 0.033, "step": 66110 }, { "grad_norm": 0.40372520685195923, "learning_rate": 2.823497157970294e-05, "loss": 0.0243, "step": 66120 }, { "grad_norm": 0.4190724492073059, "learning_rate": 2.8220086826255164e-05, "loss": 0.0225, "step": 66130 }, { "grad_norm": 0.24027897417545319, "learning_rate": 2.8205204454625934e-05, "loss": 0.0195, "step": 66140 }, { "grad_norm": 0.39597058296203613, "learning_rate": 2.8190324466442798e-05, "loss": 0.0246, "step": 66150 }, { "grad_norm": 0.5410745143890381, "learning_rate": 2.8175446863332965e-05, "loss": 0.0441, "step": 66160 }, { "grad_norm": 0.5150311589241028, "learning_rate": 2.816057164692345e-05, "loss": 0.0211, "step": 66170 }, { "grad_norm": 0.4225524067878723, "learning_rate": 2.8145698818840992e-05, "loss": 0.0278, "step": 66180 }, { "grad_norm": 0.42778217792510986, "learning_rate": 2.8130828380712065e-05, "loss": 0.038, "step": 66190 }, { "grad_norm": 0.5265972018241882, "learning_rate": 2.811596033416285e-05, "loss": 0.0192, "step": 66200 }, { "grad_norm": 0.3041941821575165, "learning_rate": 2.8101094680819307e-05, "loss": 0.0193, "step": 66210 }, { "grad_norm": 0.3983783423900604, "learning_rate": 2.808623142230715e-05, "loss": 0.0169, "step": 66220 }, { "grad_norm": 0.28048813343048096, "learning_rate": 2.8071370560251757e-05, "loss": 0.0217, "step": 66230 }, { "grad_norm": 0.3664364814758301, "learning_rate": 2.805651209627831e-05, "loss": 0.0289, "step": 66240 }, { "grad_norm": 0.320223331451416, "learning_rate": 2.804165603201171e-05, "loss": 0.023, "step": 66250 }, { "grad_norm": 0.49182629585266113, "learning_rate": 2.802680236907661e-05, "loss": 0.0203, "step": 66260 }, { "grad_norm": 0.2819865643978119, "learning_rate": 2.8011951109097344e-05, "loss": 0.0349, "step": 66270 }, { "grad_norm": 0.5182209610939026, "learning_rate": 2.7997102253698072e-05, "loss": 0.0249, "step": 66280 }, { "grad_norm": 0.38951459527015686, "learning_rate": 2.7982255804502588e-05, "loss": 0.0264, "step": 66290 }, { "grad_norm": 0.48411762714385986, "learning_rate": 2.7967411763134516e-05, "loss": 0.0198, "step": 66300 }, { "grad_norm": 0.3310903012752533, "learning_rate": 2.7952570131217166e-05, "loss": 0.0213, "step": 66310 }, { "grad_norm": 0.39598774909973145, "learning_rate": 2.7937730910373618e-05, "loss": 0.0233, "step": 66320 }, { "grad_norm": 0.4682060778141022, "learning_rate": 2.7922894102226628e-05, "loss": 0.0261, "step": 66330 }, { "grad_norm": 0.37266916036605835, "learning_rate": 2.790805970839876e-05, "loss": 0.0262, "step": 66340 }, { "grad_norm": 0.6080222129821777, "learning_rate": 2.789322773051225e-05, "loss": 0.0208, "step": 66350 }, { "grad_norm": 0.7178638577461243, "learning_rate": 2.7878398170189124e-05, "loss": 0.0351, "step": 66360 }, { "grad_norm": 0.5248498916625977, "learning_rate": 2.7863571029051112e-05, "loss": 0.0277, "step": 66370 }, { "grad_norm": 0.322159081697464, "learning_rate": 2.784874630871971e-05, "loss": 0.0246, "step": 66380 }, { "grad_norm": 0.5110290050506592, "learning_rate": 2.7833924010816086e-05, "loss": 0.0258, "step": 66390 }, { "grad_norm": 0.4127480089664459, "learning_rate": 2.7819104136961226e-05, "loss": 0.0241, "step": 66400 }, { "grad_norm": 0.39750880002975464, "learning_rate": 2.780428668877577e-05, "loss": 0.0272, "step": 66410 }, { "grad_norm": 0.5439813733100891, "learning_rate": 2.7789471667880152e-05, "loss": 0.0239, "step": 66420 }, { "grad_norm": 0.6851187348365784, "learning_rate": 2.7774659075894523e-05, "loss": 0.0325, "step": 66430 }, { "grad_norm": 0.590684711933136, "learning_rate": 2.7759848914438778e-05, "loss": 0.0185, "step": 66440 }, { "grad_norm": 0.47504380345344543, "learning_rate": 2.7745041185132496e-05, "loss": 0.0522, "step": 66450 }, { "grad_norm": 0.2490435689687729, "learning_rate": 2.773023588959506e-05, "loss": 0.0199, "step": 66460 }, { "grad_norm": 0.79209303855896, "learning_rate": 2.7715433029445555e-05, "loss": 0.0236, "step": 66470 }, { "grad_norm": 0.3231735825538635, "learning_rate": 2.7700632606302766e-05, "loss": 0.0275, "step": 66480 }, { "grad_norm": 0.2858825922012329, "learning_rate": 2.768583462178528e-05, "loss": 0.0182, "step": 66490 }, { "grad_norm": 0.4966074228286743, "learning_rate": 2.767103907751136e-05, "loss": 0.038, "step": 66500 }, { "grad_norm": 0.25950416922569275, "learning_rate": 2.7656245975099065e-05, "loss": 0.021, "step": 66510 }, { "grad_norm": 0.39245933294296265, "learning_rate": 2.7641455316166093e-05, "loss": 0.0227, "step": 66520 }, { "grad_norm": 0.4918918311595917, "learning_rate": 2.7626667102329973e-05, "loss": 0.0233, "step": 66530 }, { "grad_norm": 0.31372007727622986, "learning_rate": 2.761188133520788e-05, "loss": 0.0325, "step": 66540 }, { "grad_norm": 0.26804256439208984, "learning_rate": 2.7597098016416784e-05, "loss": 0.0271, "step": 66550 }, { "grad_norm": 0.736243486404419, "learning_rate": 2.7582317147573368e-05, "loss": 0.0315, "step": 66560 }, { "grad_norm": 0.3498404324054718, "learning_rate": 2.7567538730294064e-05, "loss": 0.0178, "step": 66570 }, { "grad_norm": 0.5243226289749146, "learning_rate": 2.7552762766194973e-05, "loss": 0.0257, "step": 66580 }, { "grad_norm": 0.5945916771888733, "learning_rate": 2.753798925689202e-05, "loss": 0.0237, "step": 66590 }, { "grad_norm": 0.30379682779312134, "learning_rate": 2.7523218204000774e-05, "loss": 0.0224, "step": 66600 }, { "grad_norm": 0.3833535313606262, "learning_rate": 2.7508449609136612e-05, "loss": 0.0294, "step": 66610 }, { "grad_norm": 0.4058331549167633, "learning_rate": 2.749368347391454e-05, "loss": 0.0312, "step": 66620 }, { "grad_norm": 0.5239313840866089, "learning_rate": 2.7478919799949443e-05, "loss": 0.0242, "step": 66630 }, { "grad_norm": 0.3060673773288727, "learning_rate": 2.7464158588855803e-05, "loss": 0.0285, "step": 66640 }, { "grad_norm": 0.3389180302619934, "learning_rate": 2.7449399842247918e-05, "loss": 0.0211, "step": 66650 }, { "grad_norm": 0.5505356192588806, "learning_rate": 2.7434643561739736e-05, "loss": 0.0254, "step": 66660 }, { "grad_norm": 0.4738669991493225, "learning_rate": 2.7419889748945016e-05, "loss": 0.0253, "step": 66670 }, { "grad_norm": 0.19414080679416656, "learning_rate": 2.7405138405477203e-05, "loss": 0.0273, "step": 66680 }, { "grad_norm": 0.3914740979671478, "learning_rate": 2.73903895329495e-05, "loss": 0.029, "step": 66690 }, { "grad_norm": 0.3297837972640991, "learning_rate": 2.737564313297478e-05, "loss": 0.0178, "step": 66700 }, { "grad_norm": 0.5543848872184753, "learning_rate": 2.736089920716572e-05, "loss": 0.0237, "step": 66710 }, { "grad_norm": 0.34339627623558044, "learning_rate": 2.7346157757134706e-05, "loss": 0.018, "step": 66720 }, { "grad_norm": 0.5711151361465454, "learning_rate": 2.733141878449379e-05, "loss": 0.0351, "step": 66730 }, { "grad_norm": 0.32198280096054077, "learning_rate": 2.7316682290854844e-05, "loss": 0.0257, "step": 66740 }, { "grad_norm": 0.6011384725570679, "learning_rate": 2.7301948277829414e-05, "loss": 0.0286, "step": 66750 }, { "grad_norm": 0.9794595837593079, "learning_rate": 2.7287216747028814e-05, "loss": 0.0254, "step": 66760 }, { "grad_norm": 0.21839728951454163, "learning_rate": 2.7272487700064025e-05, "loss": 0.0254, "step": 66770 }, { "grad_norm": 0.38629376888275146, "learning_rate": 2.725776113854583e-05, "loss": 0.0244, "step": 66780 }, { "grad_norm": 0.7896490097045898, "learning_rate": 2.7243037064084664e-05, "loss": 0.0218, "step": 66790 }, { "grad_norm": 0.5627228021621704, "learning_rate": 2.7228315478290766e-05, "loss": 0.0217, "step": 66800 }, { "grad_norm": 0.38558870553970337, "learning_rate": 2.721359638277402e-05, "loss": 0.0194, "step": 66810 }, { "grad_norm": 0.5063501000404358, "learning_rate": 2.7198879779144148e-05, "loss": 0.0287, "step": 66820 }, { "grad_norm": 0.2478543370962143, "learning_rate": 2.7184165669010485e-05, "loss": 0.0226, "step": 66830 }, { "grad_norm": 0.523903489112854, "learning_rate": 2.716945405398218e-05, "loss": 0.0227, "step": 66840 }, { "grad_norm": 0.4536110758781433, "learning_rate": 2.715474493566803e-05, "loss": 0.0215, "step": 66850 }, { "grad_norm": 0.31228527426719666, "learning_rate": 2.7140038315676652e-05, "loss": 0.0229, "step": 66860 }, { "grad_norm": 0.44603532552719116, "learning_rate": 2.7125334195616275e-05, "loss": 0.0271, "step": 66870 }, { "grad_norm": 0.47679606080055237, "learning_rate": 2.7110632577094997e-05, "loss": 0.024, "step": 66880 }, { "grad_norm": 0.31546011567115784, "learning_rate": 2.709593346172051e-05, "loss": 0.0133, "step": 66890 }, { "grad_norm": 0.42226147651672363, "learning_rate": 2.708123685110032e-05, "loss": 0.0243, "step": 66900 }, { "grad_norm": 0.5833927989006042, "learning_rate": 2.7066542746841588e-05, "loss": 0.0216, "step": 66910 }, { "grad_norm": 0.22285659611225128, "learning_rate": 2.7051851150551267e-05, "loss": 0.0224, "step": 66920 }, { "grad_norm": 0.2689083218574524, "learning_rate": 2.7037162063835996e-05, "loss": 0.0207, "step": 66930 }, { "grad_norm": 0.4422704577445984, "learning_rate": 2.702247548830218e-05, "loss": 0.0171, "step": 66940 }, { "grad_norm": 0.4427341818809509, "learning_rate": 2.7007791425555883e-05, "loss": 0.026, "step": 66950 }, { "grad_norm": 0.4317672550678253, "learning_rate": 2.6993109877202945e-05, "loss": 0.022, "step": 66960 }, { "grad_norm": 0.6262643337249756, "learning_rate": 2.6978430844848935e-05, "loss": 0.0179, "step": 66970 }, { "grad_norm": 0.32021674513816833, "learning_rate": 2.69637543300991e-05, "loss": 0.0236, "step": 66980 }, { "grad_norm": 0.5373168587684631, "learning_rate": 2.694908033455848e-05, "loss": 0.026, "step": 66990 }, { "grad_norm": 0.4319372773170471, "learning_rate": 2.6934408859831733e-05, "loss": 0.0234, "step": 67000 }, { "grad_norm": 0.5630390644073486, "learning_rate": 2.69197399075234e-05, "loss": 0.0262, "step": 67010 }, { "grad_norm": 0.31355270743370056, "learning_rate": 2.6905073479237584e-05, "loss": 0.0264, "step": 67020 }, { "grad_norm": 0.802790641784668, "learning_rate": 2.6890409576578235e-05, "loss": 0.0232, "step": 67030 }, { "grad_norm": 0.34289541840553284, "learning_rate": 2.687574820114892e-05, "loss": 0.0238, "step": 67040 }, { "grad_norm": 0.3192283809185028, "learning_rate": 2.686108935455305e-05, "loss": 0.0248, "step": 67050 }, { "grad_norm": 0.3819241523742676, "learning_rate": 2.6846433038393603e-05, "loss": 0.021, "step": 67060 }, { "grad_norm": 0.31256303191185, "learning_rate": 2.6831779254273482e-05, "loss": 0.0207, "step": 67070 }, { "grad_norm": 0.7243808507919312, "learning_rate": 2.6817128003795122e-05, "loss": 0.0242, "step": 67080 }, { "grad_norm": 1.4211026430130005, "learning_rate": 2.680247928856081e-05, "loss": 0.0359, "step": 67090 }, { "grad_norm": 0.5250127911567688, "learning_rate": 2.6787833110172466e-05, "loss": 0.019, "step": 67100 }, { "grad_norm": 1.2747968435287476, "learning_rate": 2.6773189470231807e-05, "loss": 0.0302, "step": 67110 }, { "grad_norm": 0.2706454396247864, "learning_rate": 2.675854837034022e-05, "loss": 0.0225, "step": 67120 }, { "grad_norm": 0.37843742966651917, "learning_rate": 2.6743909812098833e-05, "loss": 0.0242, "step": 67130 }, { "grad_norm": 0.40957319736480713, "learning_rate": 2.67292737971085e-05, "loss": 0.0201, "step": 67140 }, { "grad_norm": 0.4533793032169342, "learning_rate": 2.671464032696982e-05, "loss": 0.0253, "step": 67150 }, { "grad_norm": 0.37818819284439087, "learning_rate": 2.6700009403283055e-05, "loss": 0.0114, "step": 67160 }, { "grad_norm": 0.3598371148109436, "learning_rate": 2.6685381027648214e-05, "loss": 0.0226, "step": 67170 }, { "grad_norm": 0.4997674524784088, "learning_rate": 2.6670755201665086e-05, "loss": 0.0262, "step": 67180 }, { "grad_norm": 0.3227909803390503, "learning_rate": 2.6656131926933066e-05, "loss": 0.0225, "step": 67190 }, { "grad_norm": 0.4952685236930847, "learning_rate": 2.664151120505136e-05, "loss": 0.0175, "step": 67200 }, { "grad_norm": 0.2997296154499054, "learning_rate": 2.6626893037618873e-05, "loss": 0.0146, "step": 67210 }, { "grad_norm": 0.38468822836875916, "learning_rate": 2.6612277426234243e-05, "loss": 0.0215, "step": 67220 }, { "grad_norm": 0.5642328858375549, "learning_rate": 2.659766437249577e-05, "loss": 0.0322, "step": 67230 }, { "grad_norm": 0.5524678826332092, "learning_rate": 2.6583053878001557e-05, "loss": 0.0255, "step": 67240 }, { "grad_norm": 0.3344495892524719, "learning_rate": 2.6568445944349323e-05, "loss": 0.0298, "step": 67250 }, { "grad_norm": 0.4109451174736023, "learning_rate": 2.6553840573136657e-05, "loss": 0.0191, "step": 67260 }, { "grad_norm": 0.3225076198577881, "learning_rate": 2.6539237765960706e-05, "loss": 0.0259, "step": 67270 }, { "grad_norm": 0.4670450687408447, "learning_rate": 2.6524637524418465e-05, "loss": 0.0229, "step": 67280 }, { "grad_norm": 0.48432546854019165, "learning_rate": 2.651003985010655e-05, "loss": 0.0235, "step": 67290 }, { "grad_norm": 0.45632079243659973, "learning_rate": 2.649544474462138e-05, "loss": 0.0227, "step": 67300 }, { "grad_norm": 0.24571990966796875, "learning_rate": 2.6480852209559015e-05, "loss": 0.0262, "step": 67310 }, { "grad_norm": 0.2780037522315979, "learning_rate": 2.6466262246515296e-05, "loss": 0.0166, "step": 67320 }, { "grad_norm": 0.5077024102210999, "learning_rate": 2.6451674857085746e-05, "loss": 0.0294, "step": 67330 }, { "grad_norm": 0.5584097504615784, "learning_rate": 2.6437090042865655e-05, "loss": 0.0335, "step": 67340 }, { "grad_norm": 0.29512596130371094, "learning_rate": 2.6422507805449947e-05, "loss": 0.0249, "step": 67350 }, { "grad_norm": 0.33365941047668457, "learning_rate": 2.6407928146433357e-05, "loss": 0.0192, "step": 67360 }, { "grad_norm": 0.5135411620140076, "learning_rate": 2.639335106741026e-05, "loss": 0.0198, "step": 67370 }, { "grad_norm": 0.33982840180397034, "learning_rate": 2.6378776569974796e-05, "loss": 0.0175, "step": 67380 }, { "grad_norm": 0.4428439736366272, "learning_rate": 2.6364204655720814e-05, "loss": 0.0157, "step": 67390 }, { "grad_norm": 0.3080964684486389, "learning_rate": 2.6349635326241894e-05, "loss": 0.0182, "step": 67400 }, { "grad_norm": 0.5140613317489624, "learning_rate": 2.633506858313127e-05, "loss": 0.0303, "step": 67410 }, { "grad_norm": 0.37292391061782837, "learning_rate": 2.6320504427981975e-05, "loss": 0.0155, "step": 67420 }, { "grad_norm": 0.295072466135025, "learning_rate": 2.6305942862386734e-05, "loss": 0.0208, "step": 67430 }, { "grad_norm": 0.36599522829055786, "learning_rate": 2.629138388793793e-05, "loss": 0.0201, "step": 67440 }, { "grad_norm": 0.3454101085662842, "learning_rate": 2.6276827506227743e-05, "loss": 0.0264, "step": 67450 }, { "grad_norm": 0.3405727744102478, "learning_rate": 2.6262273718848037e-05, "loss": 0.0263, "step": 67460 }, { "grad_norm": 0.31507158279418945, "learning_rate": 2.6247722527390407e-05, "loss": 0.014, "step": 67470 }, { "grad_norm": 0.23223400115966797, "learning_rate": 2.6233173933446104e-05, "loss": 0.026, "step": 67480 }, { "grad_norm": 0.6315438747406006, "learning_rate": 2.6218627938606188e-05, "loss": 0.0336, "step": 67490 }, { "grad_norm": 0.4038015305995941, "learning_rate": 2.620408454446135e-05, "loss": 0.0167, "step": 67500 }, { "grad_norm": 0.8310685753822327, "learning_rate": 2.6189543752602046e-05, "loss": 0.0218, "step": 67510 }, { "grad_norm": 0.42232945561408997, "learning_rate": 2.6175005564618444e-05, "loss": 0.0296, "step": 67520 }, { "grad_norm": 0.20259512960910797, "learning_rate": 2.6160469982100428e-05, "loss": 0.0253, "step": 67530 }, { "grad_norm": 0.4965240955352783, "learning_rate": 2.6145937006637555e-05, "loss": 0.0215, "step": 67540 }, { "grad_norm": 0.5769661068916321, "learning_rate": 2.6131406639819166e-05, "loss": 0.0194, "step": 67550 }, { "grad_norm": 0.40481337904930115, "learning_rate": 2.611687888323424e-05, "loss": 0.0267, "step": 67560 }, { "grad_norm": 1.4977436065673828, "learning_rate": 2.610235373847153e-05, "loss": 0.023, "step": 67570 }, { "grad_norm": 0.47646552324295044, "learning_rate": 2.6087831207119494e-05, "loss": 0.028, "step": 67580 }, { "grad_norm": 1.1595697402954102, "learning_rate": 2.60733112907663e-05, "loss": 0.0289, "step": 67590 }, { "grad_norm": 0.5189024209976196, "learning_rate": 2.6058793990999786e-05, "loss": 0.0286, "step": 67600 }, { "grad_norm": 0.4851619601249695, "learning_rate": 2.6044279309407593e-05, "loss": 0.0209, "step": 67610 }, { "grad_norm": 0.34249112010002136, "learning_rate": 2.6029767247576976e-05, "loss": 0.0237, "step": 67620 }, { "grad_norm": 0.38649123907089233, "learning_rate": 2.601525780709497e-05, "loss": 0.0174, "step": 67630 }, { "grad_norm": 0.2312651425600052, "learning_rate": 2.6000750989548318e-05, "loss": 0.0227, "step": 67640 }, { "grad_norm": 0.31805384159088135, "learning_rate": 2.5986246796523462e-05, "loss": 0.0215, "step": 67650 }, { "grad_norm": 0.36027106642723083, "learning_rate": 2.5971745229606546e-05, "loss": 0.0197, "step": 67660 }, { "grad_norm": 0.4121243357658386, "learning_rate": 2.5957246290383442e-05, "loss": 0.0262, "step": 67670 }, { "grad_norm": 0.3875292241573334, "learning_rate": 2.594274998043975e-05, "loss": 0.0232, "step": 67680 }, { "grad_norm": 0.5144380927085876, "learning_rate": 2.5928256301360733e-05, "loss": 0.0248, "step": 67690 }, { "grad_norm": 0.36481189727783203, "learning_rate": 2.591376525473142e-05, "loss": 0.0196, "step": 67700 }, { "grad_norm": 0.393820196390152, "learning_rate": 2.5899276842136523e-05, "loss": 0.0251, "step": 67710 }, { "grad_norm": 0.45398664474487305, "learning_rate": 2.5884791065160495e-05, "loss": 0.0198, "step": 67720 }, { "grad_norm": 0.2732049226760864, "learning_rate": 2.5870307925387436e-05, "loss": 0.0186, "step": 67730 }, { "grad_norm": 0.20502662658691406, "learning_rate": 2.585582742440125e-05, "loss": 0.017, "step": 67740 }, { "grad_norm": 0.3487433195114136, "learning_rate": 2.5841349563785455e-05, "loss": 0.0232, "step": 67750 }, { "grad_norm": 0.3749742805957794, "learning_rate": 2.5826874345123355e-05, "loss": 0.0237, "step": 67760 }, { "grad_norm": 0.4131990075111389, "learning_rate": 2.581240176999792e-05, "loss": 0.0156, "step": 67770 }, { "grad_norm": 0.4321487545967102, "learning_rate": 2.579793183999189e-05, "loss": 0.029, "step": 67780 }, { "grad_norm": 0.3455876410007477, "learning_rate": 2.5783464556687627e-05, "loss": 0.0266, "step": 67790 }, { "grad_norm": 0.1868211328983307, "learning_rate": 2.576899992166729e-05, "loss": 0.0166, "step": 67800 }, { "grad_norm": 0.4379206895828247, "learning_rate": 2.5754537936512668e-05, "loss": 0.0192, "step": 67810 }, { "grad_norm": 0.25635653734207153, "learning_rate": 2.574007860280535e-05, "loss": 0.0333, "step": 67820 }, { "grad_norm": 0.4247840642929077, "learning_rate": 2.5725621922126518e-05, "loss": 0.0419, "step": 67830 }, { "grad_norm": 0.25306251645088196, "learning_rate": 2.571116789605722e-05, "loss": 0.0317, "step": 67840 }, { "grad_norm": 0.4134916067123413, "learning_rate": 2.5696716526178054e-05, "loss": 0.0259, "step": 67850 }, { "grad_norm": 0.4623168706893921, "learning_rate": 2.568226781406946e-05, "loss": 0.0199, "step": 67860 }, { "grad_norm": 0.4828008711338043, "learning_rate": 2.5667821761311474e-05, "loss": 0.0158, "step": 67870 }, { "grad_norm": 0.24953873455524445, "learning_rate": 2.5653378369483917e-05, "loss": 0.0164, "step": 67880 }, { "grad_norm": 0.3538094162940979, "learning_rate": 2.5638937640166292e-05, "loss": 0.0495, "step": 67890 }, { "grad_norm": 0.3033200204372406, "learning_rate": 2.5624499574937842e-05, "loss": 0.0223, "step": 67900 }, { "grad_norm": 0.2459130585193634, "learning_rate": 2.5610064175377456e-05, "loss": 0.0211, "step": 67910 }, { "grad_norm": 0.35202011466026306, "learning_rate": 2.5595631443063782e-05, "loss": 0.0262, "step": 67920 }, { "grad_norm": 0.35271456837654114, "learning_rate": 2.5581201379575177e-05, "loss": 0.0175, "step": 67930 }, { "grad_norm": 0.6391897201538086, "learning_rate": 2.5566773986489655e-05, "loss": 0.0215, "step": 67940 }, { "grad_norm": 0.21759089827537537, "learning_rate": 2.5552349265385e-05, "loss": 0.0128, "step": 67950 }, { "grad_norm": 0.29391342401504517, "learning_rate": 2.5537927217838675e-05, "loss": 0.0251, "step": 67960 }, { "grad_norm": 0.2075357735157013, "learning_rate": 2.5523507845427864e-05, "loss": 0.0125, "step": 67970 }, { "grad_norm": 0.5873269438743591, "learning_rate": 2.5509091149729413e-05, "loss": 0.0211, "step": 67980 }, { "grad_norm": 0.6193384528160095, "learning_rate": 2.5494677132319962e-05, "loss": 0.0254, "step": 67990 }, { "grad_norm": 0.29762277007102966, "learning_rate": 2.548026579477575e-05, "loss": 0.0302, "step": 68000 }, { "grad_norm": 0.4093847870826721, "learning_rate": 2.5465857138672834e-05, "loss": 0.0227, "step": 68010 }, { "grad_norm": 0.5304200649261475, "learning_rate": 2.5451451165586848e-05, "loss": 0.0304, "step": 68020 }, { "grad_norm": 0.46789631247520447, "learning_rate": 2.543704787709329e-05, "loss": 0.0178, "step": 68030 }, { "grad_norm": 0.5026671290397644, "learning_rate": 2.542264727476723e-05, "loss": 0.0278, "step": 68040 }, { "grad_norm": 0.39758986234664917, "learning_rate": 2.5408249360183535e-05, "loss": 0.0261, "step": 68050 }, { "grad_norm": 0.3140036463737488, "learning_rate": 2.5393854134916694e-05, "loss": 0.0206, "step": 68060 }, { "grad_norm": 0.3394707441329956, "learning_rate": 2.5379461600540993e-05, "loss": 0.0276, "step": 68070 }, { "grad_norm": 0.3250548541545868, "learning_rate": 2.5365071758630317e-05, "loss": 0.0304, "step": 68080 }, { "grad_norm": 0.24672462046146393, "learning_rate": 2.535068461075839e-05, "loss": 0.0254, "step": 68090 }, { "grad_norm": 0.3516257405281067, "learning_rate": 2.5336300158498516e-05, "loss": 0.0223, "step": 68100 }, { "grad_norm": 0.28903084993362427, "learning_rate": 2.5321918403423793e-05, "loss": 0.0191, "step": 68110 }, { "grad_norm": 0.6674867868423462, "learning_rate": 2.5307539347106957e-05, "loss": 0.0264, "step": 68120 }, { "grad_norm": 0.5366861820220947, "learning_rate": 2.529316299112049e-05, "loss": 0.0232, "step": 68130 }, { "grad_norm": 0.3881434500217438, "learning_rate": 2.5278789337036567e-05, "loss": 0.0241, "step": 68140 }, { "grad_norm": 0.46904605627059937, "learning_rate": 2.52644183864271e-05, "loss": 0.0183, "step": 68150 }, { "grad_norm": 0.7175247669219971, "learning_rate": 2.525005014086363e-05, "loss": 0.0262, "step": 68160 }, { "grad_norm": 0.3588079810142517, "learning_rate": 2.5235684601917465e-05, "loss": 0.0184, "step": 68170 }, { "grad_norm": 0.5849226117134094, "learning_rate": 2.5221321771159618e-05, "loss": 0.0129, "step": 68180 }, { "grad_norm": 0.41366004943847656, "learning_rate": 2.5206961650160743e-05, "loss": 0.0245, "step": 68190 }, { "grad_norm": 0.3090498149394989, "learning_rate": 2.519260424049129e-05, "loss": 0.0308, "step": 68200 }, { "grad_norm": 0.6009929776191711, "learning_rate": 2.5178249543721304e-05, "loss": 0.0217, "step": 68210 }, { "grad_norm": 0.4999106228351593, "learning_rate": 2.5163897561420668e-05, "loss": 0.0249, "step": 68220 }, { "grad_norm": 0.39074915647506714, "learning_rate": 2.514954829515883e-05, "loss": 0.0211, "step": 68230 }, { "grad_norm": 0.4483858644962311, "learning_rate": 2.513520174650505e-05, "loss": 0.0213, "step": 68240 }, { "grad_norm": 0.29797568917274475, "learning_rate": 2.5120857917028207e-05, "loss": 0.0195, "step": 68250 }, { "grad_norm": 0.29879340529441833, "learning_rate": 2.5106516808296953e-05, "loss": 0.0278, "step": 68260 }, { "grad_norm": 0.9018714427947998, "learning_rate": 2.5092178421879552e-05, "loss": 0.015, "step": 68270 }, { "grad_norm": 0.5690357089042664, "learning_rate": 2.5077842759344116e-05, "loss": 0.0164, "step": 68280 }, { "grad_norm": 0.29867658019065857, "learning_rate": 2.5063509822258304e-05, "loss": 0.021, "step": 68290 }, { "grad_norm": 0.2683553993701935, "learning_rate": 2.5049179612189576e-05, "loss": 0.0287, "step": 68300 }, { "grad_norm": 0.3401447832584381, "learning_rate": 2.5034852130705043e-05, "loss": 0.0178, "step": 68310 }, { "grad_norm": 0.3930397033691406, "learning_rate": 2.5020527379371572e-05, "loss": 0.0212, "step": 68320 }, { "grad_norm": 0.5452980399131775, "learning_rate": 2.500620535975564e-05, "loss": 0.0345, "step": 68330 }, { "grad_norm": 0.35408443212509155, "learning_rate": 2.4991886073423515e-05, "loss": 0.0219, "step": 68340 }, { "grad_norm": 0.4206700325012207, "learning_rate": 2.4977569521941134e-05, "loss": 0.0232, "step": 68350 }, { "grad_norm": 0.3485111892223358, "learning_rate": 2.4963255706874143e-05, "loss": 0.0191, "step": 68360 }, { "grad_norm": 0.43224114179611206, "learning_rate": 2.4948944629787853e-05, "loss": 0.0181, "step": 68370 }, { "grad_norm": 0.4695822298526764, "learning_rate": 2.493463629224731e-05, "loss": 0.0272, "step": 68380 }, { "grad_norm": 0.28866055607795715, "learning_rate": 2.492033069581728e-05, "loss": 0.0245, "step": 68390 }, { "grad_norm": 0.2571967542171478, "learning_rate": 2.4906027842062137e-05, "loss": 0.0233, "step": 68400 }, { "grad_norm": 0.33500993251800537, "learning_rate": 2.48917277325461e-05, "loss": 0.0179, "step": 68410 }, { "grad_norm": 0.24366387724876404, "learning_rate": 2.487743036883295e-05, "loss": 0.0209, "step": 68420 }, { "grad_norm": 0.3051843047142029, "learning_rate": 2.486313575248626e-05, "loss": 0.0261, "step": 68430 }, { "grad_norm": 0.5907734632492065, "learning_rate": 2.4848843885069235e-05, "loss": 0.0206, "step": 68440 }, { "grad_norm": 0.16073141992092133, "learning_rate": 2.483455476814484e-05, "loss": 0.0201, "step": 68450 }, { "grad_norm": 0.3429287374019623, "learning_rate": 2.4820268403275672e-05, "loss": 0.0201, "step": 68460 }, { "grad_norm": 0.6332221627235413, "learning_rate": 2.480598479202413e-05, "loss": 0.0262, "step": 68470 }, { "grad_norm": 0.16434554755687714, "learning_rate": 2.4791703935952193e-05, "loss": 0.0142, "step": 68480 }, { "grad_norm": 4.517096996307373, "learning_rate": 2.477742583662163e-05, "loss": 0.0245, "step": 68490 }, { "grad_norm": 0.23914356529712677, "learning_rate": 2.476315049559384e-05, "loss": 0.0194, "step": 68500 }, { "grad_norm": 0.4442862272262573, "learning_rate": 2.4748877914429987e-05, "loss": 0.0264, "step": 68510 }, { "grad_norm": 0.44486355781555176, "learning_rate": 2.4734608094690857e-05, "loss": 0.0189, "step": 68520 }, { "grad_norm": 0.770881175994873, "learning_rate": 2.4720341037937006e-05, "loss": 0.0187, "step": 68530 }, { "grad_norm": 0.727027416229248, "learning_rate": 2.4706076745728645e-05, "loss": 0.0394, "step": 68540 }, { "grad_norm": 0.3057388365268707, "learning_rate": 2.469181521962573e-05, "loss": 0.0335, "step": 68550 }, { "grad_norm": 0.4289341866970062, "learning_rate": 2.467755646118783e-05, "loss": 0.0187, "step": 68560 }, { "grad_norm": 0.255001962184906, "learning_rate": 2.46633004719743e-05, "loss": 0.0226, "step": 68570 }, { "grad_norm": 0.5725656747817993, "learning_rate": 2.4649047253544123e-05, "loss": 0.0248, "step": 68580 }, { "grad_norm": 0.5557990074157715, "learning_rate": 2.4634796807456016e-05, "loss": 0.0231, "step": 68590 }, { "grad_norm": 0.3570966124534607, "learning_rate": 2.46205491352684e-05, "loss": 0.0177, "step": 68600 }, { "grad_norm": 0.9734482169151306, "learning_rate": 2.4606304238539375e-05, "loss": 0.0319, "step": 68610 }, { "grad_norm": 0.41128483414649963, "learning_rate": 2.4592062118826753e-05, "loss": 0.0211, "step": 68620 }, { "grad_norm": 0.4294279217720032, "learning_rate": 2.4577822777688004e-05, "loss": 0.0174, "step": 68630 }, { "grad_norm": 0.5358169078826904, "learning_rate": 2.4563586216680347e-05, "loss": 0.0169, "step": 68640 }, { "grad_norm": 0.32354536652565, "learning_rate": 2.4549352437360622e-05, "loss": 0.0191, "step": 68650 }, { "grad_norm": 0.5563029646873474, "learning_rate": 2.4535121441285493e-05, "loss": 0.0201, "step": 68660 }, { "grad_norm": 0.4803687632083893, "learning_rate": 2.4520893230011172e-05, "loss": 0.0221, "step": 68670 }, { "grad_norm": 0.38188236951828003, "learning_rate": 2.450666780509368e-05, "loss": 0.0161, "step": 68680 }, { "grad_norm": 0.2721633315086365, "learning_rate": 2.4492445168088656e-05, "loss": 0.0252, "step": 68690 }, { "grad_norm": 1.0359889268875122, "learning_rate": 2.447822532055149e-05, "loss": 0.0312, "step": 68700 }, { "grad_norm": 0.443739652633667, "learning_rate": 2.4464008264037212e-05, "loss": 0.0214, "step": 68710 }, { "grad_norm": 0.4898330867290497, "learning_rate": 2.4449794000100605e-05, "loss": 0.0224, "step": 68720 }, { "grad_norm": 0.5896316766738892, "learning_rate": 2.4435582530296108e-05, "loss": 0.0301, "step": 68730 }, { "grad_norm": 0.5402125716209412, "learning_rate": 2.4421373856177887e-05, "loss": 0.0285, "step": 68740 }, { "grad_norm": 0.43430373072624207, "learning_rate": 2.440716797929975e-05, "loss": 0.0202, "step": 68750 }, { "grad_norm": 0.49426543712615967, "learning_rate": 2.439296490121526e-05, "loss": 0.0264, "step": 68760 }, { "grad_norm": 0.3909512162208557, "learning_rate": 2.4378764623477608e-05, "loss": 0.027, "step": 68770 }, { "grad_norm": 0.36281833052635193, "learning_rate": 2.436456714763974e-05, "loss": 0.0244, "step": 68780 }, { "grad_norm": 0.3230171203613281, "learning_rate": 2.4350372475254262e-05, "loss": 0.0225, "step": 68790 }, { "grad_norm": 0.4422355890274048, "learning_rate": 2.433618060787351e-05, "loss": 0.0255, "step": 68800 }, { "grad_norm": 0.7661280035972595, "learning_rate": 2.432199154704944e-05, "loss": 0.0172, "step": 68810 }, { "grad_norm": 0.4180275797843933, "learning_rate": 2.430780529433377e-05, "loss": 0.0267, "step": 68820 }, { "grad_norm": 0.35815611481666565, "learning_rate": 2.42936218512779e-05, "loss": 0.0209, "step": 68830 }, { "grad_norm": 0.1968649923801422, "learning_rate": 2.427944121943288e-05, "loss": 0.0196, "step": 68840 }, { "grad_norm": 0.48163163661956787, "learning_rate": 2.4265263400349497e-05, "loss": 0.0198, "step": 68850 }, { "grad_norm": 0.5582569241523743, "learning_rate": 2.4251088395578213e-05, "loss": 0.027, "step": 68860 }, { "grad_norm": 0.5925424098968506, "learning_rate": 2.4236916206669214e-05, "loss": 0.0259, "step": 68870 }, { "grad_norm": 0.19950075447559357, "learning_rate": 2.4222746835172304e-05, "loss": 0.0262, "step": 68880 }, { "grad_norm": 0.6870796084403992, "learning_rate": 2.4208580282637065e-05, "loss": 0.0387, "step": 68890 }, { "grad_norm": 0.6134009957313538, "learning_rate": 2.4194416550612687e-05, "loss": 0.029, "step": 68900 }, { "grad_norm": 0.32458895444869995, "learning_rate": 2.418025564064812e-05, "loss": 0.0214, "step": 68910 }, { "grad_norm": 0.3855271339416504, "learning_rate": 2.4166097554291978e-05, "loss": 0.0209, "step": 68920 }, { "grad_norm": 0.4289919435977936, "learning_rate": 2.4151942293092583e-05, "loss": 0.0355, "step": 68930 }, { "grad_norm": 0.4063750207424164, "learning_rate": 2.41377898585979e-05, "loss": 0.0242, "step": 68940 }, { "grad_norm": 0.45007506012916565, "learning_rate": 2.4123640252355652e-05, "loss": 0.0201, "step": 68950 }, { "grad_norm": 0.47682175040245056, "learning_rate": 2.410949347591318e-05, "loss": 0.0205, "step": 68960 }, { "grad_norm": 0.2053399682044983, "learning_rate": 2.4095349530817578e-05, "loss": 0.0281, "step": 68970 }, { "grad_norm": 0.2479851096868515, "learning_rate": 2.4081208418615603e-05, "loss": 0.0217, "step": 68980 }, { "grad_norm": 1.3215388059616089, "learning_rate": 2.4067070140853725e-05, "loss": 0.0266, "step": 68990 }, { "grad_norm": 0.23922297358512878, "learning_rate": 2.405293469907805e-05, "loss": 0.0188, "step": 69000 }, { "grad_norm": 0.3877456486225128, "learning_rate": 2.4038802094834446e-05, "loss": 0.0262, "step": 69010 }, { "grad_norm": 0.47765377163887024, "learning_rate": 2.4024672329668395e-05, "loss": 0.0193, "step": 69020 }, { "grad_norm": 0.3140150010585785, "learning_rate": 2.4010545405125125e-05, "loss": 0.0191, "step": 69030 }, { "grad_norm": 0.3565734624862671, "learning_rate": 2.3996421322749528e-05, "loss": 0.021, "step": 69040 }, { "grad_norm": 0.3857012987136841, "learning_rate": 2.3982300084086224e-05, "loss": 0.0289, "step": 69050 }, { "grad_norm": 0.44199278950691223, "learning_rate": 2.396818169067944e-05, "loss": 0.0259, "step": 69060 }, { "grad_norm": 0.6504522562026978, "learning_rate": 2.3954066144073178e-05, "loss": 0.0233, "step": 69070 }, { "grad_norm": 0.48564091324806213, "learning_rate": 2.39399534458111e-05, "loss": 0.0163, "step": 69080 }, { "grad_norm": 0.3165345788002014, "learning_rate": 2.392584359743651e-05, "loss": 0.0162, "step": 69090 }, { "grad_norm": 0.3366410434246063, "learning_rate": 2.3911736600492463e-05, "loss": 0.0154, "step": 69100 }, { "grad_norm": 0.29974105954170227, "learning_rate": 2.3897632456521672e-05, "loss": 0.0157, "step": 69110 }, { "grad_norm": 0.36765390634536743, "learning_rate": 2.388353116706658e-05, "loss": 0.0186, "step": 69120 }, { "grad_norm": 0.4640134572982788, "learning_rate": 2.3869432733669232e-05, "loss": 0.0232, "step": 69130 }, { "grad_norm": 0.3031480014324188, "learning_rate": 2.385533715787145e-05, "loss": 0.02, "step": 69140 }, { "grad_norm": 0.6314582228660583, "learning_rate": 2.3841244441214673e-05, "loss": 0.0286, "step": 69150 }, { "grad_norm": 0.23421929776668549, "learning_rate": 2.3827154585240073e-05, "loss": 0.0213, "step": 69160 }, { "grad_norm": 0.3388530910015106, "learning_rate": 2.3813067591488498e-05, "loss": 0.0283, "step": 69170 }, { "grad_norm": 0.29327821731567383, "learning_rate": 2.3798983461500494e-05, "loss": 0.0218, "step": 69180 }, { "grad_norm": 0.22530324757099152, "learning_rate": 2.3784902196816256e-05, "loss": 0.0322, "step": 69190 }, { "grad_norm": 0.455477774143219, "learning_rate": 2.3770823798975716e-05, "loss": 0.0264, "step": 69200 }, { "grad_norm": 0.5961194634437561, "learning_rate": 2.3756748269518424e-05, "loss": 0.0187, "step": 69210 }, { "grad_norm": 0.3777901828289032, "learning_rate": 2.374267560998371e-05, "loss": 0.0266, "step": 69220 }, { "grad_norm": 0.30592259764671326, "learning_rate": 2.3728605821910477e-05, "loss": 0.0162, "step": 69230 }, { "grad_norm": 0.43855130672454834, "learning_rate": 2.3714538906837453e-05, "loss": 0.015, "step": 69240 }, { "grad_norm": 0.7065961956977844, "learning_rate": 2.3700474866302913e-05, "loss": 0.0186, "step": 69250 }, { "grad_norm": 0.48860201239585876, "learning_rate": 2.3686413701844917e-05, "loss": 0.023, "step": 69260 }, { "grad_norm": 0.8852399587631226, "learning_rate": 2.367235541500114e-05, "loss": 0.0246, "step": 69270 }, { "grad_norm": 0.5417636632919312, "learning_rate": 2.3658300007308993e-05, "loss": 0.0274, "step": 69280 }, { "grad_norm": 0.4216534495353699, "learning_rate": 2.364424748030555e-05, "loss": 0.0278, "step": 69290 }, { "grad_norm": 0.33783864974975586, "learning_rate": 2.3630197835527595e-05, "loss": 0.0205, "step": 69300 }, { "grad_norm": 0.47777870297431946, "learning_rate": 2.3616151074511546e-05, "loss": 0.0219, "step": 69310 }, { "grad_norm": 0.35529476404190063, "learning_rate": 2.3602107198793545e-05, "loss": 0.019, "step": 69320 }, { "grad_norm": 0.19626016914844513, "learning_rate": 2.3588066209909427e-05, "loss": 0.0172, "step": 69330 }, { "grad_norm": 0.19646048545837402, "learning_rate": 2.3574028109394662e-05, "loss": 0.0196, "step": 69340 }, { "grad_norm": 0.3781728446483612, "learning_rate": 2.3559992898784454e-05, "loss": 0.0233, "step": 69350 }, { "grad_norm": 0.2984166741371155, "learning_rate": 2.354596057961367e-05, "loss": 0.0186, "step": 69360 }, { "grad_norm": 0.4390251636505127, "learning_rate": 2.353193115341688e-05, "loss": 0.0256, "step": 69370 }, { "grad_norm": 0.19197368621826172, "learning_rate": 2.3517904621728294e-05, "loss": 0.0241, "step": 69380 }, { "grad_norm": 0.8524041771888733, "learning_rate": 2.3503880986081855e-05, "loss": 0.0279, "step": 69390 }, { "grad_norm": 0.27316105365753174, "learning_rate": 2.3489860248011136e-05, "loss": 0.0225, "step": 69400 }, { "grad_norm": 0.5209732055664062, "learning_rate": 2.3475842409049464e-05, "loss": 0.0221, "step": 69410 }, { "grad_norm": 0.3139890134334564, "learning_rate": 2.346182747072975e-05, "loss": 0.0232, "step": 69420 }, { "grad_norm": 0.3862137496471405, "learning_rate": 2.3447815434584718e-05, "loss": 0.0382, "step": 69430 }, { "grad_norm": 0.3366240859031677, "learning_rate": 2.3433806302146656e-05, "loss": 0.0231, "step": 69440 }, { "grad_norm": 0.299465149641037, "learning_rate": 2.3419800074947612e-05, "loss": 0.0166, "step": 69450 }, { "grad_norm": 0.45813554525375366, "learning_rate": 2.3405796754519248e-05, "loss": 0.0169, "step": 69460 }, { "grad_norm": 0.2606757879257202, "learning_rate": 2.3391796342392992e-05, "loss": 0.0201, "step": 69470 }, { "grad_norm": 0.3516906499862671, "learning_rate": 2.337779884009984e-05, "loss": 0.0166, "step": 69480 }, { "grad_norm": 0.45484304428100586, "learning_rate": 2.3363804249170624e-05, "loss": 0.0209, "step": 69490 }, { "grad_norm": 0.31867384910583496, "learning_rate": 2.3349812571135703e-05, "loss": 0.0192, "step": 69500 }, { "grad_norm": 0.46364694833755493, "learning_rate": 2.333582380752523e-05, "loss": 0.0337, "step": 69510 }, { "grad_norm": 0.4264422059059143, "learning_rate": 2.3321837959868963e-05, "loss": 0.0217, "step": 69520 }, { "grad_norm": 0.41851484775543213, "learning_rate": 2.3307855029696383e-05, "loss": 0.0197, "step": 69530 }, { "grad_norm": 0.3105546832084656, "learning_rate": 2.3293875018536648e-05, "loss": 0.0224, "step": 69540 }, { "grad_norm": 0.31748124957084656, "learning_rate": 2.327989792791861e-05, "loss": 0.0452, "step": 69550 }, { "grad_norm": 0.2667838931083679, "learning_rate": 2.326592375937074e-05, "loss": 0.0249, "step": 69560 }, { "grad_norm": 0.2972293496131897, "learning_rate": 2.3251952514421248e-05, "loss": 0.0212, "step": 69570 }, { "grad_norm": 0.5144262909889221, "learning_rate": 2.323798419459804e-05, "loss": 0.026, "step": 69580 }, { "grad_norm": 0.3289845883846283, "learning_rate": 2.3224018801428625e-05, "loss": 0.0203, "step": 69590 }, { "grad_norm": 0.33853679895401, "learning_rate": 2.3210056336440273e-05, "loss": 0.017, "step": 69600 }, { "grad_norm": 0.34427231550216675, "learning_rate": 2.3196096801159844e-05, "loss": 0.0172, "step": 69610 }, { "grad_norm": 0.31371837854385376, "learning_rate": 2.3182140197114012e-05, "loss": 0.0215, "step": 69620 }, { "grad_norm": 0.29109683632850647, "learning_rate": 2.3168186525828985e-05, "loss": 0.0224, "step": 69630 }, { "grad_norm": 0.3676668405532837, "learning_rate": 2.3154235788830757e-05, "loss": 0.0167, "step": 69640 }, { "grad_norm": 0.6000643372535706, "learning_rate": 2.3140287987644927e-05, "loss": 0.0219, "step": 69650 }, { "grad_norm": 0.37453100085258484, "learning_rate": 2.3126343123796833e-05, "loss": 0.0205, "step": 69660 }, { "grad_norm": 0.34715336561203003, "learning_rate": 2.3112401198811418e-05, "loss": 0.018, "step": 69670 }, { "grad_norm": 0.39154669642448425, "learning_rate": 2.3098462214213418e-05, "loss": 0.0225, "step": 69680 }, { "grad_norm": 0.3996085226535797, "learning_rate": 2.308452617152712e-05, "loss": 0.0248, "step": 69690 }, { "grad_norm": 0.3446179926395416, "learning_rate": 2.3070593072276603e-05, "loss": 0.0213, "step": 69700 }, { "grad_norm": 0.32858583331108093, "learning_rate": 2.3056662917985518e-05, "loss": 0.0275, "step": 69710 }, { "grad_norm": 0.39939451217651367, "learning_rate": 2.3042735710177283e-05, "loss": 0.0185, "step": 69720 }, { "grad_norm": 0.580695390701294, "learning_rate": 2.302881145037492e-05, "loss": 0.0307, "step": 69730 }, { "grad_norm": 0.4073522686958313, "learning_rate": 2.3014890140101192e-05, "loss": 0.0152, "step": 69740 }, { "grad_norm": 0.41125383973121643, "learning_rate": 2.30009717808785e-05, "loss": 0.0166, "step": 69750 }, { "grad_norm": 0.44116437435150146, "learning_rate": 2.2987056374228967e-05, "loss": 0.0219, "step": 69760 }, { "grad_norm": 0.35447317361831665, "learning_rate": 2.297314392167431e-05, "loss": 0.0222, "step": 69770 }, { "grad_norm": 0.7115516662597656, "learning_rate": 2.295923442473601e-05, "loss": 0.033, "step": 69780 }, { "grad_norm": 0.46183308959007263, "learning_rate": 2.294532788493519e-05, "loss": 0.0214, "step": 69790 }, { "grad_norm": 0.32460400462150574, "learning_rate": 2.293142430379262e-05, "loss": 0.0195, "step": 69800 }, { "grad_norm": 1.1800190210342407, "learning_rate": 2.291752368282879e-05, "loss": 0.0243, "step": 69810 }, { "grad_norm": 0.27175602316856384, "learning_rate": 2.290362602356385e-05, "loss": 0.0214, "step": 69820 }, { "grad_norm": 0.4865763485431671, "learning_rate": 2.288973132751765e-05, "loss": 0.0206, "step": 69830 }, { "grad_norm": 0.4707678258419037, "learning_rate": 2.287583959620965e-05, "loss": 0.0162, "step": 69840 }, { "grad_norm": 0.28250429034233093, "learning_rate": 2.2861950831159074e-05, "loss": 0.0205, "step": 69850 }, { "grad_norm": 0.38571256399154663, "learning_rate": 2.284806503388472e-05, "loss": 0.0184, "step": 69860 }, { "grad_norm": 0.24458648264408112, "learning_rate": 2.283418220590518e-05, "loss": 0.0178, "step": 69870 }, { "grad_norm": 0.4101516008377075, "learning_rate": 2.282030234873862e-05, "loss": 0.0209, "step": 69880 }, { "grad_norm": 0.29917198419570923, "learning_rate": 2.2806425463902948e-05, "loss": 0.026, "step": 69890 }, { "grad_norm": 0.4059514105319977, "learning_rate": 2.279255155291568e-05, "loss": 0.0195, "step": 69900 }, { "grad_norm": 0.42241227626800537, "learning_rate": 2.277868061729409e-05, "loss": 0.0199, "step": 69910 }, { "grad_norm": 0.438859224319458, "learning_rate": 2.276481265855504e-05, "loss": 0.0279, "step": 69920 }, { "grad_norm": 0.4200059771537781, "learning_rate": 2.2750947678215128e-05, "loss": 0.0278, "step": 69930 }, { "grad_norm": 0.415371835231781, "learning_rate": 2.2737085677790616e-05, "loss": 0.0316, "step": 69940 }, { "grad_norm": 0.17200063169002533, "learning_rate": 2.2723226658797443e-05, "loss": 0.0198, "step": 69950 }, { "grad_norm": 0.22639696300029755, "learning_rate": 2.270937062275117e-05, "loss": 0.0212, "step": 69960 }, { "grad_norm": 0.419181227684021, "learning_rate": 2.2695517571167113e-05, "loss": 0.0313, "step": 69970 }, { "grad_norm": 0.5758766531944275, "learning_rate": 2.268166750556019e-05, "loss": 0.0299, "step": 69980 }, { "grad_norm": 0.5524380207061768, "learning_rate": 2.266782042744503e-05, "loss": 0.0195, "step": 69990 }, { "grad_norm": 0.36064186692237854, "learning_rate": 2.2653976338335935e-05, "loss": 0.0253, "step": 70000 }, { "grad_norm": 0.4791671335697174, "learning_rate": 2.2640135239746895e-05, "loss": 0.0265, "step": 70010 }, { "grad_norm": 0.3567493259906769, "learning_rate": 2.2626297133191515e-05, "loss": 0.0225, "step": 70020 }, { "grad_norm": 0.3005402088165283, "learning_rate": 2.261246202018312e-05, "loss": 0.0193, "step": 70030 }, { "grad_norm": 0.609366774559021, "learning_rate": 2.2598629902234718e-05, "loss": 0.0165, "step": 70040 }, { "grad_norm": 0.7580344080924988, "learning_rate": 2.258480078085894e-05, "loss": 0.0292, "step": 70050 }, { "grad_norm": 0.4395211338996887, "learning_rate": 2.2570974657568126e-05, "loss": 0.022, "step": 70060 }, { "grad_norm": 0.41542765498161316, "learning_rate": 2.2557151533874286e-05, "loss": 0.0245, "step": 70070 }, { "grad_norm": 0.4084853529930115, "learning_rate": 2.2543331411289116e-05, "loss": 0.0208, "step": 70080 }, { "grad_norm": 0.47556135058403015, "learning_rate": 2.252951429132392e-05, "loss": 0.0227, "step": 70090 }, { "grad_norm": 0.5551711916923523, "learning_rate": 2.2515700175489762e-05, "loss": 0.0238, "step": 70100 }, { "grad_norm": 0.5691690444946289, "learning_rate": 2.2501889065297288e-05, "loss": 0.0224, "step": 70110 }, { "grad_norm": 0.34950146079063416, "learning_rate": 2.2488080962256887e-05, "loss": 0.0192, "step": 70120 }, { "grad_norm": 0.559798538684845, "learning_rate": 2.2474275867878585e-05, "loss": 0.0205, "step": 70130 }, { "grad_norm": 0.31646353006362915, "learning_rate": 2.246047378367211e-05, "loss": 0.0214, "step": 70140 }, { "grad_norm": 0.3019666075706482, "learning_rate": 2.2446674711146798e-05, "loss": 0.024, "step": 70150 }, { "grad_norm": 0.5604095458984375, "learning_rate": 2.2432878651811734e-05, "loss": 0.0256, "step": 70160 }, { "grad_norm": 0.45412003993988037, "learning_rate": 2.2419085607175594e-05, "loss": 0.0288, "step": 70170 }, { "grad_norm": 0.31909623742103577, "learning_rate": 2.240529557874678e-05, "loss": 0.0235, "step": 70180 }, { "grad_norm": 0.2821885645389557, "learning_rate": 2.2391508568033358e-05, "loss": 0.017, "step": 70190 }, { "grad_norm": 0.3074028491973877, "learning_rate": 2.237772457654307e-05, "loss": 0.0379, "step": 70200 }, { "grad_norm": 1.6132479906082153, "learning_rate": 2.236394360578327e-05, "loss": 0.0234, "step": 70210 }, { "grad_norm": 0.22112838923931122, "learning_rate": 2.2350165657261073e-05, "loss": 0.0295, "step": 70220 }, { "grad_norm": 0.45085784792900085, "learning_rate": 2.2336390732483164e-05, "loss": 0.0258, "step": 70230 }, { "grad_norm": 0.40533486008644104, "learning_rate": 2.2322618832955976e-05, "loss": 0.0249, "step": 70240 }, { "grad_norm": 0.36005839705467224, "learning_rate": 2.2308849960185578e-05, "loss": 0.0235, "step": 70250 }, { "grad_norm": 0.49823319911956787, "learning_rate": 2.229508411567774e-05, "loss": 0.0211, "step": 70260 }, { "grad_norm": 0.42998361587524414, "learning_rate": 2.2281321300937823e-05, "loss": 0.0295, "step": 70270 }, { "grad_norm": 0.5095300674438477, "learning_rate": 2.2267561517470937e-05, "loss": 0.0431, "step": 70280 }, { "grad_norm": 0.17059412598609924, "learning_rate": 2.2253804766781845e-05, "loss": 0.0226, "step": 70290 }, { "grad_norm": 0.8288387060165405, "learning_rate": 2.2240051050374933e-05, "loss": 0.0234, "step": 70300 }, { "grad_norm": 0.7554763555526733, "learning_rate": 2.22263003697543e-05, "loss": 0.0182, "step": 70310 }, { "grad_norm": 0.43211716413497925, "learning_rate": 2.2212552726423702e-05, "loss": 0.0202, "step": 70320 }, { "grad_norm": 0.3618253171443939, "learning_rate": 2.2198808121886578e-05, "loss": 0.0246, "step": 70330 }, { "grad_norm": 0.2409130185842514, "learning_rate": 2.218506655764598e-05, "loss": 0.0165, "step": 70340 }, { "grad_norm": 0.4901457130908966, "learning_rate": 2.2171328035204704e-05, "loss": 0.0178, "step": 70350 }, { "grad_norm": 0.30753985047340393, "learning_rate": 2.2157592556065133e-05, "loss": 0.0198, "step": 70360 }, { "grad_norm": 0.4351752996444702, "learning_rate": 2.2143860121729382e-05, "loss": 0.0242, "step": 70370 }, { "grad_norm": 0.355276495218277, "learning_rate": 2.2130130733699206e-05, "loss": 0.0176, "step": 70380 }, { "grad_norm": 0.2503798007965088, "learning_rate": 2.2116404393476052e-05, "loss": 0.0189, "step": 70390 }, { "grad_norm": 0.2776252329349518, "learning_rate": 2.2102681102560967e-05, "loss": 0.0225, "step": 70400 }, { "grad_norm": 0.44494131207466125, "learning_rate": 2.208896086245476e-05, "loss": 0.0347, "step": 70410 }, { "grad_norm": 0.2954006493091583, "learning_rate": 2.2075243674657808e-05, "loss": 0.0403, "step": 70420 }, { "grad_norm": 0.39196932315826416, "learning_rate": 2.206152954067024e-05, "loss": 0.0173, "step": 70430 }, { "grad_norm": 0.32896313071250916, "learning_rate": 2.2047818461991758e-05, "loss": 0.0186, "step": 70440 }, { "grad_norm": 0.3304055631160736, "learning_rate": 2.2034110440121864e-05, "loss": 0.0259, "step": 70450 }, { "grad_norm": 0.23417899012565613, "learning_rate": 2.2020405476559585e-05, "loss": 0.0197, "step": 70460 }, { "grad_norm": 0.34349340200424194, "learning_rate": 2.2006703572803722e-05, "loss": 0.0228, "step": 70470 }, { "grad_norm": 0.2867255210876465, "learning_rate": 2.1993004730352646e-05, "loss": 0.0203, "step": 70480 }, { "grad_norm": 0.3187393844127655, "learning_rate": 2.1979308950704464e-05, "loss": 0.0165, "step": 70490 }, { "grad_norm": 0.17831987142562866, "learning_rate": 2.1965616235356923e-05, "loss": 0.0232, "step": 70500 }, { "grad_norm": 0.513755202293396, "learning_rate": 2.1951926585807465e-05, "loss": 0.0188, "step": 70510 }, { "grad_norm": 0.2776276171207428, "learning_rate": 2.193824000355313e-05, "loss": 0.0181, "step": 70520 }, { "grad_norm": 0.5786274671554565, "learning_rate": 2.192455649009067e-05, "loss": 0.0176, "step": 70530 }, { "grad_norm": 0.4938475489616394, "learning_rate": 2.1910876046916523e-05, "loss": 0.0272, "step": 70540 }, { "grad_norm": 0.5000971555709839, "learning_rate": 2.1897198675526725e-05, "loss": 0.0326, "step": 70550 }, { "grad_norm": 0.5455718636512756, "learning_rate": 2.1883524377417024e-05, "loss": 0.0335, "step": 70560 }, { "grad_norm": 0.2606504261493683, "learning_rate": 2.1869853154082825e-05, "loss": 0.0122, "step": 70570 }, { "grad_norm": 0.5212387442588806, "learning_rate": 2.1856185007019215e-05, "loss": 0.0215, "step": 70580 }, { "grad_norm": 1.0109984874725342, "learning_rate": 2.1842519937720874e-05, "loss": 0.0269, "step": 70590 }, { "grad_norm": 0.5193662643432617, "learning_rate": 2.182885794768224e-05, "loss": 0.0174, "step": 70600 }, { "grad_norm": 0.8494640588760376, "learning_rate": 2.1815199038397326e-05, "loss": 0.0227, "step": 70610 }, { "grad_norm": 0.2211652249097824, "learning_rate": 2.1801543211359893e-05, "loss": 0.0252, "step": 70620 }, { "grad_norm": 0.2900364100933075, "learning_rate": 2.178789046806326e-05, "loss": 0.0185, "step": 70630 }, { "grad_norm": 0.7225984334945679, "learning_rate": 2.1774240810000547e-05, "loss": 0.0264, "step": 70640 }, { "grad_norm": 0.2716333270072937, "learning_rate": 2.17605942386644e-05, "loss": 0.0225, "step": 70650 }, { "grad_norm": 0.8002699613571167, "learning_rate": 2.174695075554723e-05, "loss": 0.0202, "step": 70660 }, { "grad_norm": 0.3489826023578644, "learning_rate": 2.173331036214103e-05, "loss": 0.0158, "step": 70670 }, { "grad_norm": 0.5148196220397949, "learning_rate": 2.171967305993753e-05, "loss": 0.0159, "step": 70680 }, { "grad_norm": 0.36261627078056335, "learning_rate": 2.1706038850428024e-05, "loss": 0.0176, "step": 70690 }, { "grad_norm": 0.4459214508533478, "learning_rate": 2.169240773510361e-05, "loss": 0.0241, "step": 70700 }, { "grad_norm": 0.24443607032299042, "learning_rate": 2.1678779715454906e-05, "loss": 0.0146, "step": 70710 }, { "grad_norm": 0.5102013349533081, "learning_rate": 2.166515479297229e-05, "loss": 0.0243, "step": 70720 }, { "grad_norm": 0.3120249807834625, "learning_rate": 2.1651532969145732e-05, "loss": 0.0188, "step": 70730 }, { "grad_norm": 0.4572942852973938, "learning_rate": 2.1637914245464903e-05, "loss": 0.021, "step": 70740 }, { "grad_norm": 0.5506894588470459, "learning_rate": 2.1624298623419133e-05, "loss": 0.0205, "step": 70750 }, { "grad_norm": 0.3184676766395569, "learning_rate": 2.1610686104497413e-05, "loss": 0.0222, "step": 70760 }, { "grad_norm": 0.581393301486969, "learning_rate": 2.1597076690188366e-05, "loss": 0.0222, "step": 70770 }, { "grad_norm": 0.45259982347488403, "learning_rate": 2.1583470381980315e-05, "loss": 0.0208, "step": 70780 }, { "grad_norm": 0.6403977870941162, "learning_rate": 2.156986718136123e-05, "loss": 0.0272, "step": 70790 }, { "grad_norm": 0.4349345862865448, "learning_rate": 2.155626708981871e-05, "loss": 0.0302, "step": 70800 }, { "grad_norm": 0.3569570481777191, "learning_rate": 2.1542670108840075e-05, "loss": 0.0195, "step": 70810 }, { "grad_norm": 0.3414336144924164, "learning_rate": 2.152907623991221e-05, "loss": 0.0235, "step": 70820 }, { "grad_norm": 0.2922308146953583, "learning_rate": 2.1515485484521802e-05, "loss": 0.0277, "step": 70830 }, { "grad_norm": 0.34904852509498596, "learning_rate": 2.150189784415506e-05, "loss": 0.0222, "step": 70840 }, { "grad_norm": 0.33320537209510803, "learning_rate": 2.148831332029794e-05, "loss": 0.0308, "step": 70850 }, { "grad_norm": 0.5306043028831482, "learning_rate": 2.147473191443599e-05, "loss": 0.0252, "step": 70860 }, { "grad_norm": 0.32608991861343384, "learning_rate": 2.1461153628054492e-05, "loss": 0.0208, "step": 70870 }, { "grad_norm": 0.30224305391311646, "learning_rate": 2.1447578462638284e-05, "loss": 0.0215, "step": 70880 }, { "grad_norm": 0.3507006764411926, "learning_rate": 2.143400641967201e-05, "loss": 0.0209, "step": 70890 }, { "grad_norm": 0.3219364881515503, "learning_rate": 2.1420437500639823e-05, "loss": 0.0161, "step": 70900 }, { "grad_norm": 0.44347941875457764, "learning_rate": 2.1406871707025646e-05, "loss": 0.0275, "step": 70910 }, { "grad_norm": 0.3145621418952942, "learning_rate": 2.1393309040312963e-05, "loss": 0.0193, "step": 70920 }, { "grad_norm": 0.23975226283073425, "learning_rate": 2.137974950198502e-05, "loss": 0.0151, "step": 70930 }, { "grad_norm": 0.8007020354270935, "learning_rate": 2.1366193093524617e-05, "loss": 0.0208, "step": 70940 }, { "grad_norm": 0.3477359414100647, "learning_rate": 2.1352639816414288e-05, "loss": 0.0169, "step": 70950 }, { "grad_norm": 0.39396947622299194, "learning_rate": 2.1339089672136202e-05, "loss": 0.0172, "step": 70960 }, { "grad_norm": 0.5942368507385254, "learning_rate": 2.1325542662172193e-05, "loss": 0.0261, "step": 70970 }, { "grad_norm": 0.4666416347026825, "learning_rate": 2.1311998788003712e-05, "loss": 0.0181, "step": 70980 }, { "grad_norm": 0.27898508310317993, "learning_rate": 2.1298458051111907e-05, "loss": 0.0218, "step": 70990 }, { "grad_norm": 0.5490089058876038, "learning_rate": 2.1284920452977602e-05, "loss": 0.021, "step": 71000 }, { "grad_norm": 0.4103148281574249, "learning_rate": 2.12713859950812e-05, "loss": 0.0167, "step": 71010 }, { "grad_norm": 0.3900330364704132, "learning_rate": 2.1257854678902832e-05, "loss": 0.0223, "step": 71020 }, { "grad_norm": 0.30955347418785095, "learning_rate": 2.124432650592227e-05, "loss": 0.0215, "step": 71030 }, { "grad_norm": 0.30853086709976196, "learning_rate": 2.1230801477618937e-05, "loss": 0.027, "step": 71040 }, { "grad_norm": 0.26327094435691833, "learning_rate": 2.121727959547189e-05, "loss": 0.0147, "step": 71050 }, { "grad_norm": 0.6106817722320557, "learning_rate": 2.1203760860959886e-05, "loss": 0.0244, "step": 71060 }, { "grad_norm": 0.4536515474319458, "learning_rate": 2.119024527556127e-05, "loss": 0.0209, "step": 71070 }, { "grad_norm": 0.3208129405975342, "learning_rate": 2.117673284075415e-05, "loss": 0.0257, "step": 71080 }, { "grad_norm": 0.40881288051605225, "learning_rate": 2.1163223558016165e-05, "loss": 0.0195, "step": 71090 }, { "grad_norm": 0.26995375752449036, "learning_rate": 2.1149717428824722e-05, "loss": 0.0239, "step": 71100 }, { "grad_norm": 0.48251762986183167, "learning_rate": 2.1136214454656784e-05, "loss": 0.0254, "step": 71110 }, { "grad_norm": 0.45472365617752075, "learning_rate": 2.1122714636989056e-05, "loss": 0.0173, "step": 71120 }, { "grad_norm": 0.3897867500782013, "learning_rate": 2.1109217977297823e-05, "loss": 0.014, "step": 71130 }, { "grad_norm": 0.37458154559135437, "learning_rate": 2.1095724477059077e-05, "loss": 0.0192, "step": 71140 }, { "grad_norm": 0.5426631569862366, "learning_rate": 2.1082234137748434e-05, "loss": 0.0204, "step": 71150 }, { "grad_norm": 0.20378315448760986, "learning_rate": 2.106874696084122e-05, "loss": 0.023, "step": 71160 }, { "grad_norm": 0.5305080413818359, "learning_rate": 2.1055262947812315e-05, "loss": 0.0249, "step": 71170 }, { "grad_norm": 0.6327890753746033, "learning_rate": 2.104178210013636e-05, "loss": 0.0266, "step": 71180 }, { "grad_norm": 0.3829801082611084, "learning_rate": 2.1028304419287554e-05, "loss": 0.0224, "step": 71190 }, { "grad_norm": 0.4124414920806885, "learning_rate": 2.101482990673982e-05, "loss": 0.047, "step": 71200 }, { "grad_norm": 0.2572578489780426, "learning_rate": 2.1001358563966707e-05, "loss": 0.0182, "step": 71210 }, { "grad_norm": 0.24468913674354553, "learning_rate": 2.098789039244144e-05, "loss": 0.0169, "step": 71220 }, { "grad_norm": 0.20269230008125305, "learning_rate": 2.097442539363685e-05, "loss": 0.0145, "step": 71230 }, { "grad_norm": 0.30112797021865845, "learning_rate": 2.0960963569025448e-05, "loss": 0.0133, "step": 71240 }, { "grad_norm": 0.24684810638427734, "learning_rate": 2.0947504920079435e-05, "loss": 0.0175, "step": 71250 }, { "grad_norm": 0.32777801156044006, "learning_rate": 2.093404944827059e-05, "loss": 0.0214, "step": 71260 }, { "grad_norm": 0.6090935468673706, "learning_rate": 2.0920597155070394e-05, "loss": 0.0213, "step": 71270 }, { "grad_norm": 0.4873449504375458, "learning_rate": 2.090714804194997e-05, "loss": 0.0233, "step": 71280 }, { "grad_norm": 0.7983710169792175, "learning_rate": 2.0893702110380114e-05, "loss": 0.0205, "step": 71290 }, { "grad_norm": 0.558174192905426, "learning_rate": 2.0880259361831216e-05, "loss": 0.0206, "step": 71300 }, { "grad_norm": 0.46387457847595215, "learning_rate": 2.0866819797773397e-05, "loss": 0.0237, "step": 71310 }, { "grad_norm": 0.3811648488044739, "learning_rate": 2.085338341967634e-05, "loss": 0.0219, "step": 71320 }, { "grad_norm": 0.7107480764389038, "learning_rate": 2.083995022900946e-05, "loss": 0.0246, "step": 71330 }, { "grad_norm": 0.41717594861984253, "learning_rate": 2.0826520227241768e-05, "loss": 0.0237, "step": 71340 }, { "grad_norm": 0.41576963663101196, "learning_rate": 2.081309341584199e-05, "loss": 0.0166, "step": 71350 }, { "grad_norm": 0.26824241876602173, "learning_rate": 2.0799669796278413e-05, "loss": 0.0174, "step": 71360 }, { "grad_norm": 0.7796371579170227, "learning_rate": 2.078624937001906e-05, "loss": 0.0204, "step": 71370 }, { "grad_norm": 0.42842766642570496, "learning_rate": 2.0772832138531544e-05, "loss": 0.0239, "step": 71380 }, { "grad_norm": 0.2703316807746887, "learning_rate": 2.0759418103283158e-05, "loss": 0.0321, "step": 71390 }, { "grad_norm": 0.4160032570362091, "learning_rate": 2.0746007265740846e-05, "loss": 0.0249, "step": 71400 }, { "grad_norm": 0.4372067451477051, "learning_rate": 2.0732599627371215e-05, "loss": 0.0172, "step": 71410 }, { "grad_norm": 0.30845946073532104, "learning_rate": 2.0719195189640468e-05, "loss": 0.0153, "step": 71420 }, { "grad_norm": 0.2784569263458252, "learning_rate": 2.0705793954014528e-05, "loss": 0.0232, "step": 71430 }, { "grad_norm": 0.3991762399673462, "learning_rate": 2.06923959219589e-05, "loss": 0.0151, "step": 71440 }, { "grad_norm": 0.3255990445613861, "learning_rate": 2.0679001094938793e-05, "loss": 0.0178, "step": 71450 }, { "grad_norm": 0.3210333585739136, "learning_rate": 2.066560947441904e-05, "loss": 0.0251, "step": 71460 }, { "grad_norm": 0.43138691782951355, "learning_rate": 2.0652221061864152e-05, "loss": 0.0166, "step": 71470 }, { "grad_norm": 0.3746892809867859, "learning_rate": 2.0638835858738226e-05, "loss": 0.0161, "step": 71480 }, { "grad_norm": 0.28161266446113586, "learning_rate": 2.0625453866505068e-05, "loss": 0.0241, "step": 71490 }, { "grad_norm": 0.36884772777557373, "learning_rate": 2.0612075086628123e-05, "loss": 0.0189, "step": 71500 }, { "grad_norm": 0.368317574262619, "learning_rate": 2.059869952057045e-05, "loss": 0.0242, "step": 71510 }, { "grad_norm": 0.3109011948108673, "learning_rate": 2.0585327169794793e-05, "loss": 0.0288, "step": 71520 }, { "grad_norm": 0.49372434616088867, "learning_rate": 2.0571958035763534e-05, "loss": 0.0183, "step": 71530 }, { "grad_norm": 0.3510999083518982, "learning_rate": 2.0558592119938715e-05, "loss": 0.0254, "step": 71540 }, { "grad_norm": 0.30957695841789246, "learning_rate": 2.054522942378198e-05, "loss": 0.0262, "step": 71550 }, { "grad_norm": 0.1994648575782776, "learning_rate": 2.053186994875469e-05, "loss": 0.0186, "step": 71560 }, { "grad_norm": 0.2839697003364563, "learning_rate": 2.0518513696317775e-05, "loss": 0.0165, "step": 71570 }, { "grad_norm": 0.3365190029144287, "learning_rate": 2.050516066793188e-05, "loss": 0.0181, "step": 71580 }, { "grad_norm": 0.26546064019203186, "learning_rate": 2.0491810865057276e-05, "loss": 0.0285, "step": 71590 }, { "grad_norm": 0.33927151560783386, "learning_rate": 2.0478464289153882e-05, "loss": 0.027, "step": 71600 }, { "grad_norm": 0.46652138233184814, "learning_rate": 2.0465120941681232e-05, "loss": 0.0298, "step": 71610 }, { "grad_norm": 0.24174994230270386, "learning_rate": 2.0451780824098566e-05, "loss": 0.0175, "step": 71620 }, { "grad_norm": 0.2094026356935501, "learning_rate": 2.043844393786471e-05, "loss": 0.0214, "step": 71630 }, { "grad_norm": 0.6914349794387817, "learning_rate": 2.0425110284438193e-05, "loss": 0.0182, "step": 71640 }, { "grad_norm": 0.3364335596561432, "learning_rate": 2.041177986527712e-05, "loss": 0.0203, "step": 71650 }, { "grad_norm": 0.5979913473129272, "learning_rate": 2.0398452681839347e-05, "loss": 0.0255, "step": 71660 }, { "grad_norm": 0.34701642394065857, "learning_rate": 2.0385128735582272e-05, "loss": 0.021, "step": 71670 }, { "grad_norm": 0.18047969043254852, "learning_rate": 2.0371808027963003e-05, "loss": 0.0229, "step": 71680 }, { "grad_norm": 0.463185578584671, "learning_rate": 2.035849056043825e-05, "loss": 0.0227, "step": 71690 }, { "grad_norm": 0.6046944856643677, "learning_rate": 2.03451763344644e-05, "loss": 0.0186, "step": 71700 }, { "grad_norm": 0.2879403829574585, "learning_rate": 2.0331865351497482e-05, "loss": 0.0249, "step": 71710 }, { "grad_norm": 0.5673694014549255, "learning_rate": 2.0318557612993185e-05, "loss": 0.021, "step": 71720 }, { "grad_norm": 0.36953771114349365, "learning_rate": 2.0305253120406782e-05, "loss": 0.0189, "step": 71730 }, { "grad_norm": 0.30941376090049744, "learning_rate": 2.029195187519325e-05, "loss": 0.0177, "step": 71740 }, { "grad_norm": 0.3403233289718628, "learning_rate": 2.0278653878807217e-05, "loss": 0.0178, "step": 71750 }, { "grad_norm": 0.5207770466804504, "learning_rate": 2.0265359132702892e-05, "loss": 0.0186, "step": 71760 }, { "grad_norm": 0.5552080273628235, "learning_rate": 2.0252067638334182e-05, "loss": 0.0149, "step": 71770 }, { "grad_norm": 0.2986392378807068, "learning_rate": 2.023877939715464e-05, "loss": 0.0217, "step": 71780 }, { "grad_norm": 0.1858637034893036, "learning_rate": 2.0225494410617448e-05, "loss": 0.0238, "step": 71790 }, { "grad_norm": 0.37482205033302307, "learning_rate": 2.02122126801754e-05, "loss": 0.0247, "step": 71800 }, { "grad_norm": 0.3568757474422455, "learning_rate": 2.0198934207281018e-05, "loss": 0.0473, "step": 71810 }, { "grad_norm": 0.4641575515270233, "learning_rate": 2.0185658993386357e-05, "loss": 0.0232, "step": 71820 }, { "grad_norm": 0.6999392509460449, "learning_rate": 2.0172387039943226e-05, "loss": 0.034, "step": 71830 }, { "grad_norm": 0.30552470684051514, "learning_rate": 2.0159118348402968e-05, "loss": 0.0177, "step": 71840 }, { "grad_norm": 0.45400160551071167, "learning_rate": 2.0145852920216697e-05, "loss": 0.0144, "step": 71850 }, { "grad_norm": 0.3565146327018738, "learning_rate": 2.0132590756835045e-05, "loss": 0.0229, "step": 71860 }, { "grad_norm": 0.5626718997955322, "learning_rate": 2.0119331859708386e-05, "loss": 0.0269, "step": 71870 }, { "grad_norm": 0.3408813178539276, "learning_rate": 2.0106076230286654e-05, "loss": 0.0241, "step": 71880 }, { "grad_norm": 0.5942689776420593, "learning_rate": 2.0092823870019488e-05, "loss": 0.0202, "step": 71890 }, { "grad_norm": 0.49778059124946594, "learning_rate": 2.0079574780356115e-05, "loss": 0.0283, "step": 71900 }, { "grad_norm": 0.5861255526542664, "learning_rate": 2.0066328962745486e-05, "loss": 0.0122, "step": 71910 }, { "grad_norm": 0.16734179854393005, "learning_rate": 2.00530864186361e-05, "loss": 0.024, "step": 71920 }, { "grad_norm": 0.4890558421611786, "learning_rate": 2.003984714947618e-05, "loss": 0.0287, "step": 71930 }, { "grad_norm": 0.38118723034858704, "learning_rate": 2.0026611156713515e-05, "loss": 0.0197, "step": 71940 }, { "grad_norm": 0.26039695739746094, "learning_rate": 2.0013378441795584e-05, "loss": 0.0183, "step": 71950 }, { "grad_norm": 0.6922745108604431, "learning_rate": 2.0000149006169506e-05, "loss": 0.0283, "step": 71960 }, { "grad_norm": 0.3007993698120117, "learning_rate": 1.9986922851282048e-05, "loss": 0.0276, "step": 71970 }, { "grad_norm": 0.5891326665878296, "learning_rate": 1.997369997857956e-05, "loss": 0.018, "step": 71980 }, { "grad_norm": 0.20123045146465302, "learning_rate": 1.9960480389508103e-05, "loss": 0.0153, "step": 71990 }, { "grad_norm": 0.30427056550979614, "learning_rate": 1.9947264085513363e-05, "loss": 0.0409, "step": 72000 }, { "grad_norm": 0.5962011814117432, "learning_rate": 1.993405106804062e-05, "loss": 0.0156, "step": 72010 }, { "grad_norm": 0.7151709198951721, "learning_rate": 1.9920841338534867e-05, "loss": 0.022, "step": 72020 }, { "grad_norm": 0.27817219495773315, "learning_rate": 1.9907634898440646e-05, "loss": 0.0201, "step": 72030 }, { "grad_norm": 0.5610421895980835, "learning_rate": 1.989443174920227e-05, "loss": 0.0259, "step": 72040 }, { "grad_norm": 0.37357282638549805, "learning_rate": 1.988123189226355e-05, "loss": 0.0249, "step": 72050 }, { "grad_norm": 0.2982602119445801, "learning_rate": 1.986803532906805e-05, "loss": 0.0237, "step": 72060 }, { "grad_norm": 0.47554126381874084, "learning_rate": 1.985484206105889e-05, "loss": 0.0211, "step": 72070 }, { "grad_norm": 0.383675217628479, "learning_rate": 1.9841652089678887e-05, "loss": 0.0171, "step": 72080 }, { "grad_norm": 0.31205692887306213, "learning_rate": 1.9828465416370435e-05, "loss": 0.0257, "step": 72090 }, { "grad_norm": 0.33654695749282837, "learning_rate": 1.9815282042575685e-05, "loss": 0.0158, "step": 72100 }, { "grad_norm": 0.46852588653564453, "learning_rate": 1.9802101969736285e-05, "loss": 0.0238, "step": 72110 }, { "grad_norm": 0.4735877215862274, "learning_rate": 1.9788925199293633e-05, "loss": 0.0152, "step": 72120 }, { "grad_norm": 0.2679841220378876, "learning_rate": 1.977575173268868e-05, "loss": 0.0226, "step": 72130 }, { "grad_norm": 0.4393831193447113, "learning_rate": 1.9762581571362094e-05, "loss": 0.0214, "step": 72140 }, { "grad_norm": 0.3369457423686981, "learning_rate": 1.9749414716754112e-05, "loss": 0.0153, "step": 72150 }, { "grad_norm": 0.5140268802642822, "learning_rate": 1.973625117030465e-05, "loss": 0.0226, "step": 72160 }, { "grad_norm": 0.4005593955516815, "learning_rate": 1.972309093345326e-05, "loss": 0.0179, "step": 72170 }, { "grad_norm": 0.13721893727779388, "learning_rate": 1.970993400763915e-05, "loss": 0.0207, "step": 72180 }, { "grad_norm": 0.4955657124519348, "learning_rate": 1.969678039430109e-05, "loss": 0.0206, "step": 72190 }, { "grad_norm": 0.5615071654319763, "learning_rate": 1.9683630094877564e-05, "loss": 0.0291, "step": 72200 }, { "grad_norm": 0.45188140869140625, "learning_rate": 1.9670483110806696e-05, "loss": 0.0198, "step": 72210 }, { "grad_norm": 0.31492069363594055, "learning_rate": 1.9657339443526172e-05, "loss": 0.0148, "step": 72220 }, { "grad_norm": 0.5828408598899841, "learning_rate": 1.964419909447339e-05, "loss": 0.0153, "step": 72230 }, { "grad_norm": 0.3155880868434906, "learning_rate": 1.963106206508536e-05, "loss": 0.0138, "step": 72240 }, { "grad_norm": 0.7268244624137878, "learning_rate": 1.9617928356798738e-05, "loss": 0.0217, "step": 72250 }, { "grad_norm": 0.22677382826805115, "learning_rate": 1.960479797104977e-05, "loss": 0.0224, "step": 72260 }, { "grad_norm": 0.40992605686187744, "learning_rate": 1.9591670909274418e-05, "loss": 0.0202, "step": 72270 }, { "grad_norm": 0.4544620215892792, "learning_rate": 1.9578547172908184e-05, "loss": 0.0195, "step": 72280 }, { "grad_norm": 0.4433149993419647, "learning_rate": 1.9565426763386325e-05, "loss": 0.0185, "step": 72290 }, { "grad_norm": 0.2519432306289673, "learning_rate": 1.9552309682143622e-05, "loss": 0.0162, "step": 72300 }, { "grad_norm": 0.7774105668067932, "learning_rate": 1.9539195930614574e-05, "loss": 0.022, "step": 72310 }, { "grad_norm": 0.2671387791633606, "learning_rate": 1.952608551023324e-05, "loss": 0.0169, "step": 72320 }, { "grad_norm": 0.26749610900878906, "learning_rate": 1.9512978422433397e-05, "loss": 0.013, "step": 72330 }, { "grad_norm": 0.4464133083820343, "learning_rate": 1.9499874668648376e-05, "loss": 0.0163, "step": 72340 }, { "grad_norm": 0.2931791841983795, "learning_rate": 1.9486774250311213e-05, "loss": 0.025, "step": 72350 }, { "grad_norm": 0.3201873302459717, "learning_rate": 1.9473677168854537e-05, "loss": 0.0267, "step": 72360 }, { "grad_norm": 0.43408897519111633, "learning_rate": 1.946058342571065e-05, "loss": 0.0317, "step": 72370 }, { "grad_norm": 0.20477429032325745, "learning_rate": 1.9447493022311424e-05, "loss": 0.0171, "step": 72380 }, { "grad_norm": 0.34679746627807617, "learning_rate": 1.9434405960088442e-05, "loss": 0.0258, "step": 72390 }, { "grad_norm": 0.4905996024608612, "learning_rate": 1.9421322240472857e-05, "loss": 0.0197, "step": 72400 }, { "grad_norm": 0.4072250723838806, "learning_rate": 1.9408241864895494e-05, "loss": 0.0227, "step": 72410 }, { "grad_norm": 0.8482547402381897, "learning_rate": 1.939516483478681e-05, "loss": 0.0269, "step": 72420 }, { "grad_norm": 0.38608160614967346, "learning_rate": 1.9382091151576902e-05, "loss": 0.0401, "step": 72430 }, { "grad_norm": 0.4939761757850647, "learning_rate": 1.9369020816695454e-05, "loss": 0.0216, "step": 72440 }, { "grad_norm": 0.390544056892395, "learning_rate": 1.935595383157184e-05, "loss": 0.0207, "step": 72450 }, { "grad_norm": 0.3383190631866455, "learning_rate": 1.9342890197635057e-05, "loss": 0.0189, "step": 72460 }, { "grad_norm": 0.2683970332145691, "learning_rate": 1.9329829916313685e-05, "loss": 0.0185, "step": 72470 }, { "grad_norm": 0.27749112248420715, "learning_rate": 1.9316772989036013e-05, "loss": 0.0175, "step": 72480 }, { "grad_norm": 0.3815130889415741, "learning_rate": 1.930371941722991e-05, "loss": 0.0192, "step": 72490 }, { "grad_norm": 0.2758406400680542, "learning_rate": 1.929066920232292e-05, "loss": 0.0114, "step": 72500 }, { "grad_norm": 0.932163655757904, "learning_rate": 1.9277622345742152e-05, "loss": 0.0269, "step": 72510 }, { "grad_norm": 0.17013947665691376, "learning_rate": 1.926457884891444e-05, "loss": 0.0171, "step": 72520 }, { "grad_norm": 0.5962952971458435, "learning_rate": 1.9251538713266147e-05, "loss": 0.0208, "step": 72530 }, { "grad_norm": 0.4145893454551697, "learning_rate": 1.9238501940223358e-05, "loss": 0.0238, "step": 72540 }, { "grad_norm": 0.33607521653175354, "learning_rate": 1.922546853121175e-05, "loss": 0.0215, "step": 72550 }, { "grad_norm": 0.49202898144721985, "learning_rate": 1.9212438487656647e-05, "loss": 0.0291, "step": 72560 }, { "grad_norm": 0.5832868814468384, "learning_rate": 1.9199411810982965e-05, "loss": 0.0383, "step": 72570 }, { "grad_norm": 0.3658351004123688, "learning_rate": 1.918638850261532e-05, "loss": 0.0318, "step": 72580 }, { "grad_norm": 0.1718740165233612, "learning_rate": 1.9173368563977885e-05, "loss": 0.0129, "step": 72590 }, { "grad_norm": 0.2034463733434677, "learning_rate": 1.9160351996494513e-05, "loss": 0.016, "step": 72600 }, { "grad_norm": 0.33894065022468567, "learning_rate": 1.9147338801588677e-05, "loss": 0.0157, "step": 72610 }, { "grad_norm": 0.38012680411338806, "learning_rate": 1.913432898068351e-05, "loss": 0.0218, "step": 72620 }, { "grad_norm": 0.3530241549015045, "learning_rate": 1.912132253520169e-05, "loss": 0.0164, "step": 72630 }, { "grad_norm": 0.2781125009059906, "learning_rate": 1.910831946656564e-05, "loss": 0.0224, "step": 72640 }, { "grad_norm": 0.3346594274044037, "learning_rate": 1.909531977619731e-05, "loss": 0.0241, "step": 72650 }, { "grad_norm": 0.38223978877067566, "learning_rate": 1.908232346551834e-05, "loss": 0.0174, "step": 72660 }, { "grad_norm": 0.3398454189300537, "learning_rate": 1.9069330535949993e-05, "loss": 0.0129, "step": 72670 }, { "grad_norm": 0.2452649623155594, "learning_rate": 1.9056340988913167e-05, "loss": 0.0188, "step": 72680 }, { "grad_norm": 0.5849760174751282, "learning_rate": 1.9043354825828357e-05, "loss": 0.0216, "step": 72690 }, { "grad_norm": 0.4670294523239136, "learning_rate": 1.9030372048115724e-05, "loss": 0.0323, "step": 72700 }, { "grad_norm": 0.8558078408241272, "learning_rate": 1.9017392657195056e-05, "loss": 0.0344, "step": 72710 }, { "grad_norm": 0.30160436034202576, "learning_rate": 1.9004416654485724e-05, "loss": 0.0157, "step": 72720 }, { "grad_norm": 0.5589337944984436, "learning_rate": 1.8991444041406785e-05, "loss": 0.0275, "step": 72730 }, { "grad_norm": 0.3775467574596405, "learning_rate": 1.8978474819376912e-05, "loss": 0.0178, "step": 72740 }, { "grad_norm": 0.6267972588539124, "learning_rate": 1.8965508989814406e-05, "loss": 0.0237, "step": 72750 }, { "grad_norm": 0.6529532670974731, "learning_rate": 1.8952546554137158e-05, "loss": 0.0407, "step": 72760 }, { "grad_norm": 0.3824616074562073, "learning_rate": 1.8939587513762757e-05, "loss": 0.0179, "step": 72770 }, { "grad_norm": 0.9678862690925598, "learning_rate": 1.892663187010835e-05, "loss": 0.0327, "step": 72780 }, { "grad_norm": 0.4570462703704834, "learning_rate": 1.891367962459077e-05, "loss": 0.022, "step": 72790 }, { "grad_norm": 0.31746867299079895, "learning_rate": 1.890073077862644e-05, "loss": 0.0174, "step": 72800 }, { "grad_norm": 0.5297000408172607, "learning_rate": 1.888778533363146e-05, "loss": 0.0298, "step": 72810 }, { "grad_norm": 0.3485260009765625, "learning_rate": 1.8874843291021478e-05, "loss": 0.0143, "step": 72820 }, { "grad_norm": 0.21278858184814453, "learning_rate": 1.8861904652211853e-05, "loss": 0.0181, "step": 72830 }, { "grad_norm": 0.43655329942703247, "learning_rate": 1.8848969418617507e-05, "loss": 0.0139, "step": 72840 }, { "grad_norm": 0.21135298907756805, "learning_rate": 1.8836037591653044e-05, "loss": 0.014, "step": 72850 }, { "grad_norm": 0.5705932974815369, "learning_rate": 1.882310917273262e-05, "loss": 0.025, "step": 72860 }, { "grad_norm": 0.6740062236785889, "learning_rate": 1.881018416327014e-05, "loss": 0.0314, "step": 72870 }, { "grad_norm": 0.21704745292663574, "learning_rate": 1.8797262564679014e-05, "loss": 0.0171, "step": 72880 }, { "grad_norm": 0.2941944897174835, "learning_rate": 1.8784344378372355e-05, "loss": 0.0222, "step": 72890 }, { "grad_norm": 0.5782750844955444, "learning_rate": 1.8771429605762843e-05, "loss": 0.022, "step": 72900 }, { "grad_norm": 0.3182998299598694, "learning_rate": 1.8758518248262835e-05, "loss": 0.0222, "step": 72910 }, { "grad_norm": 0.33224794268608093, "learning_rate": 1.8745610307284305e-05, "loss": 0.0176, "step": 72920 }, { "grad_norm": 0.2609078586101532, "learning_rate": 1.8732705784238862e-05, "loss": 0.0149, "step": 72930 }, { "grad_norm": 0.5771978497505188, "learning_rate": 1.8719804680537682e-05, "loss": 0.0163, "step": 72940 }, { "grad_norm": 0.24400274455547333, "learning_rate": 1.870690699759164e-05, "loss": 0.0186, "step": 72950 }, { "grad_norm": 0.43388691544532776, "learning_rate": 1.8694012736811213e-05, "loss": 0.0189, "step": 72960 }, { "grad_norm": 0.17605765163898468, "learning_rate": 1.8681121899606463e-05, "loss": 0.0122, "step": 72970 }, { "grad_norm": 0.3077296018600464, "learning_rate": 1.8668234487387144e-05, "loss": 0.0187, "step": 72980 }, { "grad_norm": 0.7306345701217651, "learning_rate": 1.8655350501562592e-05, "loss": 0.0185, "step": 72990 }, { "grad_norm": 0.2738783061504364, "learning_rate": 1.8642469943541796e-05, "loss": 0.0221, "step": 73000 }, { "grad_norm": 0.8797098398208618, "learning_rate": 1.8629592814733328e-05, "loss": 0.0202, "step": 73010 }, { "grad_norm": 0.2129351943731308, "learning_rate": 1.8616719116545438e-05, "loss": 0.0154, "step": 73020 }, { "grad_norm": 0.37167221307754517, "learning_rate": 1.8603848850385952e-05, "loss": 0.0165, "step": 73030 }, { "grad_norm": 0.2271718531847, "learning_rate": 1.859098201766236e-05, "loss": 0.0144, "step": 73040 }, { "grad_norm": 0.5040579438209534, "learning_rate": 1.8578118619781726e-05, "loss": 0.0241, "step": 73050 }, { "grad_norm": 0.5155370831489563, "learning_rate": 1.856525865815083e-05, "loss": 0.0223, "step": 73060 }, { "grad_norm": 0.2583709955215454, "learning_rate": 1.8552402134175966e-05, "loss": 0.0189, "step": 73070 }, { "grad_norm": 0.35383617877960205, "learning_rate": 1.8539549049263143e-05, "loss": 0.022, "step": 73080 }, { "grad_norm": 0.18915797770023346, "learning_rate": 1.8526699404817917e-05, "loss": 0.0187, "step": 73090 }, { "grad_norm": 0.3389815390110016, "learning_rate": 1.8513853202245546e-05, "loss": 0.0114, "step": 73100 }, { "grad_norm": 0.5809349417686462, "learning_rate": 1.850101044295081e-05, "loss": 0.0289, "step": 73110 }, { "grad_norm": 0.38596585392951965, "learning_rate": 1.848817112833825e-05, "loss": 0.0286, "step": 73120 }, { "grad_norm": 0.3001735508441925, "learning_rate": 1.8475335259811905e-05, "loss": 0.013, "step": 73130 }, { "grad_norm": 0.7481369972229004, "learning_rate": 1.8462502838775508e-05, "loss": 0.0183, "step": 73140 }, { "grad_norm": 0.25749069452285767, "learning_rate": 1.8449673866632378e-05, "loss": 0.0159, "step": 73150 }, { "grad_norm": 0.2444990873336792, "learning_rate": 1.8436848344785473e-05, "loss": 0.0463, "step": 73160 }, { "grad_norm": 0.27371615171432495, "learning_rate": 1.8424026274637378e-05, "loss": 0.0242, "step": 73170 }, { "grad_norm": 0.5556023120880127, "learning_rate": 1.8411207657590312e-05, "loss": 0.0277, "step": 73180 }, { "grad_norm": 0.47490766644477844, "learning_rate": 1.839839249504607e-05, "loss": 0.0179, "step": 73190 }, { "grad_norm": 0.12294953316450119, "learning_rate": 1.8385580788406108e-05, "loss": 0.0245, "step": 73200 }, { "grad_norm": 0.40652695298194885, "learning_rate": 1.837277253907152e-05, "loss": 0.0261, "step": 73210 }, { "grad_norm": 0.31974196434020996, "learning_rate": 1.8359967748442957e-05, "loss": 0.0211, "step": 73220 }, { "grad_norm": 0.3912845551967621, "learning_rate": 1.8347166417920774e-05, "loss": 0.0193, "step": 73230 }, { "grad_norm": 0.2611333727836609, "learning_rate": 1.833436854890484e-05, "loss": 0.0185, "step": 73240 }, { "grad_norm": 0.5114876627922058, "learning_rate": 1.8321574142794796e-05, "loss": 0.0223, "step": 73250 }, { "grad_norm": 0.6288888454437256, "learning_rate": 1.8308783200989748e-05, "loss": 0.0228, "step": 73260 }, { "grad_norm": 0.20906126499176025, "learning_rate": 1.829599572488855e-05, "loss": 0.0132, "step": 73270 }, { "grad_norm": 0.18209543824195862, "learning_rate": 1.828321171588957e-05, "loss": 0.0275, "step": 73280 }, { "grad_norm": 0.5010460615158081, "learning_rate": 1.827043117539089e-05, "loss": 0.0221, "step": 73290 }, { "grad_norm": 0.35908815264701843, "learning_rate": 1.8257654104790116e-05, "loss": 0.024, "step": 73300 }, { "grad_norm": 0.4846605062484741, "learning_rate": 1.82448805054846e-05, "loss": 0.0231, "step": 73310 }, { "grad_norm": 0.3169860541820526, "learning_rate": 1.823211037887119e-05, "loss": 0.0195, "step": 73320 }, { "grad_norm": 0.2968665063381195, "learning_rate": 1.821934372634645e-05, "loss": 0.0241, "step": 73330 }, { "grad_norm": 0.4296887218952179, "learning_rate": 1.8206580549306472e-05, "loss": 0.0208, "step": 73340 }, { "grad_norm": 0.6513416171073914, "learning_rate": 1.819382084914707e-05, "loss": 0.0211, "step": 73350 }, { "grad_norm": 0.6500809192657471, "learning_rate": 1.8181064627263577e-05, "loss": 0.039, "step": 73360 }, { "grad_norm": 0.24919186532497406, "learning_rate": 1.8168311885051022e-05, "loss": 0.0279, "step": 73370 }, { "grad_norm": 0.1504155397415161, "learning_rate": 1.815556262390402e-05, "loss": 0.0248, "step": 73380 }, { "grad_norm": 0.3026255667209625, "learning_rate": 1.8142816845216826e-05, "loss": 0.0154, "step": 73390 }, { "grad_norm": 0.8584738969802856, "learning_rate": 1.8130074550383276e-05, "loss": 0.0179, "step": 73400 }, { "grad_norm": 0.3687184453010559, "learning_rate": 1.8117335740796848e-05, "loss": 0.0152, "step": 73410 }, { "grad_norm": 0.4176861345767975, "learning_rate": 1.810460041785067e-05, "loss": 0.0296, "step": 73420 }, { "grad_norm": 0.22732792794704437, "learning_rate": 1.809186858293742e-05, "loss": 0.0206, "step": 73430 }, { "grad_norm": 0.6165385246276855, "learning_rate": 1.8079140237449448e-05, "loss": 0.0273, "step": 73440 }, { "grad_norm": 0.3132425844669342, "learning_rate": 1.8066415382778707e-05, "loss": 0.0157, "step": 73450 }, { "grad_norm": 0.4875866174697876, "learning_rate": 1.8053694020316786e-05, "loss": 0.0191, "step": 73460 }, { "grad_norm": 0.6621169447898865, "learning_rate": 1.8040976151454837e-05, "loss": 0.0237, "step": 73470 }, { "grad_norm": 0.45347869396209717, "learning_rate": 1.8028261777583704e-05, "loss": 0.0215, "step": 73480 }, { "grad_norm": 0.3861374258995056, "learning_rate": 1.8015550900093758e-05, "loss": 0.017, "step": 73490 }, { "grad_norm": 0.534959077835083, "learning_rate": 1.800284352037512e-05, "loss": 0.0204, "step": 73500 }, { "grad_norm": 0.4622260630130768, "learning_rate": 1.799013963981738e-05, "loss": 0.0272, "step": 73510 }, { "grad_norm": 0.27648961544036865, "learning_rate": 1.797743925980987e-05, "loss": 0.0122, "step": 73520 }, { "grad_norm": 0.3573148548603058, "learning_rate": 1.7964742381741435e-05, "loss": 0.0154, "step": 73530 }, { "grad_norm": 0.42993083596229553, "learning_rate": 1.7952049007000627e-05, "loss": 0.0205, "step": 73540 }, { "grad_norm": 0.3495863378047943, "learning_rate": 1.7939359136975535e-05, "loss": 0.0235, "step": 73550 }, { "grad_norm": 0.23736687004566193, "learning_rate": 1.7926672773053932e-05, "loss": 0.0336, "step": 73560 }, { "grad_norm": 0.39534425735473633, "learning_rate": 1.791398991662317e-05, "loss": 0.019, "step": 73570 }, { "grad_norm": 0.5140237212181091, "learning_rate": 1.7901310569070244e-05, "loss": 0.0261, "step": 73580 }, { "grad_norm": 0.46522751450538635, "learning_rate": 1.7888634731781724e-05, "loss": 0.0202, "step": 73590 }, { "grad_norm": 0.5460008382797241, "learning_rate": 1.7875962406143844e-05, "loss": 0.0156, "step": 73600 }, { "grad_norm": 0.19096387922763824, "learning_rate": 1.7863293593542403e-05, "loss": 0.0171, "step": 73610 }, { "grad_norm": 0.344078928232193, "learning_rate": 1.7850628295362853e-05, "loss": 0.0236, "step": 73620 }, { "grad_norm": 0.29534152150154114, "learning_rate": 1.7837966512990263e-05, "loss": 0.0148, "step": 73630 }, { "grad_norm": 0.32152602076530457, "learning_rate": 1.782530824780931e-05, "loss": 0.0241, "step": 73640 }, { "grad_norm": 0.31178808212280273, "learning_rate": 1.7812653501204263e-05, "loss": 0.0242, "step": 73650 }, { "grad_norm": 0.4965670108795166, "learning_rate": 1.780000227455903e-05, "loss": 0.0232, "step": 73660 }, { "grad_norm": 0.21314401924610138, "learning_rate": 1.778735456925716e-05, "loss": 0.0118, "step": 73670 }, { "grad_norm": 0.24681365489959717, "learning_rate": 1.7774710386681737e-05, "loss": 0.0206, "step": 73680 }, { "grad_norm": 0.23184731602668762, "learning_rate": 1.7762069728215545e-05, "loss": 0.0265, "step": 73690 }, { "grad_norm": 0.43689751625061035, "learning_rate": 1.7749432595240932e-05, "loss": 0.0185, "step": 73700 }, { "grad_norm": 0.2441178560256958, "learning_rate": 1.773679898913991e-05, "loss": 0.0323, "step": 73710 }, { "grad_norm": 0.39545294642448425, "learning_rate": 1.7724168911294015e-05, "loss": 0.019, "step": 73720 }, { "grad_norm": 0.9425083994865417, "learning_rate": 1.7711542363084504e-05, "loss": 0.0225, "step": 73730 }, { "grad_norm": 0.413830429315567, "learning_rate": 1.7698919345892158e-05, "loss": 0.0214, "step": 73740 }, { "grad_norm": 0.2163299173116684, "learning_rate": 1.768629986109743e-05, "loss": 0.0177, "step": 73750 }, { "grad_norm": 0.27123111486434937, "learning_rate": 1.7673683910080357e-05, "loss": 0.0333, "step": 73760 }, { "grad_norm": 0.23494933545589447, "learning_rate": 1.7661071494220638e-05, "loss": 0.0198, "step": 73770 }, { "grad_norm": 0.634262204170227, "learning_rate": 1.7648462614897498e-05, "loss": 0.0208, "step": 73780 }, { "grad_norm": 0.3443475663661957, "learning_rate": 1.7635857273489863e-05, "loss": 0.0256, "step": 73790 }, { "grad_norm": 0.3070146441459656, "learning_rate": 1.7623255471376198e-05, "loss": 0.0236, "step": 73800 }, { "grad_norm": 0.37944889068603516, "learning_rate": 1.7610657209934646e-05, "loss": 0.0199, "step": 73810 }, { "grad_norm": 0.2698383927345276, "learning_rate": 1.7598062490542915e-05, "loss": 0.0193, "step": 73820 }, { "grad_norm": 0.2939474880695343, "learning_rate": 1.7585471314578377e-05, "loss": 0.0142, "step": 73830 }, { "grad_norm": 0.26197579503059387, "learning_rate": 1.7572883683417945e-05, "loss": 0.0219, "step": 73840 }, { "grad_norm": 0.5926030874252319, "learning_rate": 1.7560299598438212e-05, "loss": 0.0178, "step": 73850 }, { "grad_norm": 0.5691366195678711, "learning_rate": 1.7547719061015334e-05, "loss": 0.0269, "step": 73860 }, { "grad_norm": 0.27060991525650024, "learning_rate": 1.75351420725251e-05, "loss": 0.0195, "step": 73870 }, { "grad_norm": 0.42937880754470825, "learning_rate": 1.7522568634342923e-05, "loss": 0.0279, "step": 73880 }, { "grad_norm": 1.0906001329421997, "learning_rate": 1.7509998747843825e-05, "loss": 0.0162, "step": 73890 }, { "grad_norm": 0.295718789100647, "learning_rate": 1.7497432414402403e-05, "loss": 0.018, "step": 73900 }, { "grad_norm": 0.35404953360557556, "learning_rate": 1.7484869635392903e-05, "loss": 0.0212, "step": 73910 }, { "grad_norm": 0.31389880180358887, "learning_rate": 1.747231041218919e-05, "loss": 0.0206, "step": 73920 }, { "grad_norm": 0.3278457820415497, "learning_rate": 1.7459754746164692e-05, "loss": 0.0151, "step": 73930 }, { "grad_norm": 0.1884077489376068, "learning_rate": 1.744720263869249e-05, "loss": 0.017, "step": 73940 }, { "grad_norm": 0.5090070366859436, "learning_rate": 1.7434654091145263e-05, "loss": 0.0325, "step": 73950 }, { "grad_norm": 0.257027804851532, "learning_rate": 1.742210910489532e-05, "loss": 0.026, "step": 73960 }, { "grad_norm": 0.24900759756565094, "learning_rate": 1.7409567681314532e-05, "loss": 0.0192, "step": 73970 }, { "grad_norm": 0.42310285568237305, "learning_rate": 1.7397029821774434e-05, "loss": 0.0159, "step": 73980 }, { "grad_norm": 0.4193348288536072, "learning_rate": 1.7384495527646126e-05, "loss": 0.0203, "step": 73990 }, { "grad_norm": 0.26417386531829834, "learning_rate": 1.737196480030035e-05, "loss": 0.0234, "step": 74000 }, { "grad_norm": 0.35009071230888367, "learning_rate": 1.7359437641107447e-05, "loss": 0.0196, "step": 74010 }, { "grad_norm": 0.20936141908168793, "learning_rate": 1.7346914051437387e-05, "loss": 0.0198, "step": 74020 }, { "grad_norm": 1.3474088907241821, "learning_rate": 1.733439403265969e-05, "loss": 0.0232, "step": 74030 }, { "grad_norm": 0.37989604473114014, "learning_rate": 1.732187758614357e-05, "loss": 0.0178, "step": 74040 }, { "grad_norm": 0.29524773359298706, "learning_rate": 1.730936471325777e-05, "loss": 0.0174, "step": 74050 }, { "grad_norm": 0.3763261139392853, "learning_rate": 1.7296855415370718e-05, "loss": 0.0169, "step": 74060 }, { "grad_norm": 0.23553290963172913, "learning_rate": 1.728434969385035e-05, "loss": 0.0188, "step": 74070 }, { "grad_norm": 0.5379644632339478, "learning_rate": 1.7271847550064346e-05, "loss": 0.0141, "step": 74080 }, { "grad_norm": 0.24971328675746918, "learning_rate": 1.725934898537987e-05, "loss": 0.0254, "step": 74090 }, { "grad_norm": 0.21124164760112762, "learning_rate": 1.7246854001163783e-05, "loss": 0.0256, "step": 74100 }, { "grad_norm": 0.3979785144329071, "learning_rate": 1.7234362598782476e-05, "loss": 0.0156, "step": 74110 }, { "grad_norm": 0.5979927182197571, "learning_rate": 1.722187477960202e-05, "loss": 0.0302, "step": 74120 }, { "grad_norm": 0.3711460530757904, "learning_rate": 1.7209390544988057e-05, "loss": 0.0207, "step": 74130 }, { "grad_norm": 0.36888694763183594, "learning_rate": 1.7196909896305856e-05, "loss": 0.0166, "step": 74140 }, { "grad_norm": 0.5426792502403259, "learning_rate": 1.718443283492026e-05, "loss": 0.0236, "step": 74150 }, { "grad_norm": 0.35508257150650024, "learning_rate": 1.7171959362195746e-05, "loss": 0.0185, "step": 74160 }, { "grad_norm": 0.25496935844421387, "learning_rate": 1.715948947949642e-05, "loss": 0.013, "step": 74170 }, { "grad_norm": 0.24748440086841583, "learning_rate": 1.714702318818593e-05, "loss": 0.0141, "step": 74180 }, { "grad_norm": 0.3227177560329437, "learning_rate": 1.7134560489627592e-05, "loss": 0.0184, "step": 74190 }, { "grad_norm": 0.5186265110969543, "learning_rate": 1.712210138518431e-05, "loss": 0.0198, "step": 74200 }, { "grad_norm": 0.44305214285850525, "learning_rate": 1.71096458762186e-05, "loss": 0.0174, "step": 74210 }, { "grad_norm": 0.2683337330818176, "learning_rate": 1.7097193964092552e-05, "loss": 0.0194, "step": 74220 }, { "grad_norm": 0.4189726710319519, "learning_rate": 1.708474565016792e-05, "loss": 0.025, "step": 74230 }, { "grad_norm": 0.293203741312027, "learning_rate": 1.7072300935806e-05, "loss": 0.0174, "step": 74240 }, { "grad_norm": 0.3526265621185303, "learning_rate": 1.705985982236776e-05, "loss": 0.019, "step": 74250 }, { "grad_norm": 0.32985755801200867, "learning_rate": 1.7047422311213697e-05, "loss": 0.0224, "step": 74260 }, { "grad_norm": 0.4969634711742401, "learning_rate": 1.703498840370401e-05, "loss": 0.0286, "step": 74270 }, { "grad_norm": 0.3873468041419983, "learning_rate": 1.702255810119842e-05, "loss": 0.0215, "step": 74280 }, { "grad_norm": 0.3277508020401001, "learning_rate": 1.7010131405056305e-05, "loss": 0.0244, "step": 74290 }, { "grad_norm": 0.3175014555454254, "learning_rate": 1.6997708316636613e-05, "loss": 0.0141, "step": 74300 }, { "grad_norm": 0.38212859630584717, "learning_rate": 1.6985288837297932e-05, "loss": 0.0258, "step": 74310 }, { "grad_norm": 0.4423624873161316, "learning_rate": 1.6972872968398394e-05, "loss": 0.0234, "step": 74320 }, { "grad_norm": 0.5000932216644287, "learning_rate": 1.6960460711295845e-05, "loss": 0.025, "step": 74330 }, { "grad_norm": 0.4509434401988983, "learning_rate": 1.6948052067347623e-05, "loss": 0.0164, "step": 74340 }, { "grad_norm": 0.37368226051330566, "learning_rate": 1.693564703791075e-05, "loss": 0.0232, "step": 74350 }, { "grad_norm": 0.36483439803123474, "learning_rate": 1.6923245624341793e-05, "loss": 0.0415, "step": 74360 }, { "grad_norm": 0.26304128766059875, "learning_rate": 1.6910847827996962e-05, "loss": 0.0173, "step": 74370 }, { "grad_norm": 0.6253136992454529, "learning_rate": 1.689845365023206e-05, "loss": 0.0197, "step": 74380 }, { "grad_norm": 0.3677963614463806, "learning_rate": 1.6886063092402505e-05, "loss": 0.0148, "step": 74390 }, { "grad_norm": 0.2304171621799469, "learning_rate": 1.6873676155863317e-05, "loss": 0.0175, "step": 74400 }, { "grad_norm": 0.3769577443599701, "learning_rate": 1.6861292841969085e-05, "loss": 0.0163, "step": 74410 }, { "grad_norm": 0.33908283710479736, "learning_rate": 1.684891315207406e-05, "loss": 0.0152, "step": 74420 }, { "grad_norm": 0.2260962724685669, "learning_rate": 1.6836537087532034e-05, "loss": 0.0142, "step": 74430 }, { "grad_norm": 0.15976278483867645, "learning_rate": 1.6824164649696465e-05, "loss": 0.0183, "step": 74440 }, { "grad_norm": 0.4156275689601898, "learning_rate": 1.681179583992034e-05, "loss": 0.0259, "step": 74450 }, { "grad_norm": 0.35131797194480896, "learning_rate": 1.679943065955636e-05, "loss": 0.0114, "step": 74460 }, { "grad_norm": 0.32170459628105164, "learning_rate": 1.678706910995671e-05, "loss": 0.0138, "step": 74470 }, { "grad_norm": 0.516772449016571, "learning_rate": 1.6774711192473262e-05, "loss": 0.0232, "step": 74480 }, { "grad_norm": 0.4587346315383911, "learning_rate": 1.676235690845743e-05, "loss": 0.0194, "step": 74490 }, { "grad_norm": 0.35948988795280457, "learning_rate": 1.6750006259260285e-05, "loss": 0.0209, "step": 74500 }, { "grad_norm": 0.5302704572677612, "learning_rate": 1.673765924623244e-05, "loss": 0.0194, "step": 74510 }, { "grad_norm": 0.3068152666091919, "learning_rate": 1.67253158707242e-05, "loss": 0.0156, "step": 74520 }, { "grad_norm": 0.4063524305820465, "learning_rate": 1.671297613408537e-05, "loss": 0.0182, "step": 74530 }, { "grad_norm": 0.45646536350250244, "learning_rate": 1.6700640037665444e-05, "loss": 0.0147, "step": 74540 }, { "grad_norm": 0.4413850009441376, "learning_rate": 1.668830758281344e-05, "loss": 0.0281, "step": 74550 }, { "grad_norm": 0.2762502431869507, "learning_rate": 1.6675978770878052e-05, "loss": 0.0253, "step": 74560 }, { "grad_norm": 0.32364436984062195, "learning_rate": 1.6663653603207506e-05, "loss": 0.0161, "step": 74570 }, { "grad_norm": 0.27780666947364807, "learning_rate": 1.6651332081149685e-05, "loss": 0.031, "step": 74580 }, { "grad_norm": 0.29256919026374817, "learning_rate": 1.6639014206052046e-05, "loss": 0.0214, "step": 74590 }, { "grad_norm": 0.6295137405395508, "learning_rate": 1.6626699979261657e-05, "loss": 0.0161, "step": 74600 }, { "grad_norm": 0.5579106211662292, "learning_rate": 1.6614389402125197e-05, "loss": 0.0207, "step": 74610 }, { "grad_norm": 0.5329152345657349, "learning_rate": 1.66020824759889e-05, "loss": 0.0225, "step": 74620 }, { "grad_norm": 0.40937793254852295, "learning_rate": 1.6589779202198675e-05, "loss": 0.0192, "step": 74630 }, { "grad_norm": 0.3001285195350647, "learning_rate": 1.6577479582099924e-05, "loss": 0.0303, "step": 74640 }, { "grad_norm": 0.9684721231460571, "learning_rate": 1.6565183617037788e-05, "loss": 0.026, "step": 74650 }, { "grad_norm": 0.6559624671936035, "learning_rate": 1.6552891308356894e-05, "loss": 0.0197, "step": 74660 }, { "grad_norm": 0.35880565643310547, "learning_rate": 1.6540602657401526e-05, "loss": 0.0205, "step": 74670 }, { "grad_norm": 0.33091771602630615, "learning_rate": 1.6528317665515537e-05, "loss": 0.0194, "step": 74680 }, { "grad_norm": 0.49098190665245056, "learning_rate": 1.6516036334042416e-05, "loss": 0.017, "step": 74690 }, { "grad_norm": 0.3373318612575531, "learning_rate": 1.650375866432519e-05, "loss": 0.0181, "step": 74700 }, { "grad_norm": 0.548123836517334, "learning_rate": 1.649148465770659e-05, "loss": 0.0279, "step": 74710 }, { "grad_norm": 0.24558132886886597, "learning_rate": 1.647921431552883e-05, "loss": 0.0184, "step": 74720 }, { "grad_norm": 0.16005244851112366, "learning_rate": 1.646694763913381e-05, "loss": 0.02, "step": 74730 }, { "grad_norm": 0.36763995885849, "learning_rate": 1.6454684629862966e-05, "loss": 0.0171, "step": 74740 }, { "grad_norm": 0.33097976446151733, "learning_rate": 1.6442425289057388e-05, "loss": 0.0208, "step": 74750 }, { "grad_norm": 0.39021965861320496, "learning_rate": 1.6430169618057718e-05, "loss": 0.0194, "step": 74760 }, { "grad_norm": 0.17494668066501617, "learning_rate": 1.641791761820422e-05, "loss": 0.0172, "step": 74770 }, { "grad_norm": 0.4095337986946106, "learning_rate": 1.640566929083676e-05, "loss": 0.0187, "step": 74780 }, { "grad_norm": 0.550984799861908, "learning_rate": 1.6393424637294823e-05, "loss": 0.0183, "step": 74790 }, { "grad_norm": 0.47200262546539307, "learning_rate": 1.638118365891742e-05, "loss": 0.0199, "step": 74800 }, { "grad_norm": 0.32023605704307556, "learning_rate": 1.636894635704323e-05, "loss": 0.0182, "step": 74810 }, { "grad_norm": 0.5061972737312317, "learning_rate": 1.6356712733010517e-05, "loss": 0.0261, "step": 74820 }, { "grad_norm": 0.688629150390625, "learning_rate": 1.6344482788157096e-05, "loss": 0.0224, "step": 74830 }, { "grad_norm": 0.3387283384799957, "learning_rate": 1.6332256523820444e-05, "loss": 0.0157, "step": 74840 }, { "grad_norm": 0.33497941493988037, "learning_rate": 1.6320033941337597e-05, "loss": 0.0212, "step": 74850 }, { "grad_norm": 0.30167001485824585, "learning_rate": 1.6307815042045216e-05, "loss": 0.0142, "step": 74860 }, { "grad_norm": 0.40296587347984314, "learning_rate": 1.6295599827279513e-05, "loss": 0.0422, "step": 74870 }, { "grad_norm": 0.2073844075202942, "learning_rate": 1.6283388298376356e-05, "loss": 0.0224, "step": 74880 }, { "grad_norm": 0.4528641402721405, "learning_rate": 1.6271180456671125e-05, "loss": 0.0255, "step": 74890 }, { "grad_norm": 0.35097792744636536, "learning_rate": 1.625897630349893e-05, "loss": 0.0361, "step": 74900 }, { "grad_norm": 0.3721325099468231, "learning_rate": 1.6246775840194335e-05, "loss": 0.0212, "step": 74910 }, { "grad_norm": 0.30334895849227905, "learning_rate": 1.6234579068091604e-05, "loss": 0.0172, "step": 74920 }, { "grad_norm": 0.45490923523902893, "learning_rate": 1.6222385988524524e-05, "loss": 0.0369, "step": 74930 }, { "grad_norm": 0.36126577854156494, "learning_rate": 1.6210196602826544e-05, "loss": 0.032, "step": 74940 }, { "grad_norm": 0.4591818153858185, "learning_rate": 1.6198010912330646e-05, "loss": 0.0154, "step": 74950 }, { "grad_norm": 0.5250828266143799, "learning_rate": 1.6185828918369456e-05, "loss": 0.0199, "step": 74960 }, { "grad_norm": 0.6008971333503723, "learning_rate": 1.617365062227517e-05, "loss": 0.026, "step": 74970 }, { "grad_norm": 0.5881208181381226, "learning_rate": 1.616147602537962e-05, "loss": 0.0248, "step": 74980 }, { "grad_norm": 0.45459190011024475, "learning_rate": 1.614930512901415e-05, "loss": 0.021, "step": 74990 }, { "grad_norm": 0.5284317135810852, "learning_rate": 1.6137137934509804e-05, "loss": 0.0197, "step": 75000 }, { "grad_norm": 0.23824544250965118, "learning_rate": 1.6124974443197123e-05, "loss": 0.0261, "step": 75010 }, { "grad_norm": 0.6950571537017822, "learning_rate": 1.6112814656406306e-05, "loss": 0.0269, "step": 75020 }, { "grad_norm": 0.33351612091064453, "learning_rate": 1.6100658575467125e-05, "loss": 0.0272, "step": 75030 }, { "grad_norm": 0.512940526008606, "learning_rate": 1.6088506201708974e-05, "loss": 0.0258, "step": 75040 }, { "grad_norm": 0.25279468297958374, "learning_rate": 1.6076357536460785e-05, "loss": 0.0185, "step": 75050 }, { "grad_norm": 0.37155681848526, "learning_rate": 1.606421258105113e-05, "loss": 0.0218, "step": 75060 }, { "grad_norm": 0.3915630280971527, "learning_rate": 1.6052071336808183e-05, "loss": 0.0218, "step": 75070 }, { "grad_norm": 0.3928394317626953, "learning_rate": 1.6039933805059658e-05, "loss": 0.0221, "step": 75080 }, { "grad_norm": 0.21107490360736847, "learning_rate": 1.6027799987132915e-05, "loss": 0.0276, "step": 75090 }, { "grad_norm": 0.4675652086734772, "learning_rate": 1.6015669884354888e-05, "loss": 0.0219, "step": 75100 }, { "grad_norm": 0.29580509662628174, "learning_rate": 1.6003543498052125e-05, "loss": 0.0205, "step": 75110 }, { "grad_norm": 0.6700116991996765, "learning_rate": 1.5991420829550715e-05, "loss": 0.0349, "step": 75120 }, { "grad_norm": 0.48959627747535706, "learning_rate": 1.5979301880176405e-05, "loss": 0.0222, "step": 75130 }, { "grad_norm": 0.29768434166908264, "learning_rate": 1.5967186651254485e-05, "loss": 0.0169, "step": 75140 }, { "grad_norm": 0.2959638237953186, "learning_rate": 1.5955075144109853e-05, "loss": 0.0201, "step": 75150 }, { "grad_norm": 0.12065466493368149, "learning_rate": 1.594296736006703e-05, "loss": 0.0156, "step": 75160 }, { "grad_norm": 0.7225145101547241, "learning_rate": 1.59308633004501e-05, "loss": 0.0166, "step": 75170 }, { "grad_norm": 0.5016928911209106, "learning_rate": 1.5918762966582724e-05, "loss": 0.0288, "step": 75180 }, { "grad_norm": 0.34793218970298767, "learning_rate": 1.5906666359788203e-05, "loss": 0.017, "step": 75190 }, { "grad_norm": 0.5509600043296814, "learning_rate": 1.589457348138938e-05, "loss": 0.0198, "step": 75200 }, { "grad_norm": 0.50275057554245, "learning_rate": 1.588248433270872e-05, "loss": 0.0187, "step": 75210 }, { "grad_norm": 0.3143766522407532, "learning_rate": 1.587039891506828e-05, "loss": 0.02, "step": 75220 }, { "grad_norm": 0.3288251757621765, "learning_rate": 1.5858317229789715e-05, "loss": 0.02, "step": 75230 }, { "grad_norm": 0.2849523425102234, "learning_rate": 1.584623927819423e-05, "loss": 0.0181, "step": 75240 }, { "grad_norm": 0.33072707056999207, "learning_rate": 1.583416506160269e-05, "loss": 0.0165, "step": 75250 }, { "grad_norm": 0.5317420363426208, "learning_rate": 1.5822094581335472e-05, "loss": 0.0231, "step": 75260 }, { "grad_norm": 0.398743599653244, "learning_rate": 1.5810027838712604e-05, "loss": 0.0198, "step": 75270 }, { "grad_norm": 0.37592124938964844, "learning_rate": 1.5797964835053687e-05, "loss": 0.0181, "step": 75280 }, { "grad_norm": 0.3121987283229828, "learning_rate": 1.5785905571677933e-05, "loss": 0.0177, "step": 75290 }, { "grad_norm": 0.30664560198783875, "learning_rate": 1.5773850049904086e-05, "loss": 0.0164, "step": 75300 }, { "grad_norm": 0.9569337368011475, "learning_rate": 1.5761798271050543e-05, "loss": 0.0368, "step": 75310 }, { "grad_norm": 0.5133716464042664, "learning_rate": 1.5749750236435277e-05, "loss": 0.0208, "step": 75320 }, { "grad_norm": 0.33977898955345154, "learning_rate": 1.5737705947375815e-05, "loss": 0.0148, "step": 75330 }, { "grad_norm": 0.49905121326446533, "learning_rate": 1.5725665405189322e-05, "loss": 0.017, "step": 75340 }, { "grad_norm": 0.20182885229587555, "learning_rate": 1.571362861119253e-05, "loss": 0.0217, "step": 75350 }, { "grad_norm": 0.45024022459983826, "learning_rate": 1.570159556670177e-05, "loss": 0.0213, "step": 75360 }, { "grad_norm": 0.35438162088394165, "learning_rate": 1.5689566273032946e-05, "loss": 0.0217, "step": 75370 }, { "grad_norm": 0.4745042622089386, "learning_rate": 1.5677540731501583e-05, "loss": 0.0197, "step": 75380 }, { "grad_norm": 0.296451210975647, "learning_rate": 1.5665518943422746e-05, "loss": 0.0171, "step": 75390 }, { "grad_norm": 0.2461397349834442, "learning_rate": 1.5653500910111134e-05, "loss": 0.0178, "step": 75400 }, { "grad_norm": 0.1895069032907486, "learning_rate": 1.5641486632881023e-05, "loss": 0.021, "step": 75410 }, { "grad_norm": 0.4462536573410034, "learning_rate": 1.56294761130463e-05, "loss": 0.0167, "step": 75420 }, { "grad_norm": 0.20273318886756897, "learning_rate": 1.5617469351920373e-05, "loss": 0.0259, "step": 75430 }, { "grad_norm": 0.41695651412010193, "learning_rate": 1.5605466350816315e-05, "loss": 0.0235, "step": 75440 }, { "grad_norm": 0.37109068036079407, "learning_rate": 1.559346711104674e-05, "loss": 0.0197, "step": 75450 }, { "grad_norm": 0.24544702470302582, "learning_rate": 1.5581471633923883e-05, "loss": 0.0165, "step": 75460 }, { "grad_norm": 0.36518436670303345, "learning_rate": 1.5569479920759505e-05, "loss": 0.0189, "step": 75470 }, { "grad_norm": 0.5046368837356567, "learning_rate": 1.5557491972865072e-05, "loss": 0.0156, "step": 75480 }, { "grad_norm": 0.3431212604045868, "learning_rate": 1.554550779155152e-05, "loss": 0.0222, "step": 75490 }, { "grad_norm": 0.3410235047340393, "learning_rate": 1.5533527378129453e-05, "loss": 0.0287, "step": 75500 }, { "grad_norm": 0.4729028046131134, "learning_rate": 1.552155073390899e-05, "loss": 0.0448, "step": 75510 }, { "grad_norm": 0.478875070810318, "learning_rate": 1.5509577860199903e-05, "loss": 0.0271, "step": 75520 }, { "grad_norm": 0.34929102659225464, "learning_rate": 1.5497608758311525e-05, "loss": 0.0182, "step": 75530 }, { "grad_norm": 0.5967898368835449, "learning_rate": 1.5485643429552797e-05, "loss": 0.024, "step": 75540 }, { "grad_norm": 0.26325806975364685, "learning_rate": 1.5473681875232197e-05, "loss": 0.0144, "step": 75550 }, { "grad_norm": 0.5696154236793518, "learning_rate": 1.5461724096657837e-05, "loss": 0.0191, "step": 75560 }, { "grad_norm": 0.498085081577301, "learning_rate": 1.5449770095137416e-05, "loss": 0.0163, "step": 75570 }, { "grad_norm": 0.2104104459285736, "learning_rate": 1.5437819871978178e-05, "loss": 0.0179, "step": 75580 }, { "grad_norm": 0.26511070132255554, "learning_rate": 1.5425873428486992e-05, "loss": 0.0182, "step": 75590 }, { "grad_norm": 0.15397432446479797, "learning_rate": 1.5413930765970312e-05, "loss": 0.0164, "step": 75600 }, { "grad_norm": 0.3241179585456848, "learning_rate": 1.5401991885734175e-05, "loss": 0.0251, "step": 75610 }, { "grad_norm": 0.31208348274230957, "learning_rate": 1.5390056789084174e-05, "loss": 0.0192, "step": 75620 }, { "grad_norm": 0.33813196420669556, "learning_rate": 1.5378125477325545e-05, "loss": 0.0196, "step": 75630 }, { "grad_norm": 0.4626425504684448, "learning_rate": 1.5366197951763046e-05, "loss": 0.0237, "step": 75640 }, { "grad_norm": 0.17420555651187897, "learning_rate": 1.5354274213701075e-05, "loss": 0.0155, "step": 75650 }, { "grad_norm": 0.29186153411865234, "learning_rate": 1.5342354264443566e-05, "loss": 0.0126, "step": 75660 }, { "grad_norm": 0.40170928835868835, "learning_rate": 1.533043810529411e-05, "loss": 0.0289, "step": 75670 }, { "grad_norm": 0.6884006857872009, "learning_rate": 1.5318525737555807e-05, "loss": 0.0237, "step": 75680 }, { "grad_norm": 0.32580167055130005, "learning_rate": 1.53066171625314e-05, "loss": 0.0234, "step": 75690 }, { "grad_norm": 0.20379941165447235, "learning_rate": 1.529471238152317e-05, "loss": 0.015, "step": 75700 }, { "grad_norm": 0.5723612308502197, "learning_rate": 1.5282811395833023e-05, "loss": 0.0183, "step": 75710 }, { "grad_norm": 0.5305019021034241, "learning_rate": 1.5270914206762397e-05, "loss": 0.0189, "step": 75720 }, { "grad_norm": 0.20945626497268677, "learning_rate": 1.525902081561241e-05, "loss": 0.0167, "step": 75730 }, { "grad_norm": 0.4791974127292633, "learning_rate": 1.5247131223683664e-05, "loss": 0.0191, "step": 75740 }, { "grad_norm": 0.21250009536743164, "learning_rate": 1.5235245432276408e-05, "loss": 0.0313, "step": 75750 }, { "grad_norm": 0.5452994704246521, "learning_rate": 1.5223363442690424e-05, "loss": 0.0192, "step": 75760 }, { "grad_norm": 0.36640480160713196, "learning_rate": 1.5211485256225133e-05, "loss": 0.018, "step": 75770 }, { "grad_norm": 0.2941621243953705, "learning_rate": 1.519961087417951e-05, "loss": 0.019, "step": 75780 }, { "grad_norm": 0.2686893939971924, "learning_rate": 1.5187740297852137e-05, "loss": 0.0144, "step": 75790 }, { "grad_norm": 0.44615036249160767, "learning_rate": 1.517587352854113e-05, "loss": 0.0203, "step": 75800 }, { "grad_norm": 0.4149090647697449, "learning_rate": 1.516401056754423e-05, "loss": 0.027, "step": 75810 }, { "grad_norm": 0.32209455966949463, "learning_rate": 1.5152151416158783e-05, "loss": 0.0123, "step": 75820 }, { "grad_norm": 0.30784013867378235, "learning_rate": 1.5140296075681642e-05, "loss": 0.0142, "step": 75830 }, { "grad_norm": 0.20876243710517883, "learning_rate": 1.5128444547409331e-05, "loss": 0.0219, "step": 75840 }, { "grad_norm": 0.42225295305252075, "learning_rate": 1.5116596832637864e-05, "loss": 0.016, "step": 75850 }, { "grad_norm": 0.21453745663166046, "learning_rate": 1.5104752932662952e-05, "loss": 0.0247, "step": 75860 }, { "grad_norm": 0.33150002360343933, "learning_rate": 1.5092912848779778e-05, "loss": 0.0174, "step": 75870 }, { "grad_norm": 0.3596715033054352, "learning_rate": 1.5081076582283194e-05, "loss": 0.0206, "step": 75880 }, { "grad_norm": 0.6648087501525879, "learning_rate": 1.5069244134467553e-05, "loss": 0.0241, "step": 75890 }, { "grad_norm": 0.4036506116390228, "learning_rate": 1.5057415506626876e-05, "loss": 0.0247, "step": 75900 }, { "grad_norm": 0.3625953495502472, "learning_rate": 1.5045590700054668e-05, "loss": 0.0151, "step": 75910 }, { "grad_norm": 0.4589497447013855, "learning_rate": 1.5033769716044144e-05, "loss": 0.0207, "step": 75920 }, { "grad_norm": 0.5002558827400208, "learning_rate": 1.5021952555887975e-05, "loss": 0.0242, "step": 75930 }, { "grad_norm": 0.552417516708374, "learning_rate": 1.50101392208785e-05, "loss": 0.017, "step": 75940 }, { "grad_norm": 0.22067008912563324, "learning_rate": 1.4998329712307579e-05, "loss": 0.018, "step": 75950 }, { "grad_norm": 0.3487483561038971, "learning_rate": 1.4986524031466708e-05, "loss": 0.0201, "step": 75960 }, { "grad_norm": 0.32274895906448364, "learning_rate": 1.4974722179646916e-05, "loss": 0.0149, "step": 75970 }, { "grad_norm": 0.4748222529888153, "learning_rate": 1.4962924158138847e-05, "loss": 0.0144, "step": 75980 }, { "grad_norm": 0.4461226463317871, "learning_rate": 1.4951129968232714e-05, "loss": 0.0359, "step": 75990 }, { "grad_norm": 0.21111948788166046, "learning_rate": 1.4939339611218329e-05, "loss": 0.0245, "step": 76000 }, { "grad_norm": 0.2685212790966034, "learning_rate": 1.4927553088385032e-05, "loss": 0.0233, "step": 76010 }, { "grad_norm": 0.5531975626945496, "learning_rate": 1.4915770401021801e-05, "loss": 0.0255, "step": 76020 }, { "grad_norm": 0.5324238538742065, "learning_rate": 1.4903991550417185e-05, "loss": 0.0164, "step": 76030 }, { "grad_norm": 0.22414791584014893, "learning_rate": 1.4892216537859277e-05, "loss": 0.0204, "step": 76040 }, { "grad_norm": 0.22906124591827393, "learning_rate": 1.4880445364635775e-05, "loss": 0.0254, "step": 76050 }, { "grad_norm": 0.3727772831916809, "learning_rate": 1.4868678032033973e-05, "loss": 0.0149, "step": 76060 }, { "grad_norm": 0.3155811131000519, "learning_rate": 1.4856914541340739e-05, "loss": 0.0146, "step": 76070 }, { "grad_norm": 0.23252804577350616, "learning_rate": 1.4845154893842472e-05, "loss": 0.0197, "step": 76080 }, { "grad_norm": 0.21617770195007324, "learning_rate": 1.4833399090825234e-05, "loss": 0.0381, "step": 76090 }, { "grad_norm": 0.2489594966173172, "learning_rate": 1.4821647133574562e-05, "loss": 0.026, "step": 76100 }, { "grad_norm": 0.3551657497882843, "learning_rate": 1.4809899023375706e-05, "loss": 0.0219, "step": 76110 }, { "grad_norm": 0.27614131569862366, "learning_rate": 1.4798154761513366e-05, "loss": 0.0183, "step": 76120 }, { "grad_norm": 0.309670627117157, "learning_rate": 1.4786414349271916e-05, "loss": 0.0126, "step": 76130 }, { "grad_norm": 0.2960592210292816, "learning_rate": 1.4774677787935226e-05, "loss": 0.0168, "step": 76140 }, { "grad_norm": 0.3140711188316345, "learning_rate": 1.4762945078786838e-05, "loss": 0.0262, "step": 76150 }, { "grad_norm": 0.24022504687309265, "learning_rate": 1.4751216223109777e-05, "loss": 0.0184, "step": 76160 }, { "grad_norm": 0.6531702876091003, "learning_rate": 1.4739491222186714e-05, "loss": 0.0161, "step": 76170 }, { "grad_norm": 0.43290120363235474, "learning_rate": 1.4727770077299879e-05, "loss": 0.0215, "step": 76180 }, { "grad_norm": 0.24295908212661743, "learning_rate": 1.4716052789731094e-05, "loss": 0.0242, "step": 76190 }, { "grad_norm": 0.279161274433136, "learning_rate": 1.470433936076171e-05, "loss": 0.0234, "step": 76200 }, { "grad_norm": 0.20554804801940918, "learning_rate": 1.4692629791672719e-05, "loss": 0.0147, "step": 76210 }, { "grad_norm": 0.6254004240036011, "learning_rate": 1.468092408374464e-05, "loss": 0.0215, "step": 76220 }, { "grad_norm": 0.2791551649570465, "learning_rate": 1.4669222238257602e-05, "loss": 0.0209, "step": 76230 }, { "grad_norm": 0.43629190325737, "learning_rate": 1.4657524256491301e-05, "loss": 0.0187, "step": 76240 }, { "grad_norm": 0.4897692799568176, "learning_rate": 1.464583013972502e-05, "loss": 0.0186, "step": 76250 }, { "grad_norm": 0.4623385965824127, "learning_rate": 1.4634139889237591e-05, "loss": 0.034, "step": 76260 }, { "grad_norm": 0.5598297119140625, "learning_rate": 1.4622453506307448e-05, "loss": 0.0267, "step": 76270 }, { "grad_norm": 0.7701531052589417, "learning_rate": 1.461077099221262e-05, "loss": 0.0358, "step": 76280 }, { "grad_norm": 0.1985127031803131, "learning_rate": 1.4599092348230652e-05, "loss": 0.0153, "step": 76290 }, { "grad_norm": 0.5713141560554504, "learning_rate": 1.4587417575638724e-05, "loss": 0.0133, "step": 76300 }, { "grad_norm": 0.6074168086051941, "learning_rate": 1.4575746675713563e-05, "loss": 0.0295, "step": 76310 }, { "grad_norm": 0.5162229537963867, "learning_rate": 1.4564079649731504e-05, "loss": 0.017, "step": 76320 }, { "grad_norm": 0.21061961352825165, "learning_rate": 1.4552416498968402e-05, "loss": 0.0243, "step": 76330 }, { "grad_norm": 0.35300010442733765, "learning_rate": 1.4540757224699747e-05, "loss": 0.0163, "step": 76340 }, { "grad_norm": 0.43670886754989624, "learning_rate": 1.452910182820056e-05, "loss": 0.0264, "step": 76350 }, { "grad_norm": 0.2278144508600235, "learning_rate": 1.4517450310745468e-05, "loss": 0.0184, "step": 76360 }, { "grad_norm": 0.2843414545059204, "learning_rate": 1.450580267360866e-05, "loss": 0.0319, "step": 76370 }, { "grad_norm": 0.3810505270957947, "learning_rate": 1.4494158918063922e-05, "loss": 0.0193, "step": 76380 }, { "grad_norm": 0.20250961184501648, "learning_rate": 1.4482519045384568e-05, "loss": 0.0145, "step": 76390 }, { "grad_norm": 0.3915597200393677, "learning_rate": 1.4470883056843548e-05, "loss": 0.0229, "step": 76400 }, { "grad_norm": 0.30890002846717834, "learning_rate": 1.4459250953713322e-05, "loss": 0.0186, "step": 76410 }, { "grad_norm": 0.5579456686973572, "learning_rate": 1.444762273726598e-05, "loss": 0.0237, "step": 76420 }, { "grad_norm": 0.3819269835948944, "learning_rate": 1.443599840877316e-05, "loss": 0.016, "step": 76430 }, { "grad_norm": 0.283113569021225, "learning_rate": 1.4424377969506103e-05, "loss": 0.0202, "step": 76440 }, { "grad_norm": 0.2641023099422455, "learning_rate": 1.4412761420735565e-05, "loss": 0.0181, "step": 76450 }, { "grad_norm": 0.7205040454864502, "learning_rate": 1.4401148763731953e-05, "loss": 0.0428, "step": 76460 }, { "grad_norm": 0.29087066650390625, "learning_rate": 1.4389539999765173e-05, "loss": 0.0152, "step": 76470 }, { "grad_norm": 0.3163648247718811, "learning_rate": 1.4377935130104758e-05, "loss": 0.0248, "step": 76480 }, { "grad_norm": 0.32026782631874084, "learning_rate": 1.4366334156019801e-05, "loss": 0.018, "step": 76490 }, { "grad_norm": 0.3761298954486847, "learning_rate": 1.435473707877898e-05, "loss": 0.0179, "step": 76500 }, { "grad_norm": 0.31524455547332764, "learning_rate": 1.4343143899650508e-05, "loss": 0.0161, "step": 76510 }, { "grad_norm": 0.5240085124969482, "learning_rate": 1.4331554619902204e-05, "loss": 0.0248, "step": 76520 }, { "grad_norm": 0.20421375334262848, "learning_rate": 1.4319969240801484e-05, "loss": 0.021, "step": 76530 }, { "grad_norm": 0.2648092806339264, "learning_rate": 1.430838776361526e-05, "loss": 0.0187, "step": 76540 }, { "grad_norm": 0.4179786145687103, "learning_rate": 1.4296810189610093e-05, "loss": 0.0131, "step": 76550 }, { "grad_norm": 0.5089518427848816, "learning_rate": 1.4285236520052087e-05, "loss": 0.028, "step": 76560 }, { "grad_norm": 0.20109300315380096, "learning_rate": 1.4273666756206933e-05, "loss": 0.0132, "step": 76570 }, { "grad_norm": 0.2389681041240692, "learning_rate": 1.4262100899339848e-05, "loss": 0.0193, "step": 76580 }, { "grad_norm": 0.651429295539856, "learning_rate": 1.4250538950715696e-05, "loss": 0.0258, "step": 76590 }, { "grad_norm": 0.4335779845714569, "learning_rate": 1.4238980911598842e-05, "loss": 0.0099, "step": 76600 }, { "grad_norm": 0.46594008803367615, "learning_rate": 1.4227426783253267e-05, "loss": 0.0159, "step": 76610 }, { "grad_norm": 0.8047844171524048, "learning_rate": 1.4215876566942521e-05, "loss": 0.0245, "step": 76620 }, { "grad_norm": 0.318707138299942, "learning_rate": 1.4204330263929721e-05, "loss": 0.0299, "step": 76630 }, { "grad_norm": 0.19931499660015106, "learning_rate": 1.4192787875477537e-05, "loss": 0.0127, "step": 76640 }, { "grad_norm": 0.6599230170249939, "learning_rate": 1.4181249402848246e-05, "loss": 0.0136, "step": 76650 }, { "grad_norm": 0.4911733865737915, "learning_rate": 1.4169714847303656e-05, "loss": 0.019, "step": 76660 }, { "grad_norm": 0.6515690088272095, "learning_rate": 1.4158184210105197e-05, "loss": 0.0249, "step": 76670 }, { "grad_norm": 0.3653828203678131, "learning_rate": 1.4146657492513787e-05, "loss": 0.0207, "step": 76680 }, { "grad_norm": 0.256808876991272, "learning_rate": 1.4135134695790047e-05, "loss": 0.0144, "step": 76690 }, { "grad_norm": 0.4544486105442047, "learning_rate": 1.4123615821194037e-05, "loss": 0.0131, "step": 76700 }, { "grad_norm": 0.552748441696167, "learning_rate": 1.4112100869985473e-05, "loss": 0.0194, "step": 76710 }, { "grad_norm": 0.33420658111572266, "learning_rate": 1.4100589843423589e-05, "loss": 0.013, "step": 76720 }, { "grad_norm": 0.5115941762924194, "learning_rate": 1.408908274276723e-05, "loss": 0.0218, "step": 76730 }, { "grad_norm": 0.4053557813167572, "learning_rate": 1.4077579569274786e-05, "loss": 0.0107, "step": 76740 }, { "grad_norm": 0.1877138912677765, "learning_rate": 1.4066080324204245e-05, "loss": 0.0258, "step": 76750 }, { "grad_norm": 0.4068759083747864, "learning_rate": 1.405458500881312e-05, "loss": 0.0192, "step": 76760 }, { "grad_norm": 0.29672300815582275, "learning_rate": 1.4043093624358538e-05, "loss": 0.017, "step": 76770 }, { "grad_norm": 0.3571047782897949, "learning_rate": 1.4031606172097189e-05, "loss": 0.0218, "step": 76780 }, { "grad_norm": 0.3511684834957123, "learning_rate": 1.40201226532853e-05, "loss": 0.0274, "step": 76790 }, { "grad_norm": 0.24128200113773346, "learning_rate": 1.4008643069178701e-05, "loss": 0.0233, "step": 76800 }, { "grad_norm": 0.34449177980422974, "learning_rate": 1.399716742103278e-05, "loss": 0.0155, "step": 76810 }, { "grad_norm": 0.5611186027526855, "learning_rate": 1.398569571010252e-05, "loss": 0.018, "step": 76820 }, { "grad_norm": 0.5355168581008911, "learning_rate": 1.3974227937642414e-05, "loss": 0.0315, "step": 76830 }, { "grad_norm": 0.27307626605033875, "learning_rate": 1.3962764104906595e-05, "loss": 0.0219, "step": 76840 }, { "grad_norm": 0.12015341967344284, "learning_rate": 1.3951304213148696e-05, "loss": 0.0224, "step": 76850 }, { "grad_norm": 0.42345288395881653, "learning_rate": 1.3939848263621979e-05, "loss": 0.0272, "step": 76860 }, { "grad_norm": 0.6796815991401672, "learning_rate": 1.3928396257579219e-05, "loss": 0.0152, "step": 76870 }, { "grad_norm": 0.4900607764720917, "learning_rate": 1.3916948196272833e-05, "loss": 0.0318, "step": 76880 }, { "grad_norm": 0.31126537919044495, "learning_rate": 1.3905504080954734e-05, "loss": 0.0144, "step": 76890 }, { "grad_norm": 0.4491290748119354, "learning_rate": 1.3894063912876454e-05, "loss": 0.024, "step": 76900 }, { "grad_norm": 0.2893564701080322, "learning_rate": 1.3882627693289046e-05, "loss": 0.0155, "step": 76910 }, { "grad_norm": 0.18057306110858917, "learning_rate": 1.3871195423443183e-05, "loss": 0.0158, "step": 76920 }, { "grad_norm": 0.22760413587093353, "learning_rate": 1.3859767104589044e-05, "loss": 0.0161, "step": 76930 }, { "grad_norm": 0.5009481310844421, "learning_rate": 1.3848342737976467e-05, "loss": 0.0256, "step": 76940 }, { "grad_norm": 0.32780417799949646, "learning_rate": 1.3836922324854757e-05, "loss": 0.0199, "step": 76950 }, { "grad_norm": 0.32064419984817505, "learning_rate": 1.382550586647287e-05, "loss": 0.0202, "step": 76960 }, { "grad_norm": 0.754470705986023, "learning_rate": 1.3814093364079257e-05, "loss": 0.0437, "step": 76970 }, { "grad_norm": 0.32335904240608215, "learning_rate": 1.3802684818921991e-05, "loss": 0.0182, "step": 76980 }, { "grad_norm": 0.19297628104686737, "learning_rate": 1.3791280232248687e-05, "loss": 0.0158, "step": 76990 }, { "grad_norm": 0.4469359815120697, "learning_rate": 1.3779879605306555e-05, "loss": 0.0163, "step": 77000 }, { "grad_norm": 0.41998088359832764, "learning_rate": 1.376848293934232e-05, "loss": 0.0181, "step": 77010 }, { "grad_norm": 0.3355378806591034, "learning_rate": 1.3757090235602316e-05, "loss": 0.0238, "step": 77020 }, { "grad_norm": 0.4711242914199829, "learning_rate": 1.3745701495332447e-05, "loss": 0.0248, "step": 77030 }, { "grad_norm": 0.36589568853378296, "learning_rate": 1.3734316719778145e-05, "loss": 0.0176, "step": 77040 }, { "grad_norm": 0.26635366678237915, "learning_rate": 1.3722935910184454e-05, "loss": 0.0225, "step": 77050 }, { "grad_norm": 0.39046403765678406, "learning_rate": 1.3711559067795921e-05, "loss": 0.0148, "step": 77060 }, { "grad_norm": 0.30221572518348694, "learning_rate": 1.3700186193856756e-05, "loss": 0.0216, "step": 77070 }, { "grad_norm": 0.3019445836544037, "learning_rate": 1.368881728961064e-05, "loss": 0.0192, "step": 77080 }, { "grad_norm": 0.4361954927444458, "learning_rate": 1.3677452356300885e-05, "loss": 0.017, "step": 77090 }, { "grad_norm": 0.42436355352401733, "learning_rate": 1.3666091395170321e-05, "loss": 0.022, "step": 77100 }, { "grad_norm": 0.4644588530063629, "learning_rate": 1.3654734407461384e-05, "loss": 0.015, "step": 77110 }, { "grad_norm": 0.6875295042991638, "learning_rate": 1.3643381394416022e-05, "loss": 0.0226, "step": 77120 }, { "grad_norm": 0.5461626052856445, "learning_rate": 1.3632032357275838e-05, "loss": 0.0184, "step": 77130 }, { "grad_norm": 0.3266490399837494, "learning_rate": 1.3620687297281898e-05, "loss": 0.021, "step": 77140 }, { "grad_norm": 0.39973458647727966, "learning_rate": 1.3609346215674917e-05, "loss": 0.0189, "step": 77150 }, { "grad_norm": 0.26467111706733704, "learning_rate": 1.3598009113695109e-05, "loss": 0.022, "step": 77160 }, { "grad_norm": 0.2962389886379242, "learning_rate": 1.358667599258231e-05, "loss": 0.0193, "step": 77170 }, { "grad_norm": 0.35751795768737793, "learning_rate": 1.357534685357586e-05, "loss": 0.0142, "step": 77180 }, { "grad_norm": 0.18950234353542328, "learning_rate": 1.3564021697914715e-05, "loss": 0.0167, "step": 77190 }, { "grad_norm": 0.30990102887153625, "learning_rate": 1.3552700526837381e-05, "loss": 0.0322, "step": 77200 }, { "grad_norm": 0.33066317439079285, "learning_rate": 1.3541383341581925e-05, "loss": 0.0188, "step": 77210 }, { "grad_norm": 0.28427016735076904, "learning_rate": 1.3530070143385965e-05, "loss": 0.0223, "step": 77220 }, { "grad_norm": 0.4169999659061432, "learning_rate": 1.3518760933486702e-05, "loss": 0.0193, "step": 77230 }, { "grad_norm": 0.27453291416168213, "learning_rate": 1.3507455713120908e-05, "loss": 0.016, "step": 77240 }, { "grad_norm": 0.47893545031547546, "learning_rate": 1.3496154483524876e-05, "loss": 0.0254, "step": 77250 }, { "grad_norm": 0.40636810660362244, "learning_rate": 1.3484857245934506e-05, "loss": 0.0221, "step": 77260 }, { "grad_norm": 0.612080991268158, "learning_rate": 1.3473564001585248e-05, "loss": 0.0175, "step": 77270 }, { "grad_norm": 0.9003581404685974, "learning_rate": 1.3462274751712129e-05, "loss": 0.019, "step": 77280 }, { "grad_norm": 0.3844352960586548, "learning_rate": 1.3450989497549698e-05, "loss": 0.0169, "step": 77290 }, { "grad_norm": 0.792914867401123, "learning_rate": 1.3439708240332115e-05, "loss": 0.0194, "step": 77300 }, { "grad_norm": 0.33861714601516724, "learning_rate": 1.3428430981293043e-05, "loss": 0.0223, "step": 77310 }, { "grad_norm": 0.15375672280788422, "learning_rate": 1.3417157721665808e-05, "loss": 0.0132, "step": 77320 }, { "grad_norm": 0.21956561505794525, "learning_rate": 1.3405888462683186e-05, "loss": 0.0227, "step": 77330 }, { "grad_norm": 0.21591994166374207, "learning_rate": 1.3394623205577595e-05, "loss": 0.0133, "step": 77340 }, { "grad_norm": 0.24099168181419373, "learning_rate": 1.3383361951580963e-05, "loss": 0.0225, "step": 77350 }, { "grad_norm": 0.4026108980178833, "learning_rate": 1.337210470192483e-05, "loss": 0.0213, "step": 77360 }, { "grad_norm": 0.27788689732551575, "learning_rate": 1.3360851457840245e-05, "loss": 0.0188, "step": 77370 }, { "grad_norm": 0.3594362735748291, "learning_rate": 1.3349602220557856e-05, "loss": 0.0217, "step": 77380 }, { "grad_norm": 0.5274567604064941, "learning_rate": 1.3338356991307865e-05, "loss": 0.021, "step": 77390 }, { "grad_norm": 0.34641233086586, "learning_rate": 1.3327115771320048e-05, "loss": 0.0202, "step": 77400 }, { "grad_norm": 0.25858551263809204, "learning_rate": 1.3315878561823697e-05, "loss": 0.0313, "step": 77410 }, { "grad_norm": 0.20710492134094238, "learning_rate": 1.330464536404773e-05, "loss": 0.0157, "step": 77420 }, { "grad_norm": 0.49099352955818176, "learning_rate": 1.3293416179220563e-05, "loss": 0.02, "step": 77430 }, { "grad_norm": 0.6237185001373291, "learning_rate": 1.328219100857021e-05, "loss": 0.025, "step": 77440 }, { "grad_norm": 0.25048890709877014, "learning_rate": 1.327096985332425e-05, "loss": 0.0105, "step": 77450 }, { "grad_norm": 0.22151602804660797, "learning_rate": 1.3259752714709816e-05, "loss": 0.0207, "step": 77460 }, { "grad_norm": 0.21796943247318268, "learning_rate": 1.3248539593953569e-05, "loss": 0.0137, "step": 77470 }, { "grad_norm": 0.3640630841255188, "learning_rate": 1.323733049228178e-05, "loss": 0.0153, "step": 77480 }, { "grad_norm": 0.5961496233940125, "learning_rate": 1.3226125410920276e-05, "loss": 0.0228, "step": 77490 }, { "grad_norm": 0.48285427689552307, "learning_rate": 1.3214924351094388e-05, "loss": 0.02, "step": 77500 }, { "grad_norm": 0.4593760371208191, "learning_rate": 1.3203727314029069e-05, "loss": 0.0157, "step": 77510 }, { "grad_norm": 0.39636799693107605, "learning_rate": 1.3192534300948811e-05, "loss": 0.0193, "step": 77520 }, { "grad_norm": 0.25567805767059326, "learning_rate": 1.3181345313077675e-05, "loss": 0.0172, "step": 77530 }, { "grad_norm": 0.4003337621688843, "learning_rate": 1.3170160351639243e-05, "loss": 0.0195, "step": 77540 }, { "grad_norm": 0.19517114758491516, "learning_rate": 1.3158979417856721e-05, "loss": 0.0172, "step": 77550 }, { "grad_norm": 0.3939504325389862, "learning_rate": 1.31478025129528e-05, "loss": 0.0199, "step": 77560 }, { "grad_norm": 0.40319958329200745, "learning_rate": 1.3136629638149794e-05, "loss": 0.017, "step": 77570 }, { "grad_norm": 0.652675449848175, "learning_rate": 1.3125460794669547e-05, "loss": 0.0173, "step": 77580 }, { "grad_norm": 0.2670753300189972, "learning_rate": 1.3114295983733483e-05, "loss": 0.0189, "step": 77590 }, { "grad_norm": 1.0589258670806885, "learning_rate": 1.3103135206562534e-05, "loss": 0.0179, "step": 77600 }, { "grad_norm": 0.38317012786865234, "learning_rate": 1.3091978464377264e-05, "loss": 0.0285, "step": 77610 }, { "grad_norm": 0.2946408689022064, "learning_rate": 1.3080825758397724e-05, "loss": 0.0261, "step": 77620 }, { "grad_norm": 0.5325000286102295, "learning_rate": 1.3069677089843568e-05, "loss": 0.024, "step": 77630 }, { "grad_norm": 0.4476683437824249, "learning_rate": 1.305853245993401e-05, "loss": 0.0179, "step": 77640 }, { "grad_norm": 0.5455346703529358, "learning_rate": 1.3047391869887809e-05, "loss": 0.013, "step": 77650 }, { "grad_norm": 0.32973024249076843, "learning_rate": 1.3036255320923263e-05, "loss": 0.0317, "step": 77660 }, { "grad_norm": 0.7167679071426392, "learning_rate": 1.3025122814258273e-05, "loss": 0.0187, "step": 77670 }, { "grad_norm": 0.5428513288497925, "learning_rate": 1.3013994351110249e-05, "loss": 0.0191, "step": 77680 }, { "grad_norm": 0.17075859010219574, "learning_rate": 1.3002869932696194e-05, "loss": 0.0292, "step": 77690 }, { "grad_norm": 0.22187261283397675, "learning_rate": 1.2991749560232664e-05, "loss": 0.0155, "step": 77700 }, { "grad_norm": 0.2849520444869995, "learning_rate": 1.2980633234935768e-05, "loss": 0.0148, "step": 77710 }, { "grad_norm": 0.2603054642677307, "learning_rate": 1.2969520958021153e-05, "loss": 0.0128, "step": 77720 }, { "grad_norm": 0.43227773904800415, "learning_rate": 1.295841273070405e-05, "loss": 0.0143, "step": 77730 }, { "grad_norm": 0.6964147090911865, "learning_rate": 1.2947308554199255e-05, "loss": 0.0134, "step": 77740 }, { "grad_norm": 0.1847919523715973, "learning_rate": 1.2936208429721075e-05, "loss": 0.0198, "step": 77750 }, { "grad_norm": 0.6285639405250549, "learning_rate": 1.2925112358483416e-05, "loss": 0.0213, "step": 77760 }, { "grad_norm": 0.2807776927947998, "learning_rate": 1.2914020341699728e-05, "loss": 0.014, "step": 77770 }, { "grad_norm": 0.27548888325691223, "learning_rate": 1.2902932380583032e-05, "loss": 0.0231, "step": 77780 }, { "grad_norm": 0.3723001182079315, "learning_rate": 1.2891848476345863e-05, "loss": 0.0205, "step": 77790 }, { "grad_norm": 0.5097260475158691, "learning_rate": 1.2880768630200368e-05, "loss": 0.0133, "step": 77800 }, { "grad_norm": 0.5234659910202026, "learning_rate": 1.286969284335819e-05, "loss": 0.0153, "step": 77810 }, { "grad_norm": 0.5756644010543823, "learning_rate": 1.2858621117030578e-05, "loss": 0.0221, "step": 77820 }, { "grad_norm": 0.24651168286800385, "learning_rate": 1.2847553452428324e-05, "loss": 0.0248, "step": 77830 }, { "grad_norm": 0.3416759967803955, "learning_rate": 1.283648985076178e-05, "loss": 0.0239, "step": 77840 }, { "grad_norm": 0.3706189692020416, "learning_rate": 1.2825430313240817e-05, "loss": 0.0188, "step": 77850 }, { "grad_norm": 0.3340883255004883, "learning_rate": 1.281437484107491e-05, "loss": 0.0211, "step": 77860 }, { "grad_norm": 0.7974541783332825, "learning_rate": 1.2803323435473058e-05, "loss": 0.0237, "step": 77870 }, { "grad_norm": 0.24792395532131195, "learning_rate": 1.2792276097643841e-05, "loss": 0.0232, "step": 77880 }, { "grad_norm": 0.5036256909370422, "learning_rate": 1.2781232828795342e-05, "loss": 0.0164, "step": 77890 }, { "grad_norm": 0.2842475473880768, "learning_rate": 1.2770193630135292e-05, "loss": 0.018, "step": 77900 }, { "grad_norm": 0.19301959872245789, "learning_rate": 1.2759158502870877e-05, "loss": 0.0178, "step": 77910 }, { "grad_norm": 0.2928772270679474, "learning_rate": 1.2748127448208913e-05, "loss": 0.0175, "step": 77920 }, { "grad_norm": 0.15594395995140076, "learning_rate": 1.2737100467355706e-05, "loss": 0.0141, "step": 77930 }, { "grad_norm": 0.4684505760669708, "learning_rate": 1.2726077561517169e-05, "loss": 0.0183, "step": 77940 }, { "grad_norm": 0.2804711163043976, "learning_rate": 1.2715058731898755e-05, "loss": 0.0207, "step": 77950 }, { "grad_norm": 0.2007085084915161, "learning_rate": 1.2704043979705471e-05, "loss": 0.0248, "step": 77960 }, { "grad_norm": 0.20700539648532867, "learning_rate": 1.269303330614185e-05, "loss": 0.0151, "step": 77970 }, { "grad_norm": 0.32826218008995056, "learning_rate": 1.2682026712412015e-05, "loss": 0.027, "step": 77980 }, { "grad_norm": 0.32572945952415466, "learning_rate": 1.2671024199719644e-05, "loss": 0.0198, "step": 77990 }, { "grad_norm": 0.431484580039978, "learning_rate": 1.2660025769267936e-05, "loss": 0.0142, "step": 78000 }, { "grad_norm": 0.1699695736169815, "learning_rate": 1.2649031422259667e-05, "loss": 0.0201, "step": 78010 }, { "grad_norm": 0.4123453199863434, "learning_rate": 1.2638041159897163e-05, "loss": 0.0159, "step": 78020 }, { "grad_norm": 0.3712257742881775, "learning_rate": 1.262705498338232e-05, "loss": 0.0168, "step": 78030 }, { "grad_norm": 0.3845163583755493, "learning_rate": 1.2616072893916542e-05, "loss": 0.0141, "step": 78040 }, { "grad_norm": 0.1570054441690445, "learning_rate": 1.2605094892700836e-05, "loss": 0.0168, "step": 78050 }, { "grad_norm": 0.3059375584125519, "learning_rate": 1.259412098093572e-05, "loss": 0.0149, "step": 78060 }, { "grad_norm": 0.2695600986480713, "learning_rate": 1.258315115982131e-05, "loss": 0.0141, "step": 78070 }, { "grad_norm": 0.6675609350204468, "learning_rate": 1.2572185430557203e-05, "loss": 0.0194, "step": 78080 }, { "grad_norm": 0.40767279267311096, "learning_rate": 1.2561223794342659e-05, "loss": 0.0127, "step": 78090 }, { "grad_norm": 0.2948984205722809, "learning_rate": 1.255026625237637e-05, "loss": 0.0225, "step": 78100 }, { "grad_norm": 0.2730354368686676, "learning_rate": 1.2539312805856678e-05, "loss": 0.0154, "step": 78110 }, { "grad_norm": 0.29642197489738464, "learning_rate": 1.25283634559814e-05, "loss": 0.0125, "step": 78120 }, { "grad_norm": 0.12627609074115753, "learning_rate": 1.2517418203947972e-05, "loss": 0.0187, "step": 78130 }, { "grad_norm": 0.6833050847053528, "learning_rate": 1.2506477050953308e-05, "loss": 0.0187, "step": 78140 }, { "grad_norm": 0.5052998661994934, "learning_rate": 1.2495539998193972e-05, "loss": 0.0264, "step": 78150 }, { "grad_norm": 0.5876003503799438, "learning_rate": 1.2484607046865975e-05, "loss": 0.0241, "step": 78160 }, { "grad_norm": 0.5697001218795776, "learning_rate": 1.2473678198164968e-05, "loss": 0.017, "step": 78170 }, { "grad_norm": 0.2927383780479431, "learning_rate": 1.2462753453286075e-05, "loss": 0.0166, "step": 78180 }, { "grad_norm": 0.2668634057044983, "learning_rate": 1.2451832813424031e-05, "loss": 0.0182, "step": 78190 }, { "grad_norm": 0.5089660286903381, "learning_rate": 1.2440916279773095e-05, "loss": 0.0122, "step": 78200 }, { "grad_norm": 0.8796955943107605, "learning_rate": 1.2430003853527101e-05, "loss": 0.0139, "step": 78210 }, { "grad_norm": 0.3515373170375824, "learning_rate": 1.2419095535879383e-05, "loss": 0.0233, "step": 78220 }, { "grad_norm": 0.18789979815483093, "learning_rate": 1.2408191328022878e-05, "loss": 0.0166, "step": 78230 }, { "grad_norm": 0.31287869811058044, "learning_rate": 1.2397291231150066e-05, "loss": 0.0123, "step": 78240 }, { "grad_norm": 0.3257242441177368, "learning_rate": 1.2386395246452937e-05, "loss": 0.0176, "step": 78250 }, { "grad_norm": 0.34418007731437683, "learning_rate": 1.2375503375123083e-05, "loss": 0.016, "step": 78260 }, { "grad_norm": 0.34906795620918274, "learning_rate": 1.2364615618351583e-05, "loss": 0.0136, "step": 78270 }, { "grad_norm": 0.3452180027961731, "learning_rate": 1.2353731977329169e-05, "loss": 0.0166, "step": 78280 }, { "grad_norm": 0.10819468647241592, "learning_rate": 1.2342852453246007e-05, "loss": 0.0134, "step": 78290 }, { "grad_norm": 0.5137754082679749, "learning_rate": 1.2331977047291898e-05, "loss": 0.0363, "step": 78300 }, { "grad_norm": 0.2594771683216095, "learning_rate": 1.2321105760656137e-05, "loss": 0.0177, "step": 78310 }, { "grad_norm": 0.28384843468666077, "learning_rate": 1.2310238594527613e-05, "loss": 0.0166, "step": 78320 }, { "grad_norm": 0.29872754216194153, "learning_rate": 1.2299375550094693e-05, "loss": 0.0175, "step": 78330 }, { "grad_norm": 0.22615976631641388, "learning_rate": 1.2288516628545421e-05, "loss": 0.0141, "step": 78340 }, { "grad_norm": 0.2725623548030853, "learning_rate": 1.2277661831067255e-05, "loss": 0.0169, "step": 78350 }, { "grad_norm": 0.10409832745790482, "learning_rate": 1.2266811158847286e-05, "loss": 0.0171, "step": 78360 }, { "grad_norm": 0.33285900950431824, "learning_rate": 1.2255964613072107e-05, "loss": 0.0182, "step": 78370 }, { "grad_norm": 0.2530743181705475, "learning_rate": 1.2245122194927905e-05, "loss": 0.0193, "step": 78380 }, { "grad_norm": 0.3847549259662628, "learning_rate": 1.2234283905600364e-05, "loss": 0.0128, "step": 78390 }, { "grad_norm": 0.30867916345596313, "learning_rate": 1.222344974627475e-05, "loss": 0.0129, "step": 78400 }, { "grad_norm": 1.5575456619262695, "learning_rate": 1.2212619718135875e-05, "loss": 0.0283, "step": 78410 }, { "grad_norm": 0.36770421266555786, "learning_rate": 1.2201793822368118e-05, "loss": 0.0203, "step": 78420 }, { "grad_norm": 0.3889349102973938, "learning_rate": 1.2190972060155337e-05, "loss": 0.015, "step": 78430 }, { "grad_norm": 1.232549786567688, "learning_rate": 1.2180154432681013e-05, "loss": 0.0179, "step": 78440 }, { "grad_norm": 0.4542587399482727, "learning_rate": 1.2169340941128148e-05, "loss": 0.0137, "step": 78450 }, { "grad_norm": 0.3159148097038269, "learning_rate": 1.2158531586679267e-05, "loss": 0.0193, "step": 78460 }, { "grad_norm": 0.2634536027908325, "learning_rate": 1.2147726370516477e-05, "loss": 0.0231, "step": 78470 }, { "grad_norm": 0.31126901507377625, "learning_rate": 1.2136925293821422e-05, "loss": 0.0205, "step": 78480 }, { "grad_norm": 0.39702895283699036, "learning_rate": 1.2126128357775306e-05, "loss": 0.0184, "step": 78490 }, { "grad_norm": 0.4155533015727997, "learning_rate": 1.2115335563558838e-05, "loss": 0.0119, "step": 78500 }, { "grad_norm": 0.4540148675441742, "learning_rate": 1.2104546912352327e-05, "loss": 0.0186, "step": 78510 }, { "grad_norm": 0.4188535809516907, "learning_rate": 1.2093762405335557e-05, "loss": 0.0138, "step": 78520 }, { "grad_norm": 0.25237026810646057, "learning_rate": 1.2082982043687974e-05, "loss": 0.0215, "step": 78530 }, { "grad_norm": 0.411990761756897, "learning_rate": 1.207220582858845e-05, "loss": 0.0126, "step": 78540 }, { "grad_norm": 0.41764944791793823, "learning_rate": 1.206143376121549e-05, "loss": 0.0185, "step": 78550 }, { "grad_norm": 0.6210418343544006, "learning_rate": 1.2050665842747078e-05, "loss": 0.0199, "step": 78560 }, { "grad_norm": 0.4112373888492584, "learning_rate": 1.2039902074360804e-05, "loss": 0.0329, "step": 78570 }, { "grad_norm": 0.3964926302433014, "learning_rate": 1.2029142457233755e-05, "loss": 0.0328, "step": 78580 }, { "grad_norm": 0.4251878559589386, "learning_rate": 1.2018386992542601e-05, "loss": 0.0204, "step": 78590 }, { "grad_norm": 0.5306007266044617, "learning_rate": 1.2007635681463542e-05, "loss": 0.0191, "step": 78600 }, { "grad_norm": 0.6178027987480164, "learning_rate": 1.1996888525172334e-05, "loss": 0.0235, "step": 78610 }, { "grad_norm": 0.2818060517311096, "learning_rate": 1.198614552484425e-05, "loss": 0.0139, "step": 78620 }, { "grad_norm": 0.34534725546836853, "learning_rate": 1.197540668165415e-05, "loss": 0.0172, "step": 78630 }, { "grad_norm": 0.3747949004173279, "learning_rate": 1.1964671996776395e-05, "loss": 0.0161, "step": 78640 }, { "grad_norm": 0.23555821180343628, "learning_rate": 1.1953941471384922e-05, "loss": 0.0227, "step": 78650 }, { "grad_norm": 0.1661396622657776, "learning_rate": 1.1943215106653205e-05, "loss": 0.0113, "step": 78660 }, { "grad_norm": 0.45555949211120605, "learning_rate": 1.193249290375429e-05, "loss": 0.0164, "step": 78670 }, { "grad_norm": 0.4449200928211212, "learning_rate": 1.1921774863860696e-05, "loss": 0.023, "step": 78680 }, { "grad_norm": 0.49174484610557556, "learning_rate": 1.1911060988144556e-05, "loss": 0.0247, "step": 78690 }, { "grad_norm": 0.36251989006996155, "learning_rate": 1.1900351277777533e-05, "loss": 0.0206, "step": 78700 }, { "grad_norm": 0.20270782709121704, "learning_rate": 1.1889645733930798e-05, "loss": 0.018, "step": 78710 }, { "grad_norm": 0.23815175890922546, "learning_rate": 1.1878944357775107e-05, "loss": 0.0178, "step": 78720 }, { "grad_norm": 0.22501249611377716, "learning_rate": 1.1868247150480749e-05, "loss": 0.0148, "step": 78730 }, { "grad_norm": 0.2174779772758484, "learning_rate": 1.1857554113217567e-05, "loss": 0.0165, "step": 78740 }, { "grad_norm": 0.28975048661231995, "learning_rate": 1.1846865247154903e-05, "loss": 0.0157, "step": 78750 }, { "grad_norm": 0.4090501070022583, "learning_rate": 1.1836180553461711e-05, "loss": 0.0174, "step": 78760 }, { "grad_norm": 0.5129702091217041, "learning_rate": 1.1825500033306419e-05, "loss": 0.0209, "step": 78770 }, { "grad_norm": 0.18472933769226074, "learning_rate": 1.1814823687857052e-05, "loss": 0.0166, "step": 78780 }, { "grad_norm": 0.35376545786857605, "learning_rate": 1.1804151518281153e-05, "loss": 0.0308, "step": 78790 }, { "grad_norm": 0.1518169492483139, "learning_rate": 1.179348352574583e-05, "loss": 0.0187, "step": 78800 }, { "grad_norm": 0.30677688121795654, "learning_rate": 1.1782819711417697e-05, "loss": 0.02, "step": 78810 }, { "grad_norm": 0.8148382902145386, "learning_rate": 1.1772160076462951e-05, "loss": 0.0374, "step": 78820 }, { "grad_norm": 0.3090469539165497, "learning_rate": 1.1761504622047293e-05, "loss": 0.0174, "step": 78830 }, { "grad_norm": 0.1416093409061432, "learning_rate": 1.1750853349335999e-05, "loss": 0.0119, "step": 78840 }, { "grad_norm": 0.3822934329509735, "learning_rate": 1.1740206259493874e-05, "loss": 0.0193, "step": 78850 }, { "grad_norm": 0.6153758764266968, "learning_rate": 1.1729563353685286e-05, "loss": 0.0187, "step": 78860 }, { "grad_norm": 0.26587626338005066, "learning_rate": 1.1718924633074091e-05, "loss": 0.0192, "step": 78870 }, { "grad_norm": 0.7320269346237183, "learning_rate": 1.1708290098823765e-05, "loss": 0.0226, "step": 78880 }, { "grad_norm": 0.16984789073467255, "learning_rate": 1.1697659752097245e-05, "loss": 0.0116, "step": 78890 }, { "grad_norm": 0.22397971153259277, "learning_rate": 1.1687033594057068e-05, "loss": 0.0164, "step": 78900 }, { "grad_norm": 0.4258178472518921, "learning_rate": 1.1676411625865297e-05, "loss": 0.0164, "step": 78910 }, { "grad_norm": 0.3007175624370575, "learning_rate": 1.1665793848683542e-05, "loss": 0.0209, "step": 78920 }, { "grad_norm": 0.26598837971687317, "learning_rate": 1.1655180263672927e-05, "loss": 0.019, "step": 78930 }, { "grad_norm": 0.5339335203170776, "learning_rate": 1.1644570871994142e-05, "loss": 0.0143, "step": 78940 }, { "grad_norm": 0.2832391858100891, "learning_rate": 1.163396567480744e-05, "loss": 0.021, "step": 78950 }, { "grad_norm": 0.3366535007953644, "learning_rate": 1.1623364673272552e-05, "loss": 0.0164, "step": 78960 }, { "grad_norm": 0.29668858647346497, "learning_rate": 1.1612767868548807e-05, "loss": 0.0188, "step": 78970 }, { "grad_norm": 0.42550352215766907, "learning_rate": 1.1602175261795057e-05, "loss": 0.0304, "step": 78980 }, { "grad_norm": 0.31839004158973694, "learning_rate": 1.1591586854169696e-05, "loss": 0.0157, "step": 78990 }, { "grad_norm": 0.4608873426914215, "learning_rate": 1.158100264683064e-05, "loss": 0.019, "step": 79000 }, { "grad_norm": 0.284719854593277, "learning_rate": 1.1570422640935386e-05, "loss": 0.0179, "step": 79010 }, { "grad_norm": 0.48221510648727417, "learning_rate": 1.155984683764092e-05, "loss": 0.0198, "step": 79020 }, { "grad_norm": 0.483950674533844, "learning_rate": 1.1549275238103812e-05, "loss": 0.0216, "step": 79030 }, { "grad_norm": 0.4497763216495514, "learning_rate": 1.1538707843480152e-05, "loss": 0.0194, "step": 79040 }, { "grad_norm": 0.29878729581832886, "learning_rate": 1.1528144654925593e-05, "loss": 0.0158, "step": 79050 }, { "grad_norm": 0.39268356561660767, "learning_rate": 1.1517585673595277e-05, "loss": 0.0123, "step": 79060 }, { "grad_norm": 0.39585453271865845, "learning_rate": 1.150703090064395e-05, "loss": 0.0151, "step": 79070 }, { "grad_norm": 0.19495074450969696, "learning_rate": 1.1496480337225835e-05, "loss": 0.0145, "step": 79080 }, { "grad_norm": 0.3936992883682251, "learning_rate": 1.1485933984494751e-05, "loss": 0.0203, "step": 79090 }, { "grad_norm": 0.5035050511360168, "learning_rate": 1.1475391843604e-05, "loss": 0.0402, "step": 79100 }, { "grad_norm": 0.33697831630706787, "learning_rate": 1.1464853915706503e-05, "loss": 0.0186, "step": 79110 }, { "grad_norm": 0.4510882794857025, "learning_rate": 1.1454320201954626e-05, "loss": 0.0285, "step": 79120 }, { "grad_norm": 0.37840983271598816, "learning_rate": 1.1443790703500356e-05, "loss": 0.0157, "step": 79130 }, { "grad_norm": 0.26578786969184875, "learning_rate": 1.1433265421495154e-05, "loss": 0.0119, "step": 79140 }, { "grad_norm": 0.37807387113571167, "learning_rate": 1.1422744357090065e-05, "loss": 0.0125, "step": 79150 }, { "grad_norm": 0.5466829538345337, "learning_rate": 1.1412227511435647e-05, "loss": 0.0304, "step": 79160 }, { "grad_norm": 0.17533615231513977, "learning_rate": 1.1401714885682025e-05, "loss": 0.0302, "step": 79170 }, { "grad_norm": 0.6386876702308655, "learning_rate": 1.1391206480978823e-05, "loss": 0.0267, "step": 79180 }, { "grad_norm": 0.48993349075317383, "learning_rate": 1.1380702298475227e-05, "loss": 0.0203, "step": 79190 }, { "grad_norm": 0.5733208656311035, "learning_rate": 1.1370202339319985e-05, "loss": 0.0224, "step": 79200 }, { "grad_norm": 0.5350093841552734, "learning_rate": 1.1359706604661319e-05, "loss": 0.0201, "step": 79210 }, { "grad_norm": 0.3646032512187958, "learning_rate": 1.1349215095647042e-05, "loss": 0.0163, "step": 79220 }, { "grad_norm": 0.5182647705078125, "learning_rate": 1.1338727813424493e-05, "loss": 0.0143, "step": 79230 }, { "grad_norm": 0.33496448397636414, "learning_rate": 1.132824475914056e-05, "loss": 0.0168, "step": 79240 }, { "grad_norm": 0.3164517283439636, "learning_rate": 1.131776593394162e-05, "loss": 0.0274, "step": 79250 }, { "grad_norm": 0.3308659791946411, "learning_rate": 1.1307291338973657e-05, "loss": 0.0163, "step": 79260 }, { "grad_norm": 0.3334071934223175, "learning_rate": 1.1296820975382121e-05, "loss": 0.0171, "step": 79270 }, { "grad_norm": 0.31720584630966187, "learning_rate": 1.1286354844312074e-05, "loss": 0.0443, "step": 79280 }, { "grad_norm": 0.6561388373374939, "learning_rate": 1.1275892946908023e-05, "loss": 0.0442, "step": 79290 }, { "grad_norm": 0.5173195600509644, "learning_rate": 1.126543528431413e-05, "loss": 0.0226, "step": 79300 }, { "grad_norm": 0.46205946803092957, "learning_rate": 1.125498185767398e-05, "loss": 0.0215, "step": 79310 }, { "grad_norm": 0.20226268470287323, "learning_rate": 1.1244532668130781e-05, "loss": 0.0135, "step": 79320 }, { "grad_norm": 0.3636792302131653, "learning_rate": 1.1234087716827203e-05, "loss": 0.0181, "step": 79330 }, { "grad_norm": 0.23045361042022705, "learning_rate": 1.1223647004905524e-05, "loss": 0.0208, "step": 79340 }, { "grad_norm": 0.3316347002983093, "learning_rate": 1.1213210533507485e-05, "loss": 0.0185, "step": 79350 }, { "grad_norm": 0.5035514235496521, "learning_rate": 1.1202778303774447e-05, "loss": 0.0161, "step": 79360 }, { "grad_norm": 0.6102706789970398, "learning_rate": 1.1192350316847234e-05, "loss": 0.017, "step": 79370 }, { "grad_norm": 0.41490843892097473, "learning_rate": 1.1181926573866258e-05, "loss": 0.0179, "step": 79380 }, { "grad_norm": 0.4188374876976013, "learning_rate": 1.1171507075971416e-05, "loss": 0.0164, "step": 79390 }, { "grad_norm": 0.2475927174091339, "learning_rate": 1.1161091824302177e-05, "loss": 0.014, "step": 79400 }, { "grad_norm": 0.6320706605911255, "learning_rate": 1.115068081999755e-05, "loss": 0.0243, "step": 79410 }, { "grad_norm": 0.3017667829990387, "learning_rate": 1.1140274064196071e-05, "loss": 0.015, "step": 79420 }, { "grad_norm": 0.49113938212394714, "learning_rate": 1.112987155803578e-05, "loss": 0.0148, "step": 79430 }, { "grad_norm": 0.6043796539306641, "learning_rate": 1.1119473302654298e-05, "loss": 0.0167, "step": 79440 }, { "grad_norm": 0.604714572429657, "learning_rate": 1.1109079299188769e-05, "loss": 0.0416, "step": 79450 }, { "grad_norm": 0.4213525056838989, "learning_rate": 1.1098689548775847e-05, "loss": 0.0262, "step": 79460 }, { "grad_norm": 0.4621042311191559, "learning_rate": 1.1088304052551762e-05, "loss": 0.0211, "step": 79470 }, { "grad_norm": 0.17677979171276093, "learning_rate": 1.1077922811652215e-05, "loss": 0.0139, "step": 79480 }, { "grad_norm": 0.3684217631816864, "learning_rate": 1.1067545827212534e-05, "loss": 0.0235, "step": 79490 }, { "grad_norm": 0.25137007236480713, "learning_rate": 1.1057173100367496e-05, "loss": 0.0126, "step": 79500 }, { "grad_norm": 0.38356146216392517, "learning_rate": 1.1046804632251474e-05, "loss": 0.0211, "step": 79510 }, { "grad_norm": 0.4400818943977356, "learning_rate": 1.1036440423998318e-05, "loss": 0.0189, "step": 79520 }, { "grad_norm": 0.5034945011138916, "learning_rate": 1.1026080476741469e-05, "loss": 0.0179, "step": 79530 }, { "grad_norm": 0.25411102175712585, "learning_rate": 1.1015724791613836e-05, "loss": 0.0181, "step": 79540 }, { "grad_norm": 0.29722604155540466, "learning_rate": 1.1005373369747951e-05, "loss": 0.0198, "step": 79550 }, { "grad_norm": 0.3035956621170044, "learning_rate": 1.09950262122758e-05, "loss": 0.0183, "step": 79560 }, { "grad_norm": 0.3651798963546753, "learning_rate": 1.0984683320328948e-05, "loss": 0.0238, "step": 79570 }, { "grad_norm": 0.21737435460090637, "learning_rate": 1.0974344695038458e-05, "loss": 0.0163, "step": 79580 }, { "grad_norm": 0.3806007504463196, "learning_rate": 1.0964010337534968e-05, "loss": 0.0183, "step": 79590 }, { "grad_norm": 0.42446842789649963, "learning_rate": 1.0953680248948611e-05, "loss": 0.0294, "step": 79600 }, { "grad_norm": 0.23707892000675201, "learning_rate": 1.0943354430409069e-05, "loss": 0.026, "step": 79610 }, { "grad_norm": 0.5610902905464172, "learning_rate": 1.0933032883045574e-05, "loss": 0.0175, "step": 79620 }, { "grad_norm": 0.18325987458229065, "learning_rate": 1.0922715607986872e-05, "loss": 0.0177, "step": 79630 }, { "grad_norm": 0.26413246989250183, "learning_rate": 1.0912402606361234e-05, "loss": 0.0311, "step": 79640 }, { "grad_norm": 0.28820544481277466, "learning_rate": 1.090209387929647e-05, "loss": 0.0167, "step": 79650 }, { "grad_norm": 0.2920369505882263, "learning_rate": 1.0891789427919957e-05, "loss": 0.0144, "step": 79660 }, { "grad_norm": 0.26837337017059326, "learning_rate": 1.088148925335854e-05, "loss": 0.019, "step": 79670 }, { "grad_norm": 0.17256514728069305, "learning_rate": 1.087119335673864e-05, "loss": 0.0183, "step": 79680 }, { "grad_norm": 0.4866051971912384, "learning_rate": 1.0860901739186208e-05, "loss": 0.0147, "step": 79690 }, { "grad_norm": 0.347629576921463, "learning_rate": 1.085061440182673e-05, "loss": 0.0107, "step": 79700 }, { "grad_norm": 0.394151508808136, "learning_rate": 1.0840331345785193e-05, "loss": 0.0193, "step": 79710 }, { "grad_norm": 0.19097791612148285, "learning_rate": 1.083005257218615e-05, "loss": 0.0187, "step": 79720 }, { "grad_norm": 0.18291406333446503, "learning_rate": 1.0819778082153642e-05, "loss": 0.0185, "step": 79730 }, { "grad_norm": 0.3380112051963806, "learning_rate": 1.0809507876811325e-05, "loss": 0.0243, "step": 79740 }, { "grad_norm": 0.21724924445152283, "learning_rate": 1.0799241957282292e-05, "loss": 0.0183, "step": 79750 }, { "grad_norm": 0.3441557288169861, "learning_rate": 1.078898032468923e-05, "loss": 0.0205, "step": 79760 }, { "grad_norm": 0.2641480267047882, "learning_rate": 1.0778722980154315e-05, "loss": 0.018, "step": 79770 }, { "grad_norm": 0.29515767097473145, "learning_rate": 1.0768469924799302e-05, "loss": 0.0176, "step": 79780 }, { "grad_norm": 0.5999544262886047, "learning_rate": 1.0758221159745419e-05, "loss": 0.0155, "step": 79790 }, { "grad_norm": 0.34536442160606384, "learning_rate": 1.074797668611347e-05, "loss": 0.0135, "step": 79800 }, { "grad_norm": 0.31913065910339355, "learning_rate": 1.0737736505023777e-05, "loss": 0.0228, "step": 79810 }, { "grad_norm": 0.22526536881923676, "learning_rate": 1.0727500617596203e-05, "loss": 0.0145, "step": 79820 }, { "grad_norm": 0.3038215935230255, "learning_rate": 1.0717269024950104e-05, "loss": 0.015, "step": 79830 }, { "grad_norm": 0.4748794734477997, "learning_rate": 1.070704172820442e-05, "loss": 0.0205, "step": 79840 }, { "grad_norm": 0.7015753984451294, "learning_rate": 1.0696818728477559e-05, "loss": 0.0276, "step": 79850 }, { "grad_norm": 0.24762314558029175, "learning_rate": 1.0686600026887511e-05, "loss": 0.0207, "step": 79860 }, { "grad_norm": 0.32245105504989624, "learning_rate": 1.0676385624551777e-05, "loss": 0.0151, "step": 79870 }, { "grad_norm": 0.4819428026676178, "learning_rate": 1.0666175522587402e-05, "loss": 0.0164, "step": 79880 }, { "grad_norm": 0.248176708817482, "learning_rate": 1.0655969722110925e-05, "loss": 0.0209, "step": 79890 }, { "grad_norm": 0.41624653339385986, "learning_rate": 1.0645768224238439e-05, "loss": 0.0228, "step": 79900 }, { "grad_norm": 0.41504600644111633, "learning_rate": 1.063557103008559e-05, "loss": 0.022, "step": 79910 }, { "grad_norm": 0.6014853119850159, "learning_rate": 1.0625378140767494e-05, "loss": 0.0162, "step": 79920 }, { "grad_norm": 0.7496892213821411, "learning_rate": 1.0615189557398846e-05, "loss": 0.028, "step": 79930 }, { "grad_norm": 0.3016098141670227, "learning_rate": 1.060500528109386e-05, "loss": 0.0173, "step": 79940 }, { "grad_norm": 0.13655595481395721, "learning_rate": 1.0594825312966272e-05, "loss": 0.0189, "step": 79950 }, { "grad_norm": 0.34505465626716614, "learning_rate": 1.0584649654129336e-05, "loss": 0.0201, "step": 79960 }, { "grad_norm": 0.2823081612586975, "learning_rate": 1.0574478305695863e-05, "loss": 0.0185, "step": 79970 }, { "grad_norm": 0.3388015627861023, "learning_rate": 1.0564311268778154e-05, "loss": 0.0153, "step": 79980 }, { "grad_norm": 0.3193046748638153, "learning_rate": 1.0554148544488073e-05, "loss": 0.0238, "step": 79990 }, { "grad_norm": 0.5520115494728088, "learning_rate": 1.0543990133937004e-05, "loss": 0.0193, "step": 80000 }, { "grad_norm": 0.6666634678840637, "learning_rate": 1.0533836038235868e-05, "loss": 0.0151, "step": 80010 }, { "grad_norm": 0.3973402976989746, "learning_rate": 1.052368625849507e-05, "loss": 0.0195, "step": 80020 }, { "grad_norm": 0.5340097546577454, "learning_rate": 1.05135407958246e-05, "loss": 0.0246, "step": 80030 }, { "grad_norm": 0.2492111176252365, "learning_rate": 1.0503399651333934e-05, "loss": 0.0106, "step": 80040 }, { "grad_norm": 0.2325439155101776, "learning_rate": 1.04932628261321e-05, "loss": 0.0148, "step": 80050 }, { "grad_norm": 0.48551279306411743, "learning_rate": 1.0483130321327644e-05, "loss": 0.0306, "step": 80060 }, { "grad_norm": 0.5665409564971924, "learning_rate": 1.0473002138028654e-05, "loss": 0.0135, "step": 80070 }, { "grad_norm": 0.2923957109451294, "learning_rate": 1.0462878277342713e-05, "loss": 0.0208, "step": 80080 }, { "grad_norm": 0.19683952629566193, "learning_rate": 1.0452758740376972e-05, "loss": 0.017, "step": 80090 }, { "grad_norm": 0.2791449725627899, "learning_rate": 1.0442643528238067e-05, "loss": 0.0212, "step": 80100 }, { "grad_norm": 0.5388031005859375, "learning_rate": 1.0432532642032188e-05, "loss": 0.0205, "step": 80110 }, { "grad_norm": 0.5695989727973938, "learning_rate": 1.042242608286505e-05, "loss": 0.0159, "step": 80120 }, { "grad_norm": 0.3694205582141876, "learning_rate": 1.041232385184191e-05, "loss": 0.0202, "step": 80130 }, { "grad_norm": 0.5045957565307617, "learning_rate": 1.0402225950067495e-05, "loss": 0.0281, "step": 80140 }, { "grad_norm": 0.6452594995498657, "learning_rate": 1.0392132378646119e-05, "loss": 0.0151, "step": 80150 }, { "grad_norm": 0.6037416458129883, "learning_rate": 1.0382043138681607e-05, "loss": 0.0174, "step": 80160 }, { "grad_norm": 0.9514095783233643, "learning_rate": 1.0371958231277284e-05, "loss": 0.0144, "step": 80170 }, { "grad_norm": 0.4876919090747833, "learning_rate": 1.0361877657536024e-05, "loss": 0.0278, "step": 80180 }, { "grad_norm": 0.3883875012397766, "learning_rate": 1.035180141856023e-05, "loss": 0.023, "step": 80190 }, { "grad_norm": 0.31304165720939636, "learning_rate": 1.0341729515451836e-05, "loss": 0.0195, "step": 80200 }, { "grad_norm": 0.4335811138153076, "learning_rate": 1.0331661949312265e-05, "loss": 0.0169, "step": 80210 }, { "grad_norm": 0.2564447820186615, "learning_rate": 1.0321598721242514e-05, "loss": 0.0153, "step": 80220 }, { "grad_norm": 0.4123676121234894, "learning_rate": 1.0311539832343064e-05, "loss": 0.0132, "step": 80230 }, { "grad_norm": 0.2126748263835907, "learning_rate": 1.0301485283713942e-05, "loss": 0.0206, "step": 80240 }, { "grad_norm": 0.3044528067111969, "learning_rate": 1.0291435076454703e-05, "loss": 0.0144, "step": 80250 }, { "grad_norm": 0.32885071635246277, "learning_rate": 1.028138921166444e-05, "loss": 0.0248, "step": 80260 }, { "grad_norm": 0.4581831991672516, "learning_rate": 1.0271347690441719e-05, "loss": 0.0206, "step": 80270 }, { "grad_norm": 0.29649361968040466, "learning_rate": 1.0261310513884698e-05, "loss": 0.0257, "step": 80280 }, { "grad_norm": 0.4059104323387146, "learning_rate": 1.0251277683090998e-05, "loss": 0.0251, "step": 80290 }, { "grad_norm": 0.7488739490509033, "learning_rate": 1.0241249199157826e-05, "loss": 0.0317, "step": 80300 }, { "grad_norm": 0.2037191390991211, "learning_rate": 1.0231225063181837e-05, "loss": 0.0159, "step": 80310 }, { "grad_norm": 0.3849278688430786, "learning_rate": 1.0221205276259316e-05, "loss": 0.0269, "step": 80320 }, { "grad_norm": 0.5240768790245056, "learning_rate": 1.0211189839485958e-05, "loss": 0.0126, "step": 80330 }, { "grad_norm": 0.5006541013717651, "learning_rate": 1.0201178753957075e-05, "loss": 0.01, "step": 80340 }, { "grad_norm": 0.3717540204524994, "learning_rate": 1.0191172020767432e-05, "loss": 0.0177, "step": 80350 }, { "grad_norm": 0.16699627041816711, "learning_rate": 1.0181169641011362e-05, "loss": 0.017, "step": 80360 }, { "grad_norm": 0.2701728940010071, "learning_rate": 1.0171171615782721e-05, "loss": 0.025, "step": 80370 }, { "grad_norm": 0.4052352011203766, "learning_rate": 1.0161177946174866e-05, "loss": 0.0172, "step": 80380 }, { "grad_norm": 0.3260120451450348, "learning_rate": 1.0151188633280706e-05, "loss": 0.0157, "step": 80390 }, { "grad_norm": 0.4127190113067627, "learning_rate": 1.0141203678192635e-05, "loss": 0.0171, "step": 80400 }, { "grad_norm": 0.33142873644828796, "learning_rate": 1.0131223082002612e-05, "loss": 0.0204, "step": 80410 }, { "grad_norm": 0.14372007548809052, "learning_rate": 1.012124684580208e-05, "loss": 0.0177, "step": 80420 }, { "grad_norm": 0.16893166303634644, "learning_rate": 1.0111274970682039e-05, "loss": 0.0195, "step": 80430 }, { "grad_norm": 0.3124176561832428, "learning_rate": 1.010130745773299e-05, "loss": 0.0174, "step": 80440 }, { "grad_norm": 0.2374078929424286, "learning_rate": 1.0091344308044986e-05, "loss": 0.0255, "step": 80450 }, { "grad_norm": 0.42719998955726624, "learning_rate": 1.0081385522707554e-05, "loss": 0.0154, "step": 80460 }, { "grad_norm": 0.27760887145996094, "learning_rate": 1.0071431102809803e-05, "loss": 0.0267, "step": 80470 }, { "grad_norm": 1.1241379976272583, "learning_rate": 1.0061481049440297e-05, "loss": 0.0171, "step": 80480 }, { "grad_norm": 0.2557483911514282, "learning_rate": 1.0051535363687187e-05, "loss": 0.013, "step": 80490 }, { "grad_norm": 0.3853861391544342, "learning_rate": 1.0041594046638087e-05, "loss": 0.0174, "step": 80500 }, { "grad_norm": 0.7104094624519348, "learning_rate": 1.0031657099380204e-05, "loss": 0.0188, "step": 80510 }, { "grad_norm": 0.280723512172699, "learning_rate": 1.0021724523000204e-05, "loss": 0.0212, "step": 80520 }, { "grad_norm": 0.5078167915344238, "learning_rate": 1.001179631858431e-05, "loss": 0.0374, "step": 80530 }, { "grad_norm": 0.4067346751689911, "learning_rate": 1.0001872487218245e-05, "loss": 0.0153, "step": 80540 }, { "grad_norm": 0.38668370246887207, "learning_rate": 9.991953029987272e-06, "loss": 0.0155, "step": 80550 }, { "grad_norm": 0.1587337702512741, "learning_rate": 9.982037947976147e-06, "loss": 0.0169, "step": 80560 }, { "grad_norm": 0.3591119945049286, "learning_rate": 9.972127242269208e-06, "loss": 0.0282, "step": 80570 }, { "grad_norm": 0.251051127910614, "learning_rate": 9.962220913950238e-06, "loss": 0.013, "step": 80580 }, { "grad_norm": 0.3002515435218811, "learning_rate": 9.952318964102591e-06, "loss": 0.0252, "step": 80590 }, { "grad_norm": 0.5016790628433228, "learning_rate": 9.942421393809148e-06, "loss": 0.0206, "step": 80600 }, { "grad_norm": 0.632861852645874, "learning_rate": 9.932528204152264e-06, "loss": 0.0231, "step": 80610 }, { "grad_norm": 0.3328024446964264, "learning_rate": 9.922639396213856e-06, "loss": 0.0227, "step": 80620 }, { "grad_norm": 0.4361996054649353, "learning_rate": 9.912754971075344e-06, "loss": 0.0162, "step": 80630 }, { "grad_norm": 0.2550394535064697, "learning_rate": 9.902874929817696e-06, "loss": 0.0147, "step": 80640 }, { "grad_norm": 0.468808114528656, "learning_rate": 9.892999273521342e-06, "loss": 0.0192, "step": 80650 }, { "grad_norm": 0.36077478528022766, "learning_rate": 9.883128003266302e-06, "loss": 0.0204, "step": 80660 }, { "grad_norm": 0.3229256272315979, "learning_rate": 9.87326112013206e-06, "loss": 0.0161, "step": 80670 }, { "grad_norm": 0.3866502642631531, "learning_rate": 9.863398625197662e-06, "loss": 0.016, "step": 80680 }, { "grad_norm": 0.19939737021923065, "learning_rate": 9.853540519541615e-06, "loss": 0.0228, "step": 80690 }, { "grad_norm": 0.45074188709259033, "learning_rate": 9.843686804242053e-06, "loss": 0.0156, "step": 80700 }, { "grad_norm": 0.22849996387958527, "learning_rate": 9.833837480376506e-06, "loss": 0.0158, "step": 80710 }, { "grad_norm": 0.22988836467266083, "learning_rate": 9.82399254902212e-06, "loss": 0.0135, "step": 80720 }, { "grad_norm": 0.29445403814315796, "learning_rate": 9.814152011255494e-06, "loss": 0.0139, "step": 80730 }, { "grad_norm": 0.3379943370819092, "learning_rate": 9.804315868152797e-06, "loss": 0.0214, "step": 80740 }, { "grad_norm": 0.2910797595977783, "learning_rate": 9.79448412078966e-06, "loss": 0.014, "step": 80750 }, { "grad_norm": 0.32888826727867126, "learning_rate": 9.784656770241312e-06, "loss": 0.0189, "step": 80760 }, { "grad_norm": 0.2641822397708893, "learning_rate": 9.77483381758243e-06, "loss": 0.0149, "step": 80770 }, { "grad_norm": 0.26920750737190247, "learning_rate": 9.76501526388725e-06, "loss": 0.0163, "step": 80780 }, { "grad_norm": 0.28304654359817505, "learning_rate": 9.7552011102295e-06, "loss": 0.0119, "step": 80790 }, { "grad_norm": 0.45119616389274597, "learning_rate": 9.745391357682449e-06, "loss": 0.0278, "step": 80800 }, { "grad_norm": 0.7655418515205383, "learning_rate": 9.735586007318875e-06, "loss": 0.015, "step": 80810 }, { "grad_norm": 0.46747705340385437, "learning_rate": 9.725785060211096e-06, "loss": 0.0165, "step": 80820 }, { "grad_norm": 0.5321580171585083, "learning_rate": 9.715988517430896e-06, "loss": 0.0191, "step": 80830 }, { "grad_norm": 0.19581790268421173, "learning_rate": 9.706196380049625e-06, "loss": 0.019, "step": 80840 }, { "grad_norm": 0.573164165019989, "learning_rate": 9.69640864913815e-06, "loss": 0.0248, "step": 80850 }, { "grad_norm": 0.2849571406841278, "learning_rate": 9.686625325766818e-06, "loss": 0.0173, "step": 80860 }, { "grad_norm": 0.3479829728603363, "learning_rate": 9.67684641100554e-06, "loss": 0.0156, "step": 80870 }, { "grad_norm": 0.30439263582229614, "learning_rate": 9.667071905923692e-06, "loss": 0.017, "step": 80880 }, { "grad_norm": 0.23155266046524048, "learning_rate": 9.657301811590242e-06, "loss": 0.0148, "step": 80890 }, { "grad_norm": 0.5033516883850098, "learning_rate": 9.647536129073603e-06, "loss": 0.0142, "step": 80900 }, { "grad_norm": 0.42790430784225464, "learning_rate": 9.637774859441751e-06, "loss": 0.018, "step": 80910 }, { "grad_norm": 0.22855281829833984, "learning_rate": 9.628018003762152e-06, "loss": 0.0266, "step": 80920 }, { "grad_norm": 0.32776838541030884, "learning_rate": 9.618265563101813e-06, "loss": 0.0137, "step": 80930 }, { "grad_norm": 0.4510840177536011, "learning_rate": 9.608517538527211e-06, "loss": 0.0134, "step": 80940 }, { "grad_norm": 0.36791864037513733, "learning_rate": 9.598773931104431e-06, "loss": 0.0219, "step": 80950 }, { "grad_norm": 0.24544651806354523, "learning_rate": 9.589034741898983e-06, "loss": 0.0189, "step": 80960 }, { "grad_norm": 0.3381585478782654, "learning_rate": 9.579299971975947e-06, "loss": 0.0194, "step": 80970 }, { "grad_norm": 0.5161088705062866, "learning_rate": 9.569569622399882e-06, "loss": 0.0147, "step": 80980 }, { "grad_norm": 0.2532898783683777, "learning_rate": 9.559843694234916e-06, "loss": 0.0143, "step": 80990 }, { "grad_norm": 0.48137030005455017, "learning_rate": 9.550122188544625e-06, "loss": 0.0177, "step": 81000 }, { "grad_norm": 0.40515047311782837, "learning_rate": 9.540405106392152e-06, "loss": 0.0152, "step": 81010 }, { "grad_norm": 0.38125190138816833, "learning_rate": 9.53069244884015e-06, "loss": 0.0184, "step": 81020 }, { "grad_norm": 0.43228188157081604, "learning_rate": 9.52098421695079e-06, "loss": 0.0122, "step": 81030 }, { "grad_norm": 0.34211230278015137, "learning_rate": 9.511280411785728e-06, "loss": 0.0172, "step": 81040 }, { "grad_norm": 0.36431068181991577, "learning_rate": 9.50158103440616e-06, "loss": 0.0274, "step": 81050 }, { "grad_norm": 0.18769223988056183, "learning_rate": 9.491886085872815e-06, "loss": 0.0134, "step": 81060 }, { "grad_norm": 0.37166067957878113, "learning_rate": 9.482195567245895e-06, "loss": 0.014, "step": 81070 }, { "grad_norm": 0.250235378742218, "learning_rate": 9.472509479585151e-06, "loss": 0.0147, "step": 81080 }, { "grad_norm": 0.4693199098110199, "learning_rate": 9.462827823949833e-06, "loss": 0.0253, "step": 81090 }, { "grad_norm": 0.44214314222335815, "learning_rate": 9.453150601398735e-06, "loss": 0.0166, "step": 81100 }, { "grad_norm": 0.3476523458957672, "learning_rate": 9.44347781299011e-06, "loss": 0.0138, "step": 81110 }, { "grad_norm": 0.22277358174324036, "learning_rate": 9.43380945978179e-06, "loss": 0.0148, "step": 81120 }, { "grad_norm": 0.7280287146568298, "learning_rate": 9.424145542831047e-06, "loss": 0.0182, "step": 81130 }, { "grad_norm": 0.31002604961395264, "learning_rate": 9.414486063194778e-06, "loss": 0.0157, "step": 81140 }, { "grad_norm": 0.22942177951335907, "learning_rate": 9.404831021929273e-06, "loss": 0.02, "step": 81150 }, { "grad_norm": 0.17868326604366302, "learning_rate": 9.395180420090427e-06, "loss": 0.0123, "step": 81160 }, { "grad_norm": 0.6240208148956299, "learning_rate": 9.385534258733591e-06, "loss": 0.025, "step": 81170 }, { "grad_norm": 0.5046596527099609, "learning_rate": 9.375892538913677e-06, "loss": 0.0167, "step": 81180 }, { "grad_norm": 0.27535387873649597, "learning_rate": 9.36625526168507e-06, "loss": 0.0242, "step": 81190 }, { "grad_norm": 0.4483073353767395, "learning_rate": 9.356622428101686e-06, "loss": 0.0169, "step": 81200 }, { "grad_norm": 0.2654660642147064, "learning_rate": 9.346994039216972e-06, "loss": 0.0178, "step": 81210 }, { "grad_norm": 0.4520235061645508, "learning_rate": 9.337370096083886e-06, "loss": 0.017, "step": 81220 }, { "grad_norm": 0.2550322711467743, "learning_rate": 9.327750599754854e-06, "loss": 0.0138, "step": 81230 }, { "grad_norm": 0.32943668961524963, "learning_rate": 9.318135551281875e-06, "loss": 0.0136, "step": 81240 }, { "grad_norm": 0.479049414396286, "learning_rate": 9.308524951716424e-06, "loss": 0.0141, "step": 81250 }, { "grad_norm": 0.3304324746131897, "learning_rate": 9.298918802109497e-06, "loss": 0.022, "step": 81260 }, { "grad_norm": 0.33009225130081177, "learning_rate": 9.289317103511625e-06, "loss": 0.0188, "step": 81270 }, { "grad_norm": 0.4306018352508545, "learning_rate": 9.279719856972835e-06, "loss": 0.0155, "step": 81280 }, { "grad_norm": 0.5580460429191589, "learning_rate": 9.270127063542645e-06, "loss": 0.0207, "step": 81290 }, { "grad_norm": 0.23390530049800873, "learning_rate": 9.260538724270124e-06, "loss": 0.0146, "step": 81300 }, { "grad_norm": 0.7406839728355408, "learning_rate": 9.250954840203852e-06, "loss": 0.0289, "step": 81310 }, { "grad_norm": 0.3447372019290924, "learning_rate": 9.241375412391878e-06, "loss": 0.0251, "step": 81320 }, { "grad_norm": 0.3037140667438507, "learning_rate": 9.23180044188181e-06, "loss": 0.0242, "step": 81330 }, { "grad_norm": 0.2983970642089844, "learning_rate": 9.222229929720755e-06, "loss": 0.0289, "step": 81340 }, { "grad_norm": 0.30949366092681885, "learning_rate": 9.212663876955336e-06, "loss": 0.0161, "step": 81350 }, { "grad_norm": 0.2351861447095871, "learning_rate": 9.20310228463166e-06, "loss": 0.0118, "step": 81360 }, { "grad_norm": 0.11327462643384933, "learning_rate": 9.19354515379539e-06, "loss": 0.0177, "step": 81370 }, { "grad_norm": 0.3222094178199768, "learning_rate": 9.183992485491666e-06, "loss": 0.0136, "step": 81380 }, { "grad_norm": 0.34042224287986755, "learning_rate": 9.174444280765154e-06, "loss": 0.0167, "step": 81390 }, { "grad_norm": 0.11400753259658813, "learning_rate": 9.164900540660032e-06, "loss": 0.0169, "step": 81400 }, { "grad_norm": 0.5068635940551758, "learning_rate": 9.155361266220008e-06, "loss": 0.0115, "step": 81410 }, { "grad_norm": 0.4396839439868927, "learning_rate": 9.145826458488254e-06, "loss": 0.0213, "step": 81420 }, { "grad_norm": 0.42601776123046875, "learning_rate": 9.136296118507509e-06, "loss": 0.0259, "step": 81430 }, { "grad_norm": 0.43102729320526123, "learning_rate": 9.126770247319966e-06, "loss": 0.0207, "step": 81440 }, { "grad_norm": 0.17345955967903137, "learning_rate": 9.11724884596738e-06, "loss": 0.015, "step": 81450 }, { "grad_norm": 0.8654612302780151, "learning_rate": 9.107731915490996e-06, "loss": 0.0145, "step": 81460 }, { "grad_norm": 0.18979638814926147, "learning_rate": 9.098219456931579e-06, "loss": 0.0135, "step": 81470 }, { "grad_norm": 0.3533066511154175, "learning_rate": 9.088711471329376e-06, "loss": 0.0181, "step": 81480 }, { "grad_norm": 0.49803197383880615, "learning_rate": 9.079207959724184e-06, "loss": 0.0232, "step": 81490 }, { "grad_norm": 0.20220094919204712, "learning_rate": 9.069708923155279e-06, "loss": 0.0184, "step": 81500 }, { "grad_norm": 0.19994144141674042, "learning_rate": 9.060214362661468e-06, "loss": 0.0115, "step": 81510 }, { "grad_norm": 0.41502121090888977, "learning_rate": 9.050724279281064e-06, "loss": 0.0151, "step": 81520 }, { "grad_norm": 0.5333042740821838, "learning_rate": 9.041238674051894e-06, "loss": 0.0201, "step": 81530 }, { "grad_norm": 0.17232772707939148, "learning_rate": 9.031757548011277e-06, "loss": 0.014, "step": 81540 }, { "grad_norm": 0.41450199484825134, "learning_rate": 9.022280902196051e-06, "loss": 0.0135, "step": 81550 }, { "grad_norm": 0.3527314066886902, "learning_rate": 9.012808737642592e-06, "loss": 0.0294, "step": 81560 }, { "grad_norm": 0.3869919180870056, "learning_rate": 9.003341055386739e-06, "loss": 0.0217, "step": 81570 }, { "grad_norm": 0.5062971115112305, "learning_rate": 8.993877856463862e-06, "loss": 0.0219, "step": 81580 }, { "grad_norm": 0.424622118473053, "learning_rate": 8.984419141908856e-06, "loss": 0.0161, "step": 81590 }, { "grad_norm": 0.31946754455566406, "learning_rate": 8.974964912756117e-06, "loss": 0.0107, "step": 81600 }, { "grad_norm": 0.6882067918777466, "learning_rate": 8.965515170039523e-06, "loss": 0.0298, "step": 81610 }, { "grad_norm": 0.5861371159553528, "learning_rate": 8.956069914792514e-06, "loss": 0.015, "step": 81620 }, { "grad_norm": 0.2699701189994812, "learning_rate": 8.946629148047969e-06, "loss": 0.0174, "step": 81630 }, { "grad_norm": 0.5728534460067749, "learning_rate": 8.93719287083834e-06, "loss": 0.0251, "step": 81640 }, { "grad_norm": 0.5644391775131226, "learning_rate": 8.927761084195569e-06, "loss": 0.016, "step": 81650 }, { "grad_norm": 0.21118353307247162, "learning_rate": 8.918333789151096e-06, "loss": 0.011, "step": 81660 }, { "grad_norm": 0.5070578455924988, "learning_rate": 8.90891098673587e-06, "loss": 0.0222, "step": 81670 }, { "grad_norm": 0.25811243057250977, "learning_rate": 8.89949267798037e-06, "loss": 0.0157, "step": 81680 }, { "grad_norm": 0.1783217042684555, "learning_rate": 8.890078863914542e-06, "loss": 0.0183, "step": 81690 }, { "grad_norm": 0.7505279779434204, "learning_rate": 8.880669545567894e-06, "loss": 0.0098, "step": 81700 }, { "grad_norm": 0.21616245806217194, "learning_rate": 8.871264723969381e-06, "loss": 0.0132, "step": 81710 }, { "grad_norm": 0.35295912623405457, "learning_rate": 8.86186440014754e-06, "loss": 0.0165, "step": 81720 }, { "grad_norm": 0.3760607838630676, "learning_rate": 8.852468575130351e-06, "loss": 0.0201, "step": 81730 }, { "grad_norm": 0.32614588737487793, "learning_rate": 8.843077249945347e-06, "loss": 0.0134, "step": 81740 }, { "grad_norm": 0.38384106755256653, "learning_rate": 8.833690425619523e-06, "loss": 0.0141, "step": 81750 }, { "grad_norm": 0.4638996422290802, "learning_rate": 8.82430810317942e-06, "loss": 0.0199, "step": 81760 }, { "grad_norm": 0.20379678905010223, "learning_rate": 8.814930283651074e-06, "loss": 0.0163, "step": 81770 }, { "grad_norm": 0.43802544474601746, "learning_rate": 8.805556968060048e-06, "loss": 0.0192, "step": 81780 }, { "grad_norm": 0.3002045154571533, "learning_rate": 8.796188157431362e-06, "loss": 0.0124, "step": 81790 }, { "grad_norm": 0.1134154200553894, "learning_rate": 8.786823852789595e-06, "loss": 0.0195, "step": 81800 }, { "grad_norm": 0.18820901215076447, "learning_rate": 8.777464055158813e-06, "loss": 0.0198, "step": 81810 }, { "grad_norm": 0.3899526000022888, "learning_rate": 8.768108765562578e-06, "loss": 0.0185, "step": 81820 }, { "grad_norm": 0.2540113925933838, "learning_rate": 8.758757985023975e-06, "loss": 0.0152, "step": 81830 }, { "grad_norm": 0.3000129461288452, "learning_rate": 8.749411714565598e-06, "loss": 0.0111, "step": 81840 }, { "grad_norm": 0.309528648853302, "learning_rate": 8.74006995520954e-06, "loss": 0.0166, "step": 81850 }, { "grad_norm": 1.083815097808838, "learning_rate": 8.730732707977397e-06, "loss": 0.0127, "step": 81860 }, { "grad_norm": 0.17613625526428223, "learning_rate": 8.721399973890281e-06, "loss": 0.0207, "step": 81870 }, { "grad_norm": 0.19976919889450073, "learning_rate": 8.712071753968792e-06, "loss": 0.0176, "step": 81880 }, { "grad_norm": 0.1308770328760147, "learning_rate": 8.702748049233072e-06, "loss": 0.0137, "step": 81890 }, { "grad_norm": 0.2067824751138687, "learning_rate": 8.693428860702718e-06, "loss": 0.0172, "step": 81900 }, { "grad_norm": 0.1155957281589508, "learning_rate": 8.684114189396898e-06, "loss": 0.0148, "step": 81910 }, { "grad_norm": 0.5286649465560913, "learning_rate": 8.674804036334222e-06, "loss": 0.0197, "step": 81920 }, { "grad_norm": 0.31332671642303467, "learning_rate": 8.665498402532857e-06, "loss": 0.0178, "step": 81930 }, { "grad_norm": 0.5609434843063354, "learning_rate": 8.656197289010432e-06, "loss": 0.0185, "step": 81940 }, { "grad_norm": 0.3642835021018982, "learning_rate": 8.646900696784122e-06, "loss": 0.0166, "step": 81950 }, { "grad_norm": 0.29970577359199524, "learning_rate": 8.637608626870553e-06, "loss": 0.0193, "step": 81960 }, { "grad_norm": 0.17347019910812378, "learning_rate": 8.628321080285944e-06, "loss": 0.0124, "step": 81970 }, { "grad_norm": 0.37875741720199585, "learning_rate": 8.619038058045925e-06, "loss": 0.0205, "step": 81980 }, { "grad_norm": 0.6113789677619934, "learning_rate": 8.609759561165704e-06, "loss": 0.0191, "step": 81990 }, { "grad_norm": 0.4385148882865906, "learning_rate": 8.600485590659934e-06, "loss": 0.0209, "step": 82000 }, { "grad_norm": 0.35036700963974, "learning_rate": 8.591216147542813e-06, "loss": 0.017, "step": 82010 }, { "grad_norm": 0.17166011035442352, "learning_rate": 8.581951232828045e-06, "loss": 0.0156, "step": 82020 }, { "grad_norm": 0.34763407707214355, "learning_rate": 8.572690847528825e-06, "loss": 0.0185, "step": 82030 }, { "grad_norm": 0.28757330775260925, "learning_rate": 8.563434992657838e-06, "loss": 0.0107, "step": 82040 }, { "grad_norm": 0.21689769625663757, "learning_rate": 8.554183669227307e-06, "loss": 0.0135, "step": 82050 }, { "grad_norm": 0.8403670191764832, "learning_rate": 8.544936878248937e-06, "loss": 0.0316, "step": 82060 }, { "grad_norm": 0.5139225721359253, "learning_rate": 8.535694620733942e-06, "loss": 0.0279, "step": 82070 }, { "grad_norm": 0.4164554476737976, "learning_rate": 8.526456897693053e-06, "loss": 0.0145, "step": 82080 }, { "grad_norm": 0.17043983936309814, "learning_rate": 8.517223710136456e-06, "loss": 0.0267, "step": 82090 }, { "grad_norm": 0.1713833063840866, "learning_rate": 8.507995059073931e-06, "loss": 0.0176, "step": 82100 }, { "grad_norm": 0.4396677017211914, "learning_rate": 8.498770945514673e-06, "loss": 0.0167, "step": 82110 }, { "grad_norm": 0.37273120880126953, "learning_rate": 8.489551370467436e-06, "loss": 0.0151, "step": 82120 }, { "grad_norm": 0.32821303606033325, "learning_rate": 8.480336334940442e-06, "loss": 0.0265, "step": 82130 }, { "grad_norm": 0.27079230546951294, "learning_rate": 8.471125839941458e-06, "loss": 0.0231, "step": 82140 }, { "grad_norm": 0.5139828324317932, "learning_rate": 8.461919886477682e-06, "loss": 0.0162, "step": 82150 }, { "grad_norm": 0.2741728127002716, "learning_rate": 8.452718475555927e-06, "loss": 0.0144, "step": 82160 }, { "grad_norm": 0.3346966505050659, "learning_rate": 8.443521608182398e-06, "loss": 0.0249, "step": 82170 }, { "grad_norm": 0.5938899517059326, "learning_rate": 8.43432928536288e-06, "loss": 0.0236, "step": 82180 }, { "grad_norm": 0.18995527923107147, "learning_rate": 8.425141508102607e-06, "loss": 0.0151, "step": 82190 }, { "grad_norm": 0.16057422757148743, "learning_rate": 8.415958277406365e-06, "loss": 0.0178, "step": 82200 }, { "grad_norm": 0.3661581873893738, "learning_rate": 8.406779594278392e-06, "loss": 0.0164, "step": 82210 }, { "grad_norm": 0.27426084876060486, "learning_rate": 8.397605459722468e-06, "loss": 0.0168, "step": 82220 }, { "grad_norm": 0.4338720440864563, "learning_rate": 8.388435874741863e-06, "loss": 0.0147, "step": 82230 }, { "grad_norm": 0.2279062718153, "learning_rate": 8.37927084033936e-06, "loss": 0.0199, "step": 82240 }, { "grad_norm": 0.5352409482002258, "learning_rate": 8.370110357517208e-06, "loss": 0.0204, "step": 82250 }, { "grad_norm": 0.39159923791885376, "learning_rate": 8.360954427277201e-06, "loss": 0.0236, "step": 82260 }, { "grad_norm": 0.30996042490005493, "learning_rate": 8.351803050620627e-06, "loss": 0.0172, "step": 82270 }, { "grad_norm": 0.5491981506347656, "learning_rate": 8.34265622854824e-06, "loss": 0.0187, "step": 82280 }, { "grad_norm": 0.43868520855903625, "learning_rate": 8.333513962060335e-06, "loss": 0.0137, "step": 82290 }, { "grad_norm": 0.44625723361968994, "learning_rate": 8.324376252156701e-06, "loss": 0.0166, "step": 82300 }, { "grad_norm": 0.4777933359146118, "learning_rate": 8.315243099836633e-06, "loss": 0.0274, "step": 82310 }, { "grad_norm": 0.25712764263153076, "learning_rate": 8.3061145060989e-06, "loss": 0.0187, "step": 82320 }, { "grad_norm": 0.25192806124687195, "learning_rate": 8.296990471941806e-06, "loss": 0.0125, "step": 82330 }, { "grad_norm": 0.41825172305107117, "learning_rate": 8.287870998363112e-06, "loss": 0.0223, "step": 82340 }, { "grad_norm": 0.2798709571361542, "learning_rate": 8.278756086360156e-06, "loss": 0.0133, "step": 82350 }, { "grad_norm": 0.3367806673049927, "learning_rate": 8.269645736929693e-06, "loss": 0.0137, "step": 82360 }, { "grad_norm": 0.25731682777404785, "learning_rate": 8.260539951068048e-06, "loss": 0.014, "step": 82370 }, { "grad_norm": 0.40258699655532837, "learning_rate": 8.251438729770983e-06, "loss": 0.0152, "step": 82380 }, { "grad_norm": 0.4455309212207794, "learning_rate": 8.242342074033826e-06, "loss": 0.0157, "step": 82390 }, { "grad_norm": 0.35424357652664185, "learning_rate": 8.233249984851344e-06, "loss": 0.0112, "step": 82400 }, { "grad_norm": 0.5578779578208923, "learning_rate": 8.224162463217849e-06, "loss": 0.0211, "step": 82410 }, { "grad_norm": 0.1509818583726883, "learning_rate": 8.215079510127139e-06, "loss": 0.0125, "step": 82420 }, { "grad_norm": 0.2512648403644562, "learning_rate": 8.206001126572521e-06, "loss": 0.0189, "step": 82430 }, { "grad_norm": 0.18836280703544617, "learning_rate": 8.196927313546776e-06, "loss": 0.0102, "step": 82440 }, { "grad_norm": 0.3216821849346161, "learning_rate": 8.18785807204222e-06, "loss": 0.0203, "step": 82450 }, { "grad_norm": 0.3405193090438843, "learning_rate": 8.178793403050627e-06, "loss": 0.0282, "step": 82460 }, { "grad_norm": 0.47309839725494385, "learning_rate": 8.16973330756331e-06, "loss": 0.0168, "step": 82470 }, { "grad_norm": 0.2904690206050873, "learning_rate": 8.16067778657107e-06, "loss": 0.0114, "step": 82480 }, { "grad_norm": 0.297101765871048, "learning_rate": 8.151626841064214e-06, "loss": 0.0141, "step": 82490 }, { "grad_norm": 0.15142475068569183, "learning_rate": 8.142580472032518e-06, "loss": 0.0202, "step": 82500 }, { "grad_norm": 0.24560324847698212, "learning_rate": 8.133538680465285e-06, "loss": 0.033, "step": 82510 }, { "grad_norm": 0.14191414415836334, "learning_rate": 8.124501467351336e-06, "loss": 0.0179, "step": 82520 }, { "grad_norm": 0.3946243226528168, "learning_rate": 8.11546883367893e-06, "loss": 0.0131, "step": 82530 }, { "grad_norm": 0.177127867937088, "learning_rate": 8.106440780435882e-06, "loss": 0.0206, "step": 82540 }, { "grad_norm": 0.2523236870765686, "learning_rate": 8.097417308609485e-06, "loss": 0.0179, "step": 82550 }, { "grad_norm": 0.2353014051914215, "learning_rate": 8.088398419186544e-06, "loss": 0.0145, "step": 82560 }, { "grad_norm": 0.24409091472625732, "learning_rate": 8.079384113153321e-06, "loss": 0.0098, "step": 82570 }, { "grad_norm": 0.3416798412799835, "learning_rate": 8.070374391495644e-06, "loss": 0.0153, "step": 82580 }, { "grad_norm": 0.3046371042728424, "learning_rate": 8.061369255198764e-06, "loss": 0.0168, "step": 82590 }, { "grad_norm": 0.6020464301109314, "learning_rate": 8.052368705247493e-06, "loss": 0.0148, "step": 82600 }, { "grad_norm": 0.25699669122695923, "learning_rate": 8.043372742626114e-06, "loss": 0.018, "step": 82610 }, { "grad_norm": 0.324523389339447, "learning_rate": 8.034381368318422e-06, "loss": 0.0176, "step": 82620 }, { "grad_norm": 0.23263385891914368, "learning_rate": 8.025394583307677e-06, "loss": 0.014, "step": 82630 }, { "grad_norm": 0.2329302877187729, "learning_rate": 8.016412388576683e-06, "loss": 0.0139, "step": 82640 }, { "grad_norm": 0.6492869853973389, "learning_rate": 8.007434785107699e-06, "loss": 0.0144, "step": 82650 }, { "grad_norm": 0.5927948355674744, "learning_rate": 7.998461773882509e-06, "loss": 0.0173, "step": 82660 }, { "grad_norm": 0.17413011193275452, "learning_rate": 7.989493355882393e-06, "loss": 0.016, "step": 82670 }, { "grad_norm": 0.4575504660606384, "learning_rate": 7.980529532088137e-06, "loss": 0.0165, "step": 82680 }, { "grad_norm": 0.331098347902298, "learning_rate": 7.97157030347998e-06, "loss": 0.0191, "step": 82690 }, { "grad_norm": 0.4583914875984192, "learning_rate": 7.96261567103772e-06, "loss": 0.0181, "step": 82700 }, { "grad_norm": 0.2151535451412201, "learning_rate": 7.953665635740593e-06, "loss": 0.0149, "step": 82710 }, { "grad_norm": 0.2797674834728241, "learning_rate": 7.94472019856738e-06, "loss": 0.016, "step": 82720 }, { "grad_norm": 0.47084566950798035, "learning_rate": 7.935779360496337e-06, "loss": 0.0135, "step": 82730 }, { "grad_norm": 0.3584079444408417, "learning_rate": 7.92684312250523e-06, "loss": 0.0234, "step": 82740 }, { "grad_norm": 0.2918575704097748, "learning_rate": 7.917911485571289e-06, "loss": 0.0128, "step": 82750 }, { "grad_norm": 0.596538245677948, "learning_rate": 7.90898445067128e-06, "loss": 0.0179, "step": 82760 }, { "grad_norm": 0.4617188274860382, "learning_rate": 7.900062018781462e-06, "loss": 0.0148, "step": 82770 }, { "grad_norm": 0.5487786531448364, "learning_rate": 7.891144190877553e-06, "loss": 0.0142, "step": 82780 }, { "grad_norm": 0.34290459752082825, "learning_rate": 7.882230967934806e-06, "loss": 0.0104, "step": 82790 }, { "grad_norm": 0.4247382581233978, "learning_rate": 7.873322350927959e-06, "loss": 0.0118, "step": 82800 }, { "grad_norm": 0.2651377022266388, "learning_rate": 7.864418340831253e-06, "loss": 0.0126, "step": 82810 }, { "grad_norm": 0.2544041872024536, "learning_rate": 7.855518938618395e-06, "loss": 0.0136, "step": 82820 }, { "grad_norm": 0.7961647510528564, "learning_rate": 7.846624145262633e-06, "loss": 0.0171, "step": 82830 }, { "grad_norm": 0.21793082356452942, "learning_rate": 7.837733961736672e-06, "loss": 0.0106, "step": 82840 }, { "grad_norm": 0.4942707419395447, "learning_rate": 7.828848389012727e-06, "loss": 0.0139, "step": 82850 }, { "grad_norm": 0.23958124220371246, "learning_rate": 7.819967428062524e-06, "loss": 0.0148, "step": 82860 }, { "grad_norm": 0.3547373116016388, "learning_rate": 7.811091079857274e-06, "loss": 0.0159, "step": 82870 }, { "grad_norm": 0.22424697875976562, "learning_rate": 7.802219345367667e-06, "loss": 0.0084, "step": 82880 }, { "grad_norm": 0.6690734624862671, "learning_rate": 7.793352225563916e-06, "loss": 0.0187, "step": 82890 }, { "grad_norm": 0.21288058161735535, "learning_rate": 7.7844897214157e-06, "loss": 0.0152, "step": 82900 }, { "grad_norm": 0.49979883432388306, "learning_rate": 7.775631833892228e-06, "loss": 0.0187, "step": 82910 }, { "grad_norm": 0.26738980412483215, "learning_rate": 7.76677856396215e-06, "loss": 0.0186, "step": 82920 }, { "grad_norm": 0.4116550087928772, "learning_rate": 7.757929912593697e-06, "loss": 0.0137, "step": 82930 }, { "grad_norm": 0.47242891788482666, "learning_rate": 7.749085880754509e-06, "loss": 0.0141, "step": 82940 }, { "grad_norm": 0.5238950848579407, "learning_rate": 7.740246469411777e-06, "loss": 0.0173, "step": 82950 }, { "grad_norm": 0.7053255438804626, "learning_rate": 7.731411679532142e-06, "loss": 0.0212, "step": 82960 }, { "grad_norm": 0.3292190134525299, "learning_rate": 7.722581512081779e-06, "loss": 0.0161, "step": 82970 }, { "grad_norm": 0.5039373636245728, "learning_rate": 7.713755968026331e-06, "loss": 0.012, "step": 82980 }, { "grad_norm": 0.21475186944007874, "learning_rate": 7.704935048330969e-06, "loss": 0.0115, "step": 82990 }, { "grad_norm": 0.4688771963119507, "learning_rate": 7.696118753960313e-06, "loss": 0.0168, "step": 83000 }, { "grad_norm": 0.3197193741798401, "learning_rate": 7.687307085878503e-06, "loss": 0.0166, "step": 83010 }, { "grad_norm": 0.4664428234100342, "learning_rate": 7.678500045049186e-06, "loss": 0.0212, "step": 83020 }, { "grad_norm": 0.46909913420677185, "learning_rate": 7.669697632435463e-06, "loss": 0.0234, "step": 83030 }, { "grad_norm": 0.53355872631073, "learning_rate": 7.660899848999963e-06, "loss": 0.0197, "step": 83040 }, { "grad_norm": 0.3811146318912506, "learning_rate": 7.652106695704803e-06, "loss": 0.0248, "step": 83050 }, { "grad_norm": 0.197108656167984, "learning_rate": 7.643318173511598e-06, "loss": 0.0156, "step": 83060 }, { "grad_norm": 0.4319280683994293, "learning_rate": 7.634534283381423e-06, "loss": 0.0139, "step": 83070 }, { "grad_norm": 0.4368375837802887, "learning_rate": 7.625755026274889e-06, "loss": 0.0131, "step": 83080 }, { "grad_norm": 0.29216620326042175, "learning_rate": 7.616980403152069e-06, "loss": 0.0129, "step": 83090 }, { "grad_norm": 0.1613045632839203, "learning_rate": 7.608210414972561e-06, "loss": 0.0118, "step": 83100 }, { "grad_norm": 0.25649839639663696, "learning_rate": 7.599445062695404e-06, "loss": 0.0231, "step": 83110 }, { "grad_norm": 0.181540846824646, "learning_rate": 7.590684347279214e-06, "loss": 0.0138, "step": 83120 }, { "grad_norm": 0.7301566004753113, "learning_rate": 7.581928269682004e-06, "loss": 0.0173, "step": 83130 }, { "grad_norm": 0.8477617502212524, "learning_rate": 7.573176830861361e-06, "loss": 0.03, "step": 83140 }, { "grad_norm": 0.4460063874721527, "learning_rate": 7.5644300317743e-06, "loss": 0.0188, "step": 83150 }, { "grad_norm": 0.2746797204017639, "learning_rate": 7.555687873377376e-06, "loss": 0.0232, "step": 83160 }, { "grad_norm": 0.15786324441432953, "learning_rate": 7.546950356626592e-06, "loss": 0.014, "step": 83170 }, { "grad_norm": 0.3392334282398224, "learning_rate": 7.538217482477516e-06, "loss": 0.0164, "step": 83180 }, { "grad_norm": 0.4166202247142792, "learning_rate": 7.529489251885119e-06, "loss": 0.0183, "step": 83190 }, { "grad_norm": 0.23644590377807617, "learning_rate": 7.520765665803941e-06, "loss": 0.0134, "step": 83200 }, { "grad_norm": 0.3910173177719116, "learning_rate": 7.512046725187949e-06, "loss": 0.0264, "step": 83210 }, { "grad_norm": 0.24141879379749298, "learning_rate": 7.503332430990645e-06, "loss": 0.0133, "step": 83220 }, { "grad_norm": 0.2725639045238495, "learning_rate": 7.494622784165017e-06, "loss": 0.0161, "step": 83230 }, { "grad_norm": 0.4246138036251068, "learning_rate": 7.485917785663543e-06, "loss": 0.0186, "step": 83240 }, { "grad_norm": 0.5307580828666687, "learning_rate": 7.477217436438172e-06, "loss": 0.0186, "step": 83250 }, { "grad_norm": 0.1656322479248047, "learning_rate": 7.46852173744037e-06, "loss": 0.0156, "step": 83260 }, { "grad_norm": 0.5158258080482483, "learning_rate": 7.4598306896211004e-06, "loss": 0.0137, "step": 83270 }, { "grad_norm": 0.3851783871650696, "learning_rate": 7.451144293930773e-06, "loss": 0.015, "step": 83280 }, { "grad_norm": 0.3530905544757843, "learning_rate": 7.4424625513193455e-06, "loss": 0.0141, "step": 83290 }, { "grad_norm": 0.5994076728820801, "learning_rate": 7.433785462736209e-06, "loss": 0.0187, "step": 83300 }, { "grad_norm": 0.2930745780467987, "learning_rate": 7.425113029130315e-06, "loss": 0.0161, "step": 83310 }, { "grad_norm": 0.38977041840553284, "learning_rate": 7.416445251450044e-06, "loss": 0.0262, "step": 83320 }, { "grad_norm": 0.653849184513092, "learning_rate": 7.4077821306433035e-06, "loss": 0.0125, "step": 83330 }, { "grad_norm": 0.23422816395759583, "learning_rate": 7.399123667657465e-06, "loss": 0.0149, "step": 83340 }, { "grad_norm": 0.37425604462623596, "learning_rate": 7.390469863439426e-06, "loss": 0.0139, "step": 83350 }, { "grad_norm": 0.2409379929304123, "learning_rate": 7.381820718935512e-06, "loss": 0.0198, "step": 83360 }, { "grad_norm": 0.2004496306180954, "learning_rate": 7.373176235091633e-06, "loss": 0.017, "step": 83370 }, { "grad_norm": 0.26008039712905884, "learning_rate": 7.364536412853101e-06, "loss": 0.0113, "step": 83380 }, { "grad_norm": 0.4442370533943176, "learning_rate": 7.355901253164776e-06, "loss": 0.0137, "step": 83390 }, { "grad_norm": 0.2584497928619385, "learning_rate": 7.34727075697097e-06, "loss": 0.0153, "step": 83400 }, { "grad_norm": 0.35792380571365356, "learning_rate": 7.3386449252155095e-06, "loss": 0.0161, "step": 83410 }, { "grad_norm": 0.5379095673561096, "learning_rate": 7.330023758841692e-06, "loss": 0.0134, "step": 83420 }, { "grad_norm": 0.5686967968940735, "learning_rate": 7.321407258792323e-06, "loss": 0.0209, "step": 83430 }, { "grad_norm": 0.3732726275920868, "learning_rate": 7.312795426009694e-06, "loss": 0.0193, "step": 83440 }, { "grad_norm": 0.3450073003768921, "learning_rate": 7.304188261435585e-06, "loss": 0.0229, "step": 83450 }, { "grad_norm": 0.3015486001968384, "learning_rate": 7.295585766011242e-06, "loss": 0.0186, "step": 83460 }, { "grad_norm": 0.5308105945587158, "learning_rate": 7.286987940677436e-06, "loss": 0.0161, "step": 83470 }, { "grad_norm": 0.4229375123977661, "learning_rate": 7.278394786374426e-06, "loss": 0.0175, "step": 83480 }, { "grad_norm": 0.4992729127407074, "learning_rate": 7.269806304041915e-06, "loss": 0.0186, "step": 83490 }, { "grad_norm": 0.4635560214519501, "learning_rate": 7.261222494619147e-06, "loss": 0.0159, "step": 83500 }, { "grad_norm": 0.3120400309562683, "learning_rate": 7.252643359044825e-06, "loss": 0.015, "step": 83510 }, { "grad_norm": 0.1661655753850937, "learning_rate": 7.244068898257172e-06, "loss": 0.0193, "step": 83520 }, { "grad_norm": 0.2669239342212677, "learning_rate": 7.2354991131938495e-06, "loss": 0.019, "step": 83530 }, { "grad_norm": 0.46622854471206665, "learning_rate": 7.2269340047920574e-06, "loss": 0.0156, "step": 83540 }, { "grad_norm": 0.35824817419052124, "learning_rate": 7.218373573988435e-06, "loss": 0.0156, "step": 83550 }, { "grad_norm": 0.2946150302886963, "learning_rate": 7.209817821719178e-06, "loss": 0.0193, "step": 83560 }, { "grad_norm": 0.16818048059940338, "learning_rate": 7.201266748919894e-06, "loss": 0.0194, "step": 83570 }, { "grad_norm": 0.3498770594596863, "learning_rate": 7.19272035652575e-06, "loss": 0.0134, "step": 83580 }, { "grad_norm": 0.5377230048179626, "learning_rate": 7.184178645471334e-06, "loss": 0.0154, "step": 83590 }, { "grad_norm": 0.4608483910560608, "learning_rate": 7.175641616690775e-06, "loss": 0.0214, "step": 83600 }, { "grad_norm": 0.2695919871330261, "learning_rate": 7.167109271117661e-06, "loss": 0.0167, "step": 83610 }, { "grad_norm": 0.23334117233753204, "learning_rate": 7.158581609685072e-06, "loss": 0.0134, "step": 83620 }, { "grad_norm": 0.2133646011352539, "learning_rate": 7.150058633325591e-06, "loss": 0.011, "step": 83630 }, { "grad_norm": 0.3951152563095093, "learning_rate": 7.141540342971287e-06, "loss": 0.0129, "step": 83640 }, { "grad_norm": 0.42301100492477417, "learning_rate": 7.133026739553677e-06, "loss": 0.0213, "step": 83650 }, { "grad_norm": 0.37987416982650757, "learning_rate": 7.124517824003829e-06, "loss": 0.0167, "step": 83660 }, { "grad_norm": 0.36806753277778625, "learning_rate": 7.116013597252236e-06, "loss": 0.0174, "step": 83670 }, { "grad_norm": 0.38189834356307983, "learning_rate": 7.107514060228921e-06, "loss": 0.0143, "step": 83680 }, { "grad_norm": 0.5169095993041992, "learning_rate": 7.099019213863378e-06, "loss": 0.0191, "step": 83690 }, { "grad_norm": 0.39829742908477783, "learning_rate": 7.090529059084611e-06, "loss": 0.015, "step": 83700 }, { "grad_norm": 0.2338411659002304, "learning_rate": 7.082043596821058e-06, "loss": 0.0126, "step": 83710 }, { "grad_norm": 0.2527497708797455, "learning_rate": 7.07356282800069e-06, "loss": 0.035, "step": 83720 }, { "grad_norm": 0.3222312927246094, "learning_rate": 7.0650867535509645e-06, "loss": 0.0174, "step": 83730 }, { "grad_norm": 0.2171505093574524, "learning_rate": 7.0566153743987895e-06, "loss": 0.0168, "step": 83740 }, { "grad_norm": 0.13880357146263123, "learning_rate": 7.04814869147059e-06, "loss": 0.0151, "step": 83750 }, { "grad_norm": 0.2281380295753479, "learning_rate": 7.0396867056922725e-06, "loss": 0.0183, "step": 83760 }, { "grad_norm": 0.18841718137264252, "learning_rate": 7.031229417989244e-06, "loss": 0.0169, "step": 83770 }, { "grad_norm": 0.22790361940860748, "learning_rate": 7.022776829286348e-06, "loss": 0.0076, "step": 83780 }, { "grad_norm": 0.2525108754634857, "learning_rate": 7.014328940507975e-06, "loss": 0.0197, "step": 83790 }, { "grad_norm": 0.4033471643924713, "learning_rate": 7.005885752577945e-06, "loss": 0.0164, "step": 83800 }, { "grad_norm": 0.4223637282848358, "learning_rate": 6.997447266419615e-06, "loss": 0.0204, "step": 83810 }, { "grad_norm": 0.3764050602912903, "learning_rate": 6.989013482955797e-06, "loss": 0.0135, "step": 83820 }, { "grad_norm": 0.23522354662418365, "learning_rate": 6.980584403108808e-06, "loss": 0.0124, "step": 83830 }, { "grad_norm": 0.3075510263442993, "learning_rate": 6.972160027800417e-06, "loss": 0.0123, "step": 83840 }, { "grad_norm": 0.22887764871120453, "learning_rate": 6.963740357951931e-06, "loss": 0.0165, "step": 83850 }, { "grad_norm": 0.35345879197120667, "learning_rate": 6.95532539448408e-06, "loss": 0.0161, "step": 83860 }, { "grad_norm": 0.2927854359149933, "learning_rate": 6.946915138317128e-06, "loss": 0.0142, "step": 83870 }, { "grad_norm": 0.47958528995513916, "learning_rate": 6.9385095903708065e-06, "loss": 0.0145, "step": 83880 }, { "grad_norm": 0.3751552999019623, "learning_rate": 6.930108751564346e-06, "loss": 0.0182, "step": 83890 }, { "grad_norm": 0.22109630703926086, "learning_rate": 6.921712622816428e-06, "loss": 0.0273, "step": 83900 }, { "grad_norm": 0.38220277428627014, "learning_rate": 6.913321205045259e-06, "loss": 0.0177, "step": 83910 }, { "grad_norm": 0.25545626878738403, "learning_rate": 6.904934499168491e-06, "loss": 0.0136, "step": 83920 }, { "grad_norm": 0.5427999496459961, "learning_rate": 6.896552506103299e-06, "loss": 0.0202, "step": 83930 }, { "grad_norm": 0.20561961829662323, "learning_rate": 6.888175226766313e-06, "loss": 0.0152, "step": 83940 }, { "grad_norm": 0.4520810842514038, "learning_rate": 6.879802662073686e-06, "loss": 0.0127, "step": 83950 }, { "grad_norm": 0.423750102519989, "learning_rate": 6.871434812940991e-06, "loss": 0.0187, "step": 83960 }, { "grad_norm": 0.34136250615119934, "learning_rate": 6.863071680283345e-06, "loss": 0.0196, "step": 83970 }, { "grad_norm": 0.4665721654891968, "learning_rate": 6.854713265015339e-06, "loss": 0.019, "step": 83980 }, { "grad_norm": 0.3565571904182434, "learning_rate": 6.8463595680510025e-06, "loss": 0.02, "step": 83990 }, { "grad_norm": 0.3877408802509308, "learning_rate": 6.838010590303911e-06, "loss": 0.0139, "step": 84000 }, { "grad_norm": 0.2033834308385849, "learning_rate": 6.829666332687085e-06, "loss": 0.0139, "step": 84010 }, { "grad_norm": 0.2722867727279663, "learning_rate": 6.821326796113059e-06, "loss": 0.0128, "step": 84020 }, { "grad_norm": 0.17556487023830414, "learning_rate": 6.812991981493799e-06, "loss": 0.0142, "step": 84030 }, { "grad_norm": 0.4728371500968933, "learning_rate": 6.804661889740821e-06, "loss": 0.0209, "step": 84040 }, { "grad_norm": 0.3224708139896393, "learning_rate": 6.7963365217650625e-06, "loss": 0.0161, "step": 84050 }, { "grad_norm": 0.39368733763694763, "learning_rate": 6.788015878476983e-06, "loss": 0.0181, "step": 84060 }, { "grad_norm": 0.39277392625808716, "learning_rate": 6.779699960786523e-06, "loss": 0.0147, "step": 84070 }, { "grad_norm": 0.13282495737075806, "learning_rate": 6.771388769603104e-06, "loss": 0.0361, "step": 84080 }, { "grad_norm": 0.6896288394927979, "learning_rate": 6.763082305835605e-06, "loss": 0.0225, "step": 84090 }, { "grad_norm": 0.5399630665779114, "learning_rate": 6.754780570392433e-06, "loss": 0.0267, "step": 84100 }, { "grad_norm": 0.25778329372406006, "learning_rate": 6.7464835641814215e-06, "loss": 0.0166, "step": 84110 }, { "grad_norm": 0.36111342906951904, "learning_rate": 6.7381912881099515e-06, "loss": 0.0207, "step": 84120 }, { "grad_norm": 0.815133273601532, "learning_rate": 6.72990374308482e-06, "loss": 0.0147, "step": 84130 }, { "grad_norm": 0.3187391459941864, "learning_rate": 6.721620930012379e-06, "loss": 0.0168, "step": 84140 }, { "grad_norm": 0.44902849197387695, "learning_rate": 6.713342849798393e-06, "loss": 0.0145, "step": 84150 }, { "grad_norm": 0.655761182308197, "learning_rate": 6.705069503348166e-06, "loss": 0.0232, "step": 84160 }, { "grad_norm": 0.3620799779891968, "learning_rate": 6.696800891566435e-06, "loss": 0.014, "step": 84170 }, { "grad_norm": 0.5963062047958374, "learning_rate": 6.688537015357449e-06, "loss": 0.016, "step": 84180 }, { "grad_norm": 0.1721278578042984, "learning_rate": 6.680277875624941e-06, "loss": 0.0183, "step": 84190 }, { "grad_norm": 0.5294497609138489, "learning_rate": 6.672023473272121e-06, "loss": 0.0269, "step": 84200 }, { "grad_norm": 0.2132055014371872, "learning_rate": 6.663773809201667e-06, "loss": 0.0183, "step": 84210 }, { "grad_norm": 0.22927312552928925, "learning_rate": 6.655528884315748e-06, "loss": 0.0188, "step": 84220 }, { "grad_norm": 0.8047236204147339, "learning_rate": 6.647288699516036e-06, "loss": 0.0282, "step": 84230 }, { "grad_norm": 0.32365062832832336, "learning_rate": 6.639053255703642e-06, "loss": 0.0139, "step": 84240 }, { "grad_norm": 0.3829210102558136, "learning_rate": 6.6308225537791925e-06, "loss": 0.0142, "step": 84250 }, { "grad_norm": 0.3562696576118469, "learning_rate": 6.622596594642777e-06, "loss": 0.0177, "step": 84260 }, { "grad_norm": 0.26419034600257874, "learning_rate": 6.614375379193999e-06, "loss": 0.0173, "step": 84270 }, { "grad_norm": 0.4089340567588806, "learning_rate": 6.606158908331889e-06, "loss": 0.0194, "step": 84280 }, { "grad_norm": 0.5151782035827637, "learning_rate": 6.597947182955005e-06, "loss": 0.0186, "step": 84290 }, { "grad_norm": 0.335407555103302, "learning_rate": 6.589740203961359e-06, "loss": 0.0179, "step": 84300 }, { "grad_norm": 0.6874114871025085, "learning_rate": 6.581537972248464e-06, "loss": 0.0155, "step": 84310 }, { "grad_norm": 0.47351720929145813, "learning_rate": 6.573340488713275e-06, "loss": 0.0152, "step": 84320 }, { "grad_norm": 0.39964559674263, "learning_rate": 6.565147754252304e-06, "loss": 0.0179, "step": 84330 }, { "grad_norm": 0.5909122228622437, "learning_rate": 6.556959769761462e-06, "loss": 0.0141, "step": 84340 }, { "grad_norm": 0.3144915997982025, "learning_rate": 6.548776536136192e-06, "loss": 0.0113, "step": 84350 }, { "grad_norm": 0.6226850152015686, "learning_rate": 6.54059805427138e-06, "loss": 0.0142, "step": 84360 }, { "grad_norm": 0.3197662830352783, "learning_rate": 6.532424325061437e-06, "loss": 0.0169, "step": 84370 }, { "grad_norm": 0.3655736446380615, "learning_rate": 6.5242553494001935e-06, "loss": 0.0142, "step": 84380 }, { "grad_norm": 0.294994980096817, "learning_rate": 6.516091128181046e-06, "loss": 0.0132, "step": 84390 }, { "grad_norm": 0.37831127643585205, "learning_rate": 6.507931662296785e-06, "loss": 0.0173, "step": 84400 }, { "grad_norm": 0.3274916112422943, "learning_rate": 6.4997769526397354e-06, "loss": 0.0149, "step": 84410 }, { "grad_norm": 0.35867592692375183, "learning_rate": 6.491627000101663e-06, "loss": 0.0185, "step": 84420 }, { "grad_norm": 0.2548252046108246, "learning_rate": 6.4834818055738545e-06, "loss": 0.0158, "step": 84430 }, { "grad_norm": 0.5175645351409912, "learning_rate": 6.4753413699470465e-06, "loss": 0.021, "step": 84440 }, { "grad_norm": 0.47143831849098206, "learning_rate": 6.467205694111478e-06, "loss": 0.0128, "step": 84450 }, { "grad_norm": 0.24113456904888153, "learning_rate": 6.459074778956836e-06, "loss": 0.0195, "step": 84460 }, { "grad_norm": 0.2915436923503876, "learning_rate": 6.450948625372311e-06, "loss": 0.0169, "step": 84470 }, { "grad_norm": 0.3945898115634918, "learning_rate": 6.442827234246585e-06, "loss": 0.0178, "step": 84480 }, { "grad_norm": 0.3256470263004303, "learning_rate": 6.43471060646777e-06, "loss": 0.0141, "step": 84490 }, { "grad_norm": 0.3817158341407776, "learning_rate": 6.426598742923512e-06, "loss": 0.0159, "step": 84500 }, { "grad_norm": 0.23692435026168823, "learning_rate": 6.418491644500885e-06, "loss": 0.0257, "step": 84510 }, { "grad_norm": 0.2340994030237198, "learning_rate": 6.410389312086512e-06, "loss": 0.0162, "step": 84520 }, { "grad_norm": 0.16138069331645966, "learning_rate": 6.4022917465664135e-06, "loss": 0.016, "step": 84530 }, { "grad_norm": 0.5246076583862305, "learning_rate": 6.3941989488261575e-06, "loss": 0.0213, "step": 84540 }, { "grad_norm": 0.38675588369369507, "learning_rate": 6.386110919750726e-06, "loss": 0.016, "step": 84550 }, { "grad_norm": 1.6640832424163818, "learning_rate": 6.378027660224651e-06, "loss": 0.0201, "step": 84560 }, { "grad_norm": 0.1321662962436676, "learning_rate": 6.369949171131856e-06, "loss": 0.017, "step": 84570 }, { "grad_norm": 0.5557442903518677, "learning_rate": 6.361875453355848e-06, "loss": 0.0191, "step": 84580 }, { "grad_norm": 0.2534303367137909, "learning_rate": 6.353806507779525e-06, "loss": 0.0171, "step": 84590 }, { "grad_norm": 0.34638646245002747, "learning_rate": 6.3457423352853044e-06, "loss": 0.0128, "step": 84600 }, { "grad_norm": 0.25771793723106384, "learning_rate": 6.337682936755062e-06, "loss": 0.0145, "step": 84610 }, { "grad_norm": 0.43237802386283875, "learning_rate": 6.329628313070179e-06, "loss": 0.013, "step": 84620 }, { "grad_norm": 0.312430739402771, "learning_rate": 6.321578465111477e-06, "loss": 0.0144, "step": 84630 }, { "grad_norm": 0.3360458016395569, "learning_rate": 6.31353339375928e-06, "loss": 0.0165, "step": 84640 }, { "grad_norm": 0.2799482047557831, "learning_rate": 6.305493099893395e-06, "loss": 0.0147, "step": 84650 }, { "grad_norm": 0.3675019145011902, "learning_rate": 6.297457584393096e-06, "loss": 0.0137, "step": 84660 }, { "grad_norm": 0.5134354829788208, "learning_rate": 6.289426848137126e-06, "loss": 0.0183, "step": 84670 }, { "grad_norm": 0.35880085825920105, "learning_rate": 6.281400892003713e-06, "loss": 0.028, "step": 84680 }, { "grad_norm": 0.20089396834373474, "learning_rate": 6.273379716870581e-06, "loss": 0.0118, "step": 84690 }, { "grad_norm": 0.624733030796051, "learning_rate": 6.265363323614887e-06, "loss": 0.0191, "step": 84700 }, { "grad_norm": 0.372372567653656, "learning_rate": 6.257351713113308e-06, "loss": 0.0118, "step": 84710 }, { "grad_norm": 0.20008009672164917, "learning_rate": 6.249344886241981e-06, "loss": 0.0186, "step": 84720 }, { "grad_norm": 0.275078684091568, "learning_rate": 6.2413428438765295e-06, "loss": 0.0201, "step": 84730 }, { "grad_norm": 0.25288233160972595, "learning_rate": 6.233345586892026e-06, "loss": 0.0159, "step": 84740 }, { "grad_norm": 0.245784193277359, "learning_rate": 6.225353116163052e-06, "loss": 0.0208, "step": 84750 }, { "grad_norm": 0.9607012867927551, "learning_rate": 6.217365432563627e-06, "loss": 0.0164, "step": 84760 }, { "grad_norm": 0.633060097694397, "learning_rate": 6.209382536967318e-06, "loss": 0.0185, "step": 84770 }, { "grad_norm": 0.44689038395881653, "learning_rate": 6.201404430247082e-06, "loss": 0.02, "step": 84780 }, { "grad_norm": 0.3023066520690918, "learning_rate": 6.1934311132754165e-06, "loss": 0.0184, "step": 84790 }, { "grad_norm": 0.3149014115333557, "learning_rate": 6.185462586924257e-06, "loss": 0.0134, "step": 84800 }, { "grad_norm": 0.32693514227867126, "learning_rate": 6.177498852065039e-06, "loss": 0.0144, "step": 84810 }, { "grad_norm": 0.18504361808300018, "learning_rate": 6.169539909568655e-06, "loss": 0.0163, "step": 84820 }, { "grad_norm": 0.5025947690010071, "learning_rate": 6.161585760305488e-06, "loss": 0.0146, "step": 84830 }, { "grad_norm": 0.3761153817176819, "learning_rate": 6.153636405145391e-06, "loss": 0.0143, "step": 84840 }, { "grad_norm": 0.30399566888809204, "learning_rate": 6.145691844957702e-06, "loss": 0.0212, "step": 84850 }, { "grad_norm": 0.288965106010437, "learning_rate": 6.1377520806112145e-06, "loss": 0.0166, "step": 84860 }, { "grad_norm": 0.3747032880783081, "learning_rate": 6.129817112974218e-06, "loss": 0.0155, "step": 84870 }, { "grad_norm": 0.37474125623703003, "learning_rate": 6.121886942914457e-06, "loss": 0.0143, "step": 84880 }, { "grad_norm": 0.22005689144134521, "learning_rate": 6.11396157129917e-06, "loss": 0.0117, "step": 84890 }, { "grad_norm": 0.3007165193557739, "learning_rate": 6.106040998995066e-06, "loss": 0.0116, "step": 84900 }, { "grad_norm": 0.4891507625579834, "learning_rate": 6.098125226868329e-06, "loss": 0.0158, "step": 84910 }, { "grad_norm": 0.4865717887878418, "learning_rate": 6.090214255784604e-06, "loss": 0.0136, "step": 84920 }, { "grad_norm": 0.47265392541885376, "learning_rate": 6.0823080866090346e-06, "loss": 0.0125, "step": 84930 }, { "grad_norm": 0.27573490142822266, "learning_rate": 6.074406720206233e-06, "loss": 0.0187, "step": 84940 }, { "grad_norm": 0.2775052487850189, "learning_rate": 6.066510157440259e-06, "loss": 0.0111, "step": 84950 }, { "grad_norm": 0.20996688306331635, "learning_rate": 6.058618399174676e-06, "loss": 0.0087, "step": 84960 }, { "grad_norm": 0.4146011769771576, "learning_rate": 6.050731446272528e-06, "loss": 0.0302, "step": 84970 }, { "grad_norm": 0.5124209523200989, "learning_rate": 6.042849299596321e-06, "loss": 0.0174, "step": 84980 }, { "grad_norm": 0.46733805537223816, "learning_rate": 6.034971960008013e-06, "loss": 0.0173, "step": 84990 }, { "grad_norm": 0.4977037012577057, "learning_rate": 6.027099428369082e-06, "loss": 0.0213, "step": 85000 }, { "grad_norm": 0.3361358642578125, "learning_rate": 6.019231705540435e-06, "loss": 0.0176, "step": 85010 }, { "grad_norm": 0.34569382667541504, "learning_rate": 6.011368792382488e-06, "loss": 0.0146, "step": 85020 }, { "grad_norm": 0.6208165884017944, "learning_rate": 6.00351068975511e-06, "loss": 0.0175, "step": 85030 }, { "grad_norm": 0.20959137380123138, "learning_rate": 5.9956573985176674e-06, "loss": 0.0198, "step": 85040 }, { "grad_norm": 0.22174400091171265, "learning_rate": 5.987808919528959e-06, "loss": 0.0128, "step": 85050 }, { "grad_norm": 0.28373411297798157, "learning_rate": 5.979965253647307e-06, "loss": 0.019, "step": 85060 }, { "grad_norm": 0.15136173367500305, "learning_rate": 5.9721264017304655e-06, "loss": 0.0221, "step": 85070 }, { "grad_norm": 0.15909238159656525, "learning_rate": 5.964292364635682e-06, "loss": 0.0126, "step": 85080 }, { "grad_norm": 0.12280955910682678, "learning_rate": 5.956463143219682e-06, "loss": 0.0135, "step": 85090 }, { "grad_norm": 0.3740738034248352, "learning_rate": 5.948638738338663e-06, "loss": 0.018, "step": 85100 }, { "grad_norm": 0.28955143690109253, "learning_rate": 5.940819150848271e-06, "loss": 0.0144, "step": 85110 }, { "grad_norm": 0.3149174153804779, "learning_rate": 5.933004381603663e-06, "loss": 0.0123, "step": 85120 }, { "grad_norm": 0.40085506439208984, "learning_rate": 5.92519443145943e-06, "loss": 0.0121, "step": 85130 }, { "grad_norm": 0.41343820095062256, "learning_rate": 5.917389301269671e-06, "loss": 0.0159, "step": 85140 }, { "grad_norm": 0.2309570014476776, "learning_rate": 5.9095889918879435e-06, "loss": 0.0083, "step": 85150 }, { "grad_norm": 0.30416184663772583, "learning_rate": 5.901793504167286e-06, "loss": 0.0144, "step": 85160 }, { "grad_norm": 0.22234094142913818, "learning_rate": 5.894002838960172e-06, "loss": 0.0124, "step": 85170 }, { "grad_norm": 0.47182998061180115, "learning_rate": 5.886216997118604e-06, "loss": 0.0287, "step": 85180 }, { "grad_norm": 0.20043165981769562, "learning_rate": 5.878435979494023e-06, "loss": 0.0138, "step": 85190 }, { "grad_norm": 0.3111123740673065, "learning_rate": 5.870659786937344e-06, "loss": 0.0133, "step": 85200 }, { "grad_norm": 0.23781020939350128, "learning_rate": 5.862888420298962e-06, "loss": 0.0126, "step": 85210 }, { "grad_norm": 0.5048614740371704, "learning_rate": 5.855121880428743e-06, "loss": 0.0139, "step": 85220 }, { "grad_norm": 0.4132964015007019, "learning_rate": 5.84736016817603e-06, "loss": 0.0292, "step": 85230 }, { "grad_norm": 0.40942758321762085, "learning_rate": 5.839603284389622e-06, "loss": 0.0126, "step": 85240 }, { "grad_norm": 0.2658119797706604, "learning_rate": 5.831851229917812e-06, "loss": 0.0152, "step": 85250 }, { "grad_norm": 0.49553126096725464, "learning_rate": 5.824104005608338e-06, "loss": 0.0145, "step": 85260 }, { "grad_norm": 0.28971874713897705, "learning_rate": 5.81636161230843e-06, "loss": 0.0146, "step": 85270 }, { "grad_norm": 0.14103445410728455, "learning_rate": 5.808624050864791e-06, "loss": 0.014, "step": 85280 }, { "grad_norm": 0.2603238821029663, "learning_rate": 5.800891322123592e-06, "loss": 0.0199, "step": 85290 }, { "grad_norm": 0.2977811098098755, "learning_rate": 5.793163426930454e-06, "loss": 0.0162, "step": 85300 }, { "grad_norm": 0.3219243586063385, "learning_rate": 5.785440366130512e-06, "loss": 0.0221, "step": 85310 }, { "grad_norm": 0.3444174826145172, "learning_rate": 5.777722140568326e-06, "loss": 0.0122, "step": 85320 }, { "grad_norm": 0.3346364498138428, "learning_rate": 5.7700087510879685e-06, "loss": 0.0148, "step": 85330 }, { "grad_norm": 0.2192470282316208, "learning_rate": 5.762300198532933e-06, "loss": 0.0096, "step": 85340 }, { "grad_norm": 0.35351261496543884, "learning_rate": 5.7545964837462606e-06, "loss": 0.0177, "step": 85350 }, { "grad_norm": 0.4716376066207886, "learning_rate": 5.746897607570384e-06, "loss": 0.0165, "step": 85360 }, { "grad_norm": 0.2908468544483185, "learning_rate": 5.7392035708472626e-06, "loss": 0.0233, "step": 85370 }, { "grad_norm": 0.3454611599445343, "learning_rate": 5.731514374418278e-06, "loss": 0.0129, "step": 85380 }, { "grad_norm": 0.34778326749801636, "learning_rate": 5.7238300191243325e-06, "loss": 0.0152, "step": 85390 }, { "grad_norm": 0.27896758913993835, "learning_rate": 5.716150505805762e-06, "loss": 0.0148, "step": 85400 }, { "grad_norm": 0.4855315089225769, "learning_rate": 5.708475835302407e-06, "loss": 0.0225, "step": 85410 }, { "grad_norm": 0.24888277053833008, "learning_rate": 5.70080600845353e-06, "loss": 0.0159, "step": 85420 }, { "grad_norm": 0.4578079879283905, "learning_rate": 5.693141026097909e-06, "loss": 0.0389, "step": 85430 }, { "grad_norm": 0.762321949005127, "learning_rate": 5.6854808890737745e-06, "loss": 0.0141, "step": 85440 }, { "grad_norm": 0.1921210139989853, "learning_rate": 5.6778255982188225e-06, "loss": 0.0155, "step": 85450 }, { "grad_norm": 0.19941571354866028, "learning_rate": 5.670175154370217e-06, "loss": 0.0172, "step": 85460 }, { "grad_norm": 0.3946700990200043, "learning_rate": 5.662529558364616e-06, "loss": 0.015, "step": 85470 }, { "grad_norm": 0.2727261483669281, "learning_rate": 5.654888811038128e-06, "loss": 0.0122, "step": 85480 }, { "grad_norm": 0.46494513750076294, "learning_rate": 5.647252913226314e-06, "loss": 0.0185, "step": 85490 }, { "grad_norm": 0.4674566090106964, "learning_rate": 5.639621865764255e-06, "loss": 0.0168, "step": 85500 }, { "grad_norm": 0.106594018638134, "learning_rate": 5.63199566948644e-06, "loss": 0.0126, "step": 85510 }, { "grad_norm": 0.28175365924835205, "learning_rate": 5.6243743252268835e-06, "loss": 0.0189, "step": 85520 }, { "grad_norm": 0.33569642901420593, "learning_rate": 5.616757833819009e-06, "loss": 0.0179, "step": 85530 }, { "grad_norm": 0.8385989665985107, "learning_rate": 5.609146196095794e-06, "loss": 0.0138, "step": 85540 }, { "grad_norm": 0.5454559326171875, "learning_rate": 5.6015394128896005e-06, "loss": 0.0179, "step": 85550 }, { "grad_norm": 0.20073696970939636, "learning_rate": 5.593937485032308e-06, "loss": 0.0115, "step": 85560 }, { "grad_norm": 0.3336535394191742, "learning_rate": 5.586340413355246e-06, "loss": 0.0132, "step": 85570 }, { "grad_norm": 0.15539336204528809, "learning_rate": 5.578748198689227e-06, "loss": 0.0187, "step": 85580 }, { "grad_norm": 0.49260812997817993, "learning_rate": 5.571160841864504e-06, "loss": 0.0156, "step": 85590 }, { "grad_norm": 0.12069470435380936, "learning_rate": 5.563578343710846e-06, "loss": 0.0103, "step": 85600 }, { "grad_norm": 0.3067748546600342, "learning_rate": 5.556000705057446e-06, "loss": 0.0147, "step": 85610 }, { "grad_norm": 0.23305191099643707, "learning_rate": 5.548427926732996e-06, "loss": 0.0162, "step": 85620 }, { "grad_norm": 0.5418761968612671, "learning_rate": 5.540860009565629e-06, "loss": 0.0146, "step": 85630 }, { "grad_norm": 0.6441519260406494, "learning_rate": 5.533296954382966e-06, "loss": 0.0153, "step": 85640 }, { "grad_norm": 0.15778160095214844, "learning_rate": 5.525738762012095e-06, "loss": 0.0148, "step": 85650 }, { "grad_norm": 0.44148188829421997, "learning_rate": 5.518185433279577e-06, "loss": 0.0214, "step": 85660 }, { "grad_norm": 0.1953735649585724, "learning_rate": 5.510636969011407e-06, "loss": 0.0166, "step": 85670 }, { "grad_norm": 1.1165339946746826, "learning_rate": 5.503093370033096e-06, "loss": 0.0322, "step": 85680 }, { "grad_norm": 0.4302864372730255, "learning_rate": 5.495554637169603e-06, "loss": 0.0147, "step": 85690 }, { "grad_norm": 0.3445281684398651, "learning_rate": 5.488020771245328e-06, "loss": 0.0144, "step": 85700 }, { "grad_norm": 0.21100519597530365, "learning_rate": 5.480491773084195e-06, "loss": 0.0134, "step": 85710 }, { "grad_norm": 0.3724687993526459, "learning_rate": 5.472967643509525e-06, "loss": 0.0139, "step": 85720 }, { "grad_norm": 0.23945119976997375, "learning_rate": 5.46544838334419e-06, "loss": 0.0152, "step": 85730 }, { "grad_norm": 0.26273223757743835, "learning_rate": 5.4579339934104525e-06, "loss": 0.0179, "step": 85740 }, { "grad_norm": 0.21187874674797058, "learning_rate": 5.450424474530091e-06, "loss": 0.0152, "step": 85750 }, { "grad_norm": 0.28809547424316406, "learning_rate": 5.442919827524323e-06, "loss": 0.01, "step": 85760 }, { "grad_norm": 0.240768700838089, "learning_rate": 5.435420053213863e-06, "loss": 0.0165, "step": 85770 }, { "grad_norm": 0.3600199222564697, "learning_rate": 5.427925152418845e-06, "loss": 0.0176, "step": 85780 }, { "grad_norm": 0.3152226507663727, "learning_rate": 5.420435125958934e-06, "loss": 0.0187, "step": 85790 }, { "grad_norm": 0.22440873086452484, "learning_rate": 5.412949974653209e-06, "loss": 0.0141, "step": 85800 }, { "grad_norm": 0.19365531206130981, "learning_rate": 5.405469699320248e-06, "loss": 0.0127, "step": 85810 }, { "grad_norm": 0.43408945202827454, "learning_rate": 5.397994300778064e-06, "loss": 0.01, "step": 85820 }, { "grad_norm": 0.12800319492816925, "learning_rate": 5.39052377984417e-06, "loss": 0.0177, "step": 85830 }, { "grad_norm": 0.22426995635032654, "learning_rate": 5.383058137335523e-06, "loss": 0.0135, "step": 85840 }, { "grad_norm": 0.18396250903606415, "learning_rate": 5.375597374068553e-06, "loss": 0.0249, "step": 85850 }, { "grad_norm": 0.3135053813457489, "learning_rate": 5.368141490859163e-06, "loss": 0.0143, "step": 85860 }, { "grad_norm": 0.5286843776702881, "learning_rate": 5.360690488522729e-06, "loss": 0.0162, "step": 85870 }, { "grad_norm": 0.46348366141319275, "learning_rate": 5.353244367874055e-06, "loss": 0.0193, "step": 85880 }, { "grad_norm": 0.128712460398674, "learning_rate": 5.345803129727456e-06, "loss": 0.0126, "step": 85890 }, { "grad_norm": 0.34229663014411926, "learning_rate": 5.338366774896697e-06, "loss": 0.0136, "step": 85900 }, { "grad_norm": 0.23320597410202026, "learning_rate": 5.330935304194995e-06, "loss": 0.0167, "step": 85910 }, { "grad_norm": 0.2776755094528198, "learning_rate": 5.323508718435044e-06, "loss": 0.0181, "step": 85920 }, { "grad_norm": 0.23971524834632874, "learning_rate": 5.316087018429011e-06, "loss": 0.0123, "step": 85930 }, { "grad_norm": 0.2717534303665161, "learning_rate": 5.30867020498853e-06, "loss": 0.0217, "step": 85940 }, { "grad_norm": 0.18813875317573547, "learning_rate": 5.30125827892467e-06, "loss": 0.0161, "step": 85950 }, { "grad_norm": 0.20450659096240997, "learning_rate": 5.293851241048015e-06, "loss": 0.0126, "step": 85960 }, { "grad_norm": 0.47451290488243103, "learning_rate": 5.286449092168555e-06, "loss": 0.0159, "step": 85970 }, { "grad_norm": 0.2261144518852234, "learning_rate": 5.2790518330958105e-06, "loss": 0.0214, "step": 85980 }, { "grad_norm": 0.25968068838119507, "learning_rate": 5.271659464638717e-06, "loss": 0.0137, "step": 85990 }, { "grad_norm": 0.3750380873680115, "learning_rate": 5.264271987605701e-06, "loss": 0.0191, "step": 86000 }, { "grad_norm": 0.2760358154773712, "learning_rate": 5.256889402804632e-06, "loss": 0.014, "step": 86010 }, { "grad_norm": 0.2906818091869354, "learning_rate": 5.249511711042881e-06, "loss": 0.0118, "step": 86020 }, { "grad_norm": 1.0091941356658936, "learning_rate": 5.242138913127237e-06, "loss": 0.0165, "step": 86030 }, { "grad_norm": 0.5181068181991577, "learning_rate": 5.234771009863987e-06, "loss": 0.0144, "step": 86040 }, { "grad_norm": 0.3428236246109009, "learning_rate": 5.227408002058881e-06, "loss": 0.0175, "step": 86050 }, { "grad_norm": 0.6034680604934692, "learning_rate": 5.220049890517126e-06, "loss": 0.0251, "step": 86060 }, { "grad_norm": 0.2982691526412964, "learning_rate": 5.212696676043377e-06, "loss": 0.0188, "step": 86070 }, { "grad_norm": 0.35618966817855835, "learning_rate": 5.205348359441798e-06, "loss": 0.0165, "step": 86080 }, { "grad_norm": 0.33826953172683716, "learning_rate": 5.198004941515966e-06, "loss": 0.0145, "step": 86090 }, { "grad_norm": 0.4079362452030182, "learning_rate": 5.190666423068952e-06, "loss": 0.014, "step": 86100 }, { "grad_norm": 0.22138234972953796, "learning_rate": 5.183332804903284e-06, "loss": 0.0103, "step": 86110 }, { "grad_norm": 0.374460369348526, "learning_rate": 5.176004087820979e-06, "loss": 0.0229, "step": 86120 }, { "grad_norm": 1.0223063230514526, "learning_rate": 5.168680272623455e-06, "loss": 0.019, "step": 86130 }, { "grad_norm": 0.30702218413352966, "learning_rate": 5.161361360111661e-06, "loss": 0.0152, "step": 86140 }, { "grad_norm": 0.3849972188472748, "learning_rate": 5.154047351085984e-06, "loss": 0.0194, "step": 86150 }, { "grad_norm": 0.16742199659347534, "learning_rate": 5.1467382463462385e-06, "loss": 0.0172, "step": 86160 }, { "grad_norm": 0.25566333532333374, "learning_rate": 5.139434046691788e-06, "loss": 0.0193, "step": 86170 }, { "grad_norm": 0.29304853081703186, "learning_rate": 5.13213475292137e-06, "loss": 0.0135, "step": 86180 }, { "grad_norm": 0.12031390517950058, "learning_rate": 5.12484036583325e-06, "loss": 0.0131, "step": 86190 }, { "grad_norm": 0.7410661578178406, "learning_rate": 5.117550886225109e-06, "loss": 0.0175, "step": 86200 }, { "grad_norm": 0.9226475954055786, "learning_rate": 5.11026631489413e-06, "loss": 0.018, "step": 86210 }, { "grad_norm": 0.41845378279685974, "learning_rate": 5.102986652636926e-06, "loss": 0.0207, "step": 86220 }, { "grad_norm": 0.319674015045166, "learning_rate": 5.095711900249606e-06, "loss": 0.0164, "step": 86230 }, { "grad_norm": 0.4009476602077484, "learning_rate": 5.088442058527715e-06, "loss": 0.0136, "step": 86240 }, { "grad_norm": 1.045952558517456, "learning_rate": 5.08117712826629e-06, "loss": 0.0222, "step": 86250 }, { "grad_norm": 0.4562382102012634, "learning_rate": 5.073917110259785e-06, "loss": 0.012, "step": 86260 }, { "grad_norm": 0.32392436265945435, "learning_rate": 5.066662005302175e-06, "loss": 0.0233, "step": 86270 }, { "grad_norm": 0.24022285640239716, "learning_rate": 5.059411814186838e-06, "loss": 0.0127, "step": 86280 }, { "grad_norm": 0.42097708582878113, "learning_rate": 5.05216653770666e-06, "loss": 0.0139, "step": 86290 }, { "grad_norm": 0.3528353273868561, "learning_rate": 5.044926176653969e-06, "loss": 0.0164, "step": 86300 }, { "grad_norm": 0.282487690448761, "learning_rate": 5.037690731820571e-06, "loss": 0.0452, "step": 86310 }, { "grad_norm": 0.3866817355155945, "learning_rate": 5.030460203997711e-06, "loss": 0.0115, "step": 86320 }, { "grad_norm": 0.1331433355808258, "learning_rate": 5.0232345939761174e-06, "loss": 0.0159, "step": 86330 }, { "grad_norm": 0.21293774247169495, "learning_rate": 5.016013902545957e-06, "loss": 0.0193, "step": 86340 }, { "grad_norm": 0.3063444495201111, "learning_rate": 5.008798130496889e-06, "loss": 0.0155, "step": 86350 }, { "grad_norm": 0.5386898517608643, "learning_rate": 5.001587278618008e-06, "loss": 0.0174, "step": 86360 }, { "grad_norm": 0.22418002784252167, "learning_rate": 4.994381347697891e-06, "loss": 0.0175, "step": 86370 }, { "grad_norm": 0.3399147093296051, "learning_rate": 4.987180338524572e-06, "loss": 0.0126, "step": 86380 }, { "grad_norm": 0.34206071496009827, "learning_rate": 4.9799842518855275e-06, "loss": 0.0142, "step": 86390 }, { "grad_norm": 1.4305063486099243, "learning_rate": 4.9727930885677285e-06, "loss": 0.0244, "step": 86400 }, { "grad_norm": 0.4537632465362549, "learning_rate": 4.965606849357573e-06, "loss": 0.0145, "step": 86410 }, { "grad_norm": 0.3741951584815979, "learning_rate": 4.9584255350409445e-06, "loss": 0.0241, "step": 86420 }, { "grad_norm": 0.1752268522977829, "learning_rate": 4.951249146403175e-06, "loss": 0.011, "step": 86430 }, { "grad_norm": 0.6080299615859985, "learning_rate": 4.9440776842290815e-06, "loss": 0.0123, "step": 86440 }, { "grad_norm": 0.5437232851982117, "learning_rate": 4.936911149302903e-06, "loss": 0.0296, "step": 86450 }, { "grad_norm": 0.31155407428741455, "learning_rate": 4.929749542408379e-06, "loss": 0.0124, "step": 86460 }, { "grad_norm": 0.631215512752533, "learning_rate": 4.922592864328673e-06, "loss": 0.0212, "step": 86470 }, { "grad_norm": 0.26807501912117004, "learning_rate": 4.915441115846436e-06, "loss": 0.0107, "step": 86480 }, { "grad_norm": 0.3004392087459564, "learning_rate": 4.908294297743782e-06, "loss": 0.0161, "step": 86490 }, { "grad_norm": 0.3469091057777405, "learning_rate": 4.901152410802273e-06, "loss": 0.0288, "step": 86500 }, { "grad_norm": 0.43451786041259766, "learning_rate": 4.894015455802919e-06, "loss": 0.0204, "step": 86510 }, { "grad_norm": 0.8166971206665039, "learning_rate": 4.8868834335262346e-06, "loss": 0.0168, "step": 86520 }, { "grad_norm": 0.2623136639595032, "learning_rate": 4.87975634475214e-06, "loss": 0.0177, "step": 86530 }, { "grad_norm": 0.28141745924949646, "learning_rate": 4.8726341902600614e-06, "loss": 0.0139, "step": 86540 }, { "grad_norm": 0.27646684646606445, "learning_rate": 4.865516970828843e-06, "loss": 0.0211, "step": 86550 }, { "grad_norm": 0.2285197675228119, "learning_rate": 4.85840468723685e-06, "loss": 0.0148, "step": 86560 }, { "grad_norm": 0.33056533336639404, "learning_rate": 4.85129734026184e-06, "loss": 0.0173, "step": 86570 }, { "grad_norm": 0.2457515150308609, "learning_rate": 4.844194930681073e-06, "loss": 0.0235, "step": 86580 }, { "grad_norm": 0.2310999482870102, "learning_rate": 4.837097459271267e-06, "loss": 0.0114, "step": 86590 }, { "grad_norm": 0.1432468444108963, "learning_rate": 4.830004926808573e-06, "loss": 0.014, "step": 86600 }, { "grad_norm": 0.3251205384731293, "learning_rate": 4.822917334068627e-06, "loss": 0.0112, "step": 86610 }, { "grad_norm": 0.43176448345184326, "learning_rate": 4.81583468182652e-06, "loss": 0.0188, "step": 86620 }, { "grad_norm": 0.20158854126930237, "learning_rate": 4.808756970856809e-06, "loss": 0.017, "step": 86630 }, { "grad_norm": 0.08904366940259933, "learning_rate": 4.801684201933476e-06, "loss": 0.0134, "step": 86640 }, { "grad_norm": 0.4055947959423065, "learning_rate": 4.794616375830019e-06, "loss": 0.0181, "step": 86650 }, { "grad_norm": 0.2981471121311188, "learning_rate": 4.787553493319341e-06, "loss": 0.0195, "step": 86660 }, { "grad_norm": 0.39583462476730347, "learning_rate": 4.780495555173836e-06, "loss": 0.0176, "step": 86670 }, { "grad_norm": 0.19750948250293732, "learning_rate": 4.7734425621653475e-06, "loss": 0.0188, "step": 86680 }, { "grad_norm": 0.36341339349746704, "learning_rate": 4.766394515065198e-06, "loss": 0.017, "step": 86690 }, { "grad_norm": 0.23543831706047058, "learning_rate": 4.75935141464412e-06, "loss": 0.0133, "step": 86700 }, { "grad_norm": 0.422596275806427, "learning_rate": 4.75231326167237e-06, "loss": 0.018, "step": 86710 }, { "grad_norm": 0.43889641761779785, "learning_rate": 4.745280056919599e-06, "loss": 0.0118, "step": 86720 }, { "grad_norm": 0.1196305900812149, "learning_rate": 4.738251801154975e-06, "loss": 0.0225, "step": 86730 }, { "grad_norm": 0.4104042053222656, "learning_rate": 4.731228495147061e-06, "loss": 0.0149, "step": 86740 }, { "grad_norm": 0.1569184958934784, "learning_rate": 4.7242101396639595e-06, "loss": 0.0157, "step": 86750 }, { "grad_norm": 0.41813865303993225, "learning_rate": 4.717196735473151e-06, "loss": 0.0188, "step": 86760 }, { "grad_norm": 0.29823851585388184, "learning_rate": 4.710188283341643e-06, "loss": 0.0142, "step": 86770 }, { "grad_norm": 0.6090856790542603, "learning_rate": 4.703184784035835e-06, "loss": 0.017, "step": 86780 }, { "grad_norm": 0.23026829957962036, "learning_rate": 4.696186238321642e-06, "loss": 0.014, "step": 86790 }, { "grad_norm": 0.1895504742860794, "learning_rate": 4.689192646964408e-06, "loss": 0.01, "step": 86800 }, { "grad_norm": 0.7481330633163452, "learning_rate": 4.682204010728952e-06, "loss": 0.0197, "step": 86810 }, { "grad_norm": 0.40149229764938354, "learning_rate": 4.67522033037952e-06, "loss": 0.0139, "step": 86820 }, { "grad_norm": 0.2222023606300354, "learning_rate": 4.668241606679846e-06, "loss": 0.0159, "step": 86830 }, { "grad_norm": 0.683735728263855, "learning_rate": 4.66126784039313e-06, "loss": 0.0102, "step": 86840 }, { "grad_norm": 0.1463589072227478, "learning_rate": 4.6542990322819836e-06, "loss": 0.0135, "step": 86850 }, { "grad_norm": 0.1960441619157791, "learning_rate": 4.64733518310852e-06, "loss": 0.0182, "step": 86860 }, { "grad_norm": 0.295740008354187, "learning_rate": 4.640376293634291e-06, "loss": 0.0247, "step": 86870 }, { "grad_norm": 0.3613068461418152, "learning_rate": 4.633422364620327e-06, "loss": 0.0116, "step": 86880 }, { "grad_norm": 0.3486234247684479, "learning_rate": 4.6264733968270745e-06, "loss": 0.0138, "step": 86890 }, { "grad_norm": 0.5259890556335449, "learning_rate": 4.619529391014482e-06, "loss": 0.0265, "step": 86900 }, { "grad_norm": 0.6062293648719788, "learning_rate": 4.61259034794192e-06, "loss": 0.0116, "step": 86910 }, { "grad_norm": 0.37320762872695923, "learning_rate": 4.605656268368247e-06, "loss": 0.0196, "step": 86920 }, { "grad_norm": 0.4021220803260803, "learning_rate": 4.598727153051735e-06, "loss": 0.0177, "step": 86930 }, { "grad_norm": 0.49717211723327637, "learning_rate": 4.591803002750178e-06, "loss": 0.0121, "step": 86940 }, { "grad_norm": 0.1158113032579422, "learning_rate": 4.58488381822077e-06, "loss": 0.01, "step": 86950 }, { "grad_norm": 0.45683565735816956, "learning_rate": 4.5779696002201946e-06, "loss": 0.0138, "step": 86960 }, { "grad_norm": 0.2923758327960968, "learning_rate": 4.571060349504563e-06, "loss": 0.0136, "step": 86970 }, { "grad_norm": 0.24856022000312805, "learning_rate": 4.564156066829478e-06, "loss": 0.0121, "step": 86980 }, { "grad_norm": 0.48314300179481506, "learning_rate": 4.557256752949957e-06, "loss": 0.0163, "step": 86990 }, { "grad_norm": 0.47809138894081116, "learning_rate": 4.550362408620534e-06, "loss": 0.0189, "step": 87000 }, { "grad_norm": 0.796026885509491, "learning_rate": 4.543473034595136e-06, "loss": 0.0212, "step": 87010 }, { "grad_norm": 0.58269864320755, "learning_rate": 4.536588631627192e-06, "loss": 0.0188, "step": 87020 }, { "grad_norm": 0.1326724886894226, "learning_rate": 4.5297092004695565e-06, "loss": 0.0123, "step": 87030 }, { "grad_norm": 0.3531653881072998, "learning_rate": 4.522834741874555e-06, "loss": 0.018, "step": 87040 }, { "grad_norm": 0.2952461540699005, "learning_rate": 4.515965256593974e-06, "loss": 0.0113, "step": 87050 }, { "grad_norm": 0.20479702949523926, "learning_rate": 4.509100745379058e-06, "loss": 0.0164, "step": 87060 }, { "grad_norm": 1.478226661682129, "learning_rate": 4.502241208980479e-06, "loss": 0.0165, "step": 87070 }, { "grad_norm": 0.3607443869113922, "learning_rate": 4.495386648148398e-06, "loss": 0.0111, "step": 87080 }, { "grad_norm": 0.2651248872280121, "learning_rate": 4.488537063632431e-06, "loss": 0.0116, "step": 87090 }, { "grad_norm": 0.28473636507987976, "learning_rate": 4.481692456181608e-06, "loss": 0.0156, "step": 87100 }, { "grad_norm": 0.20604759454727173, "learning_rate": 4.474852826544479e-06, "loss": 0.0166, "step": 87110 }, { "grad_norm": 0.2104010432958603, "learning_rate": 4.468018175468974e-06, "loss": 0.0207, "step": 87120 }, { "grad_norm": 0.26346519589424133, "learning_rate": 4.461188503702568e-06, "loss": 0.0145, "step": 87130 }, { "grad_norm": 0.3326166272163391, "learning_rate": 4.454363811992107e-06, "loss": 0.0227, "step": 87140 }, { "grad_norm": 0.31083303689956665, "learning_rate": 4.447544101083956e-06, "loss": 0.0185, "step": 87150 }, { "grad_norm": 0.25030839443206787, "learning_rate": 4.440729371723884e-06, "loss": 0.0101, "step": 87160 }, { "grad_norm": 0.1774625927209854, "learning_rate": 4.433919624657162e-06, "loss": 0.0093, "step": 87170 }, { "grad_norm": 0.8821080923080444, "learning_rate": 4.427114860628462e-06, "loss": 0.0153, "step": 87180 }, { "grad_norm": 0.2986215353012085, "learning_rate": 4.420315080381982e-06, "loss": 0.0138, "step": 87190 }, { "grad_norm": 0.33947399258613586, "learning_rate": 4.4135202846613045e-06, "loss": 0.0192, "step": 87200 }, { "grad_norm": 0.16544099152088165, "learning_rate": 4.406730474209525e-06, "loss": 0.0161, "step": 87210 }, { "grad_norm": 0.26260754466056824, "learning_rate": 4.399945649769138e-06, "loss": 0.0162, "step": 87220 }, { "grad_norm": 0.16177402436733246, "learning_rate": 4.393165812082151e-06, "loss": 0.0119, "step": 87230 }, { "grad_norm": 0.5010260939598083, "learning_rate": 4.386390961889969e-06, "loss": 0.016, "step": 87240 }, { "grad_norm": 0.2828253507614136, "learning_rate": 4.3796210999334895e-06, "loss": 0.0176, "step": 87250 }, { "grad_norm": 0.21009638905525208, "learning_rate": 4.372856226953065e-06, "loss": 0.0132, "step": 87260 }, { "grad_norm": 0.3244439661502838, "learning_rate": 4.366096343688486e-06, "loss": 0.0186, "step": 87270 }, { "grad_norm": 0.22692133486270905, "learning_rate": 4.359341450878995e-06, "loss": 0.0141, "step": 87280 }, { "grad_norm": 0.2067480981349945, "learning_rate": 4.352591549263302e-06, "loss": 0.0127, "step": 87290 }, { "grad_norm": 0.2940447926521301, "learning_rate": 4.345846639579582e-06, "loss": 0.0159, "step": 87300 }, { "grad_norm": 0.26477161049842834, "learning_rate": 4.339106722565417e-06, "loss": 0.0106, "step": 87310 }, { "grad_norm": 0.2130216807126999, "learning_rate": 4.332371798957896e-06, "loss": 0.017, "step": 87320 }, { "grad_norm": 0.28168848156929016, "learning_rate": 4.3256418694935295e-06, "loss": 0.0188, "step": 87330 }, { "grad_norm": 0.4008724093437195, "learning_rate": 4.3189169349083125e-06, "loss": 0.0187, "step": 87340 }, { "grad_norm": 0.4784872531890869, "learning_rate": 4.312196995937645e-06, "loss": 0.0187, "step": 87350 }, { "grad_norm": 0.5873874425888062, "learning_rate": 4.305482053316434e-06, "loss": 0.0154, "step": 87360 }, { "grad_norm": 0.35076460242271423, "learning_rate": 4.2987721077789925e-06, "loss": 0.0274, "step": 87370 }, { "grad_norm": 0.22929967939853668, "learning_rate": 4.2920671600591325e-06, "loss": 0.014, "step": 87380 }, { "grad_norm": 0.4678526222705841, "learning_rate": 4.285367210890084e-06, "loss": 0.0163, "step": 87390 }, { "grad_norm": 0.6326373815536499, "learning_rate": 4.278672261004551e-06, "loss": 0.0222, "step": 87400 }, { "grad_norm": 0.22205621004104614, "learning_rate": 4.2719823111346704e-06, "loss": 0.0127, "step": 87410 }, { "grad_norm": 0.6999756693840027, "learning_rate": 4.265297362012061e-06, "loss": 0.0246, "step": 87420 }, { "grad_norm": 0.7390348315238953, "learning_rate": 4.258617414367766e-06, "loss": 0.0131, "step": 87430 }, { "grad_norm": 0.38668331503868103, "learning_rate": 4.251942468932296e-06, "loss": 0.0128, "step": 87440 }, { "grad_norm": 0.09514004737138748, "learning_rate": 4.2452725264356175e-06, "loss": 0.0116, "step": 87450 }, { "grad_norm": 0.310021311044693, "learning_rate": 4.2386075876071575e-06, "loss": 0.0117, "step": 87460 }, { "grad_norm": 0.2563135027885437, "learning_rate": 4.231947653175755e-06, "loss": 0.0226, "step": 87470 }, { "grad_norm": 0.35536864399909973, "learning_rate": 4.225292723869761e-06, "loss": 0.0163, "step": 87480 }, { "grad_norm": 0.4910714030265808, "learning_rate": 4.218642800416916e-06, "loss": 0.0167, "step": 87490 }, { "grad_norm": 1.235685110092163, "learning_rate": 4.211997883544472e-06, "loss": 0.0226, "step": 87500 }, { "grad_norm": 0.4531555771827698, "learning_rate": 4.205357973979096e-06, "loss": 0.0112, "step": 87510 }, { "grad_norm": 0.28256627917289734, "learning_rate": 4.198723072446925e-06, "loss": 0.0156, "step": 87520 }, { "grad_norm": 0.35901010036468506, "learning_rate": 4.192093179673534e-06, "loss": 0.0185, "step": 87530 }, { "grad_norm": 0.518644392490387, "learning_rate": 4.185468296383959e-06, "loss": 0.0203, "step": 87540 }, { "grad_norm": 0.14919808506965637, "learning_rate": 4.1788484233027005e-06, "loss": 0.018, "step": 87550 }, { "grad_norm": 0.25505903363227844, "learning_rate": 4.172233561153677e-06, "loss": 0.0171, "step": 87560 }, { "grad_norm": 0.2996392548084259, "learning_rate": 4.165623710660288e-06, "loss": 0.0125, "step": 87570 }, { "grad_norm": 0.2523448169231415, "learning_rate": 4.1590188725453835e-06, "loss": 0.0148, "step": 87580 }, { "grad_norm": 0.2827841639518738, "learning_rate": 4.152419047531259e-06, "loss": 0.0281, "step": 87590 }, { "grad_norm": 0.3120778501033783, "learning_rate": 4.145824236339651e-06, "loss": 0.0138, "step": 87600 }, { "grad_norm": 0.430001437664032, "learning_rate": 4.1392344396917695e-06, "loss": 0.0129, "step": 87610 }, { "grad_norm": 0.5108615756034851, "learning_rate": 4.132649658308252e-06, "loss": 0.018, "step": 87620 }, { "grad_norm": 0.4540693461894989, "learning_rate": 4.126069892909213e-06, "loss": 0.0148, "step": 87630 }, { "grad_norm": 0.6583656668663025, "learning_rate": 4.119495144214191e-06, "loss": 0.0156, "step": 87640 }, { "grad_norm": 0.24474604427814484, "learning_rate": 4.112925412942214e-06, "loss": 0.0172, "step": 87650 }, { "grad_norm": 0.39885830879211426, "learning_rate": 4.106360699811712e-06, "loss": 0.0151, "step": 87660 }, { "grad_norm": 0.30137649178504944, "learning_rate": 4.099801005540616e-06, "loss": 0.0169, "step": 87670 }, { "grad_norm": 0.25463220477104187, "learning_rate": 4.093246330846256e-06, "loss": 0.0177, "step": 87680 }, { "grad_norm": 0.1323879063129425, "learning_rate": 4.086696676445467e-06, "loss": 0.0154, "step": 87690 }, { "grad_norm": 0.25571590662002563, "learning_rate": 4.0801520430544914e-06, "loss": 0.0194, "step": 87700 }, { "grad_norm": 0.10778016597032547, "learning_rate": 4.073612431389062e-06, "loss": 0.0114, "step": 87710 }, { "grad_norm": 0.511495053768158, "learning_rate": 4.067077842164319e-06, "loss": 0.0176, "step": 87720 }, { "grad_norm": 0.4613565504550934, "learning_rate": 4.0605482760948945e-06, "loss": 0.0188, "step": 87730 }, { "grad_norm": 0.4092501401901245, "learning_rate": 4.054023733894829e-06, "loss": 0.0126, "step": 87740 }, { "grad_norm": 0.2475568801164627, "learning_rate": 4.047504216277653e-06, "loss": 0.013, "step": 87750 }, { "grad_norm": 0.18666903674602509, "learning_rate": 4.040989723956323e-06, "loss": 0.0194, "step": 87760 }, { "grad_norm": 0.6441752910614014, "learning_rate": 4.03448025764327e-06, "loss": 0.0197, "step": 87770 }, { "grad_norm": 0.2670697867870331, "learning_rate": 4.027975818050339e-06, "loss": 0.0144, "step": 87780 }, { "grad_norm": 0.14652416110038757, "learning_rate": 4.021476405888858e-06, "loss": 0.0136, "step": 87790 }, { "grad_norm": 0.36311689019203186, "learning_rate": 4.01498202186959e-06, "loss": 0.0185, "step": 87800 }, { "grad_norm": 0.19505208730697632, "learning_rate": 4.008492666702751e-06, "loss": 0.0153, "step": 87810 }, { "grad_norm": 0.16786079108715057, "learning_rate": 4.002008341098001e-06, "loss": 0.0126, "step": 87820 }, { "grad_norm": 0.3771957755088806, "learning_rate": 3.995529045764462e-06, "loss": 0.012, "step": 87830 }, { "grad_norm": 0.250304251909256, "learning_rate": 3.989054781410712e-06, "loss": 0.0158, "step": 87840 }, { "grad_norm": 0.2625412344932556, "learning_rate": 3.982585548744744e-06, "loss": 0.0107, "step": 87850 }, { "grad_norm": 0.3814496397972107, "learning_rate": 3.9761213484740435e-06, "loss": 0.0155, "step": 87860 }, { "grad_norm": 0.3603876531124115, "learning_rate": 3.9696621813055055e-06, "loss": 0.0122, "step": 87870 }, { "grad_norm": 0.2922929525375366, "learning_rate": 3.9632080479455044e-06, "loss": 0.0144, "step": 87880 }, { "grad_norm": 0.4945550560951233, "learning_rate": 3.956758949099859e-06, "loss": 0.0205, "step": 87890 }, { "grad_norm": 0.2240832895040512, "learning_rate": 3.9503148854738385e-06, "loss": 0.0131, "step": 87900 }, { "grad_norm": 0.2634623348712921, "learning_rate": 3.943875857772139e-06, "loss": 0.0228, "step": 87910 }, { "grad_norm": 0.3755810856819153, "learning_rate": 3.9374418666989375e-06, "loss": 0.0181, "step": 87920 }, { "grad_norm": 0.5184426307678223, "learning_rate": 3.931012912957838e-06, "loss": 0.0135, "step": 87930 }, { "grad_norm": 0.511512279510498, "learning_rate": 3.9245889972519045e-06, "loss": 0.0215, "step": 87940 }, { "grad_norm": 0.3464917242527008, "learning_rate": 3.9181701202836265e-06, "loss": 0.016, "step": 87950 }, { "grad_norm": 0.5278235673904419, "learning_rate": 3.911756282755003e-06, "loss": 0.0217, "step": 87960 }, { "grad_norm": 0.34795045852661133, "learning_rate": 3.905347485367411e-06, "loss": 0.0165, "step": 87970 }, { "grad_norm": 0.3981402516365051, "learning_rate": 3.898943728821725e-06, "loss": 0.029, "step": 87980 }, { "grad_norm": 0.46507173776626587, "learning_rate": 3.892545013818227e-06, "loss": 0.0163, "step": 87990 }, { "grad_norm": 0.2250429391860962, "learning_rate": 3.886151341056687e-06, "loss": 0.017, "step": 88000 }, { "grad_norm": 0.08198603242635727, "learning_rate": 3.8797627112363066e-06, "loss": 0.0117, "step": 88010 }, { "grad_norm": 0.3919433653354645, "learning_rate": 3.873379125055743e-06, "loss": 0.0139, "step": 88020 }, { "grad_norm": 0.4295509159564972, "learning_rate": 3.867000583213082e-06, "loss": 0.0166, "step": 88030 }, { "grad_norm": 0.2318364381790161, "learning_rate": 3.860627086405871e-06, "loss": 0.0142, "step": 88040 }, { "grad_norm": 1.0553150177001953, "learning_rate": 3.854258635331126e-06, "loss": 0.0135, "step": 88050 }, { "grad_norm": 0.6333070397377014, "learning_rate": 3.8478952306852665e-06, "loss": 0.0187, "step": 88060 }, { "grad_norm": 0.42402777075767517, "learning_rate": 3.841536873164198e-06, "loss": 0.0111, "step": 88070 }, { "grad_norm": 0.21451301872730255, "learning_rate": 3.8351835634632626e-06, "loss": 0.0118, "step": 88080 }, { "grad_norm": 0.34976303577423096, "learning_rate": 3.828835302277245e-06, "loss": 0.0181, "step": 88090 }, { "grad_norm": 0.4666508436203003, "learning_rate": 3.822492090300378e-06, "loss": 0.0157, "step": 88100 }, { "grad_norm": 0.2780391275882721, "learning_rate": 3.8161539282263566e-06, "loss": 0.0171, "step": 88110 }, { "grad_norm": 1.8580909967422485, "learning_rate": 3.8098208167482995e-06, "loss": 0.0206, "step": 88120 }, { "grad_norm": 0.27321794629096985, "learning_rate": 3.803492756558796e-06, "loss": 0.0144, "step": 88130 }, { "grad_norm": 0.21985997259616852, "learning_rate": 3.797169748349855e-06, "loss": 0.0154, "step": 88140 }, { "grad_norm": 0.33749544620513916, "learning_rate": 3.790851792812977e-06, "loss": 0.0131, "step": 88150 }, { "grad_norm": 0.43561992049217224, "learning_rate": 3.784538890639072e-06, "loss": 0.016, "step": 88160 }, { "grad_norm": 0.28404685854911804, "learning_rate": 3.7782310425185153e-06, "loss": 0.0138, "step": 88170 }, { "grad_norm": 0.1643313765525818, "learning_rate": 3.77192824914111e-06, "loss": 0.0131, "step": 88180 }, { "grad_norm": 0.44683244824409485, "learning_rate": 3.7656305111961333e-06, "loss": 0.0216, "step": 88190 }, { "grad_norm": 0.2670958638191223, "learning_rate": 3.7593378293722782e-06, "loss": 0.0221, "step": 88200 }, { "grad_norm": 0.3866192698478699, "learning_rate": 3.7530502043577333e-06, "loss": 0.0268, "step": 88210 }, { "grad_norm": 0.17579077184200287, "learning_rate": 3.746767636840076e-06, "loss": 0.0103, "step": 88220 }, { "grad_norm": 0.23823760449886322, "learning_rate": 3.74049012750638e-06, "loss": 0.0152, "step": 88230 }, { "grad_norm": 0.30997633934020996, "learning_rate": 3.734217677043128e-06, "loss": 0.0175, "step": 88240 }, { "grad_norm": 0.3353891968727112, "learning_rate": 3.727950286136267e-06, "loss": 0.0156, "step": 88250 }, { "grad_norm": 0.3964097797870636, "learning_rate": 3.7216879554711925e-06, "loss": 0.0129, "step": 88260 }, { "grad_norm": 0.12332382798194885, "learning_rate": 3.715430685732757e-06, "loss": 0.0091, "step": 88270 }, { "grad_norm": 0.19609405100345612, "learning_rate": 3.7091784776052243e-06, "loss": 0.0181, "step": 88280 }, { "grad_norm": 0.24313953518867493, "learning_rate": 3.7029313317723356e-06, "loss": 0.0206, "step": 88290 }, { "grad_norm": 0.2654992938041687, "learning_rate": 3.6966892489172787e-06, "loss": 0.0161, "step": 88300 }, { "grad_norm": 0.932927668094635, "learning_rate": 3.6904522297226575e-06, "loss": 0.0199, "step": 88310 }, { "grad_norm": 0.25957462191581726, "learning_rate": 3.6842202748705647e-06, "loss": 0.0134, "step": 88320 }, { "grad_norm": 0.3313872814178467, "learning_rate": 3.677993385042494e-06, "loss": 0.0186, "step": 88330 }, { "grad_norm": 0.2299758791923523, "learning_rate": 3.67177156091944e-06, "loss": 0.0117, "step": 88340 }, { "grad_norm": 0.5257580280303955, "learning_rate": 3.6655548031817866e-06, "loss": 0.0148, "step": 88350 }, { "grad_norm": 0.18349048495292664, "learning_rate": 3.6593431125094e-06, "loss": 0.017, "step": 88360 }, { "grad_norm": 0.1301613450050354, "learning_rate": 3.6531364895815766e-06, "loss": 0.0209, "step": 88370 }, { "grad_norm": 0.5349246859550476, "learning_rate": 3.6469349350770677e-06, "loss": 0.0212, "step": 88380 }, { "grad_norm": 0.18027275800704956, "learning_rate": 3.6407384496740525e-06, "loss": 0.0193, "step": 88390 }, { "grad_norm": 0.35398945212364197, "learning_rate": 3.6345470340501954e-06, "loss": 0.0153, "step": 88400 }, { "grad_norm": 0.2113792896270752, "learning_rate": 3.6283606888825537e-06, "loss": 0.0147, "step": 88410 }, { "grad_norm": 0.34047749638557434, "learning_rate": 3.622179414847676e-06, "loss": 0.0112, "step": 88420 }, { "grad_norm": 0.253952294588089, "learning_rate": 3.616003212621527e-06, "loss": 0.0182, "step": 88430 }, { "grad_norm": 0.4486689269542694, "learning_rate": 3.609832082879533e-06, "loss": 0.0152, "step": 88440 }, { "grad_norm": 0.6514151096343994, "learning_rate": 3.6036660262965485e-06, "loss": 0.0139, "step": 88450 }, { "grad_norm": 0.2846716046333313, "learning_rate": 3.597505043546895e-06, "loss": 0.0112, "step": 88460 }, { "grad_norm": 0.10825479030609131, "learning_rate": 3.5913491353043227e-06, "loss": 0.0109, "step": 88470 }, { "grad_norm": 0.3204769790172577, "learning_rate": 3.5851983022420375e-06, "loss": 0.016, "step": 88480 }, { "grad_norm": 0.24106596410274506, "learning_rate": 3.5790525450326797e-06, "loss": 0.0207, "step": 88490 }, { "grad_norm": 0.6593013405799866, "learning_rate": 3.572911864348344e-06, "loss": 0.0169, "step": 88500 }, { "grad_norm": 0.8515650033950806, "learning_rate": 3.5667762608605717e-06, "loss": 0.0192, "step": 88510 }, { "grad_norm": 0.12742094695568085, "learning_rate": 3.5606457352403255e-06, "loss": 0.0122, "step": 88520 }, { "grad_norm": 0.30382367968559265, "learning_rate": 3.554520288158042e-06, "loss": 0.0171, "step": 88530 }, { "grad_norm": 0.40846750140190125, "learning_rate": 3.5483999202835903e-06, "loss": 0.0139, "step": 88540 }, { "grad_norm": 0.2537393271923065, "learning_rate": 3.5422846322862913e-06, "loss": 0.0177, "step": 88550 }, { "grad_norm": 0.2826226055622101, "learning_rate": 3.5361744248348925e-06, "loss": 0.0177, "step": 88560 }, { "grad_norm": 0.47541457414627075, "learning_rate": 3.5300692985976047e-06, "loss": 0.0221, "step": 88570 }, { "grad_norm": 0.721731424331665, "learning_rate": 3.52396925424206e-06, "loss": 0.0236, "step": 88580 }, { "grad_norm": 0.14501510560512543, "learning_rate": 3.5178742924353747e-06, "loss": 0.0095, "step": 88590 }, { "grad_norm": 0.19911302626132965, "learning_rate": 3.511784413844066e-06, "loss": 0.0197, "step": 88600 }, { "grad_norm": 0.37879133224487305, "learning_rate": 3.5056996191341287e-06, "loss": 0.0177, "step": 88610 }, { "grad_norm": 0.4116770327091217, "learning_rate": 3.4996199089709692e-06, "loss": 0.0108, "step": 88620 }, { "grad_norm": 0.29479941725730896, "learning_rate": 3.4935452840194725e-06, "loss": 0.013, "step": 88630 }, { "grad_norm": 0.1668376624584198, "learning_rate": 3.48747574494393e-06, "loss": 0.016, "step": 88640 }, { "grad_norm": 0.8392866849899292, "learning_rate": 3.4814112924081154e-06, "loss": 0.0139, "step": 88650 }, { "grad_norm": 0.24980680644512177, "learning_rate": 3.475351927075221e-06, "loss": 0.013, "step": 88660 }, { "grad_norm": 0.34988221526145935, "learning_rate": 3.469297649607894e-06, "loss": 0.0144, "step": 88670 }, { "grad_norm": 0.8342552781105042, "learning_rate": 3.463248460668217e-06, "loss": 0.0195, "step": 88680 }, { "grad_norm": 0.34837424755096436, "learning_rate": 3.457204360917726e-06, "loss": 0.0101, "step": 88690 }, { "grad_norm": 0.4147915542125702, "learning_rate": 3.4511653510173827e-06, "loss": 0.0212, "step": 88700 }, { "grad_norm": 0.6134654879570007, "learning_rate": 3.4451314316276074e-06, "loss": 0.0207, "step": 88710 }, { "grad_norm": 0.21301116049289703, "learning_rate": 3.4391026034082674e-06, "loss": 0.0136, "step": 88720 }, { "grad_norm": 0.3912278115749359, "learning_rate": 3.433078867018674e-06, "loss": 0.0149, "step": 88730 }, { "grad_norm": 0.3589131534099579, "learning_rate": 3.427060223117551e-06, "loss": 0.0111, "step": 88740 }, { "grad_norm": 0.2858092188835144, "learning_rate": 3.4210466723631042e-06, "loss": 0.019, "step": 88750 }, { "grad_norm": 0.3769989609718323, "learning_rate": 3.415038215412969e-06, "loss": 0.0181, "step": 88760 }, { "grad_norm": 0.28725460171699524, "learning_rate": 3.4090348529242088e-06, "loss": 0.0223, "step": 88770 }, { "grad_norm": 0.4445350766181946, "learning_rate": 3.4030365855533487e-06, "loss": 0.0128, "step": 88780 }, { "grad_norm": 0.1271190345287323, "learning_rate": 3.397043413956347e-06, "loss": 0.0177, "step": 88790 }, { "grad_norm": 0.5959786176681519, "learning_rate": 3.3910553387886237e-06, "loss": 0.0126, "step": 88800 }, { "grad_norm": 0.3177793323993683, "learning_rate": 3.3850723607049996e-06, "loss": 0.0377, "step": 88810 }, { "grad_norm": 0.49063658714294434, "learning_rate": 3.3790944803597902e-06, "loss": 0.0201, "step": 88820 }, { "grad_norm": 0.31068655848503113, "learning_rate": 3.373121698406706e-06, "loss": 0.0134, "step": 88830 }, { "grad_norm": 0.26415398716926575, "learning_rate": 3.3671540154989302e-06, "loss": 0.0179, "step": 88840 }, { "grad_norm": 0.3413807451725006, "learning_rate": 3.361191432289079e-06, "loss": 0.0186, "step": 88850 }, { "grad_norm": 0.2817416489124298, "learning_rate": 3.3552339494292207e-06, "loss": 0.015, "step": 88860 }, { "grad_norm": 0.29910707473754883, "learning_rate": 3.349281567570839e-06, "loss": 0.0112, "step": 88870 }, { "grad_norm": 0.4883621335029602, "learning_rate": 3.3433342873648965e-06, "loss": 0.0146, "step": 88880 }, { "grad_norm": 0.5134246349334717, "learning_rate": 3.337392109461762e-06, "loss": 0.0143, "step": 88890 }, { "grad_norm": 0.3121020793914795, "learning_rate": 3.331455034511266e-06, "loss": 0.0173, "step": 88900 }, { "grad_norm": 0.5353327393531799, "learning_rate": 3.325523063162689e-06, "loss": 0.0172, "step": 88910 }, { "grad_norm": 0.15720370411872864, "learning_rate": 3.3195961960647393e-06, "loss": 0.023, "step": 88920 }, { "grad_norm": 0.33807530999183655, "learning_rate": 3.31367443386556e-06, "loss": 0.0161, "step": 88930 }, { "grad_norm": 0.2526197135448456, "learning_rate": 3.3077577772127556e-06, "loss": 0.0227, "step": 88940 }, { "grad_norm": 0.2874830365180969, "learning_rate": 3.301846226753358e-06, "loss": 0.0136, "step": 88950 }, { "grad_norm": 0.16809287667274475, "learning_rate": 3.2959397831338444e-06, "loss": 0.0207, "step": 88960 }, { "grad_norm": 0.33251872658729553, "learning_rate": 3.2900384470001432e-06, "loss": 0.012, "step": 88970 }, { "grad_norm": 0.262249231338501, "learning_rate": 3.28414221899761e-06, "loss": 0.0117, "step": 88980 }, { "grad_norm": 0.3156430423259735, "learning_rate": 3.278251099771046e-06, "loss": 0.0117, "step": 88990 }, { "grad_norm": 0.17606016993522644, "learning_rate": 3.2723650899646906e-06, "loss": 0.015, "step": 89000 }, { "grad_norm": 0.3921619653701782, "learning_rate": 3.2664841902222466e-06, "loss": 0.0169, "step": 89010 }, { "grad_norm": 0.2894797623157501, "learning_rate": 3.2606084011868208e-06, "loss": 0.0163, "step": 89020 }, { "grad_norm": 0.3024865388870239, "learning_rate": 3.254737723500989e-06, "loss": 0.0147, "step": 89030 }, { "grad_norm": 0.45440196990966797, "learning_rate": 3.2488721578067595e-06, "loss": 0.0141, "step": 89040 }, { "grad_norm": 0.4822641909122467, "learning_rate": 3.243011704745591e-06, "loss": 0.0155, "step": 89050 }, { "grad_norm": 0.37544554471969604, "learning_rate": 3.237156364958355e-06, "loss": 0.0167, "step": 89060 }, { "grad_norm": 0.548848032951355, "learning_rate": 3.231306139085394e-06, "loss": 0.0252, "step": 89070 }, { "grad_norm": 0.24100260436534882, "learning_rate": 3.2254610277664797e-06, "loss": 0.0103, "step": 89080 }, { "grad_norm": 0.3695225119590759, "learning_rate": 3.2196210316408183e-06, "loss": 0.0292, "step": 89090 }, { "grad_norm": 0.6223540902137756, "learning_rate": 3.2137861513470644e-06, "loss": 0.0197, "step": 89100 }, { "grad_norm": 0.3441425561904907, "learning_rate": 3.2079563875233252e-06, "loss": 0.0153, "step": 89110 }, { "grad_norm": 0.11800703406333923, "learning_rate": 3.2021317408071184e-06, "loss": 0.017, "step": 89120 }, { "grad_norm": 0.3072391748428345, "learning_rate": 3.196312211835434e-06, "loss": 0.0195, "step": 89130 }, { "grad_norm": 0.17483067512512207, "learning_rate": 3.1904978012446686e-06, "loss": 0.0194, "step": 89140 }, { "grad_norm": 0.2677009403705597, "learning_rate": 3.184688509670691e-06, "loss": 0.0448, "step": 89150 }, { "grad_norm": 0.4349493682384491, "learning_rate": 3.178884337748783e-06, "loss": 0.0123, "step": 89160 }, { "grad_norm": 0.39719247817993164, "learning_rate": 3.173085286113697e-06, "loss": 0.0147, "step": 89170 }, { "grad_norm": 0.3197576701641083, "learning_rate": 3.1672913553995986e-06, "loss": 0.0187, "step": 89180 }, { "grad_norm": 0.13074125349521637, "learning_rate": 3.1615025462401138e-06, "loss": 0.0172, "step": 89190 }, { "grad_norm": 0.23459236323833466, "learning_rate": 3.1557188592682808e-06, "loss": 0.0107, "step": 89200 }, { "grad_norm": 0.33027106523513794, "learning_rate": 3.1499402951166045e-06, "loss": 0.0141, "step": 89210 }, { "grad_norm": 0.714766263961792, "learning_rate": 3.144166854417019e-06, "loss": 0.0173, "step": 89220 }, { "grad_norm": 0.4927211403846741, "learning_rate": 3.1383985378009073e-06, "loss": 0.0099, "step": 89230 }, { "grad_norm": 0.2545992434024811, "learning_rate": 3.1326353458990654e-06, "loss": 0.0162, "step": 89240 }, { "grad_norm": 0.291014701128006, "learning_rate": 3.126877279341761e-06, "loss": 0.017, "step": 89250 }, { "grad_norm": 0.5114170908927917, "learning_rate": 3.1211243387586964e-06, "loss": 0.0209, "step": 89260 }, { "grad_norm": 0.28497663140296936, "learning_rate": 3.115376524778979e-06, "loss": 0.0136, "step": 89270 }, { "grad_norm": 0.2816149592399597, "learning_rate": 3.109633838031195e-06, "loss": 0.0232, "step": 89280 }, { "grad_norm": 0.49891385436058044, "learning_rate": 3.103896279143359e-06, "loss": 0.0139, "step": 89290 }, { "grad_norm": 0.3413050174713135, "learning_rate": 3.098163848742924e-06, "loss": 0.0162, "step": 89300 }, { "grad_norm": 0.3051816523075104, "learning_rate": 3.0924365474567675e-06, "loss": 0.0159, "step": 89310 }, { "grad_norm": 0.4368646740913391, "learning_rate": 3.086714375911237e-06, "loss": 0.0176, "step": 89320 }, { "grad_norm": 0.27214035391807556, "learning_rate": 3.0809973347320775e-06, "loss": 0.0214, "step": 89330 }, { "grad_norm": 0.28143587708473206, "learning_rate": 3.075285424544516e-06, "loss": 0.0131, "step": 89340 }, { "grad_norm": 0.2273920625448227, "learning_rate": 3.069578645973176e-06, "loss": 0.0171, "step": 89350 }, { "grad_norm": 0.22903208434581757, "learning_rate": 3.0638769996421735e-06, "loss": 0.0166, "step": 89360 }, { "grad_norm": 0.25782594084739685, "learning_rate": 3.058180486175005e-06, "loss": 0.018, "step": 89370 }, { "grad_norm": 0.15901589393615723, "learning_rate": 3.052489106194645e-06, "loss": 0.0111, "step": 89380 }, { "grad_norm": 0.4520516097545624, "learning_rate": 3.0468028603234946e-06, "loss": 0.0157, "step": 89390 }, { "grad_norm": 0.3540620803833008, "learning_rate": 3.04112174918339e-06, "loss": 0.0253, "step": 89400 }, { "grad_norm": 0.1404857486486435, "learning_rate": 3.0354457733955954e-06, "loss": 0.0128, "step": 89410 }, { "grad_norm": 0.2805942893028259, "learning_rate": 3.029774933580859e-06, "loss": 0.0158, "step": 89420 }, { "grad_norm": 0.43786492943763733, "learning_rate": 3.0241092303593065e-06, "loss": 0.0106, "step": 89430 }, { "grad_norm": 0.2061878740787506, "learning_rate": 3.018448664350554e-06, "loss": 0.0112, "step": 89440 }, { "grad_norm": 0.15856440365314484, "learning_rate": 3.0127932361736055e-06, "loss": 0.0173, "step": 89450 }, { "grad_norm": 0.28599363565444946, "learning_rate": 3.00714294644695e-06, "loss": 0.0136, "step": 89460 }, { "grad_norm": 0.2436543107032776, "learning_rate": 3.0014977957884828e-06, "loss": 0.0329, "step": 89470 }, { "grad_norm": 0.2934218645095825, "learning_rate": 2.99585778481557e-06, "loss": 0.0125, "step": 89480 }, { "grad_norm": 0.4420866072177887, "learning_rate": 2.990222914144969e-06, "loss": 0.0166, "step": 89490 }, { "grad_norm": 0.45768049359321594, "learning_rate": 2.984593184392914e-06, "loss": 0.0243, "step": 89500 }, { "grad_norm": 0.6768931150436401, "learning_rate": 2.978968596175069e-06, "loss": 0.0199, "step": 89510 }, { "grad_norm": 0.30847394466400146, "learning_rate": 2.9733491501065137e-06, "loss": 0.0167, "step": 89520 }, { "grad_norm": 0.33641237020492554, "learning_rate": 2.9677348468018005e-06, "loss": 0.0189, "step": 89530 }, { "grad_norm": 0.2724337875843048, "learning_rate": 2.9621256868748726e-06, "loss": 0.017, "step": 89540 }, { "grad_norm": 0.47861793637275696, "learning_rate": 2.956521670939183e-06, "loss": 0.0195, "step": 89550 }, { "grad_norm": 0.23034271597862244, "learning_rate": 2.950922799607542e-06, "loss": 0.0191, "step": 89560 }, { "grad_norm": 0.4615974724292755, "learning_rate": 2.9453290734922535e-06, "loss": 0.043, "step": 89570 }, { "grad_norm": 0.24567551910877228, "learning_rate": 2.939740493205029e-06, "loss": 0.0186, "step": 89580 }, { "grad_norm": 0.36856111884117126, "learning_rate": 2.9341570593570345e-06, "loss": 0.0222, "step": 89590 }, { "grad_norm": 0.3313816785812378, "learning_rate": 2.928578772558843e-06, "loss": 0.014, "step": 89600 }, { "grad_norm": 0.4333721101284027, "learning_rate": 2.923005633420528e-06, "loss": 0.0123, "step": 89610 }, { "grad_norm": 0.7233840823173523, "learning_rate": 2.9174376425515237e-06, "loss": 0.0188, "step": 89620 }, { "grad_norm": 0.3003902733325958, "learning_rate": 2.91187480056076e-06, "loss": 0.016, "step": 89630 }, { "grad_norm": 0.1035694107413292, "learning_rate": 2.9063171080565664e-06, "loss": 0.03, "step": 89640 }, { "grad_norm": 0.28078019618988037, "learning_rate": 2.900764565646741e-06, "loss": 0.0133, "step": 89650 }, { "grad_norm": 0.15973511338233948, "learning_rate": 2.895217173938475e-06, "loss": 0.025, "step": 89660 }, { "grad_norm": 0.15022523701190948, "learning_rate": 2.8896749335384445e-06, "loss": 0.0103, "step": 89670 }, { "grad_norm": 0.41812872886657715, "learning_rate": 2.884137845052731e-06, "loss": 0.0163, "step": 89680 }, { "grad_norm": 0.3995759189128876, "learning_rate": 2.8786059090868734e-06, "loss": 0.0199, "step": 89690 }, { "grad_norm": 0.29071763157844543, "learning_rate": 2.87307912624582e-06, "loss": 0.016, "step": 89700 }, { "grad_norm": 0.3144381642341614, "learning_rate": 2.867557497133977e-06, "loss": 0.0096, "step": 89710 }, { "grad_norm": 0.4918665587902069, "learning_rate": 2.862041022355194e-06, "loss": 0.0115, "step": 89720 }, { "grad_norm": 0.37063416838645935, "learning_rate": 2.8565297025127226e-06, "loss": 0.0226, "step": 89730 }, { "grad_norm": 0.2049020528793335, "learning_rate": 2.8510235382092853e-06, "loss": 0.0213, "step": 89740 }, { "grad_norm": 0.2048480212688446, "learning_rate": 2.845522530047029e-06, "loss": 0.0122, "step": 89750 }, { "grad_norm": 0.43344807624816895, "learning_rate": 2.8400266786275387e-06, "loss": 0.0153, "step": 89760 }, { "grad_norm": 0.4064908027648926, "learning_rate": 2.834535984551817e-06, "loss": 0.0165, "step": 89770 }, { "grad_norm": 0.382325142621994, "learning_rate": 2.8290504484203397e-06, "loss": 0.017, "step": 89780 }, { "grad_norm": 0.29169896245002747, "learning_rate": 2.8235700708329705e-06, "loss": 0.0118, "step": 89790 }, { "grad_norm": 0.41966181993484497, "learning_rate": 2.818094852389058e-06, "loss": 0.0212, "step": 89800 }, { "grad_norm": 0.08631222695112228, "learning_rate": 2.8126247936873516e-06, "loss": 0.0136, "step": 89810 }, { "grad_norm": 0.10935146361589432, "learning_rate": 2.8071598953260614e-06, "loss": 0.0169, "step": 89820 }, { "grad_norm": 0.319817453622818, "learning_rate": 2.8017001579028035e-06, "loss": 0.0182, "step": 89830 }, { "grad_norm": 1.1090737581253052, "learning_rate": 2.7962455820146617e-06, "loss": 0.0201, "step": 89840 }, { "grad_norm": 0.31905168294906616, "learning_rate": 2.7907961682581253e-06, "loss": 0.0183, "step": 89850 }, { "grad_norm": 0.12569770216941833, "learning_rate": 2.7853519172291453e-06, "loss": 0.0078, "step": 89860 }, { "grad_norm": 0.6469892263412476, "learning_rate": 2.7799128295230904e-06, "loss": 0.0189, "step": 89870 }, { "grad_norm": 0.38291868567466736, "learning_rate": 2.774478905734784e-06, "loss": 0.0171, "step": 89880 }, { "grad_norm": 0.5846491456031799, "learning_rate": 2.7690501464584516e-06, "loss": 0.02, "step": 89890 }, { "grad_norm": 0.4691046476364136, "learning_rate": 2.763626552287796e-06, "loss": 0.0129, "step": 89900 }, { "grad_norm": 0.6746562123298645, "learning_rate": 2.7582081238159083e-06, "loss": 0.0169, "step": 89910 }, { "grad_norm": 0.25784769654273987, "learning_rate": 2.7527948616353548e-06, "loss": 0.0141, "step": 89920 }, { "grad_norm": 0.22079019248485565, "learning_rate": 2.7473867663381226e-06, "loss": 0.0115, "step": 89930 }, { "grad_norm": 0.30912351608276367, "learning_rate": 2.7419838385156383e-06, "loss": 0.026, "step": 89940 }, { "grad_norm": 0.591823935508728, "learning_rate": 2.7365860787587407e-06, "loss": 0.0148, "step": 89950 }, { "grad_norm": 0.2440595030784607, "learning_rate": 2.7311934876577295e-06, "loss": 0.0215, "step": 89960 }, { "grad_norm": 0.39816850423812866, "learning_rate": 2.725806065802339e-06, "loss": 0.0119, "step": 89970 }, { "grad_norm": 0.23479948937892914, "learning_rate": 2.720423813781714e-06, "loss": 0.0167, "step": 89980 }, { "grad_norm": 0.4143769145011902, "learning_rate": 2.715046732184462e-06, "loss": 0.0122, "step": 89990 }, { "grad_norm": 0.4160683751106262, "learning_rate": 2.7096748215986013e-06, "loss": 0.0251, "step": 90000 }, { "grad_norm": 0.9331037402153015, "learning_rate": 2.704308082611612e-06, "loss": 0.01, "step": 90010 }, { "grad_norm": 0.4215265214443207, "learning_rate": 2.698946515810374e-06, "loss": 0.0117, "step": 90020 }, { "grad_norm": 0.5356293320655823, "learning_rate": 2.6935901217812367e-06, "loss": 0.016, "step": 90030 }, { "grad_norm": 0.43065765500068665, "learning_rate": 2.688238901109952e-06, "loss": 0.0238, "step": 90040 }, { "grad_norm": 0.25579598546028137, "learning_rate": 2.68289285438173e-06, "loss": 0.0118, "step": 90050 }, { "grad_norm": 0.4267379939556122, "learning_rate": 2.6775519821812033e-06, "loss": 0.0143, "step": 90060 }, { "grad_norm": 0.3422040343284607, "learning_rate": 2.672216285092449e-06, "loss": 0.014, "step": 90070 }, { "grad_norm": 0.27009281516075134, "learning_rate": 2.666885763698962e-06, "loss": 0.0161, "step": 90080 }, { "grad_norm": 0.3192499876022339, "learning_rate": 2.661560418583692e-06, "loss": 0.0096, "step": 90090 }, { "grad_norm": 0.23286913335323334, "learning_rate": 2.656240250328995e-06, "loss": 0.0149, "step": 90100 }, { "grad_norm": 0.3736872971057892, "learning_rate": 2.650925259516679e-06, "loss": 0.0205, "step": 90110 }, { "grad_norm": 0.4292900562286377, "learning_rate": 2.645615446727995e-06, "loss": 0.0133, "step": 90120 }, { "grad_norm": 0.3330083191394806, "learning_rate": 2.6403108125436115e-06, "loss": 0.0113, "step": 90130 }, { "grad_norm": 0.3068373203277588, "learning_rate": 2.6350113575436264e-06, "loss": 0.0174, "step": 90140 }, { "grad_norm": 0.24678963422775269, "learning_rate": 2.629717082307598e-06, "loss": 0.0103, "step": 90150 }, { "grad_norm": 0.3146621882915497, "learning_rate": 2.6244279874144793e-06, "loss": 0.0127, "step": 90160 }, { "grad_norm": 0.29852399230003357, "learning_rate": 2.6191440734426862e-06, "loss": 0.0184, "step": 90170 }, { "grad_norm": 0.2653844356536865, "learning_rate": 2.613865340970062e-06, "loss": 0.0134, "step": 90180 }, { "grad_norm": 0.34425556659698486, "learning_rate": 2.608591790573889e-06, "loss": 0.016, "step": 90190 }, { "grad_norm": 0.6365183591842651, "learning_rate": 2.603323422830861e-06, "loss": 0.0131, "step": 90200 }, { "grad_norm": 0.5576565265655518, "learning_rate": 2.5980602383171183e-06, "loss": 0.0161, "step": 90210 }, { "grad_norm": 0.27450400590896606, "learning_rate": 2.5928022376082496e-06, "loss": 0.0135, "step": 90220 }, { "grad_norm": 0.633855402469635, "learning_rate": 2.5875494212792505e-06, "loss": 0.0148, "step": 90230 }, { "grad_norm": 0.09807470440864563, "learning_rate": 2.5823017899045565e-06, "loss": 0.0159, "step": 90240 }, { "grad_norm": 0.2803958058357239, "learning_rate": 2.577059344058047e-06, "loss": 0.0144, "step": 90250 }, { "grad_norm": 0.2539590895175934, "learning_rate": 2.5718220843130415e-06, "loss": 0.0133, "step": 90260 }, { "grad_norm": 0.200160413980484, "learning_rate": 2.5665900112422535e-06, "loss": 0.0164, "step": 90270 }, { "grad_norm": 0.3192153871059418, "learning_rate": 2.561363125417876e-06, "loss": 0.0116, "step": 90280 }, { "grad_norm": 0.3660360276699066, "learning_rate": 2.556141427411496e-06, "loss": 0.0087, "step": 90290 }, { "grad_norm": 0.32268550992012024, "learning_rate": 2.550924917794156e-06, "loss": 0.0142, "step": 90300 }, { "grad_norm": 0.23940475285053253, "learning_rate": 2.545713597136329e-06, "loss": 0.0133, "step": 90310 }, { "grad_norm": 0.15441346168518066, "learning_rate": 2.5405074660079298e-06, "loss": 0.0331, "step": 90320 }, { "grad_norm": 0.8401882648468018, "learning_rate": 2.5353065249782647e-06, "loss": 0.0224, "step": 90330 }, { "grad_norm": 0.42450210452079773, "learning_rate": 2.530110774616129e-06, "loss": 0.019, "step": 90340 }, { "grad_norm": 0.6416400671005249, "learning_rate": 2.524920215489701e-06, "loss": 0.0186, "step": 90350 }, { "grad_norm": 0.10547710210084915, "learning_rate": 2.519734848166627e-06, "loss": 0.0143, "step": 90360 }, { "grad_norm": 0.7316929697990417, "learning_rate": 2.514554673213948e-06, "loss": 0.0235, "step": 90370 }, { "grad_norm": 0.4909874498844147, "learning_rate": 2.5093796911981936e-06, "loss": 0.0147, "step": 90380 }, { "grad_norm": 0.31534692645072937, "learning_rate": 2.5042099026852672e-06, "loss": 0.015, "step": 90390 }, { "grad_norm": 0.1869734823703766, "learning_rate": 2.499045308240544e-06, "loss": 0.012, "step": 90400 }, { "grad_norm": 0.3498381972312927, "learning_rate": 2.4938859084288003e-06, "loss": 0.0159, "step": 90410 }, { "grad_norm": 5.3375630378723145, "learning_rate": 2.4887317038142734e-06, "loss": 0.0158, "step": 90420 }, { "grad_norm": 0.3066959083080292, "learning_rate": 2.4835826949606124e-06, "loss": 0.0111, "step": 90430 }, { "grad_norm": 0.44190433621406555, "learning_rate": 2.4784388824309167e-06, "loss": 0.0133, "step": 90440 }, { "grad_norm": 0.11932653933763504, "learning_rate": 2.473300266787687e-06, "loss": 0.015, "step": 90450 }, { "grad_norm": 0.3877980411052704, "learning_rate": 2.46816684859289e-06, "loss": 0.0122, "step": 90460 }, { "grad_norm": 0.20986512303352356, "learning_rate": 2.4630386284079043e-06, "loss": 0.0114, "step": 90470 }, { "grad_norm": 0.23502595722675323, "learning_rate": 2.4579156067935427e-06, "loss": 0.0217, "step": 90480 }, { "grad_norm": 0.43290776014328003, "learning_rate": 2.452797784310046e-06, "loss": 0.0162, "step": 90490 }, { "grad_norm": 0.2129841148853302, "learning_rate": 2.447685161517105e-06, "loss": 0.0087, "step": 90500 }, { "grad_norm": 0.4167634844779968, "learning_rate": 2.4425777389738224e-06, "loss": 0.0112, "step": 90510 }, { "grad_norm": 0.1592540144920349, "learning_rate": 2.43747551723873e-06, "loss": 0.0153, "step": 90520 }, { "grad_norm": 0.3315993845462799, "learning_rate": 2.4323784968698136e-06, "loss": 0.0192, "step": 90530 }, { "grad_norm": 0.25838586688041687, "learning_rate": 2.4272866784244565e-06, "loss": 0.0099, "step": 90540 }, { "grad_norm": 0.4220009446144104, "learning_rate": 2.422200062459518e-06, "loss": 0.0154, "step": 90550 }, { "grad_norm": 0.4310469329357147, "learning_rate": 2.41711864953123e-06, "loss": 0.0132, "step": 90560 }, { "grad_norm": 0.13528181612491608, "learning_rate": 2.412042440195322e-06, "loss": 0.0158, "step": 90570 }, { "grad_norm": 0.22839418053627014, "learning_rate": 2.4069714350068993e-06, "loss": 0.0211, "step": 90580 }, { "grad_norm": 0.3075699508190155, "learning_rate": 2.4019056345205294e-06, "loss": 0.0267, "step": 90590 }, { "grad_norm": 0.6518600583076477, "learning_rate": 2.396845039290191e-06, "loss": 0.0191, "step": 90600 }, { "grad_norm": 0.3045485317707062, "learning_rate": 2.3917896498693194e-06, "loss": 0.0153, "step": 90610 }, { "grad_norm": 0.2175264209508896, "learning_rate": 2.3867394668107335e-06, "loss": 0.0277, "step": 90620 }, { "grad_norm": 0.33453381061553955, "learning_rate": 2.3816944906667527e-06, "loss": 0.0112, "step": 90630 }, { "grad_norm": 0.20049723982810974, "learning_rate": 2.3766547219890634e-06, "loss": 0.011, "step": 90640 }, { "grad_norm": 0.33016958832740784, "learning_rate": 2.371620161328819e-06, "loss": 0.0146, "step": 90650 }, { "grad_norm": 0.25987014174461365, "learning_rate": 2.366590809236574e-06, "loss": 0.0093, "step": 90660 }, { "grad_norm": 0.22172027826309204, "learning_rate": 2.3615666662623494e-06, "loss": 0.0186, "step": 90670 }, { "grad_norm": 0.3358973264694214, "learning_rate": 2.3565477329555664e-06, "loss": 0.0157, "step": 90680 }, { "grad_norm": 0.309407502412796, "learning_rate": 2.3515340098650975e-06, "loss": 0.0216, "step": 90690 }, { "grad_norm": 0.22188593447208405, "learning_rate": 2.3465254975392258e-06, "loss": 0.0136, "step": 90700 }, { "grad_norm": 0.4055653214454651, "learning_rate": 2.3415221965256805e-06, "loss": 0.0276, "step": 90710 }, { "grad_norm": 0.49908336997032166, "learning_rate": 2.3365241073716237e-06, "loss": 0.026, "step": 90720 }, { "grad_norm": 0.20530422031879425, "learning_rate": 2.3315312306236183e-06, "loss": 0.0144, "step": 90730 }, { "grad_norm": 0.32738757133483887, "learning_rate": 2.326543566827699e-06, "loss": 0.0151, "step": 90740 }, { "grad_norm": 0.37407901883125305, "learning_rate": 2.3215611165292862e-06, "loss": 0.0128, "step": 90750 }, { "grad_norm": 0.3427391052246094, "learning_rate": 2.3165838802732765e-06, "loss": 0.0131, "step": 90760 }, { "grad_norm": 0.6990312337875366, "learning_rate": 2.311611858603957e-06, "loss": 0.0155, "step": 90770 }, { "grad_norm": 0.3656933009624481, "learning_rate": 2.3066450520650763e-06, "loss": 0.0194, "step": 90780 }, { "grad_norm": 0.2608727216720581, "learning_rate": 2.3016834611997773e-06, "loss": 0.0145, "step": 90790 }, { "grad_norm": 0.0874376893043518, "learning_rate": 2.296727086550665e-06, "loss": 0.0115, "step": 90800 }, { "grad_norm": 0.31711727380752563, "learning_rate": 2.2917759286597507e-06, "loss": 0.0139, "step": 90810 }, { "grad_norm": 0.14692114293575287, "learning_rate": 2.2868299880685006e-06, "loss": 0.0189, "step": 90820 }, { "grad_norm": 0.2658417820930481, "learning_rate": 2.2818892653177827e-06, "loss": 0.0153, "step": 90830 }, { "grad_norm": 0.16928616166114807, "learning_rate": 2.2769537609479196e-06, "loss": 0.0087, "step": 90840 }, { "grad_norm": 0.21593356132507324, "learning_rate": 2.2720234754986357e-06, "loss": 0.0115, "step": 90850 }, { "grad_norm": 0.6607963442802429, "learning_rate": 2.2670984095091108e-06, "loss": 0.0224, "step": 90860 }, { "grad_norm": 0.45715266466140747, "learning_rate": 2.262178563517936e-06, "loss": 0.0201, "step": 90870 }, { "grad_norm": 0.34097135066986084, "learning_rate": 2.257263938063137e-06, "loss": 0.012, "step": 90880 }, { "grad_norm": 0.22061316668987274, "learning_rate": 2.2523545336821715e-06, "loss": 0.0194, "step": 90890 }, { "grad_norm": 0.3214077651500702, "learning_rate": 2.247450350911939e-06, "loss": 0.0189, "step": 90900 }, { "grad_norm": 0.799622654914856, "learning_rate": 2.2425513902887328e-06, "loss": 0.0213, "step": 90910 }, { "grad_norm": 0.38643863797187805, "learning_rate": 2.2376576523483015e-06, "loss": 0.0187, "step": 90920 }, { "grad_norm": 0.40030649304389954, "learning_rate": 2.232769137625829e-06, "loss": 0.0149, "step": 90930 }, { "grad_norm": 0.16212019324302673, "learning_rate": 2.227885846655903e-06, "loss": 0.0171, "step": 90940 }, { "grad_norm": 0.5647504329681396, "learning_rate": 2.2230077799725523e-06, "loss": 0.0179, "step": 90950 }, { "grad_norm": 0.35598447918891907, "learning_rate": 2.2181349381092396e-06, "loss": 0.0117, "step": 90960 }, { "grad_norm": 0.8139737844467163, "learning_rate": 2.2132673215988552e-06, "loss": 0.0187, "step": 90970 }, { "grad_norm": 0.6070607304573059, "learning_rate": 2.2084049309737065e-06, "loss": 0.0134, "step": 90980 }, { "grad_norm": 0.38866111636161804, "learning_rate": 2.2035477667655513e-06, "loss": 0.0115, "step": 90990 }, { "grad_norm": 0.4440558850765228, "learning_rate": 2.1986958295055315e-06, "loss": 0.0197, "step": 91000 }, { "grad_norm": 1.7487146854400635, "learning_rate": 2.193849119724284e-06, "loss": 0.0132, "step": 91010 }, { "grad_norm": 0.5014232397079468, "learning_rate": 2.1890076379518175e-06, "loss": 0.0155, "step": 91020 }, { "grad_norm": 0.16104568541049957, "learning_rate": 2.1841713847175925e-06, "loss": 0.0122, "step": 91030 }, { "grad_norm": 0.14214690029621124, "learning_rate": 2.1793403605504968e-06, "loss": 0.0118, "step": 91040 }, { "grad_norm": 0.23797087371349335, "learning_rate": 2.1745145659788413e-06, "loss": 0.014, "step": 91050 }, { "grad_norm": 0.49646466970443726, "learning_rate": 2.1696940015303645e-06, "loss": 0.0113, "step": 91060 }, { "grad_norm": 0.17293745279312134, "learning_rate": 2.1648786677322397e-06, "loss": 0.0122, "step": 91070 }, { "grad_norm": 0.20188170671463013, "learning_rate": 2.1600685651110673e-06, "loss": 0.0162, "step": 91080 }, { "grad_norm": 0.32966819405555725, "learning_rate": 2.1552636941928717e-06, "loss": 0.0167, "step": 91090 }, { "grad_norm": 0.2781447768211365, "learning_rate": 2.1504640555031043e-06, "loss": 0.0177, "step": 91100 }, { "grad_norm": 0.36006960272789, "learning_rate": 2.1456696495666506e-06, "loss": 0.0215, "step": 91110 }, { "grad_norm": 0.23336610198020935, "learning_rate": 2.1408804769078084e-06, "loss": 0.0161, "step": 91120 }, { "grad_norm": 0.09715385735034943, "learning_rate": 2.136096538050325e-06, "loss": 0.011, "step": 91130 }, { "grad_norm": 0.1986570805311203, "learning_rate": 2.13131783351736e-06, "loss": 0.0175, "step": 91140 }, { "grad_norm": 0.33938363194465637, "learning_rate": 2.126544363831512e-06, "loss": 0.0202, "step": 91150 }, { "grad_norm": 0.5297418236732483, "learning_rate": 2.121776129514791e-06, "loss": 0.0146, "step": 91160 }, { "grad_norm": 0.11489226669073105, "learning_rate": 2.1170131310886523e-06, "loss": 0.0096, "step": 91170 }, { "grad_norm": 0.3789939284324646, "learning_rate": 2.112255369073968e-06, "loss": 0.0106, "step": 91180 }, { "grad_norm": 0.34329068660736084, "learning_rate": 2.1075028439910328e-06, "loss": 0.014, "step": 91190 }, { "grad_norm": 0.3131581246852875, "learning_rate": 2.1027555563595813e-06, "loss": 0.0143, "step": 91200 }, { "grad_norm": 0.32867029309272766, "learning_rate": 2.0980135066987706e-06, "loss": 0.0117, "step": 91210 }, { "grad_norm": 0.31584101915359497, "learning_rate": 2.093276695527191e-06, "loss": 0.0139, "step": 91220 }, { "grad_norm": 0.5656781196594238, "learning_rate": 2.0885451233628396e-06, "loss": 0.0243, "step": 91230 }, { "grad_norm": 0.3414847254753113, "learning_rate": 2.0838187907231633e-06, "loss": 0.0199, "step": 91240 }, { "grad_norm": 0.3699414134025574, "learning_rate": 2.0790976981250153e-06, "loss": 0.0088, "step": 91250 }, { "grad_norm": 0.4606313407421112, "learning_rate": 2.0743818460847e-06, "loss": 0.0193, "step": 91260 }, { "grad_norm": 0.18534576892852783, "learning_rate": 2.069671235117926e-06, "loss": 0.0166, "step": 91270 }, { "grad_norm": 0.2766864001750946, "learning_rate": 2.0649658657398484e-06, "loss": 0.0126, "step": 91280 }, { "grad_norm": 0.48478570580482483, "learning_rate": 2.060265738465034e-06, "loss": 0.0149, "step": 91290 }, { "grad_norm": 0.1926380693912506, "learning_rate": 2.055570853807487e-06, "loss": 0.015, "step": 91300 }, { "grad_norm": 0.28463491797447205, "learning_rate": 2.0508812122806196e-06, "loss": 0.0113, "step": 91310 }, { "grad_norm": 0.27171429991722107, "learning_rate": 2.046196814397294e-06, "loss": 0.0138, "step": 91320 }, { "grad_norm": 0.28697097301483154, "learning_rate": 2.0415176606697896e-06, "loss": 0.0161, "step": 91330 }, { "grad_norm": 0.347298800945282, "learning_rate": 2.0368437516098126e-06, "loss": 0.0167, "step": 91340 }, { "grad_norm": 0.2092285454273224, "learning_rate": 2.032175087728483e-06, "loss": 0.0185, "step": 91350 }, { "grad_norm": 0.2790646255016327, "learning_rate": 2.027511669536375e-06, "loss": 0.0182, "step": 91360 }, { "grad_norm": 0.39610555768013, "learning_rate": 2.0228534975434643e-06, "loss": 0.0174, "step": 91370 }, { "grad_norm": 0.2791581153869629, "learning_rate": 2.0182005722591548e-06, "loss": 0.0171, "step": 91380 }, { "grad_norm": 0.17924804985523224, "learning_rate": 2.0135528941922944e-06, "loss": 0.019, "step": 91390 }, { "grad_norm": 0.2510846257209778, "learning_rate": 2.008910463851149e-06, "loss": 0.0145, "step": 91400 }, { "grad_norm": 0.5004599094390869, "learning_rate": 2.0042732817433963e-06, "loss": 0.0191, "step": 91410 }, { "grad_norm": 0.2856985032558441, "learning_rate": 1.999641348376158e-06, "loss": 0.0162, "step": 91420 }, { "grad_norm": 0.18354865908622742, "learning_rate": 1.9950146642559787e-06, "loss": 0.0145, "step": 91430 }, { "grad_norm": 0.39527907967567444, "learning_rate": 1.990393229888815e-06, "loss": 0.018, "step": 91440 }, { "grad_norm": 0.2476101815700531, "learning_rate": 1.9857770457800684e-06, "loss": 0.0146, "step": 91450 }, { "grad_norm": 0.19223377108573914, "learning_rate": 1.9811661124345517e-06, "loss": 0.0116, "step": 91460 }, { "grad_norm": 0.3705977201461792, "learning_rate": 1.976560430356522e-06, "loss": 0.0189, "step": 91470 }, { "grad_norm": 0.1286117434501648, "learning_rate": 1.9719600000496383e-06, "loss": 0.0112, "step": 91480 }, { "grad_norm": 0.3643127679824829, "learning_rate": 1.967364822017004e-06, "loss": 0.0145, "step": 91490 }, { "grad_norm": 0.23903299868106842, "learning_rate": 1.962774896761127e-06, "loss": 0.014, "step": 91500 }, { "grad_norm": 0.13808470964431763, "learning_rate": 1.958190224783968e-06, "loss": 0.0184, "step": 91510 }, { "grad_norm": 0.13532990217208862, "learning_rate": 1.9536108065868975e-06, "loss": 0.0143, "step": 91520 }, { "grad_norm": 0.14179931581020355, "learning_rate": 1.949036642670715e-06, "loss": 0.0085, "step": 91530 }, { "grad_norm": 0.3397161662578583, "learning_rate": 1.9444677335356377e-06, "loss": 0.0133, "step": 91540 }, { "grad_norm": 0.2289067506790161, "learning_rate": 1.939904079681326e-06, "loss": 0.0141, "step": 91550 }, { "grad_norm": 0.10275295376777649, "learning_rate": 1.935345681606837e-06, "loss": 0.0211, "step": 91560 }, { "grad_norm": 0.4010847508907318, "learning_rate": 1.930792539810694e-06, "loss": 0.0234, "step": 91570 }, { "grad_norm": 0.2862720489501953, "learning_rate": 1.9262446547907877e-06, "loss": 0.019, "step": 91580 }, { "grad_norm": 0.36197343468666077, "learning_rate": 1.921702027044503e-06, "loss": 0.0149, "step": 91590 }, { "grad_norm": 0.19091875851154327, "learning_rate": 1.917164657068593e-06, "loss": 0.0164, "step": 91600 }, { "grad_norm": 0.3657260835170746, "learning_rate": 1.912632545359272e-06, "loss": 0.0177, "step": 91610 }, { "grad_norm": 0.6483680605888367, "learning_rate": 1.9081056924121488e-06, "loss": 0.0255, "step": 91620 }, { "grad_norm": 0.33157429099082947, "learning_rate": 1.9035840987222887e-06, "loss": 0.0143, "step": 91630 }, { "grad_norm": 0.4399746060371399, "learning_rate": 1.8990677647841515e-06, "loss": 0.0106, "step": 91640 }, { "grad_norm": 0.14582358300685883, "learning_rate": 1.8945566910916534e-06, "loss": 0.0302, "step": 91650 }, { "grad_norm": 0.10932883620262146, "learning_rate": 1.8900508781381054e-06, "loss": 0.0147, "step": 91660 }, { "grad_norm": 0.42422768473625183, "learning_rate": 1.8855503264162633e-06, "loss": 0.0199, "step": 91670 }, { "grad_norm": 0.12738516926765442, "learning_rate": 1.8810550364182998e-06, "loss": 0.0311, "step": 91680 }, { "grad_norm": 1.0259729623794556, "learning_rate": 1.8765650086358112e-06, "loss": 0.0174, "step": 91690 }, { "grad_norm": 0.5139380693435669, "learning_rate": 1.8720802435598151e-06, "loss": 0.0247, "step": 91700 }, { "grad_norm": 0.3363646864891052, "learning_rate": 1.8676007416807694e-06, "loss": 0.0177, "step": 91710 }, { "grad_norm": 0.17084625363349915, "learning_rate": 1.8631265034885437e-06, "loss": 0.0094, "step": 91720 }, { "grad_norm": 0.5372887253761292, "learning_rate": 1.858657529472424e-06, "loss": 0.0211, "step": 91730 }, { "grad_norm": 0.5921220779418945, "learning_rate": 1.8541938201211473e-06, "loss": 0.0161, "step": 91740 }, { "grad_norm": 0.13402974605560303, "learning_rate": 1.8497353759228398e-06, "loss": 0.0204, "step": 91750 }, { "grad_norm": 0.3667226731777191, "learning_rate": 1.8452821973650835e-06, "loss": 0.0157, "step": 91760 }, { "grad_norm": 0.16257067024707794, "learning_rate": 1.840834284934856e-06, "loss": 0.03, "step": 91770 }, { "grad_norm": 0.2603406608104706, "learning_rate": 1.8363916391185954e-06, "loss": 0.0211, "step": 91780 }, { "grad_norm": 0.3085324168205261, "learning_rate": 1.8319542604021245e-06, "loss": 0.0124, "step": 91790 }, { "grad_norm": 0.24291378259658813, "learning_rate": 1.8275221492707273e-06, "loss": 0.0134, "step": 91800 }, { "grad_norm": 0.20138902962207794, "learning_rate": 1.8230953062090717e-06, "loss": 0.0217, "step": 91810 }, { "grad_norm": 0.26465868949890137, "learning_rate": 1.8186737317012926e-06, "loss": 0.014, "step": 91820 }, { "grad_norm": 0.4800887405872345, "learning_rate": 1.8142574262308977e-06, "loss": 0.0121, "step": 91830 }, { "grad_norm": 0.6608021259307861, "learning_rate": 1.8098463902808727e-06, "loss": 0.0157, "step": 91840 }, { "grad_norm": 0.14729315042495728, "learning_rate": 1.805440624333593e-06, "loss": 0.0108, "step": 91850 }, { "grad_norm": 0.3188687860965729, "learning_rate": 1.8010401288708734e-06, "loss": 0.0148, "step": 91860 }, { "grad_norm": 0.08389965444803238, "learning_rate": 1.7966449043739343e-06, "loss": 0.0124, "step": 91870 }, { "grad_norm": 0.3324873745441437, "learning_rate": 1.79225495132343e-06, "loss": 0.0087, "step": 91880 }, { "grad_norm": 0.4498043656349182, "learning_rate": 1.7878702701994487e-06, "loss": 0.0193, "step": 91890 }, { "grad_norm": 0.3550713062286377, "learning_rate": 1.783490861481496e-06, "loss": 0.0182, "step": 91900 }, { "grad_norm": 0.18842794001102448, "learning_rate": 1.7791167256484832e-06, "loss": 0.0173, "step": 91910 }, { "grad_norm": 0.1545444130897522, "learning_rate": 1.774747863178766e-06, "loss": 0.0096, "step": 91920 }, { "grad_norm": 0.18694184720516205, "learning_rate": 1.7703842745501186e-06, "loss": 0.0116, "step": 91930 }, { "grad_norm": 0.27861252427101135, "learning_rate": 1.7660259602397366e-06, "loss": 0.0139, "step": 91940 }, { "grad_norm": 0.37992873787879944, "learning_rate": 1.7616729207242388e-06, "loss": 0.011, "step": 91950 }, { "grad_norm": 0.38531872630119324, "learning_rate": 1.7573251564796556e-06, "loss": 0.0125, "step": 91960 }, { "grad_norm": 0.4210188090801239, "learning_rate": 1.7529826679814676e-06, "loss": 0.034, "step": 91970 }, { "grad_norm": 0.5249161124229431, "learning_rate": 1.7486454557045562e-06, "loss": 0.0133, "step": 91980 }, { "grad_norm": 0.29507020115852356, "learning_rate": 1.7443135201232364e-06, "loss": 0.0112, "step": 91990 }, { "grad_norm": 0.2747679352760315, "learning_rate": 1.7399868617112347e-06, "loss": 0.0152, "step": 92000 }, { "grad_norm": 0.37998560070991516, "learning_rate": 1.7356654809417171e-06, "loss": 0.0124, "step": 92010 }, { "grad_norm": 0.21503840386867523, "learning_rate": 1.7313493782872448e-06, "loss": 0.012, "step": 92020 }, { "grad_norm": 0.29625099897384644, "learning_rate": 1.7270385542198453e-06, "loss": 0.0144, "step": 92030 }, { "grad_norm": 0.19471338391304016, "learning_rate": 1.7227330092109307e-06, "loss": 0.0171, "step": 92040 }, { "grad_norm": 0.16551047563552856, "learning_rate": 1.718432743731352e-06, "loss": 0.0118, "step": 92050 }, { "grad_norm": 0.5177841186523438, "learning_rate": 1.7141377582513718e-06, "loss": 0.0276, "step": 92060 }, { "grad_norm": 0.5803162455558777, "learning_rate": 1.709848053240698e-06, "loss": 0.021, "step": 92070 }, { "grad_norm": 0.22952136397361755, "learning_rate": 1.7055636291684328e-06, "loss": 0.009, "step": 92080 }, { "grad_norm": 0.23277176916599274, "learning_rate": 1.7012844865031185e-06, "loss": 0.0134, "step": 92090 }, { "grad_norm": 0.329841673374176, "learning_rate": 1.6970106257127138e-06, "loss": 0.0153, "step": 92100 }, { "grad_norm": 0.14984130859375, "learning_rate": 1.6927420472646117e-06, "loss": 0.013, "step": 92110 }, { "grad_norm": 0.25034961104393005, "learning_rate": 1.6884787516256106e-06, "loss": 0.0117, "step": 92120 }, { "grad_norm": 0.3184327483177185, "learning_rate": 1.6842207392619325e-06, "loss": 0.0138, "step": 92130 }, { "grad_norm": 0.12814082205295563, "learning_rate": 1.679968010639238e-06, "loss": 0.0345, "step": 92140 }, { "grad_norm": 0.13781176507472992, "learning_rate": 1.6757205662225829e-06, "loss": 0.0286, "step": 92150 }, { "grad_norm": 0.11722254753112793, "learning_rate": 1.6714784064764844e-06, "loss": 0.0133, "step": 92160 }, { "grad_norm": 0.3470225930213928, "learning_rate": 1.6672415318648383e-06, "loss": 0.0153, "step": 92170 }, { "grad_norm": 0.20337961614131927, "learning_rate": 1.6630099428510016e-06, "loss": 0.0254, "step": 92180 }, { "grad_norm": 0.17797107994556427, "learning_rate": 1.6587836398977153e-06, "loss": 0.0121, "step": 92190 }, { "grad_norm": 0.14775174856185913, "learning_rate": 1.6545626234671762e-06, "loss": 0.0182, "step": 92200 }, { "grad_norm": 0.6037396788597107, "learning_rate": 1.650346894020971e-06, "loss": 0.0169, "step": 92210 }, { "grad_norm": 0.4916912615299225, "learning_rate": 1.6461364520201472e-06, "loss": 0.0217, "step": 92220 }, { "grad_norm": 0.2594758868217468, "learning_rate": 1.6419312979251366e-06, "loss": 0.0195, "step": 92230 }, { "grad_norm": 0.2536679804325104, "learning_rate": 1.6377314321958214e-06, "loss": 0.0114, "step": 92240 }, { "grad_norm": 0.29565685987472534, "learning_rate": 1.6335368552914787e-06, "loss": 0.0136, "step": 92250 }, { "grad_norm": 0.16751521825790405, "learning_rate": 1.6293475676708358e-06, "loss": 0.0138, "step": 92260 }, { "grad_norm": 0.21917451918125153, "learning_rate": 1.6251635697920097e-06, "loss": 0.0099, "step": 92270 }, { "grad_norm": 0.6488094329833984, "learning_rate": 1.6209848621125624e-06, "loss": 0.0155, "step": 92280 }, { "grad_norm": 0.5000308752059937, "learning_rate": 1.6168114450894778e-06, "loss": 0.0131, "step": 92290 }, { "grad_norm": 0.15371009707450867, "learning_rate": 1.612643319179158e-06, "loss": 0.0133, "step": 92300 }, { "grad_norm": 0.5871496200561523, "learning_rate": 1.6084804848374046e-06, "loss": 0.0114, "step": 92310 }, { "grad_norm": 0.3929280638694763, "learning_rate": 1.6043229425194817e-06, "loss": 0.0202, "step": 92320 }, { "grad_norm": 0.2861257493495941, "learning_rate": 1.6001706926800308e-06, "loss": 0.0135, "step": 92330 }, { "grad_norm": 0.15343546867370605, "learning_rate": 1.5960237357731444e-06, "loss": 0.021, "step": 92340 }, { "grad_norm": 0.14996811747550964, "learning_rate": 1.5918820722523264e-06, "loss": 0.01, "step": 92350 }, { "grad_norm": 0.6091962456703186, "learning_rate": 1.587745702570509e-06, "loss": 0.0207, "step": 92360 }, { "grad_norm": 0.2550487816333771, "learning_rate": 1.5836146271800356e-06, "loss": 0.0114, "step": 92370 }, { "grad_norm": 0.3524458110332489, "learning_rate": 1.5794888465326728e-06, "loss": 0.0166, "step": 92380 }, { "grad_norm": 0.3161852955818176, "learning_rate": 1.5753683610796155e-06, "loss": 0.0159, "step": 92390 }, { "grad_norm": 0.49731937050819397, "learning_rate": 1.5712531712714528e-06, "loss": 0.0189, "step": 92400 }, { "grad_norm": 0.30215558409690857, "learning_rate": 1.5671432775582474e-06, "loss": 0.0174, "step": 92410 }, { "grad_norm": 0.31239089369773865, "learning_rate": 1.563038680389428e-06, "loss": 0.0139, "step": 92420 }, { "grad_norm": 0.14509841799736023, "learning_rate": 1.5589393802138808e-06, "loss": 0.018, "step": 92430 }, { "grad_norm": 0.20844495296478271, "learning_rate": 1.554845377479891e-06, "loss": 0.0104, "step": 92440 }, { "grad_norm": 0.4067531228065491, "learning_rate": 1.550756672635173e-06, "loss": 0.012, "step": 92450 }, { "grad_norm": 0.23737457394599915, "learning_rate": 1.5466732661268635e-06, "loss": 0.0141, "step": 92460 }, { "grad_norm": 0.2950778305530548, "learning_rate": 1.5425951584015163e-06, "loss": 0.0255, "step": 92470 }, { "grad_norm": 0.18210235238075256, "learning_rate": 1.5385223499051026e-06, "loss": 0.0166, "step": 92480 }, { "grad_norm": 0.15481936931610107, "learning_rate": 1.5344548410830328e-06, "loss": 0.0108, "step": 92490 }, { "grad_norm": 0.4282439351081848, "learning_rate": 1.5303926323801121e-06, "loss": 0.0192, "step": 92500 }, { "grad_norm": 0.15959346294403076, "learning_rate": 1.5263357242405795e-06, "loss": 0.0197, "step": 92510 }, { "grad_norm": 0.31779658794403076, "learning_rate": 1.5222841171080915e-06, "loss": 0.0185, "step": 92520 }, { "grad_norm": 0.39132270216941833, "learning_rate": 1.5182378114257268e-06, "loss": 0.0134, "step": 92530 }, { "grad_norm": 0.10312378406524658, "learning_rate": 1.5141968076359869e-06, "loss": 0.0144, "step": 92540 }, { "grad_norm": 0.16867150366306305, "learning_rate": 1.5101611061807908e-06, "loss": 0.0212, "step": 92550 }, { "grad_norm": 0.16123194992542267, "learning_rate": 1.506130707501463e-06, "loss": 0.0125, "step": 92560 }, { "grad_norm": 0.3239705264568329, "learning_rate": 1.5021056120387734e-06, "loss": 0.0092, "step": 92570 }, { "grad_norm": 0.39346829056739807, "learning_rate": 1.4980858202329084e-06, "loss": 0.0164, "step": 92580 }, { "grad_norm": 0.3653404116630554, "learning_rate": 1.4940713325234445e-06, "loss": 0.0154, "step": 92590 }, { "grad_norm": 0.13317367434501648, "learning_rate": 1.4900621493494137e-06, "loss": 0.0158, "step": 92600 }, { "grad_norm": 0.5514213442802429, "learning_rate": 1.4860582711492544e-06, "loss": 0.0159, "step": 92610 }, { "grad_norm": 0.20627762377262115, "learning_rate": 1.4820596983608325e-06, "loss": 0.0177, "step": 92620 }, { "grad_norm": 0.31674620509147644, "learning_rate": 1.4780664314214044e-06, "loss": 0.0139, "step": 92630 }, { "grad_norm": 0.2434127777814865, "learning_rate": 1.474078470767687e-06, "loss": 0.0105, "step": 92640 }, { "grad_norm": 0.24229907989501953, "learning_rate": 1.470095816835787e-06, "loss": 0.0125, "step": 92650 }, { "grad_norm": 0.3687850832939148, "learning_rate": 1.4661184700612395e-06, "loss": 0.0134, "step": 92660 }, { "grad_norm": 0.4144454300403595, "learning_rate": 1.4621464308790133e-06, "loss": 0.0204, "step": 92670 }, { "grad_norm": 0.8471382856369019, "learning_rate": 1.4581796997234775e-06, "loss": 0.0251, "step": 92680 }, { "grad_norm": 0.3271695077419281, "learning_rate": 1.4542182770284297e-06, "loss": 0.0113, "step": 92690 }, { "grad_norm": 0.21269160509109497, "learning_rate": 1.450262163227084e-06, "loss": 0.0133, "step": 92700 }, { "grad_norm": 0.21436478197574615, "learning_rate": 1.4463113587520671e-06, "loss": 0.0145, "step": 92710 }, { "grad_norm": 0.17046122252941132, "learning_rate": 1.4423658640354443e-06, "loss": 0.0174, "step": 92720 }, { "grad_norm": 0.4706082046031952, "learning_rate": 1.438425679508687e-06, "loss": 0.012, "step": 92730 }, { "grad_norm": 0.17816129326820374, "learning_rate": 1.4344908056026895e-06, "loss": 0.0122, "step": 92740 }, { "grad_norm": 0.5623806118965149, "learning_rate": 1.4305612427477521e-06, "loss": 0.0126, "step": 92750 }, { "grad_norm": 0.22944475710391998, "learning_rate": 1.42663699137362e-06, "loss": 0.0199, "step": 92760 }, { "grad_norm": 0.2950439155101776, "learning_rate": 1.4227180519094386e-06, "loss": 0.0221, "step": 92770 }, { "grad_norm": 0.32212263345718384, "learning_rate": 1.4188044247837706e-06, "loss": 0.0208, "step": 92780 }, { "grad_norm": 0.3419407308101654, "learning_rate": 1.4148961104246128e-06, "loss": 0.0157, "step": 92790 }, { "grad_norm": 0.25254055857658386, "learning_rate": 1.4109931092593731e-06, "loss": 0.0097, "step": 92800 }, { "grad_norm": 0.1620892584323883, "learning_rate": 1.4070954217148658e-06, "loss": 0.017, "step": 92810 }, { "grad_norm": 0.34431692957878113, "learning_rate": 1.4032030482173497e-06, "loss": 0.0137, "step": 92820 }, { "grad_norm": 0.24806706607341766, "learning_rate": 1.3993159891924901e-06, "loss": 0.0153, "step": 92830 }, { "grad_norm": 0.6058416366577148, "learning_rate": 1.3954342450653524e-06, "loss": 0.0156, "step": 92840 }, { "grad_norm": 0.29471391439437866, "learning_rate": 1.3915578162604582e-06, "loss": 0.014, "step": 92850 }, { "grad_norm": 0.3184433579444885, "learning_rate": 1.3876867032017126e-06, "loss": 0.0105, "step": 92860 }, { "grad_norm": 0.12306101620197296, "learning_rate": 1.383820906312472e-06, "loss": 0.0192, "step": 92870 }, { "grad_norm": 0.3154807388782501, "learning_rate": 1.3799604260154752e-06, "loss": 0.0107, "step": 92880 }, { "grad_norm": 0.46295589208602905, "learning_rate": 1.376105262732913e-06, "loss": 0.0167, "step": 92890 }, { "grad_norm": 0.4015447497367859, "learning_rate": 1.37225541688637e-06, "loss": 0.0169, "step": 92900 }, { "grad_norm": 0.39230239391326904, "learning_rate": 1.3684108888968594e-06, "loss": 0.0127, "step": 92910 }, { "grad_norm": 0.3580128848552704, "learning_rate": 1.364571679184823e-06, "loss": 0.0143, "step": 92920 }, { "grad_norm": 0.2837655246257782, "learning_rate": 1.3607377881701077e-06, "loss": 0.012, "step": 92930 }, { "grad_norm": 0.23235243558883667, "learning_rate": 1.3569092162719787e-06, "loss": 0.0139, "step": 92940 }, { "grad_norm": 0.34078237414360046, "learning_rate": 1.353085963909123e-06, "loss": 0.0217, "step": 92950 }, { "grad_norm": 0.23276634514331818, "learning_rate": 1.3492680314996454e-06, "loss": 0.0162, "step": 92960 }, { "grad_norm": 0.42898228764533997, "learning_rate": 1.345455419461078e-06, "loss": 0.0153, "step": 92970 }, { "grad_norm": 0.45260217785835266, "learning_rate": 1.3416481282103431e-06, "loss": 0.0113, "step": 92980 }, { "grad_norm": 0.28375881910324097, "learning_rate": 1.337846158163819e-06, "loss": 0.0177, "step": 92990 }, { "grad_norm": 0.23381638526916504, "learning_rate": 1.334049509737273e-06, "loss": 0.0128, "step": 93000 }, { "grad_norm": 0.21696698665618896, "learning_rate": 1.3302581833459116e-06, "loss": 0.0209, "step": 93010 }, { "grad_norm": 0.29489022493362427, "learning_rate": 1.3264721794043366e-06, "loss": 0.0143, "step": 93020 }, { "grad_norm": 0.25419172644615173, "learning_rate": 1.3226914983265836e-06, "loss": 0.0116, "step": 93030 }, { "grad_norm": 0.46341535449028015, "learning_rate": 1.3189161405260997e-06, "loss": 0.0122, "step": 93040 }, { "grad_norm": 0.48862531781196594, "learning_rate": 1.3151461064157656e-06, "loss": 0.0332, "step": 93050 }, { "grad_norm": 0.45376497507095337, "learning_rate": 1.3113813964078459e-06, "loss": 0.0265, "step": 93060 }, { "grad_norm": 0.2767418920993805, "learning_rate": 1.3076220109140614e-06, "loss": 0.0123, "step": 93070 }, { "grad_norm": 0.20109112560749054, "learning_rate": 1.3038679503455275e-06, "loss": 0.015, "step": 93080 }, { "grad_norm": 0.13239839673042297, "learning_rate": 1.3001192151127773e-06, "loss": 0.0098, "step": 93090 }, { "grad_norm": 0.4727623462677002, "learning_rate": 1.2963758056257657e-06, "loss": 0.0145, "step": 93100 }, { "grad_norm": 0.1615690439939499, "learning_rate": 1.2926377222938768e-06, "loss": 0.008, "step": 93110 }, { "grad_norm": 0.3245272934436798, "learning_rate": 1.2889049655259e-06, "loss": 0.0117, "step": 93120 }, { "grad_norm": 0.2924211919307709, "learning_rate": 1.2851775357300312e-06, "loss": 0.0138, "step": 93130 }, { "grad_norm": 0.25412309169769287, "learning_rate": 1.281455433313916e-06, "loss": 0.0156, "step": 93140 }, { "grad_norm": 0.3565068244934082, "learning_rate": 1.2777386586845796e-06, "loss": 0.0168, "step": 93150 }, { "grad_norm": 0.2805905044078827, "learning_rate": 1.2740272122484964e-06, "loss": 0.0165, "step": 93160 }, { "grad_norm": 0.3158860206604004, "learning_rate": 1.2703210944115308e-06, "loss": 0.011, "step": 93170 }, { "grad_norm": 0.22975113987922668, "learning_rate": 1.2666203055789916e-06, "loss": 0.0154, "step": 93180 }, { "grad_norm": 0.2154586762189865, "learning_rate": 1.2629248461555832e-06, "loss": 0.0153, "step": 93190 }, { "grad_norm": 0.27718374133110046, "learning_rate": 1.2592347165454431e-06, "loss": 0.0129, "step": 93200 }, { "grad_norm": 0.2608130872249603, "learning_rate": 1.2555499171521102e-06, "loss": 0.0157, "step": 93210 }, { "grad_norm": 0.18895554542541504, "learning_rate": 1.251870448378556e-06, "loss": 0.0134, "step": 93220 }, { "grad_norm": 0.4787291884422302, "learning_rate": 1.2481963106271478e-06, "loss": 0.0168, "step": 93230 }, { "grad_norm": 0.2710442543029785, "learning_rate": 1.2445275042997084e-06, "loss": 0.0136, "step": 93240 }, { "grad_norm": 0.35446518659591675, "learning_rate": 1.2408640297974283e-06, "loss": 0.0154, "step": 93250 }, { "grad_norm": 0.35803765058517456, "learning_rate": 1.2372058875209535e-06, "loss": 0.0162, "step": 93260 }, { "grad_norm": 0.3086032271385193, "learning_rate": 1.2335530778703252e-06, "loss": 0.012, "step": 93270 }, { "grad_norm": 0.23547714948654175, "learning_rate": 1.2299056012450182e-06, "loss": 0.0119, "step": 93280 }, { "grad_norm": 0.25823265314102173, "learning_rate": 1.2262634580439026e-06, "loss": 0.0154, "step": 93290 }, { "grad_norm": 0.3613475561141968, "learning_rate": 1.2226266486652926e-06, "loss": 0.0163, "step": 93300 }, { "grad_norm": 0.3493823707103729, "learning_rate": 1.2189951735068927e-06, "loss": 0.015, "step": 93310 }, { "grad_norm": 0.5616248846054077, "learning_rate": 1.2153690329658407e-06, "loss": 0.0239, "step": 93320 }, { "grad_norm": 0.2741842269897461, "learning_rate": 1.2117482274386859e-06, "loss": 0.008, "step": 93330 }, { "grad_norm": 0.16171979904174805, "learning_rate": 1.2081327573213953e-06, "loss": 0.0207, "step": 93340 }, { "grad_norm": 0.1589832901954651, "learning_rate": 1.2045226230093465e-06, "loss": 0.0133, "step": 93350 }, { "grad_norm": 0.13587746024131775, "learning_rate": 1.2009178248973296e-06, "loss": 0.0222, "step": 93360 }, { "grad_norm": 0.30594736337661743, "learning_rate": 1.1973183633795849e-06, "loss": 0.0134, "step": 93370 }, { "grad_norm": 0.4647330641746521, "learning_rate": 1.193724238849725e-06, "loss": 0.0167, "step": 93380 }, { "grad_norm": 0.21581345796585083, "learning_rate": 1.190135451700808e-06, "loss": 0.0124, "step": 93390 }, { "grad_norm": 0.21600216627120972, "learning_rate": 1.186552002325292e-06, "loss": 0.0136, "step": 93400 }, { "grad_norm": 0.4644722640514374, "learning_rate": 1.1829738911150578e-06, "loss": 0.0163, "step": 93410 }, { "grad_norm": 0.32851529121398926, "learning_rate": 1.1794011184613985e-06, "loss": 0.0181, "step": 93420 }, { "grad_norm": 0.2971426546573639, "learning_rate": 1.17583368475504e-06, "loss": 0.0168, "step": 93430 }, { "grad_norm": 0.6545498967170715, "learning_rate": 1.1722715903860982e-06, "loss": 0.0108, "step": 93440 }, { "grad_norm": 0.25007364153862, "learning_rate": 1.168714835744128e-06, "loss": 0.0144, "step": 93450 }, { "grad_norm": 0.4201244115829468, "learning_rate": 1.165163421218085e-06, "loss": 0.0243, "step": 93460 }, { "grad_norm": 0.3329412639141083, "learning_rate": 1.1616173471963477e-06, "loss": 0.0121, "step": 93470 }, { "grad_norm": 0.42518889904022217, "learning_rate": 1.1580766140667109e-06, "loss": 0.0124, "step": 93480 }, { "grad_norm": 1.7078346014022827, "learning_rate": 1.1545412222163764e-06, "loss": 0.02, "step": 93490 }, { "grad_norm": 0.17067644000053406, "learning_rate": 1.1510111720319795e-06, "loss": 0.0167, "step": 93500 }, { "grad_norm": 0.5790858864784241, "learning_rate": 1.1474864638995663e-06, "loss": 0.0123, "step": 93510 }, { "grad_norm": 0.26075464487075806, "learning_rate": 1.1439670982045735e-06, "loss": 0.0258, "step": 93520 }, { "grad_norm": 0.3106611967086792, "learning_rate": 1.1404530753318877e-06, "loss": 0.019, "step": 93530 }, { "grad_norm": 0.27263858914375305, "learning_rate": 1.136944395665801e-06, "loss": 0.0144, "step": 93540 }, { "grad_norm": 0.2914651930332184, "learning_rate": 1.1334410595900014e-06, "loss": 0.0174, "step": 93550 }, { "grad_norm": 0.4112565219402313, "learning_rate": 1.129943067487621e-06, "loss": 0.0163, "step": 93560 }, { "grad_norm": 0.3225384056568146, "learning_rate": 1.126450419741193e-06, "loss": 0.0091, "step": 93570 }, { "grad_norm": 0.23780108988285065, "learning_rate": 1.1229631167326671e-06, "loss": 0.0159, "step": 93580 }, { "grad_norm": 0.46607309579849243, "learning_rate": 1.1194811588434052e-06, "loss": 0.0124, "step": 93590 }, { "grad_norm": 0.365878701210022, "learning_rate": 1.1160045464542023e-06, "loss": 0.0205, "step": 93600 }, { "grad_norm": 0.27197542786598206, "learning_rate": 1.1125332799452325e-06, "loss": 0.0347, "step": 93610 }, { "grad_norm": 0.35025399923324585, "learning_rate": 1.109067359696131e-06, "loss": 0.0214, "step": 93620 }, { "grad_norm": 0.424023300409317, "learning_rate": 1.1056067860859166e-06, "loss": 0.0247, "step": 93630 }, { "grad_norm": 0.16250695288181305, "learning_rate": 1.102151559493031e-06, "loss": 0.0155, "step": 93640 }, { "grad_norm": 0.6198573708534241, "learning_rate": 1.0987016802953332e-06, "loss": 0.0177, "step": 93650 }, { "grad_norm": 0.24313509464263916, "learning_rate": 1.0952571488701048e-06, "loss": 0.0168, "step": 93660 }, { "grad_norm": 0.24908937513828278, "learning_rate": 1.0918179655940163e-06, "loss": 0.0127, "step": 93670 }, { "grad_norm": 0.28471338748931885, "learning_rate": 1.088384130843184e-06, "loss": 0.0116, "step": 93680 }, { "grad_norm": 0.2748686373233795, "learning_rate": 1.0849556449931296e-06, "loss": 0.0123, "step": 93690 }, { "grad_norm": 0.2715843915939331, "learning_rate": 1.0815325084187866e-06, "loss": 0.012, "step": 93700 }, { "grad_norm": 0.31263449788093567, "learning_rate": 1.0781147214944943e-06, "loss": 0.0111, "step": 93710 }, { "grad_norm": 0.39897969365119934, "learning_rate": 1.0747022845940313e-06, "loss": 0.0211, "step": 93720 }, { "grad_norm": 0.29551512002944946, "learning_rate": 1.0712951980905605e-06, "loss": 0.0143, "step": 93730 }, { "grad_norm": 0.2679308354854584, "learning_rate": 1.0678934623566839e-06, "loss": 0.019, "step": 93740 }, { "grad_norm": 0.10857070237398148, "learning_rate": 1.0644970777644091e-06, "loss": 0.0142, "step": 93750 }, { "grad_norm": 0.1004994735121727, "learning_rate": 1.0611060446851673e-06, "loss": 0.0106, "step": 93760 }, { "grad_norm": 0.18909934163093567, "learning_rate": 1.0577203634897836e-06, "loss": 0.0157, "step": 93770 }, { "grad_norm": 0.5177176594734192, "learning_rate": 1.0543400345485178e-06, "loss": 0.0115, "step": 93780 }, { "grad_norm": 0.383247047662735, "learning_rate": 1.0509650582310404e-06, "loss": 0.0113, "step": 93790 }, { "grad_norm": 0.26577410101890564, "learning_rate": 1.0475954349064287e-06, "loss": 0.0152, "step": 93800 }, { "grad_norm": 0.40568414330482483, "learning_rate": 1.0442311649431825e-06, "loss": 0.0114, "step": 93810 }, { "grad_norm": 0.4385981857776642, "learning_rate": 1.0408722487092071e-06, "loss": 0.0099, "step": 93820 }, { "grad_norm": 0.26353850960731506, "learning_rate": 1.0375186865718424e-06, "loss": 0.0195, "step": 93830 }, { "grad_norm": 0.7988880276679993, "learning_rate": 1.0341704788978113e-06, "loss": 0.0167, "step": 93840 }, { "grad_norm": 0.3702496290206909, "learning_rate": 1.0308276260532879e-06, "loss": 0.0181, "step": 93850 }, { "grad_norm": 0.5293326377868652, "learning_rate": 1.0274901284038241e-06, "loss": 0.0116, "step": 93860 }, { "grad_norm": 0.24609465897083282, "learning_rate": 1.0241579863144114e-06, "loss": 0.0181, "step": 93870 }, { "grad_norm": 0.28156816959381104, "learning_rate": 1.0208312001494413e-06, "loss": 0.0228, "step": 93880 }, { "grad_norm": 0.1915026754140854, "learning_rate": 1.0175097702727453e-06, "loss": 0.0163, "step": 93890 }, { "grad_norm": 0.26249465346336365, "learning_rate": 1.0141936970475275e-06, "loss": 0.012, "step": 93900 }, { "grad_norm": 0.3465437591075897, "learning_rate": 1.0108829808364417e-06, "loss": 0.0123, "step": 93910 }, { "grad_norm": 0.39025214314460754, "learning_rate": 1.0075776220015376e-06, "loss": 0.0153, "step": 93920 }, { "grad_norm": 0.2798021733760834, "learning_rate": 1.0042776209042816e-06, "loss": 0.0176, "step": 93930 }, { "grad_norm": 0.18276001513004303, "learning_rate": 1.000982977905568e-06, "loss": 0.0115, "step": 93940 }, { "grad_norm": 0.17600062489509583, "learning_rate": 9.976936933656867e-07, "loss": 0.0202, "step": 93950 }, { "grad_norm": 0.12910407781600952, "learning_rate": 9.944097676443442e-07, "loss": 0.011, "step": 93960 }, { "grad_norm": 0.2656311094760895, "learning_rate": 9.911312011006757e-07, "loss": 0.0097, "step": 93970 }, { "grad_norm": 0.24419833719730377, "learning_rate": 9.878579940932109e-07, "loss": 0.0221, "step": 93980 }, { "grad_norm": 0.1554439663887024, "learning_rate": 9.84590146979908e-07, "loss": 0.0216, "step": 93990 }, { "grad_norm": 0.29358136653900146, "learning_rate": 9.813276601181365e-07, "loss": 0.0133, "step": 94000 }, { "grad_norm": 0.44416213035583496, "learning_rate": 9.780705338646724e-07, "loss": 0.0184, "step": 94010 }, { "grad_norm": 0.5492837429046631, "learning_rate": 9.748187685757083e-07, "loss": 0.0166, "step": 94020 }, { "grad_norm": 0.22833488881587982, "learning_rate": 9.715723646068543e-07, "loss": 0.0144, "step": 94030 }, { "grad_norm": 0.15864162147045135, "learning_rate": 9.68331322313143e-07, "loss": 0.0203, "step": 94040 }, { "grad_norm": 0.34966909885406494, "learning_rate": 9.650956420489909e-07, "loss": 0.0186, "step": 94050 }, { "grad_norm": 0.29840153455734253, "learning_rate": 9.618653241682595e-07, "loss": 0.0129, "step": 94060 }, { "grad_norm": 0.1724994033575058, "learning_rate": 9.586403690242053e-07, "loss": 0.0139, "step": 94070 }, { "grad_norm": 0.10946053266525269, "learning_rate": 9.554207769695122e-07, "loss": 0.0224, "step": 94080 }, { "grad_norm": 0.12353867292404175, "learning_rate": 9.522065483562659e-07, "loss": 0.016, "step": 94090 }, { "grad_norm": 0.18517863750457764, "learning_rate": 9.489976835359681e-07, "loss": 0.0187, "step": 94100 }, { "grad_norm": 0.25491970777511597, "learning_rate": 9.457941828595329e-07, "loss": 0.0119, "step": 94110 }, { "grad_norm": 0.28102001547813416, "learning_rate": 9.425960466772965e-07, "loss": 0.0093, "step": 94120 }, { "grad_norm": 0.9553009867668152, "learning_rate": 9.394032753390014e-07, "loss": 0.0172, "step": 94130 }, { "grad_norm": 0.3183470368385315, "learning_rate": 9.362158691938017e-07, "loss": 0.0151, "step": 94140 }, { "grad_norm": 0.3483675420284271, "learning_rate": 9.330338285902685e-07, "loss": 0.0156, "step": 94150 }, { "grad_norm": 0.3589961528778076, "learning_rate": 9.298571538763845e-07, "loss": 0.0139, "step": 94160 }, { "grad_norm": 0.3016897737979889, "learning_rate": 9.266858453995441e-07, "loss": 0.0119, "step": 94170 }, { "grad_norm": 0.31177958846092224, "learning_rate": 9.235199035065645e-07, "loss": 0.0136, "step": 94180 }, { "grad_norm": 0.3612724542617798, "learning_rate": 9.203593285436574e-07, "loss": 0.0141, "step": 94190 }, { "grad_norm": 0.3781433403491974, "learning_rate": 9.172041208564686e-07, "loss": 0.0153, "step": 94200 }, { "grad_norm": 0.9701445698738098, "learning_rate": 9.140542807900332e-07, "loss": 0.014, "step": 94210 }, { "grad_norm": 0.37917226552963257, "learning_rate": 9.109098086888313e-07, "loss": 0.0112, "step": 94220 }, { "grad_norm": 0.22502587735652924, "learning_rate": 9.077707048967265e-07, "loss": 0.0081, "step": 94230 }, { "grad_norm": 0.37361031770706177, "learning_rate": 9.046369697570057e-07, "loss": 0.0166, "step": 94240 }, { "grad_norm": 0.17366519570350647, "learning_rate": 9.015086036123721e-07, "loss": 0.0164, "step": 94250 }, { "grad_norm": 0.4042292833328247, "learning_rate": 8.983856068049413e-07, "loss": 0.0161, "step": 94260 }, { "grad_norm": 0.35289594531059265, "learning_rate": 8.952679796762398e-07, "loss": 0.0135, "step": 94270 }, { "grad_norm": 0.2005125731229782, "learning_rate": 8.921557225671951e-07, "loss": 0.0175, "step": 94280 }, { "grad_norm": 0.3186320960521698, "learning_rate": 8.890488358181792e-07, "loss": 0.0121, "step": 94290 }, { "grad_norm": 0.43721553683280945, "learning_rate": 8.859473197689372e-07, "loss": 0.0218, "step": 94300 }, { "grad_norm": 0.31936532258987427, "learning_rate": 8.828511747586588e-07, "loss": 0.0166, "step": 94310 }, { "grad_norm": 0.42166611552238464, "learning_rate": 8.797604011259287e-07, "loss": 0.0127, "step": 94320 }, { "grad_norm": 0.20303183794021606, "learning_rate": 8.766749992087487e-07, "loss": 0.0185, "step": 94330 }, { "grad_norm": 0.36299756169319153, "learning_rate": 8.735949693445378e-07, "loss": 0.0201, "step": 94340 }, { "grad_norm": 0.4066343307495117, "learning_rate": 8.705203118701211e-07, "loss": 0.0172, "step": 94350 }, { "grad_norm": 0.2643726170063019, "learning_rate": 8.674510271217295e-07, "loss": 0.0154, "step": 94360 }, { "grad_norm": 0.21478745341300964, "learning_rate": 8.643871154350336e-07, "loss": 0.0142, "step": 94370 }, { "grad_norm": 0.22743099927902222, "learning_rate": 8.613285771450818e-07, "loss": 0.0144, "step": 94380 }, { "grad_norm": 0.2925209701061249, "learning_rate": 8.582754125863623e-07, "loss": 0.0147, "step": 94390 }, { "grad_norm": 0.26704809069633484, "learning_rate": 8.55227622092758e-07, "loss": 0.0113, "step": 94400 }, { "grad_norm": 0.20992642641067505, "learning_rate": 8.521852059975743e-07, "loss": 0.0109, "step": 94410 }, { "grad_norm": 0.4141513705253601, "learning_rate": 8.491481646335231e-07, "loss": 0.0124, "step": 94420 }, { "grad_norm": 0.22051763534545898, "learning_rate": 8.46116498332733e-07, "loss": 0.0199, "step": 94430 }, { "grad_norm": 0.3526204228401184, "learning_rate": 8.430902074267333e-07, "loss": 0.0158, "step": 94440 }, { "grad_norm": 0.37299758195877075, "learning_rate": 8.400692922464925e-07, "loss": 0.0156, "step": 94450 }, { "grad_norm": 0.9408591985702515, "learning_rate": 8.370537531223577e-07, "loss": 0.0188, "step": 94460 }, { "grad_norm": 0.28212881088256836, "learning_rate": 8.340435903841092e-07, "loss": 0.0144, "step": 94470 }, { "grad_norm": 0.33259403705596924, "learning_rate": 8.310388043609341e-07, "loss": 0.0157, "step": 94480 }, { "grad_norm": 0.2010311335325241, "learning_rate": 8.280393953814303e-07, "loss": 0.0158, "step": 94490 }, { "grad_norm": 0.2791956663131714, "learning_rate": 8.250453637736077e-07, "loss": 0.0156, "step": 94500 }, { "grad_norm": 0.382779985666275, "learning_rate": 8.22056709864899e-07, "loss": 0.0238, "step": 94510 }, { "grad_norm": 0.6610612869262695, "learning_rate": 8.190734339821204e-07, "loss": 0.027, "step": 94520 }, { "grad_norm": 0.3203345537185669, "learning_rate": 8.160955364515333e-07, "loss": 0.011, "step": 94530 }, { "grad_norm": 0.15833912789821625, "learning_rate": 8.131230175987936e-07, "loss": 0.018, "step": 94540 }, { "grad_norm": 0.3463096618652344, "learning_rate": 8.101558777489693e-07, "loss": 0.0158, "step": 94550 }, { "grad_norm": 0.19667284190654755, "learning_rate": 8.071941172265396e-07, "loss": 0.0139, "step": 94560 }, { "grad_norm": 0.46117445826530457, "learning_rate": 8.042377363553955e-07, "loss": 0.0127, "step": 94570 }, { "grad_norm": 0.17695771157741547, "learning_rate": 8.01286735458856e-07, "loss": 0.0175, "step": 94580 }, { "grad_norm": 0.2334192842245102, "learning_rate": 7.983411148596298e-07, "loss": 0.0143, "step": 94590 }, { "grad_norm": 0.3222559690475464, "learning_rate": 7.954008748798425e-07, "loss": 0.0159, "step": 94600 }, { "grad_norm": 0.35030606389045715, "learning_rate": 7.924660158410369e-07, "loss": 0.0166, "step": 94610 }, { "grad_norm": 0.4451291859149933, "learning_rate": 7.895365380641728e-07, "loss": 0.017, "step": 94620 }, { "grad_norm": 0.20315203070640564, "learning_rate": 7.866124418695941e-07, "loss": 0.0117, "step": 94630 }, { "grad_norm": 0.11847665160894394, "learning_rate": 7.83693727577095e-07, "loss": 0.0137, "step": 94640 }, { "grad_norm": 0.2595462501049042, "learning_rate": 7.807803955058535e-07, "loss": 0.0117, "step": 94650 }, { "grad_norm": 0.25380101799964905, "learning_rate": 7.778724459744702e-07, "loss": 0.013, "step": 94660 }, { "grad_norm": 0.31538617610931396, "learning_rate": 7.749698793009464e-07, "loss": 0.0124, "step": 94670 }, { "grad_norm": 0.2862304151058197, "learning_rate": 7.720726958027169e-07, "loss": 0.0134, "step": 94680 }, { "grad_norm": 0.14054465293884277, "learning_rate": 7.691808957965951e-07, "loss": 0.0094, "step": 94690 }, { "grad_norm": 0.27655723690986633, "learning_rate": 7.662944795988336e-07, "loss": 0.0126, "step": 94700 }, { "grad_norm": 0.15950681269168854, "learning_rate": 7.634134475250854e-07, "loss": 0.0191, "step": 94710 }, { "grad_norm": 0.38981759548187256, "learning_rate": 7.605377998904262e-07, "loss": 0.0122, "step": 94720 }, { "grad_norm": 0.4314841628074646, "learning_rate": 7.5766753700931e-07, "loss": 0.0108, "step": 94730 }, { "grad_norm": 0.34158095717430115, "learning_rate": 7.54802659195647e-07, "loss": 0.0308, "step": 94740 }, { "grad_norm": 0.17631062865257263, "learning_rate": 7.519431667627253e-07, "loss": 0.0138, "step": 94750 }, { "grad_norm": 0.49614882469177246, "learning_rate": 7.490890600232503e-07, "loss": 0.011, "step": 94760 }, { "grad_norm": 0.15649372339248657, "learning_rate": 7.462403392893502e-07, "loss": 0.0166, "step": 94770 }, { "grad_norm": 0.5609733462333679, "learning_rate": 7.433970048725536e-07, "loss": 0.0145, "step": 94780 }, { "grad_norm": 0.4958546757698059, "learning_rate": 7.405590570838062e-07, "loss": 0.0232, "step": 94790 }, { "grad_norm": 0.3573255240917206, "learning_rate": 7.377264962334596e-07, "loss": 0.0159, "step": 94800 }, { "grad_norm": 0.36845895648002625, "learning_rate": 7.348993226312773e-07, "loss": 0.0153, "step": 94810 }, { "grad_norm": 0.26294150948524475, "learning_rate": 7.32077536586434e-07, "loss": 0.0152, "step": 94820 }, { "grad_norm": 0.5241442322731018, "learning_rate": 7.292611384075221e-07, "loss": 0.0162, "step": 94830 }, { "grad_norm": 0.24273821711540222, "learning_rate": 7.264501284025338e-07, "loss": 0.0217, "step": 94840 }, { "grad_norm": 0.36200302839279175, "learning_rate": 7.236445068788789e-07, "loss": 0.0161, "step": 94850 }, { "grad_norm": 0.1771537959575653, "learning_rate": 7.20844274143373e-07, "loss": 0.0197, "step": 94860 }, { "grad_norm": 0.34876784682273865, "learning_rate": 7.180494305022489e-07, "loss": 0.0407, "step": 94870 }, { "grad_norm": 0.290885329246521, "learning_rate": 7.152599762611456e-07, "loss": 0.0099, "step": 94880 }, { "grad_norm": 0.3033311367034912, "learning_rate": 7.124759117251079e-07, "loss": 0.0178, "step": 94890 }, { "grad_norm": 0.15250812470912933, "learning_rate": 7.096972371986033e-07, "loss": 0.016, "step": 94900 }, { "grad_norm": 0.6428281664848328, "learning_rate": 7.069239529855054e-07, "loss": 0.0298, "step": 94910 }, { "grad_norm": 0.8482180833816528, "learning_rate": 7.041560593890939e-07, "loss": 0.0147, "step": 94920 }, { "grad_norm": 0.3994525671005249, "learning_rate": 7.013935567120599e-07, "loss": 0.0167, "step": 94930 }, { "grad_norm": 0.12538254261016846, "learning_rate": 6.986364452565009e-07, "loss": 0.0175, "step": 94940 }, { "grad_norm": 0.48381999135017395, "learning_rate": 6.95884725323942e-07, "loss": 0.0142, "step": 94950 }, { "grad_norm": 0.7206457853317261, "learning_rate": 6.931383972152982e-07, "loss": 0.0143, "step": 94960 }, { "grad_norm": 0.368910551071167, "learning_rate": 6.903974612309127e-07, "loss": 0.0118, "step": 94970 }, { "grad_norm": 0.24947616457939148, "learning_rate": 6.876619176705179e-07, "loss": 0.0116, "step": 94980 }, { "grad_norm": 0.5535697937011719, "learning_rate": 6.8493176683328e-07, "loss": 0.0127, "step": 94990 }, { "grad_norm": 0.20256595313549042, "learning_rate": 6.822070090177657e-07, "loss": 0.0212, "step": 95000 }, { "grad_norm": 0.22041508555412292, "learning_rate": 6.794876445219367e-07, "loss": 0.0106, "step": 95010 }, { "grad_norm": 0.22343836724758148, "learning_rate": 6.767736736431885e-07, "loss": 0.0114, "step": 95020 }, { "grad_norm": 0.25685495138168335, "learning_rate": 6.740650966783113e-07, "loss": 0.0132, "step": 95030 }, { "grad_norm": 0.2544379234313965, "learning_rate": 6.713619139235239e-07, "loss": 0.0165, "step": 95040 }, { "grad_norm": 0.4230839014053345, "learning_rate": 6.686641256744286e-07, "loss": 0.0176, "step": 95050 }, { "grad_norm": 0.631339967250824, "learning_rate": 6.659717322260562e-07, "loss": 0.0174, "step": 95060 }, { "grad_norm": 0.20831933617591858, "learning_rate": 6.632847338728377e-07, "loss": 0.0118, "step": 95070 }, { "grad_norm": 0.19122447073459625, "learning_rate": 6.60603130908627e-07, "loss": 0.013, "step": 95080 }, { "grad_norm": 0.4756566286087036, "learning_rate": 6.579269236266783e-07, "loss": 0.0181, "step": 95090 }, { "grad_norm": 0.1436365842819214, "learning_rate": 6.552561123196632e-07, "loss": 0.0116, "step": 95100 }, { "grad_norm": 0.32082369923591614, "learning_rate": 6.525906972796425e-07, "loss": 0.0255, "step": 95110 }, { "grad_norm": 0.4351022243499756, "learning_rate": 6.499306787981108e-07, "loss": 0.0198, "step": 95120 }, { "grad_norm": 0.48088568449020386, "learning_rate": 6.472760571659631e-07, "loss": 0.017, "step": 95130 }, { "grad_norm": 0.2018117755651474, "learning_rate": 6.446268326735061e-07, "loss": 0.0145, "step": 95140 }, { "grad_norm": 0.36218786239624023, "learning_rate": 6.41983005610458e-07, "loss": 0.0207, "step": 95150 }, { "grad_norm": 0.5313883423805237, "learning_rate": 6.393445762659378e-07, "loss": 0.0147, "step": 95160 }, { "grad_norm": 0.12148288637399673, "learning_rate": 6.36711544928481e-07, "loss": 0.0114, "step": 95170 }, { "grad_norm": 0.5224016308784485, "learning_rate": 6.340839118860298e-07, "loss": 0.0135, "step": 95180 }, { "grad_norm": 0.2662869989871979, "learning_rate": 6.314616774259485e-07, "loss": 0.0123, "step": 95190 }, { "grad_norm": 0.503836989402771, "learning_rate": 6.288448418349858e-07, "loss": 0.0362, "step": 95200 }, { "grad_norm": 0.3863289952278137, "learning_rate": 6.262334053993235e-07, "loss": 0.0184, "step": 95210 }, { "grad_norm": 0.2165961116552353, "learning_rate": 6.236273684045501e-07, "loss": 0.0141, "step": 95220 }, { "grad_norm": 0.4391467571258545, "learning_rate": 6.210267311356488e-07, "loss": 0.014, "step": 95230 }, { "grad_norm": 0.3818362355232239, "learning_rate": 6.184314938770197e-07, "loss": 0.0147, "step": 95240 }, { "grad_norm": 0.39951932430267334, "learning_rate": 6.158416569124858e-07, "loss": 0.0319, "step": 95250 }, { "grad_norm": 0.1532924324274063, "learning_rate": 6.132572205252541e-07, "loss": 0.0143, "step": 95260 }, { "grad_norm": 0.8724187612533569, "learning_rate": 6.106781849979648e-07, "loss": 0.0191, "step": 95270 }, { "grad_norm": 0.5254252552986145, "learning_rate": 6.081045506126482e-07, "loss": 0.0075, "step": 95280 }, { "grad_norm": 0.5756973624229431, "learning_rate": 6.055363176507678e-07, "loss": 0.0154, "step": 95290 }, { "grad_norm": 0.3757590055465698, "learning_rate": 6.029734863931602e-07, "loss": 0.0154, "step": 95300 }, { "grad_norm": 0.22753313183784485, "learning_rate": 6.004160571201179e-07, "loss": 0.0127, "step": 95310 }, { "grad_norm": 0.48165255784988403, "learning_rate": 5.978640301112947e-07, "loss": 0.0137, "step": 95320 }, { "grad_norm": 0.2448103427886963, "learning_rate": 5.9531740564579e-07, "loss": 0.0137, "step": 95330 }, { "grad_norm": 0.3959011137485504, "learning_rate": 5.927761840020918e-07, "loss": 0.0107, "step": 95340 }, { "grad_norm": 0.3850444257259369, "learning_rate": 5.902403654581113e-07, "loss": 0.0152, "step": 95350 }, { "grad_norm": 0.36373209953308105, "learning_rate": 5.877099502911542e-07, "loss": 0.0159, "step": 95360 }, { "grad_norm": 0.3967352509498596, "learning_rate": 5.851849387779496e-07, "loss": 0.0123, "step": 95370 }, { "grad_norm": 0.1680109202861786, "learning_rate": 5.826653311946262e-07, "loss": 0.009, "step": 95380 }, { "grad_norm": 0.26625633239746094, "learning_rate": 5.801511278167249e-07, "loss": 0.0139, "step": 95390 }, { "grad_norm": 0.08975637704133987, "learning_rate": 5.776423289191868e-07, "loss": 0.013, "step": 95400 }, { "grad_norm": 0.5204786062240601, "learning_rate": 5.751389347763869e-07, "loss": 0.0126, "step": 95410 }, { "grad_norm": 0.386562317609787, "learning_rate": 5.726409456620841e-07, "loss": 0.0109, "step": 95420 }, { "grad_norm": 0.21989013254642487, "learning_rate": 5.701483618494541e-07, "loss": 0.0163, "step": 95430 }, { "grad_norm": 0.2539593577384949, "learning_rate": 5.67661183611079e-07, "loss": 0.0108, "step": 95440 }, { "grad_norm": 0.19424208998680115, "learning_rate": 5.651794112189579e-07, "loss": 0.0157, "step": 95450 }, { "grad_norm": 0.35227763652801514, "learning_rate": 5.627030449444903e-07, "loss": 0.0162, "step": 95460 }, { "grad_norm": 0.8539464473724365, "learning_rate": 5.60232085058493e-07, "loss": 0.0343, "step": 95470 }, { "grad_norm": 0.2606319785118103, "learning_rate": 5.577665318311776e-07, "loss": 0.0119, "step": 95480 }, { "grad_norm": 0.37219691276550293, "learning_rate": 5.553063855321838e-07, "loss": 0.0141, "step": 95490 }, { "grad_norm": 0.15670549869537354, "learning_rate": 5.528516464305467e-07, "loss": 0.0136, "step": 95500 }, { "grad_norm": 0.2569981813430786, "learning_rate": 5.504023147947068e-07, "loss": 0.0148, "step": 95510 }, { "grad_norm": 0.1715465784072876, "learning_rate": 5.479583908925279e-07, "loss": 0.0145, "step": 95520 }, { "grad_norm": 0.15870751440525055, "learning_rate": 5.455198749912626e-07, "loss": 0.0139, "step": 95530 }, { "grad_norm": 0.5923036932945251, "learning_rate": 5.430867673575979e-07, "loss": 0.0118, "step": 95540 }, { "grad_norm": 0.2707495391368866, "learning_rate": 5.406590682575985e-07, "loss": 0.0099, "step": 95550 }, { "grad_norm": 0.27194875478744507, "learning_rate": 5.382367779567687e-07, "loss": 0.0306, "step": 95560 }, { "grad_norm": 0.20567277073860168, "learning_rate": 5.358198967200023e-07, "loss": 0.0133, "step": 95570 }, { "grad_norm": 0.24584053456783295, "learning_rate": 5.334084248115989e-07, "loss": 0.0122, "step": 95580 }, { "grad_norm": 0.16146767139434814, "learning_rate": 5.310023624952754e-07, "loss": 0.0116, "step": 95590 }, { "grad_norm": 0.24317380785942078, "learning_rate": 5.286017100341656e-07, "loss": 0.0112, "step": 95600 }, { "grad_norm": 0.24894028902053833, "learning_rate": 5.262064676907874e-07, "loss": 0.0147, "step": 95610 }, { "grad_norm": 0.205660879611969, "learning_rate": 5.238166357270924e-07, "loss": 0.0107, "step": 95620 }, { "grad_norm": 0.43527472019195557, "learning_rate": 5.214322144044215e-07, "loss": 0.0113, "step": 95630 }, { "grad_norm": 0.395786851644516, "learning_rate": 5.190532039835327e-07, "loss": 0.0227, "step": 95640 }, { "grad_norm": 0.84747713804245, "learning_rate": 5.166796047245903e-07, "loss": 0.0331, "step": 95650 }, { "grad_norm": 0.3542960286140442, "learning_rate": 5.143114168871699e-07, "loss": 0.0156, "step": 95660 }, { "grad_norm": 0.2171001434326172, "learning_rate": 5.119486407302532e-07, "loss": 0.0091, "step": 95670 }, { "grad_norm": 0.6090341210365295, "learning_rate": 5.095912765122335e-07, "loss": 0.0143, "step": 95680 }, { "grad_norm": 0.465763121843338, "learning_rate": 5.07239324490899e-07, "loss": 0.0373, "step": 95690 }, { "grad_norm": 0.5903303623199463, "learning_rate": 5.048927849234608e-07, "loss": 0.0211, "step": 95700 }, { "grad_norm": 0.18749433755874634, "learning_rate": 5.025516580665358e-07, "loss": 0.0107, "step": 95710 }, { "grad_norm": 0.32487741112709045, "learning_rate": 5.002159441761412e-07, "loss": 0.0101, "step": 95720 }, { "grad_norm": 0.32189613580703735, "learning_rate": 4.978856435077062e-07, "loss": 0.0125, "step": 95730 }, { "grad_norm": 0.16980327665805817, "learning_rate": 4.955607563160769e-07, "loss": 0.014, "step": 95740 }, { "grad_norm": 0.12354446202516556, "learning_rate": 4.932412828554944e-07, "loss": 0.015, "step": 95750 }, { "grad_norm": 0.25581640005111694, "learning_rate": 4.909272233796114e-07, "loss": 0.013, "step": 95760 }, { "grad_norm": 0.20477084815502167, "learning_rate": 4.88618578141492e-07, "loss": 0.0218, "step": 95770 }, { "grad_norm": 0.30036184191703796, "learning_rate": 4.863153473936011e-07, "loss": 0.0157, "step": 95780 }, { "grad_norm": 0.5863488912582397, "learning_rate": 4.840175313878259e-07, "loss": 0.0222, "step": 95790 }, { "grad_norm": 0.18675830960273743, "learning_rate": 4.817251303754433e-07, "loss": 0.0128, "step": 95800 }, { "grad_norm": 0.3299456834793091, "learning_rate": 4.794381446071583e-07, "loss": 0.0126, "step": 95810 }, { "grad_norm": 0.23043832182884216, "learning_rate": 4.771565743330597e-07, "loss": 0.0164, "step": 95820 }, { "grad_norm": 0.34938251972198486, "learning_rate": 4.7488041980266463e-07, "loss": 0.0123, "step": 95830 }, { "grad_norm": 0.37888258695602417, "learning_rate": 4.726096812648795e-07, "loss": 0.0194, "step": 95840 }, { "grad_norm": 0.3758976459503174, "learning_rate": 4.703443589680445e-07, "loss": 0.0134, "step": 95850 }, { "grad_norm": 0.10008635371923447, "learning_rate": 4.6808445315987826e-07, "loss": 0.0157, "step": 95860 }, { "grad_norm": 0.3165275752544403, "learning_rate": 4.658299640875274e-07, "loss": 0.019, "step": 95870 }, { "grad_norm": 0.18462248146533966, "learning_rate": 4.635808919975393e-07, "loss": 0.0109, "step": 95880 }, { "grad_norm": 0.12588068842887878, "learning_rate": 4.6133723713586706e-07, "loss": 0.0161, "step": 95890 }, { "grad_norm": 0.5288577675819397, "learning_rate": 4.590989997478756e-07, "loss": 0.0168, "step": 95900 }, { "grad_norm": 0.3764590322971344, "learning_rate": 4.5686618007833027e-07, "loss": 0.0195, "step": 95910 }, { "grad_norm": 0.5102682113647461, "learning_rate": 4.54638778371419e-07, "loss": 0.0147, "step": 95920 }, { "grad_norm": 0.2776382267475128, "learning_rate": 4.524167948707192e-07, "loss": 0.0139, "step": 95930 }, { "grad_norm": 0.2721536457538605, "learning_rate": 4.5020022981921986e-07, "loss": 0.0139, "step": 95940 }, { "grad_norm": 0.4665655493736267, "learning_rate": 4.47989083459327e-07, "loss": 0.0161, "step": 95950 }, { "grad_norm": 0.3347662687301636, "learning_rate": 4.457833560328528e-07, "loss": 0.0133, "step": 95960 }, { "grad_norm": 0.2228168249130249, "learning_rate": 4.4358304778100434e-07, "loss": 0.0194, "step": 95970 }, { "grad_norm": 0.20649881660938263, "learning_rate": 4.413881589444058e-07, "loss": 0.0151, "step": 95980 }, { "grad_norm": 0.1980368047952652, "learning_rate": 4.391986897630873e-07, "loss": 0.0111, "step": 95990 }, { "grad_norm": 0.25453832745552063, "learning_rate": 4.370146404764908e-07, "loss": 0.0118, "step": 96000 }, { "grad_norm": 0.37745577096939087, "learning_rate": 4.3483601132345285e-07, "loss": 0.0124, "step": 96010 }, { "grad_norm": 0.2481064349412918, "learning_rate": 4.32662802542233e-07, "loss": 0.0091, "step": 96020 }, { "grad_norm": 0.12740932404994965, "learning_rate": 4.3049501437047444e-07, "loss": 0.016, "step": 96030 }, { "grad_norm": 0.19704477488994598, "learning_rate": 4.283326470452653e-07, "loss": 0.0153, "step": 96040 }, { "grad_norm": 0.3173505663871765, "learning_rate": 4.2617570080306647e-07, "loss": 0.0153, "step": 96050 }, { "grad_norm": 0.7491220831871033, "learning_rate": 4.2402417587975585e-07, "loss": 0.0252, "step": 96060 }, { "grad_norm": 0.4292139708995819, "learning_rate": 4.218780725106286e-07, "loss": 0.0183, "step": 96070 }, { "grad_norm": 0.39728906750679016, "learning_rate": 4.197373909303748e-07, "loss": 0.0138, "step": 96080 }, { "grad_norm": 0.2470419555902481, "learning_rate": 4.17602131373096e-07, "loss": 0.0172, "step": 96090 }, { "grad_norm": 1.2102959156036377, "learning_rate": 4.154722940723055e-07, "loss": 0.0159, "step": 96100 }, { "grad_norm": 0.4564623534679413, "learning_rate": 4.1334787926091136e-07, "loss": 0.0211, "step": 96110 }, { "grad_norm": 0.26801976561546326, "learning_rate": 4.1122888717124444e-07, "loss": 0.0123, "step": 96120 }, { "grad_norm": 0.32684680819511414, "learning_rate": 4.091153180350249e-07, "loss": 0.026, "step": 96130 }, { "grad_norm": 0.46101582050323486, "learning_rate": 4.0700717208340125e-07, "loss": 0.0268, "step": 96140 }, { "grad_norm": 0.27172285318374634, "learning_rate": 4.049044495469112e-07, "loss": 0.0105, "step": 96150 }, { "grad_norm": 0.4628154933452606, "learning_rate": 4.0280715065549866e-07, "loss": 0.0162, "step": 96160 }, { "grad_norm": 0.5561738014221191, "learning_rate": 4.0071527563853575e-07, "loss": 0.0184, "step": 96170 }, { "grad_norm": 0.28690046072006226, "learning_rate": 3.9862882472477823e-07, "loss": 0.0196, "step": 96180 }, { "grad_norm": 0.32811111211776733, "learning_rate": 3.9654779814239375e-07, "loss": 0.0137, "step": 96190 }, { "grad_norm": 0.3660283386707306, "learning_rate": 3.944721961189668e-07, "loss": 0.012, "step": 96200 }, { "grad_norm": 0.7582077383995056, "learning_rate": 3.9240201888147697e-07, "loss": 0.0203, "step": 96210 }, { "grad_norm": 0.23779895901679993, "learning_rate": 3.9033726665632096e-07, "loss": 0.0116, "step": 96220 }, { "grad_norm": 0.20472979545593262, "learning_rate": 3.882779396692959e-07, "loss": 0.0136, "step": 96230 }, { "grad_norm": 0.28587305545806885, "learning_rate": 3.8622403814559947e-07, "loss": 0.0133, "step": 96240 }, { "grad_norm": 0.4558188021183014, "learning_rate": 3.841755623098575e-07, "loss": 0.0182, "step": 96250 }, { "grad_norm": 0.15006126463413239, "learning_rate": 3.821325123860742e-07, "loss": 0.0108, "step": 96260 }, { "grad_norm": 0.3034559488296509, "learning_rate": 3.800948885976818e-07, "loss": 0.0167, "step": 96270 }, { "grad_norm": 0.19026519358158112, "learning_rate": 3.780626911675078e-07, "loss": 0.0074, "step": 96280 }, { "grad_norm": 0.4850430488586426, "learning_rate": 3.76035920317791e-07, "loss": 0.0149, "step": 96290 }, { "grad_norm": 0.7079898715019226, "learning_rate": 3.740145762701819e-07, "loss": 0.0164, "step": 96300 }, { "grad_norm": 0.37160807847976685, "learning_rate": 3.7199865924572585e-07, "loss": 0.017, "step": 96310 }, { "grad_norm": 0.14483794569969177, "learning_rate": 3.699881694648799e-07, "loss": 0.0199, "step": 96320 }, { "grad_norm": 0.6161216497421265, "learning_rate": 3.679831071475126e-07, "loss": 0.0204, "step": 96330 }, { "grad_norm": 0.2540818750858307, "learning_rate": 3.65983472512893e-07, "loss": 0.0179, "step": 96340 }, { "grad_norm": 0.45961910486221313, "learning_rate": 3.6398926577969616e-07, "loss": 0.0139, "step": 96350 }, { "grad_norm": 0.4608214795589447, "learning_rate": 3.620004871660032e-07, "loss": 0.02, "step": 96360 }, { "grad_norm": 0.4191797077655792, "learning_rate": 3.600171368893124e-07, "loss": 0.0143, "step": 96370 }, { "grad_norm": 0.2589248716831207, "learning_rate": 3.580392151665113e-07, "loss": 0.0107, "step": 96380 }, { "grad_norm": 0.38677844405174255, "learning_rate": 3.5606672221391024e-07, "loss": 0.0155, "step": 96390 }, { "grad_norm": 0.46425822377204895, "learning_rate": 3.540996582472089e-07, "loss": 0.0194, "step": 96400 }, { "grad_norm": 0.20758968591690063, "learning_rate": 3.521380234815297e-07, "loss": 0.0183, "step": 96410 }, { "grad_norm": 0.21274808049201965, "learning_rate": 3.5018181813138985e-07, "loss": 0.0096, "step": 96420 }, { "grad_norm": 0.2783748209476471, "learning_rate": 3.482310424107238e-07, "loss": 0.0101, "step": 96430 }, { "grad_norm": 0.2334681898355484, "learning_rate": 3.4628569653285537e-07, "loss": 0.0121, "step": 96440 }, { "grad_norm": 0.2558535933494568, "learning_rate": 3.44345780710531e-07, "loss": 0.0198, "step": 96450 }, { "grad_norm": 0.3123893737792969, "learning_rate": 3.424112951558978e-07, "loss": 0.0235, "step": 96460 }, { "grad_norm": 0.40431156754493713, "learning_rate": 3.40482240080503e-07, "loss": 0.0132, "step": 96470 }, { "grad_norm": 0.6340899467468262, "learning_rate": 3.3855861569531133e-07, "loss": 0.0349, "step": 96480 }, { "grad_norm": 0.17844118177890778, "learning_rate": 3.3664042221068226e-07, "loss": 0.0146, "step": 96490 }, { "grad_norm": 0.2690035402774811, "learning_rate": 3.347276598363924e-07, "loss": 0.0207, "step": 96500 }, { "grad_norm": 0.3191603422164917, "learning_rate": 3.328203287816134e-07, "loss": 0.0186, "step": 96510 }, { "grad_norm": 0.18296799063682556, "learning_rate": 3.3091842925493386e-07, "loss": 0.0158, "step": 96520 }, { "grad_norm": 0.4119979739189148, "learning_rate": 3.2902196146433196e-07, "loss": 0.016, "step": 96530 }, { "grad_norm": 0.41371074318885803, "learning_rate": 3.2713092561720836e-07, "loss": 0.0144, "step": 96540 }, { "grad_norm": 0.29233452677726746, "learning_rate": 3.2524532192036996e-07, "loss": 0.0103, "step": 96550 }, { "grad_norm": 0.42869552969932556, "learning_rate": 3.2336515058001836e-07, "loss": 0.0145, "step": 96560 }, { "grad_norm": 0.4146599769592285, "learning_rate": 3.2149041180176143e-07, "loss": 0.0142, "step": 96570 }, { "grad_norm": 0.4610508978366852, "learning_rate": 3.1962110579062954e-07, "loss": 0.0126, "step": 96580 }, { "grad_norm": 0.3476763069629669, "learning_rate": 3.177572327510314e-07, "loss": 0.0146, "step": 96590 }, { "grad_norm": 0.20581233501434326, "learning_rate": 3.158987928868151e-07, "loss": 0.0176, "step": 96600 }, { "grad_norm": 0.48483166098594666, "learning_rate": 3.1404578640120144e-07, "loss": 0.0205, "step": 96610 }, { "grad_norm": 0.21615906059741974, "learning_rate": 3.1219821349683933e-07, "loss": 0.0107, "step": 96620 }, { "grad_norm": 0.21446774899959564, "learning_rate": 3.103560743757783e-07, "loss": 0.0122, "step": 96630 }, { "grad_norm": 0.3521297574043274, "learning_rate": 3.0851936923946834e-07, "loss": 0.0161, "step": 96640 }, { "grad_norm": 0.34366998076438904, "learning_rate": 3.06688098288771e-07, "loss": 0.0178, "step": 96650 }, { "grad_norm": 0.10677517205476761, "learning_rate": 3.0486226172394825e-07, "loss": 0.0108, "step": 96660 }, { "grad_norm": 0.2639418840408325, "learning_rate": 3.030418597446738e-07, "loss": 0.0152, "step": 96670 }, { "grad_norm": 0.2951837182044983, "learning_rate": 3.012268925500272e-07, "loss": 0.0141, "step": 96680 }, { "grad_norm": 0.29923656582832336, "learning_rate": 2.994173603384831e-07, "loss": 0.019, "step": 96690 }, { "grad_norm": 0.45146915316581726, "learning_rate": 2.9761326330793317e-07, "loss": 0.0136, "step": 96700 }, { "grad_norm": 0.32000309228897095, "learning_rate": 2.958146016556751e-07, "loss": 0.0091, "step": 96710 }, { "grad_norm": 0.3986717164516449, "learning_rate": 2.94021375578396e-07, "loss": 0.0111, "step": 96720 }, { "grad_norm": 0.4959351122379303, "learning_rate": 2.9223358527221135e-07, "loss": 0.0184, "step": 96730 }, { "grad_norm": 0.42150914669036865, "learning_rate": 2.904512309326257e-07, "loss": 0.0236, "step": 96740 }, { "grad_norm": 0.25703203678131104, "learning_rate": 2.8867431275455545e-07, "loss": 0.0116, "step": 96750 }, { "grad_norm": 0.3587130308151245, "learning_rate": 2.86902830932323e-07, "loss": 0.0205, "step": 96760 }, { "grad_norm": 0.18155111372470856, "learning_rate": 2.851367856596565e-07, "loss": 0.0175, "step": 96770 }, { "grad_norm": 0.3128410279750824, "learning_rate": 2.83376177129685e-07, "loss": 0.0109, "step": 96780 }, { "grad_norm": 0.49931809306144714, "learning_rate": 2.8162100553494884e-07, "loss": 0.0194, "step": 96790 }, { "grad_norm": 0.2069873958826065, "learning_rate": 2.798712710673834e-07, "loss": 0.0132, "step": 96800 }, { "grad_norm": 0.5947234630584717, "learning_rate": 2.7812697391834676e-07, "loss": 0.0135, "step": 96810 }, { "grad_norm": 0.430383563041687, "learning_rate": 2.7638811427858627e-07, "loss": 0.0115, "step": 96820 }, { "grad_norm": 0.7997053265571594, "learning_rate": 2.746546923382665e-07, "loss": 0.0156, "step": 96830 }, { "grad_norm": 0.35677316784858704, "learning_rate": 2.729267082869469e-07, "loss": 0.0134, "step": 96840 }, { "grad_norm": 0.3145233988761902, "learning_rate": 2.712041623135986e-07, "loss": 0.0154, "step": 96850 }, { "grad_norm": 0.34523290395736694, "learning_rate": 2.694870546065931e-07, "loss": 0.0299, "step": 96860 }, { "grad_norm": 0.1542213261127472, "learning_rate": 2.677753853537246e-07, "loss": 0.0117, "step": 96870 }, { "grad_norm": 0.2988985776901245, "learning_rate": 2.660691547421601e-07, "loss": 0.0098, "step": 96880 }, { "grad_norm": 0.24490372836589813, "learning_rate": 2.6436836295850033e-07, "loss": 0.0127, "step": 96890 }, { "grad_norm": 0.38042014837265015, "learning_rate": 2.6267301018874093e-07, "loss": 0.0186, "step": 96900 }, { "grad_norm": 0.43651798367500305, "learning_rate": 2.6098309661828355e-07, "loss": 0.0126, "step": 96910 }, { "grad_norm": 0.23892369866371155, "learning_rate": 2.592986224319305e-07, "loss": 0.0179, "step": 96920 }, { "grad_norm": 0.20917150378227234, "learning_rate": 2.576195878139009e-07, "loss": 0.0149, "step": 96930 }, { "grad_norm": 0.275376558303833, "learning_rate": 2.5594599294780364e-07, "loss": 0.0177, "step": 96940 }, { "grad_norm": 0.7577908039093018, "learning_rate": 2.542778380166644e-07, "loss": 0.0133, "step": 96950 }, { "grad_norm": 0.4038904905319214, "learning_rate": 2.526151232029095e-07, "loss": 0.0164, "step": 96960 }, { "grad_norm": 0.21593144536018372, "learning_rate": 2.5095784868836567e-07, "loss": 0.0178, "step": 96970 }, { "grad_norm": 0.22994573414325714, "learning_rate": 2.493060146542825e-07, "loss": 0.0201, "step": 96980 }, { "grad_norm": 0.18459804356098175, "learning_rate": 2.476596212812876e-07, "loss": 0.0175, "step": 96990 }, { "grad_norm": 0.2974364161491394, "learning_rate": 2.4601866874943703e-07, "loss": 0.0136, "step": 97000 }, { "grad_norm": 0.3819175660610199, "learning_rate": 2.4438315723818163e-07, "loss": 0.0139, "step": 97010 }, { "grad_norm": 0.22628480195999146, "learning_rate": 2.4275308692637833e-07, "loss": 0.0119, "step": 97020 }, { "grad_norm": 0.25269436836242676, "learning_rate": 2.411284579922901e-07, "loss": 0.0151, "step": 97030 }, { "grad_norm": 0.43992820382118225, "learning_rate": 2.395092706135804e-07, "loss": 0.0162, "step": 97040 }, { "grad_norm": 0.32022324204444885, "learning_rate": 2.3789552496731872e-07, "loss": 0.0122, "step": 97050 }, { "grad_norm": 0.4437674283981323, "learning_rate": 2.3628722122999158e-07, "loss": 0.0191, "step": 97060 }, { "grad_norm": 0.2996799945831299, "learning_rate": 2.3468435957747503e-07, "loss": 0.0158, "step": 97070 }, { "grad_norm": 0.10763507336378098, "learning_rate": 2.330869401850566e-07, "loss": 0.0098, "step": 97080 }, { "grad_norm": 1.6198031902313232, "learning_rate": 2.3149496322742437e-07, "loss": 0.0212, "step": 97090 }, { "grad_norm": 0.1444977968931198, "learning_rate": 2.2990842887868346e-07, "loss": 0.0119, "step": 97100 }, { "grad_norm": 0.6370754241943359, "learning_rate": 2.2832733731232292e-07, "loss": 0.0174, "step": 97110 }, { "grad_norm": 0.3071785867214203, "learning_rate": 2.2675168870124886e-07, "loss": 0.0116, "step": 97120 }, { "grad_norm": 0.7482567429542542, "learning_rate": 2.2518148321778455e-07, "loss": 0.0255, "step": 97130 }, { "grad_norm": 0.545238196849823, "learning_rate": 2.2361672103363706e-07, "loss": 0.0119, "step": 97140 }, { "grad_norm": 0.589910089969635, "learning_rate": 2.220574023199251e-07, "loss": 0.0187, "step": 97150 }, { "grad_norm": 0.26711368560791016, "learning_rate": 2.2050352724717894e-07, "loss": 0.0228, "step": 97160 }, { "grad_norm": 0.29977333545684814, "learning_rate": 2.1895509598532372e-07, "loss": 0.0135, "step": 97170 }, { "grad_norm": 0.2007180154323578, "learning_rate": 2.1741210870369066e-07, "loss": 0.0097, "step": 97180 }, { "grad_norm": 0.32380229234695435, "learning_rate": 2.1587456557102814e-07, "loss": 0.0094, "step": 97190 }, { "grad_norm": 0.3125370442867279, "learning_rate": 2.143424667554683e-07, "loss": 0.0109, "step": 97200 }, { "grad_norm": 0.4375799596309662, "learning_rate": 2.128158124245716e-07, "loss": 0.0399, "step": 97210 }, { "grad_norm": 0.4576386511325836, "learning_rate": 2.1129460274527668e-07, "loss": 0.058, "step": 97220 }, { "grad_norm": 0.3258086144924164, "learning_rate": 2.0977883788395046e-07, "loss": 0.0155, "step": 97230 }, { "grad_norm": 0.27349695563316345, "learning_rate": 2.0826851800634928e-07, "loss": 0.0129, "step": 97240 }, { "grad_norm": 0.3181256651878357, "learning_rate": 2.0676364327764653e-07, "loss": 0.0128, "step": 97250 }, { "grad_norm": 0.5008938312530518, "learning_rate": 2.0526421386240502e-07, "loss": 0.0142, "step": 97260 }, { "grad_norm": 0.34900835156440735, "learning_rate": 2.0377022992461025e-07, "loss": 0.0115, "step": 97270 }, { "grad_norm": 0.4890994131565094, "learning_rate": 2.0228169162763156e-07, "loss": 0.0133, "step": 97280 }, { "grad_norm": 0.22455532848834991, "learning_rate": 2.0079859913425536e-07, "loss": 0.0131, "step": 97290 }, { "grad_norm": 0.26947253942489624, "learning_rate": 1.9932095260667417e-07, "loss": 0.0222, "step": 97300 }, { "grad_norm": 0.2000734657049179, "learning_rate": 1.9784875220648092e-07, "loss": 0.0137, "step": 97310 }, { "grad_norm": 0.37399810552597046, "learning_rate": 1.9638199809466907e-07, "loss": 0.0134, "step": 97320 }, { "grad_norm": 0.47809502482414246, "learning_rate": 1.9492069043164364e-07, "loss": 0.0151, "step": 97330 }, { "grad_norm": 0.5437387824058533, "learning_rate": 1.9346482937721567e-07, "loss": 0.0216, "step": 97340 }, { "grad_norm": 0.385835200548172, "learning_rate": 1.920144150905856e-07, "loss": 0.0101, "step": 97350 }, { "grad_norm": 0.28239285945892334, "learning_rate": 1.9056944773037655e-07, "loss": 0.0128, "step": 97360 }, { "grad_norm": 0.2777438163757324, "learning_rate": 1.8912992745460102e-07, "loss": 0.0133, "step": 97370 }, { "grad_norm": 0.42319726943969727, "learning_rate": 1.876958544206886e-07, "loss": 0.014, "step": 97380 }, { "grad_norm": 0.3905908763408661, "learning_rate": 1.8626722878546942e-07, "loss": 0.019, "step": 97390 }, { "grad_norm": 0.20449697971343994, "learning_rate": 1.8484405070516851e-07, "loss": 0.0076, "step": 97400 }, { "grad_norm": 0.3476518988609314, "learning_rate": 1.8342632033542805e-07, "loss": 0.0151, "step": 97410 }, { "grad_norm": 0.226748988032341, "learning_rate": 1.8201403783129066e-07, "loss": 0.0175, "step": 97420 }, { "grad_norm": 0.26001328229904175, "learning_rate": 1.806072033471884e-07, "loss": 0.0167, "step": 97430 }, { "grad_norm": 0.3844868838787079, "learning_rate": 1.79205817036987e-07, "loss": 0.0109, "step": 97440 }, { "grad_norm": 0.5367535352706909, "learning_rate": 1.7780987905393065e-07, "loss": 0.0181, "step": 97450 }, { "grad_norm": 0.33798813819885254, "learning_rate": 1.764193895506805e-07, "loss": 0.0115, "step": 97460 }, { "grad_norm": 0.3672693371772766, "learning_rate": 1.7503434867929824e-07, "loss": 0.0138, "step": 97470 }, { "grad_norm": 0.32327473163604736, "learning_rate": 1.7365475659125165e-07, "loss": 0.0157, "step": 97480 }, { "grad_norm": 0.3117152154445648, "learning_rate": 1.7228061343740332e-07, "loss": 0.0104, "step": 97490 }, { "grad_norm": 0.9031414985656738, "learning_rate": 1.7091191936803308e-07, "loss": 0.0124, "step": 97500 }, { "grad_norm": 0.4705057144165039, "learning_rate": 1.6954867453281563e-07, "loss": 0.0172, "step": 97510 }, { "grad_norm": 0.26721692085266113, "learning_rate": 1.6819087908084285e-07, "loss": 0.0172, "step": 97520 }, { "grad_norm": 0.32776153087615967, "learning_rate": 1.6683853316058485e-07, "loss": 0.0154, "step": 97530 }, { "grad_norm": 0.26531150937080383, "learning_rate": 1.6549163691995106e-07, "loss": 0.0119, "step": 97540 }, { "grad_norm": 0.4129306674003601, "learning_rate": 1.641501905062237e-07, "loss": 0.0206, "step": 97550 }, { "grad_norm": 0.44137927889823914, "learning_rate": 1.628141940661021e-07, "loss": 0.0144, "step": 97560 }, { "grad_norm": 0.6699509620666504, "learning_rate": 1.6148364774569158e-07, "loss": 0.0155, "step": 97570 }, { "grad_norm": 0.44980955123901367, "learning_rate": 1.6015855169050355e-07, "loss": 0.0152, "step": 97580 }, { "grad_norm": 0.22517180442810059, "learning_rate": 1.5883890604543873e-07, "loss": 0.0102, "step": 97590 }, { "grad_norm": 0.2616801857948303, "learning_rate": 1.5752471095482057e-07, "loss": 0.0093, "step": 97600 }, { "grad_norm": 0.3537476658821106, "learning_rate": 1.5621596656235638e-07, "loss": 0.0145, "step": 97610 }, { "grad_norm": 0.7156422734260559, "learning_rate": 1.5491267301117608e-07, "loss": 0.022, "step": 97620 }, { "grad_norm": 0.3630688488483429, "learning_rate": 1.536148304438101e-07, "loss": 0.0164, "step": 97630 }, { "grad_norm": 0.3671429455280304, "learning_rate": 1.5232243900217823e-07, "loss": 0.0165, "step": 97640 }, { "grad_norm": 0.41313865780830383, "learning_rate": 1.51035498827623e-07, "loss": 0.0178, "step": 97650 }, { "grad_norm": 0.254109263420105, "learning_rate": 1.497540100608763e-07, "loss": 0.0176, "step": 97660 }, { "grad_norm": 0.23964405059814453, "learning_rate": 1.4847797284208153e-07, "loss": 0.0131, "step": 97670 }, { "grad_norm": 0.2628491520881653, "learning_rate": 1.472073873107882e-07, "loss": 0.0213, "step": 97680 }, { "grad_norm": 0.5321410298347473, "learning_rate": 1.459422536059407e-07, "loss": 0.0162, "step": 97690 }, { "grad_norm": 0.2465929239988327, "learning_rate": 1.4468257186589508e-07, "loss": 0.0171, "step": 97700 }, { "grad_norm": 0.30033183097839355, "learning_rate": 1.4342834222840217e-07, "loss": 0.0175, "step": 97710 }, { "grad_norm": 0.2136944681406021, "learning_rate": 1.4217956483063566e-07, "loss": 0.0122, "step": 97720 }, { "grad_norm": 0.3489849269390106, "learning_rate": 1.409362398091474e-07, "loss": 0.0133, "step": 97730 }, { "grad_norm": 0.25396299362182617, "learning_rate": 1.3969836729990638e-07, "loss": 0.0192, "step": 97740 }, { "grad_norm": 0.20044733583927155, "learning_rate": 1.384659474382932e-07, "loss": 0.0275, "step": 97750 }, { "grad_norm": 0.29723644256591797, "learning_rate": 1.372389803590779e-07, "loss": 0.0109, "step": 97760 }, { "grad_norm": 0.3042944073677063, "learning_rate": 1.360174661964364e-07, "loss": 0.0125, "step": 97770 }, { "grad_norm": 0.2863745093345642, "learning_rate": 1.3480140508396188e-07, "loss": 0.015, "step": 97780 }, { "grad_norm": 0.4570927917957306, "learning_rate": 1.335907971546313e-07, "loss": 0.0175, "step": 97790 }, { "grad_norm": 0.6567671895027161, "learning_rate": 1.3238564254083875e-07, "loss": 0.0142, "step": 97800 }, { "grad_norm": 0.22727449238300323, "learning_rate": 1.3118594137437323e-07, "loss": 0.0116, "step": 97810 }, { "grad_norm": 0.36485275626182556, "learning_rate": 1.2999169378644094e-07, "loss": 0.0242, "step": 97820 }, { "grad_norm": 0.1555250883102417, "learning_rate": 1.2880289990763738e-07, "loss": 0.0134, "step": 97830 }, { "grad_norm": 0.4638844132423401, "learning_rate": 1.2761955986796968e-07, "loss": 0.017, "step": 97840 }, { "grad_norm": 0.20242440700531006, "learning_rate": 1.264416737968399e-07, "loss": 0.0116, "step": 97850 }, { "grad_norm": 0.10295414924621582, "learning_rate": 1.2526924182307275e-07, "loss": 0.0086, "step": 97860 }, { "grad_norm": 0.2501147389411926, "learning_rate": 1.2410226407487124e-07, "loss": 0.0131, "step": 97870 }, { "grad_norm": 0.4289475083351135, "learning_rate": 1.229407406798555e-07, "loss": 0.0144, "step": 97880 }, { "grad_norm": 0.32193171977996826, "learning_rate": 1.2178467176505726e-07, "loss": 0.0086, "step": 97890 }, { "grad_norm": 0.2522185742855072, "learning_rate": 1.2063405745689205e-07, "loss": 0.0189, "step": 97900 }, { "grad_norm": 0.2874257266521454, "learning_rate": 1.1948889788119812e-07, "loss": 0.0113, "step": 97910 }, { "grad_norm": 0.4041862189769745, "learning_rate": 1.1834919316320303e-07, "loss": 0.0142, "step": 97920 }, { "grad_norm": 0.45982250571250916, "learning_rate": 1.1721494342754048e-07, "loss": 0.0146, "step": 97930 }, { "grad_norm": 0.18504659831523895, "learning_rate": 1.1608614879825563e-07, "loss": 0.0132, "step": 97940 }, { "grad_norm": 0.1434461623430252, "learning_rate": 1.1496280939879422e-07, "loss": 0.0111, "step": 97950 }, { "grad_norm": 0.3535839319229126, "learning_rate": 1.1384492535199687e-07, "loss": 0.013, "step": 97960 }, { "grad_norm": 0.1931455433368683, "learning_rate": 1.1273249678011578e-07, "loss": 0.0116, "step": 97970 }, { "grad_norm": 0.29232704639434814, "learning_rate": 1.1162552380480362e-07, "loss": 0.0164, "step": 97980 }, { "grad_norm": 0.3239060342311859, "learning_rate": 1.1052400654711359e-07, "loss": 0.0121, "step": 97990 }, { "grad_norm": 0.445549339056015, "learning_rate": 1.0942794512751597e-07, "loss": 0.0218, "step": 98000 }, { "grad_norm": 0.3244660794734955, "learning_rate": 1.0833733966587045e-07, "loss": 0.0151, "step": 98010 }, { "grad_norm": 0.43964847922325134, "learning_rate": 1.0725219028143718e-07, "loss": 0.0239, "step": 98020 }, { "grad_norm": 0.33182522654533386, "learning_rate": 1.0617249709289345e-07, "loss": 0.0158, "step": 98030 }, { "grad_norm": 0.4661836624145508, "learning_rate": 1.0509826021831703e-07, "loss": 0.0125, "step": 98040 }, { "grad_norm": 0.2521248161792755, "learning_rate": 1.0402947977517508e-07, "loss": 0.0096, "step": 98050 }, { "grad_norm": 0.3588317334651947, "learning_rate": 1.0296615588035185e-07, "loss": 0.0159, "step": 98060 }, { "grad_norm": 0.2985709607601166, "learning_rate": 1.0190828865012659e-07, "loss": 0.0176, "step": 98070 }, { "grad_norm": 0.24978090822696686, "learning_rate": 1.008558782001956e-07, "loss": 0.0228, "step": 98080 }, { "grad_norm": 0.17588552832603455, "learning_rate": 9.980892464563906e-08, "loss": 0.0178, "step": 98090 }, { "grad_norm": 0.4724918603897095, "learning_rate": 9.876742810095985e-08, "loss": 0.022, "step": 98100 }, { "grad_norm": 0.30601683259010315, "learning_rate": 9.77313886800446e-08, "loss": 0.013, "step": 98110 }, { "grad_norm": 0.1733628213405609, "learning_rate": 9.670080649619717e-08, "loss": 0.0196, "step": 98120 }, { "grad_norm": 0.7600831985473633, "learning_rate": 9.567568166212737e-08, "loss": 0.026, "step": 98130 }, { "grad_norm": 0.2557939887046814, "learning_rate": 9.465601428992888e-08, "loss": 0.0139, "step": 98140 }, { "grad_norm": 0.19884829223155975, "learning_rate": 9.364180449111803e-08, "loss": 0.011, "step": 98150 }, { "grad_norm": 0.2836974859237671, "learning_rate": 9.263305237661169e-08, "loss": 0.0196, "step": 98160 }, { "grad_norm": 0.4155578911304474, "learning_rate": 9.162975805671603e-08, "loss": 0.0106, "step": 98170 }, { "grad_norm": 0.20257821679115295, "learning_rate": 9.063192164115442e-08, "loss": 0.0187, "step": 98180 }, { "grad_norm": 0.39835116267204285, "learning_rate": 8.963954323904511e-08, "loss": 0.0262, "step": 98190 }, { "grad_norm": 0.38078224658966064, "learning_rate": 8.865262295891796e-08, "loss": 0.011, "step": 98200 }, { "grad_norm": 0.39675602316856384, "learning_rate": 8.767116090870331e-08, "loss": 0.016, "step": 98210 }, { "grad_norm": 0.35550621151924133, "learning_rate": 8.669515719572086e-08, "loss": 0.0173, "step": 98220 }, { "grad_norm": 0.23318344354629517, "learning_rate": 8.572461192671855e-08, "loss": 0.0144, "step": 98230 }, { "grad_norm": 0.6698954105377197, "learning_rate": 8.475952520782815e-08, "loss": 0.0258, "step": 98240 }, { "grad_norm": 0.4558037221431732, "learning_rate": 8.379989714458747e-08, "loss": 0.0158, "step": 98250 }, { "grad_norm": 0.19676601886749268, "learning_rate": 8.284572784194034e-08, "loss": 0.0169, "step": 98260 }, { "grad_norm": 0.1281735897064209, "learning_rate": 8.189701740424216e-08, "loss": 0.0127, "step": 98270 }, { "grad_norm": 0.19490520656108856, "learning_rate": 8.095376593522663e-08, "loss": 0.0164, "step": 98280 }, { "grad_norm": 0.12064468860626221, "learning_rate": 8.001597353806123e-08, "loss": 0.0128, "step": 98290 }, { "grad_norm": 0.28090399503707886, "learning_rate": 7.90836403152917e-08, "loss": 0.0106, "step": 98300 }, { "grad_norm": 0.40240278840065, "learning_rate": 7.815676636888092e-08, "loss": 0.0105, "step": 98310 }, { "grad_norm": 0.4401874840259552, "learning_rate": 7.723535180019226e-08, "loss": 0.0138, "step": 98320 }, { "grad_norm": 0.4044235348701477, "learning_rate": 7.6319396709984e-08, "loss": 0.0147, "step": 98330 }, { "grad_norm": 0.7577870488166809, "learning_rate": 7.540890119843158e-08, "loss": 0.0165, "step": 98340 }, { "grad_norm": 0.3596092462539673, "learning_rate": 7.450386536509979e-08, "loss": 0.0108, "step": 98350 }, { "grad_norm": 0.35956764221191406, "learning_rate": 7.360428930895947e-08, "loss": 0.013, "step": 98360 }, { "grad_norm": 0.2760455310344696, "learning_rate": 7.271017312839302e-08, "loss": 0.0141, "step": 98370 }, { "grad_norm": 0.3296409547328949, "learning_rate": 7.182151692118333e-08, "loss": 0.0211, "step": 98380 }, { "grad_norm": 0.14770327508449554, "learning_rate": 7.093832078449713e-08, "loss": 0.0102, "step": 98390 }, { "grad_norm": 0.19061055779457092, "learning_rate": 7.006058481493493e-08, "loss": 0.0181, "step": 98400 }, { "grad_norm": 0.530031144618988, "learning_rate": 6.918830910847552e-08, "loss": 0.0189, "step": 98410 }, { "grad_norm": 0.2516627907752991, "learning_rate": 6.832149376052033e-08, "loss": 0.0143, "step": 98420 }, { "grad_norm": 0.11055921763181686, "learning_rate": 6.746013886584912e-08, "loss": 0.0121, "step": 98430 }, { "grad_norm": 0.30336377024650574, "learning_rate": 6.660424451866431e-08, "loss": 0.0181, "step": 98440 }, { "grad_norm": 0.1331855058670044, "learning_rate": 6.575381081256327e-08, "loss": 0.011, "step": 98450 }, { "grad_norm": 0.6820236444473267, "learning_rate": 6.49088378405549e-08, "loss": 0.018, "step": 98460 }, { "grad_norm": 0.5016194581985474, "learning_rate": 6.406932569503755e-08, "loss": 0.0202, "step": 98470 }, { "grad_norm": 0.35361921787261963, "learning_rate": 6.323527446782662e-08, "loss": 0.0141, "step": 98480 }, { "grad_norm": 0.16839905083179474, "learning_rate": 6.240668425012142e-08, "loss": 0.0173, "step": 98490 }, { "grad_norm": 0.4023296535015106, "learning_rate": 6.158355513254388e-08, "loss": 0.0116, "step": 98500 }, { "grad_norm": 0.39459875226020813, "learning_rate": 6.076588720510534e-08, "loss": 0.0136, "step": 98510 }, { "grad_norm": 0.3359662890434265, "learning_rate": 5.995368055722316e-08, "loss": 0.0144, "step": 98520 }, { "grad_norm": 0.41586601734161377, "learning_rate": 5.914693527773185e-08, "loss": 0.0118, "step": 98530 }, { "grad_norm": 0.30307143926620483, "learning_rate": 5.834565145483861e-08, "loss": 0.0173, "step": 98540 }, { "grad_norm": 0.31040024757385254, "learning_rate": 5.754982917618446e-08, "loss": 0.0162, "step": 98550 }, { "grad_norm": 0.2117043137550354, "learning_rate": 5.675946852878866e-08, "loss": 0.0121, "step": 98560 }, { "grad_norm": 0.2918708324432373, "learning_rate": 5.59745695990932e-08, "loss": 0.0159, "step": 98570 }, { "grad_norm": 0.3214097321033478, "learning_rate": 5.51951324729294e-08, "loss": 0.0158, "step": 98580 }, { "grad_norm": 0.46415024995803833, "learning_rate": 5.4421157235529093e-08, "loss": 0.0111, "step": 98590 }, { "grad_norm": 0.7796269059181213, "learning_rate": 5.365264397154679e-08, "loss": 0.0121, "step": 98600 }, { "grad_norm": 0.16804572939872742, "learning_rate": 5.288959276501526e-08, "loss": 0.0113, "step": 98610 }, { "grad_norm": 0.32332783937454224, "learning_rate": 5.2132003699378876e-08, "loss": 0.0132, "step": 98620 }, { "grad_norm": 0.32212311029434204, "learning_rate": 5.137987685749912e-08, "loss": 0.011, "step": 98630 }, { "grad_norm": 0.3966935873031616, "learning_rate": 5.0633212321610227e-08, "loss": 0.0157, "step": 98640 }, { "grad_norm": 0.3242330849170685, "learning_rate": 4.989201017338019e-08, "loss": 0.0126, "step": 98650 }, { "grad_norm": 0.3394361138343811, "learning_rate": 4.9156270493860846e-08, "loss": 0.0139, "step": 98660 }, { "grad_norm": 0.24205590784549713, "learning_rate": 4.842599336351561e-08, "loss": 0.0199, "step": 98670 }, { "grad_norm": 0.3255287706851959, "learning_rate": 4.7701178862197274e-08, "loss": 0.0233, "step": 98680 }, { "grad_norm": 0.6028876900672913, "learning_rate": 4.698182706918131e-08, "loss": 0.0164, "step": 98690 }, { "grad_norm": 0.3124229311943054, "learning_rate": 4.6267938063121466e-08, "loss": 0.0104, "step": 98700 }, { "grad_norm": 0.4421544671058655, "learning_rate": 4.555951192209973e-08, "loss": 0.0171, "step": 98710 }, { "grad_norm": 0.21752451360225677, "learning_rate": 4.485654872358747e-08, "loss": 0.0128, "step": 98720 }, { "grad_norm": 0.362930566072464, "learning_rate": 4.415904854445097e-08, "loss": 0.0153, "step": 98730 }, { "grad_norm": 0.32137396931648254, "learning_rate": 4.34670114609792e-08, "loss": 0.0115, "step": 98740 }, { "grad_norm": 0.43763500452041626, "learning_rate": 4.278043754884498e-08, "loss": 0.0161, "step": 98750 }, { "grad_norm": 0.3288326561450958, "learning_rate": 4.209932688312712e-08, "loss": 0.0144, "step": 98760 }, { "grad_norm": 0.33230888843536377, "learning_rate": 4.14236795383216e-08, "loss": 0.0119, "step": 98770 }, { "grad_norm": 0.44552576541900635, "learning_rate": 4.075349558830821e-08, "loss": 0.0289, "step": 98780 }, { "grad_norm": 0.11045979708433151, "learning_rate": 4.008877510638387e-08, "loss": 0.013, "step": 98790 }, { "grad_norm": 0.3541596233844757, "learning_rate": 3.942951816523488e-08, "loss": 0.0119, "step": 98800 }, { "grad_norm": 0.3873484432697296, "learning_rate": 3.8775724836959125e-08, "loss": 0.0117, "step": 98810 }, { "grad_norm": 0.20518986880779266, "learning_rate": 3.812739519305497e-08, "loss": 0.0109, "step": 98820 }, { "grad_norm": 0.3934354782104492, "learning_rate": 3.748452930442126e-08, "loss": 0.016, "step": 98830 }, { "grad_norm": 0.2635137438774109, "learning_rate": 3.6847127241362855e-08, "loss": 0.0087, "step": 98840 }, { "grad_norm": 0.3408622145652771, "learning_rate": 3.621518907358512e-08, "loss": 0.011, "step": 98850 }, { "grad_norm": 0.5046584606170654, "learning_rate": 3.558871487019388e-08, "loss": 0.0165, "step": 98860 }, { "grad_norm": 0.6659818291664124, "learning_rate": 3.4967704699701007e-08, "loss": 0.016, "step": 98870 }, { "grad_norm": 0.35305750370025635, "learning_rate": 3.435215863001884e-08, "loss": 0.013, "step": 98880 }, { "grad_norm": 0.33473244309425354, "learning_rate": 3.374207672846019e-08, "loss": 0.0156, "step": 98890 }, { "grad_norm": 0.8327149748802185, "learning_rate": 3.313745906174392e-08, "loss": 0.0129, "step": 98900 }, { "grad_norm": 0.26549553871154785, "learning_rate": 3.253830569599492e-08, "loss": 0.0115, "step": 98910 }, { "grad_norm": 0.4470493793487549, "learning_rate": 3.1944616696727436e-08, "loss": 0.0106, "step": 98920 }, { "grad_norm": 0.3334558308124542, "learning_rate": 3.135639212887287e-08, "loss": 0.0103, "step": 98930 }, { "grad_norm": 0.23092056810855865, "learning_rate": 3.077363205675754e-08, "loss": 0.0083, "step": 98940 }, { "grad_norm": 0.12052073329687119, "learning_rate": 3.0196336544113804e-08, "loss": 0.0133, "step": 98950 }, { "grad_norm": 0.24671335518360138, "learning_rate": 2.9624505654063383e-08, "loss": 0.0129, "step": 98960 }, { "grad_norm": 0.4607592523097992, "learning_rate": 2.90581394491507e-08, "loss": 0.0175, "step": 98970 }, { "grad_norm": 0.5236124992370605, "learning_rate": 2.8497237991309545e-08, "loss": 0.0092, "step": 98980 }, { "grad_norm": 0.1758057326078415, "learning_rate": 2.7941801341879735e-08, "loss": 0.0254, "step": 98990 }, { "grad_norm": 0.17868255078792572, "learning_rate": 2.7391829561601578e-08, "loss": 0.0189, "step": 99000 }, { "grad_norm": 0.3573707342147827, "learning_rate": 2.6847322710621403e-08, "loss": 0.0134, "step": 99010 }, { "grad_norm": 0.3123684525489807, "learning_rate": 2.6308280848486022e-08, "loss": 0.0274, "step": 99020 }, { "grad_norm": 0.5091071128845215, "learning_rate": 2.5774704034137177e-08, "loss": 0.0138, "step": 99030 }, { "grad_norm": 0.5051783323287964, "learning_rate": 2.524659232593929e-08, "loss": 0.0165, "step": 99040 }, { "grad_norm": 0.6190138459205627, "learning_rate": 2.472394578163506e-08, "loss": 0.0186, "step": 99050 }, { "grad_norm": 0.2331896424293518, "learning_rate": 2.4206764458378772e-08, "loss": 0.0127, "step": 99060 }, { "grad_norm": 0.3415078818798065, "learning_rate": 2.3695048412736285e-08, "loss": 0.014, "step": 99070 }, { "grad_norm": 0.2912381887435913, "learning_rate": 2.3188797700668395e-08, "loss": 0.0139, "step": 99080 }, { "grad_norm": 0.32079970836639404, "learning_rate": 2.268801237753082e-08, "loss": 0.0133, "step": 99090 }, { "grad_norm": 0.2505406141281128, "learning_rate": 2.2192692498090862e-08, "loss": 0.0209, "step": 99100 }, { "grad_norm": 0.22903938591480255, "learning_rate": 2.170283811652185e-08, "loss": 0.0144, "step": 99110 }, { "grad_norm": 0.24444113671779633, "learning_rate": 2.12184492863865e-08, "loss": 0.0101, "step": 99120 }, { "grad_norm": 0.327375590801239, "learning_rate": 2.073952606066465e-08, "loss": 0.0119, "step": 99130 }, { "grad_norm": 0.4626942276954651, "learning_rate": 2.026606849171997e-08, "loss": 0.0214, "step": 99140 }, { "grad_norm": 0.40249526500701904, "learning_rate": 1.9798076631333262e-08, "loss": 0.0158, "step": 99150 }, { "grad_norm": 0.40169456601142883, "learning_rate": 1.933555053069136e-08, "loss": 0.0154, "step": 99160 }, { "grad_norm": 0.5651955604553223, "learning_rate": 1.887849024036492e-08, "loss": 0.0225, "step": 99170 }, { "grad_norm": 0.33589649200439453, "learning_rate": 1.8426895810341736e-08, "loss": 0.0125, "step": 99180 }, { "grad_norm": 0.43910789489746094, "learning_rate": 1.798076729000453e-08, "loss": 0.0148, "step": 99190 }, { "grad_norm": 0.31972816586494446, "learning_rate": 1.75401047281476e-08, "loss": 0.0181, "step": 99200 }, { "grad_norm": 0.36122190952301025, "learning_rate": 1.7104908172954626e-08, "loss": 0.0163, "step": 99210 }, { "grad_norm": 0.6440877914428711, "learning_rate": 1.6675177672020868e-08, "loss": 0.0224, "step": 99220 }, { "grad_norm": 0.25510475039482117, "learning_rate": 1.6250913272342073e-08, "loss": 0.0172, "step": 99230 }, { "grad_norm": 0.40315043926239014, "learning_rate": 1.5832115020314453e-08, "loss": 0.013, "step": 99240 }, { "grad_norm": 0.3960948586463928, "learning_rate": 1.5418782961734712e-08, "loss": 0.0172, "step": 99250 }, { "grad_norm": 0.42038100957870483, "learning_rate": 1.501091714181113e-08, "loss": 0.0168, "step": 99260 }, { "grad_norm": 0.16033710539340973, "learning_rate": 1.460851760513582e-08, "loss": 0.0089, "step": 99270 }, { "grad_norm": 0.3839127719402313, "learning_rate": 1.4211584395723566e-08, "loss": 0.0178, "step": 99280 }, { "grad_norm": 0.31142905354499817, "learning_rate": 1.382011755697854e-08, "loss": 0.0112, "step": 99290 }, { "grad_norm": 0.1749284863471985, "learning_rate": 1.3434117131716494e-08, "loss": 0.012, "step": 99300 }, { "grad_norm": 0.30604517459869385, "learning_rate": 1.3053583162142557e-08, "loss": 0.0152, "step": 99310 }, { "grad_norm": 0.08812572062015533, "learning_rate": 1.2678515689873438e-08, "loss": 0.0149, "step": 99320 }, { "grad_norm": 0.1973288208246231, "learning_rate": 1.2308914755931877e-08, "loss": 0.0131, "step": 99330 }, { "grad_norm": 0.5070013999938965, "learning_rate": 1.1944780400729993e-08, "loss": 0.0204, "step": 99340 }, { "grad_norm": 0.4714148938655853, "learning_rate": 1.1586112664085935e-08, "loss": 0.012, "step": 99350 }, { "grad_norm": 0.30465200543403625, "learning_rate": 1.1232911585234984e-08, "loss": 0.0128, "step": 99360 }, { "grad_norm": 0.2705930769443512, "learning_rate": 1.0885177202790697e-08, "loss": 0.0181, "step": 99370 }, { "grad_norm": 0.7046669721603394, "learning_rate": 1.0542909554789315e-08, "loss": 0.0161, "step": 99380 }, { "grad_norm": 0.4288870692253113, "learning_rate": 1.0206108678656456e-08, "loss": 0.0209, "step": 99390 }, { "grad_norm": 0.21874651312828064, "learning_rate": 9.874774611223769e-09, "loss": 0.0086, "step": 99400 }, { "grad_norm": 0.10065710544586182, "learning_rate": 9.548907388728933e-09, "loss": 0.0162, "step": 99410 }, { "grad_norm": 0.29062509536743164, "learning_rate": 9.228507046804557e-09, "loss": 0.0114, "step": 99420 }, { "grad_norm": 0.5787879228591919, "learning_rate": 8.91357362048928e-09, "loss": 0.0136, "step": 99430 }, { "grad_norm": 0.25678321719169617, "learning_rate": 8.604107144227769e-09, "loss": 0.0306, "step": 99440 }, { "grad_norm": 0.13946594297885895, "learning_rate": 8.300107651859623e-09, "loss": 0.0151, "step": 99450 }, { "grad_norm": 0.13789576292037964, "learning_rate": 8.001575176630472e-09, "loss": 0.0126, "step": 99460 }, { "grad_norm": 0.5298886895179749, "learning_rate": 7.708509751186422e-09, "loss": 0.0161, "step": 99470 }, { "grad_norm": 0.41172757744789124, "learning_rate": 7.420911407579611e-09, "loss": 0.0158, "step": 99480 }, { "grad_norm": 0.3949427306652069, "learning_rate": 7.13878017725711e-09, "loss": 0.0159, "step": 99490 }, { "grad_norm": 0.26106196641921997, "learning_rate": 6.8621160910720125e-09, "loss": 0.0151, "step": 99500 }, { "grad_norm": 0.3634512424468994, "learning_rate": 6.59091917928345e-09, "loss": 0.0185, "step": 99510 }, { "grad_norm": 0.27659857273101807, "learning_rate": 6.32518947155103e-09, "loss": 0.0094, "step": 99520 }, { "grad_norm": 0.2274625301361084, "learning_rate": 6.064926996929288e-09, "loss": 0.0128, "step": 99530 }, { "grad_norm": 0.08365204930305481, "learning_rate": 5.810131783884343e-09, "loss": 0.016, "step": 99540 }, { "grad_norm": 0.4079861640930176, "learning_rate": 5.5608038602772415e-09, "loss": 0.0121, "step": 99550 }, { "grad_norm": 0.30562639236450195, "learning_rate": 5.316943253375062e-09, "loss": 0.013, "step": 99560 }, { "grad_norm": 0.1661609411239624, "learning_rate": 5.07854998984536e-09, "loss": 0.0154, "step": 99570 }, { "grad_norm": 0.40448322892189026, "learning_rate": 4.845624095756174e-09, "loss": 0.0134, "step": 99580 }, { "grad_norm": 0.800746500492096, "learning_rate": 4.618165596587121e-09, "loss": 0.0227, "step": 99590 }, { "grad_norm": 0.12634624540805817, "learning_rate": 4.396174517207197e-09, "loss": 0.0133, "step": 99600 }, { "grad_norm": 0.2265925109386444, "learning_rate": 4.179650881896979e-09, "loss": 0.0202, "step": 99610 }, { "grad_norm": 0.2997151017189026, "learning_rate": 3.968594714331975e-09, "loss": 0.0102, "step": 99620 }, { "grad_norm": 0.22388939559459686, "learning_rate": 3.763006037593719e-09, "loss": 0.0169, "step": 99630 }, { "grad_norm": 0.2798932194709778, "learning_rate": 3.5628848741586786e-09, "loss": 0.0132, "step": 99640 }, { "grad_norm": 0.41193246841430664, "learning_rate": 3.368231245926001e-09, "loss": 0.0191, "step": 99650 }, { "grad_norm": 0.3095865249633789, "learning_rate": 3.179045174167561e-09, "loss": 0.0122, "step": 99660 }, { "grad_norm": 0.263746440410614, "learning_rate": 2.9953266795834657e-09, "loss": 0.0143, "step": 99670 }, { "grad_norm": 0.17709213495254517, "learning_rate": 2.817075782263201e-09, "loss": 0.022, "step": 99680 }, { "grad_norm": 0.29657822847366333, "learning_rate": 2.6442925016967325e-09, "loss": 0.0237, "step": 99690 }, { "grad_norm": 1.587585210800171, "learning_rate": 2.4769768567800555e-09, "loss": 0.0139, "step": 99700 }, { "grad_norm": 0.33025771379470825, "learning_rate": 2.315128865809646e-09, "loss": 0.0153, "step": 99710 }, { "grad_norm": 0.476405531167984, "learning_rate": 2.1587485464880097e-09, "loss": 0.0125, "step": 99720 }, { "grad_norm": 0.24942903220653534, "learning_rate": 2.0078359159125813e-09, "loss": 0.0127, "step": 99730 }, { "grad_norm": 0.9227271676063538, "learning_rate": 1.8623909905923776e-09, "loss": 0.015, "step": 99740 }, { "grad_norm": 0.23698343336582184, "learning_rate": 1.7224137864257916e-09, "loss": 0.0135, "step": 99750 }, { "grad_norm": 0.2585342228412628, "learning_rate": 1.5879043187283505e-09, "loss": 0.0208, "step": 99760 }, { "grad_norm": 0.10388155281543732, "learning_rate": 1.458862602204958e-09, "loss": 0.0206, "step": 99770 }, { "grad_norm": 0.2390098124742508, "learning_rate": 1.3352886509720996e-09, "loss": 0.0149, "step": 99780 }, { "grad_norm": 0.4287290573120117, "learning_rate": 1.2171824785356389e-09, "loss": 0.0141, "step": 99790 }, { "grad_norm": 0.24029695987701416, "learning_rate": 1.1045440978185718e-09, "loss": 0.015, "step": 99800 }, { "grad_norm": 0.17500317096710205, "learning_rate": 9.973735211332714e-10, "loss": 0.0118, "step": 99810 }, { "grad_norm": 0.3526921570301056, "learning_rate": 8.956707602036929e-10, "loss": 0.012, "step": 99820 }, { "grad_norm": 0.430417537689209, "learning_rate": 7.994358261542712e-10, "loss": 0.0148, "step": 99830 }, { "grad_norm": 0.26226019859313965, "learning_rate": 7.086687295043692e-10, "loss": 0.0171, "step": 99840 }, { "grad_norm": 0.21643958985805511, "learning_rate": 6.23369480179381e-10, "loss": 0.0091, "step": 99850 }, { "grad_norm": 0.24371793866157532, "learning_rate": 5.43538087510731e-10, "loss": 0.0125, "step": 99860 }, { "grad_norm": 0.5204387903213501, "learning_rate": 4.691745602303233e-10, "loss": 0.0186, "step": 99870 }, { "grad_norm": 0.14362774789333344, "learning_rate": 4.0027890646499034e-10, "loss": 0.0114, "step": 99880 }, { "grad_norm": 0.19254660606384277, "learning_rate": 3.368511337531466e-10, "loss": 0.0204, "step": 99890 }, { "grad_norm": 0.5837404131889343, "learning_rate": 2.788912490281348e-10, "loss": 0.0161, "step": 99900 }, { "grad_norm": 0.3719014525413513, "learning_rate": 2.2639925863487954e-10, "loss": 0.016, "step": 99910 }, { "grad_norm": 0.3009917736053467, "learning_rate": 1.7937516830213163e-10, "loss": 0.0176, "step": 99920 }, { "grad_norm": 0.6457886695861816, "learning_rate": 1.3781898318687702e-10, "loss": 0.0179, "step": 99930 }, { "grad_norm": 0.2569724917411804, "learning_rate": 1.0173070781882566e-10, "loss": 0.0123, "step": 99940 }, { "grad_norm": 0.3622763752937317, "learning_rate": 7.111034615592261e-11, "loss": 0.0189, "step": 99950 }, { "grad_norm": 0.41135019063949585, "learning_rate": 4.595790153993918e-11, "loss": 0.0209, "step": 99960 }, { "grad_norm": 0.39174672961235046, "learning_rate": 2.6273376729779585e-11, "loss": 0.0118, "step": 99970 }, { "grad_norm": 0.3653678297996521, "learning_rate": 1.2056773868174276e-11, "loss": 0.0126, "step": 99980 }, { "grad_norm": 0.3133258521556854, "learning_rate": 3.3080945149865926e-12, "loss": 0.0122, "step": 99990 }, { "grad_norm": 0.2665678560733795, "learning_rate": 2.7339625008337977e-14, "loss": 0.0184, "step": 100000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 5000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 4, "trial_name": null, "trial_params": null }