{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 20000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.3594739437103271, "learning_rate": 9e-07, "loss": 1.1913, "step": 10 }, { "grad_norm": 1.0572824478149414, "learning_rate": 1.9e-06, "loss": 1.1841, "step": 20 }, { "grad_norm": 0.5717663764953613, "learning_rate": 2.9e-06, "loss": 1.1508, "step": 30 }, { "grad_norm": 0.3898443877696991, "learning_rate": 3.9e-06, "loss": 1.1205, "step": 40 }, { "grad_norm": 0.28664326667785645, "learning_rate": 4.9000000000000005e-06, "loss": 1.0888, "step": 50 }, { "grad_norm": 0.1729290783405304, "learning_rate": 5.9e-06, "loss": 1.0782, "step": 60 }, { "grad_norm": 0.17002208530902863, "learning_rate": 6.900000000000001e-06, "loss": 1.0691, "step": 70 }, { "grad_norm": 0.2152942717075348, "learning_rate": 7.9e-06, "loss": 1.0562, "step": 80 }, { "grad_norm": 0.19103780388832092, "learning_rate": 8.9e-06, "loss": 1.0479, "step": 90 }, { "grad_norm": 0.3243984878063202, "learning_rate": 9.900000000000002e-06, "loss": 1.0372, "step": 100 }, { "grad_norm": 0.1820673942565918, "learning_rate": 1.09e-05, "loss": 1.0272, "step": 110 }, { "grad_norm": 0.21819084882736206, "learning_rate": 1.19e-05, "loss": 1.0236, "step": 120 }, { "grad_norm": 0.20377595722675323, "learning_rate": 1.29e-05, "loss": 1.0237, "step": 130 }, { "grad_norm": 0.20572194457054138, "learning_rate": 1.3900000000000002e-05, "loss": 1.0228, "step": 140 }, { "grad_norm": 0.20157840847969055, "learning_rate": 1.49e-05, "loss": 1.0217, "step": 150 }, { "grad_norm": 0.23459017276763916, "learning_rate": 1.59e-05, "loss": 1.0192, "step": 160 }, { "grad_norm": 0.32469043135643005, "learning_rate": 1.69e-05, "loss": 1.0063, "step": 170 }, { "grad_norm": 0.36008527874946594, "learning_rate": 1.79e-05, "loss": 0.9873, "step": 180 }, { "grad_norm": 0.5633573532104492, "learning_rate": 1.8900000000000002e-05, "loss": 0.9672, "step": 190 }, { "grad_norm": 0.7019369006156921, "learning_rate": 1.9900000000000003e-05, "loss": 0.9315, "step": 200 }, { "grad_norm": 0.5538105964660645, "learning_rate": 2.09e-05, "loss": 0.8958, "step": 210 }, { "grad_norm": 0.5306029319763184, "learning_rate": 2.19e-05, "loss": 0.8707, "step": 220 }, { "grad_norm": 0.6606974005699158, "learning_rate": 2.29e-05, "loss": 0.8479, "step": 230 }, { "grad_norm": 0.8058410882949829, "learning_rate": 2.39e-05, "loss": 0.8169, "step": 240 }, { "grad_norm": 0.7277475595474243, "learning_rate": 2.4900000000000002e-05, "loss": 0.77, "step": 250 }, { "grad_norm": 0.6617355942726135, "learning_rate": 2.5900000000000003e-05, "loss": 0.7456, "step": 260 }, { "grad_norm": 0.8156651258468628, "learning_rate": 2.6900000000000003e-05, "loss": 0.6984, "step": 270 }, { "grad_norm": 0.7090954780578613, "learning_rate": 2.7900000000000004e-05, "loss": 0.6774, "step": 280 }, { "grad_norm": 0.8667084574699402, "learning_rate": 2.8899999999999998e-05, "loss": 0.6429, "step": 290 }, { "grad_norm": 0.946596622467041, "learning_rate": 2.9900000000000002e-05, "loss": 0.6052, "step": 300 }, { "grad_norm": 0.8120863437652588, "learning_rate": 3.09e-05, "loss": 0.5681, "step": 310 }, { "grad_norm": 0.9630921483039856, "learning_rate": 3.19e-05, "loss": 0.5267, "step": 320 }, { "grad_norm": 0.9185823798179626, "learning_rate": 3.29e-05, "loss": 0.497, "step": 330 }, { "grad_norm": 0.9909350872039795, "learning_rate": 3.3900000000000004e-05, "loss": 0.4704, "step": 340 }, { "grad_norm": 0.7408623695373535, "learning_rate": 3.49e-05, "loss": 0.4463, "step": 350 }, { "grad_norm": 0.8417967557907104, "learning_rate": 3.59e-05, "loss": 0.4515, "step": 360 }, { "grad_norm": 0.9200495481491089, "learning_rate": 3.69e-05, "loss": 0.417, "step": 370 }, { "grad_norm": 1.146302342414856, "learning_rate": 3.79e-05, "loss": 0.3937, "step": 380 }, { "grad_norm": 1.0057293176651, "learning_rate": 3.8900000000000004e-05, "loss": 0.3773, "step": 390 }, { "grad_norm": 1.112216591835022, "learning_rate": 3.99e-05, "loss": 0.348, "step": 400 }, { "grad_norm": 1.0176512002944946, "learning_rate": 4.09e-05, "loss": 0.3392, "step": 410 }, { "grad_norm": 1.0310163497924805, "learning_rate": 4.19e-05, "loss": 0.3065, "step": 420 }, { "grad_norm": 1.022374153137207, "learning_rate": 4.29e-05, "loss": 0.2808, "step": 430 }, { "grad_norm": 1.368080735206604, "learning_rate": 4.39e-05, "loss": 0.2624, "step": 440 }, { "grad_norm": 1.1092591285705566, "learning_rate": 4.49e-05, "loss": 0.2405, "step": 450 }, { "grad_norm": 0.9738430380821228, "learning_rate": 4.5900000000000004e-05, "loss": 0.2254, "step": 460 }, { "grad_norm": 1.033246636390686, "learning_rate": 4.69e-05, "loss": 0.2162, "step": 470 }, { "grad_norm": 0.9855560064315796, "learning_rate": 4.79e-05, "loss": 0.2088, "step": 480 }, { "grad_norm": 1.0313360691070557, "learning_rate": 4.89e-05, "loss": 0.2188, "step": 490 }, { "grad_norm": 1.100176215171814, "learning_rate": 4.99e-05, "loss": 0.2007, "step": 500 }, { "grad_norm": 1.0784265995025635, "learning_rate": 5.0900000000000004e-05, "loss": 0.2016, "step": 510 }, { "grad_norm": 1.0822303295135498, "learning_rate": 5.19e-05, "loss": 0.1961, "step": 520 }, { "grad_norm": 1.067589282989502, "learning_rate": 5.2900000000000005e-05, "loss": 0.1801, "step": 530 }, { "grad_norm": 1.1917147636413574, "learning_rate": 5.390000000000001e-05, "loss": 0.1705, "step": 540 }, { "grad_norm": 1.3141072988510132, "learning_rate": 5.4900000000000006e-05, "loss": 0.1851, "step": 550 }, { "grad_norm": 1.002855658531189, "learning_rate": 5.590000000000001e-05, "loss": 0.1663, "step": 560 }, { "grad_norm": 1.167011022567749, "learning_rate": 5.69e-05, "loss": 0.1741, "step": 570 }, { "grad_norm": 1.0936863422393799, "learning_rate": 5.79e-05, "loss": 0.1661, "step": 580 }, { "grad_norm": 0.9669778347015381, "learning_rate": 5.89e-05, "loss": 0.1648, "step": 590 }, { "grad_norm": 0.9405611753463745, "learning_rate": 5.99e-05, "loss": 0.1627, "step": 600 }, { "grad_norm": 1.0284767150878906, "learning_rate": 6.09e-05, "loss": 0.1496, "step": 610 }, { "grad_norm": 1.1097605228424072, "learning_rate": 6.19e-05, "loss": 0.1628, "step": 620 }, { "grad_norm": 0.9104214310646057, "learning_rate": 6.29e-05, "loss": 0.1302, "step": 630 }, { "grad_norm": 0.8578998446464539, "learning_rate": 6.390000000000001e-05, "loss": 0.1326, "step": 640 }, { "grad_norm": 1.1287304162979126, "learning_rate": 6.49e-05, "loss": 0.1127, "step": 650 }, { "grad_norm": 0.8655268549919128, "learning_rate": 6.59e-05, "loss": 0.1202, "step": 660 }, { "grad_norm": 0.9937160015106201, "learning_rate": 6.690000000000001e-05, "loss": 0.1198, "step": 670 }, { "grad_norm": 0.9691420197486877, "learning_rate": 6.790000000000001e-05, "loss": 0.1096, "step": 680 }, { "grad_norm": 1.0945252180099487, "learning_rate": 6.89e-05, "loss": 0.105, "step": 690 }, { "grad_norm": 1.0388752222061157, "learning_rate": 6.99e-05, "loss": 0.1027, "step": 700 }, { "grad_norm": 0.881949245929718, "learning_rate": 7.09e-05, "loss": 0.1044, "step": 710 }, { "grad_norm": 0.8678519129753113, "learning_rate": 7.19e-05, "loss": 0.0842, "step": 720 }, { "grad_norm": 1.2314260005950928, "learning_rate": 7.29e-05, "loss": 0.0841, "step": 730 }, { "grad_norm": 0.7337191700935364, "learning_rate": 7.390000000000001e-05, "loss": 0.0771, "step": 740 }, { "grad_norm": 1.194354772567749, "learning_rate": 7.49e-05, "loss": 0.0791, "step": 750 }, { "grad_norm": 1.0703870058059692, "learning_rate": 7.59e-05, "loss": 0.0697, "step": 760 }, { "grad_norm": 0.9820927977561951, "learning_rate": 7.69e-05, "loss": 0.0798, "step": 770 }, { "grad_norm": 1.099042534828186, "learning_rate": 7.790000000000001e-05, "loss": 0.0736, "step": 780 }, { "grad_norm": 0.9056155681610107, "learning_rate": 7.890000000000001e-05, "loss": 0.0756, "step": 790 }, { "grad_norm": 0.8292648792266846, "learning_rate": 7.99e-05, "loss": 0.0796, "step": 800 }, { "grad_norm": 0.9507290720939636, "learning_rate": 8.090000000000001e-05, "loss": 0.0829, "step": 810 }, { "grad_norm": 0.9466397762298584, "learning_rate": 8.19e-05, "loss": 0.0688, "step": 820 }, { "grad_norm": 0.7956731915473938, "learning_rate": 8.29e-05, "loss": 0.0747, "step": 830 }, { "grad_norm": 0.7995853424072266, "learning_rate": 8.39e-05, "loss": 0.0634, "step": 840 }, { "grad_norm": 0.7665478587150574, "learning_rate": 8.49e-05, "loss": 0.0661, "step": 850 }, { "grad_norm": 0.9283880591392517, "learning_rate": 8.59e-05, "loss": 0.0702, "step": 860 }, { "grad_norm": 1.126967191696167, "learning_rate": 8.69e-05, "loss": 0.0716, "step": 870 }, { "grad_norm": 0.8662194609642029, "learning_rate": 8.790000000000001e-05, "loss": 0.0667, "step": 880 }, { "grad_norm": 0.9572857022285461, "learning_rate": 8.89e-05, "loss": 0.0791, "step": 890 }, { "grad_norm": 0.9036967158317566, "learning_rate": 8.99e-05, "loss": 0.0745, "step": 900 }, { "grad_norm": 0.7550048828125, "learning_rate": 9.090000000000001e-05, "loss": 0.0746, "step": 910 }, { "grad_norm": 0.9990408420562744, "learning_rate": 9.190000000000001e-05, "loss": 0.0648, "step": 920 }, { "grad_norm": 0.8286410570144653, "learning_rate": 9.290000000000001e-05, "loss": 0.0697, "step": 930 }, { "grad_norm": 0.9783310890197754, "learning_rate": 9.39e-05, "loss": 0.0749, "step": 940 }, { "grad_norm": 0.9899768233299255, "learning_rate": 9.49e-05, "loss": 0.0722, "step": 950 }, { "grad_norm": 0.7450554370880127, "learning_rate": 9.59e-05, "loss": 0.0599, "step": 960 }, { "grad_norm": 0.7791635394096375, "learning_rate": 9.69e-05, "loss": 0.0654, "step": 970 }, { "grad_norm": 0.7614015340805054, "learning_rate": 9.790000000000001e-05, "loss": 0.0558, "step": 980 }, { "grad_norm": 0.9096309542655945, "learning_rate": 9.89e-05, "loss": 0.0581, "step": 990 }, { "grad_norm": 0.668950080871582, "learning_rate": 9.99e-05, "loss": 0.0652, "step": 1000 }, { "grad_norm": 0.8658283948898315, "learning_rate": 9.999994463727085e-05, "loss": 0.0529, "step": 1010 }, { "grad_norm": 0.7495288848876953, "learning_rate": 9.999975326009292e-05, "loss": 0.059, "step": 1020 }, { "grad_norm": 0.9980189204216003, "learning_rate": 9.999942518549879e-05, "loss": 0.0638, "step": 1030 }, { "grad_norm": 0.7826606035232544, "learning_rate": 9.999896041438544e-05, "loss": 0.0546, "step": 1040 }, { "grad_norm": 0.6360778212547302, "learning_rate": 9.999835894802353e-05, "loss": 0.054, "step": 1050 }, { "grad_norm": 0.7757160067558289, "learning_rate": 9.999762078805743e-05, "loss": 0.0591, "step": 1060 }, { "grad_norm": 0.7390689849853516, "learning_rate": 9.999674593650526e-05, "loss": 0.0595, "step": 1070 }, { "grad_norm": 0.6460424065589905, "learning_rate": 9.99957343957588e-05, "loss": 0.0658, "step": 1080 }, { "grad_norm": 0.8082983493804932, "learning_rate": 9.99945861685836e-05, "loss": 0.0596, "step": 1090 }, { "grad_norm": 0.7415626645088196, "learning_rate": 9.999330125811884e-05, "loss": 0.0483, "step": 1100 }, { "grad_norm": 0.8829818367958069, "learning_rate": 9.999187966787744e-05, "loss": 0.0619, "step": 1110 }, { "grad_norm": 0.8239393830299377, "learning_rate": 9.999032140174595e-05, "loss": 0.0528, "step": 1120 }, { "grad_norm": 0.8529507517814636, "learning_rate": 9.998862646398464e-05, "loss": 0.0654, "step": 1130 }, { "grad_norm": 0.7502208948135376, "learning_rate": 9.998679485922739e-05, "loss": 0.0526, "step": 1140 }, { "grad_norm": 0.6970030069351196, "learning_rate": 9.998482659248174e-05, "loss": 0.0547, "step": 1150 }, { "grad_norm": 0.9376399517059326, "learning_rate": 9.998272166912883e-05, "loss": 0.0557, "step": 1160 }, { "grad_norm": 0.7249330282211304, "learning_rate": 9.998048009492347e-05, "loss": 0.0504, "step": 1170 }, { "grad_norm": 0.8968970775604248, "learning_rate": 9.997810187599403e-05, "loss": 0.0526, "step": 1180 }, { "grad_norm": 0.7676458358764648, "learning_rate": 9.997558701884249e-05, "loss": 0.0506, "step": 1190 }, { "grad_norm": 0.6501711010932922, "learning_rate": 9.997293553034433e-05, "loss": 0.061, "step": 1200 }, { "grad_norm": 0.677116870880127, "learning_rate": 9.997014741774866e-05, "loss": 0.0462, "step": 1210 }, { "grad_norm": 0.8147766590118408, "learning_rate": 9.996722268867803e-05, "loss": 0.0486, "step": 1220 }, { "grad_norm": 0.706069827079773, "learning_rate": 9.996416135112858e-05, "loss": 0.0511, "step": 1230 }, { "grad_norm": 0.6159539818763733, "learning_rate": 9.996096341346988e-05, "loss": 0.0492, "step": 1240 }, { "grad_norm": 0.6369336843490601, "learning_rate": 9.995762888444495e-05, "loss": 0.0479, "step": 1250 }, { "grad_norm": 0.7543830275535583, "learning_rate": 9.995415777317027e-05, "loss": 0.0493, "step": 1260 }, { "grad_norm": 0.7505154609680176, "learning_rate": 9.995055008913574e-05, "loss": 0.053, "step": 1270 }, { "grad_norm": 0.5397493243217468, "learning_rate": 9.994680584220463e-05, "loss": 0.0432, "step": 1280 }, { "grad_norm": 0.6707198619842529, "learning_rate": 9.994292504261355e-05, "loss": 0.0472, "step": 1290 }, { "grad_norm": 0.8792182803153992, "learning_rate": 9.993890770097247e-05, "loss": 0.0453, "step": 1300 }, { "grad_norm": 0.7324561476707458, "learning_rate": 9.993475382826467e-05, "loss": 0.0479, "step": 1310 }, { "grad_norm": 0.8385289907455444, "learning_rate": 9.993046343584664e-05, "loss": 0.0549, "step": 1320 }, { "grad_norm": 0.5908923745155334, "learning_rate": 9.992603653544816e-05, "loss": 0.0483, "step": 1330 }, { "grad_norm": 0.63700932264328, "learning_rate": 9.992147313917222e-05, "loss": 0.0485, "step": 1340 }, { "grad_norm": 0.7525864839553833, "learning_rate": 9.991677325949497e-05, "loss": 0.0469, "step": 1350 }, { "grad_norm": 0.5628486275672913, "learning_rate": 9.991193690926568e-05, "loss": 0.0459, "step": 1360 }, { "grad_norm": 0.795554518699646, "learning_rate": 9.990696410170678e-05, "loss": 0.0467, "step": 1370 }, { "grad_norm": 0.7957155704498291, "learning_rate": 9.990185485041371e-05, "loss": 0.0481, "step": 1380 }, { "grad_norm": 0.5773254632949829, "learning_rate": 9.989660916935498e-05, "loss": 0.0471, "step": 1390 }, { "grad_norm": 0.6150880455970764, "learning_rate": 9.989122707287208e-05, "loss": 0.0426, "step": 1400 }, { "grad_norm": 0.7106145620346069, "learning_rate": 9.988570857567945e-05, "loss": 0.0537, "step": 1410 }, { "grad_norm": 0.9491516947746277, "learning_rate": 9.988005369286446e-05, "loss": 0.0525, "step": 1420 }, { "grad_norm": 0.6860232353210449, "learning_rate": 9.987426243988734e-05, "loss": 0.0429, "step": 1430 }, { "grad_norm": 0.7841853499412537, "learning_rate": 9.986833483258114e-05, "loss": 0.0524, "step": 1440 }, { "grad_norm": 0.6175568103790283, "learning_rate": 9.986227088715173e-05, "loss": 0.0385, "step": 1450 }, { "grad_norm": 0.5932314991950989, "learning_rate": 9.98560706201777e-05, "loss": 0.0408, "step": 1460 }, { "grad_norm": 0.7410153150558472, "learning_rate": 9.984973404861036e-05, "loss": 0.043, "step": 1470 }, { "grad_norm": 0.8330276608467102, "learning_rate": 9.984326118977361e-05, "loss": 0.051, "step": 1480 }, { "grad_norm": 0.7202706933021545, "learning_rate": 9.983665206136406e-05, "loss": 0.0493, "step": 1490 }, { "grad_norm": 0.574433445930481, "learning_rate": 9.982990668145075e-05, "loss": 0.0466, "step": 1500 }, { "grad_norm": 0.7351802587509155, "learning_rate": 9.982302506847534e-05, "loss": 0.057, "step": 1510 }, { "grad_norm": 0.819564163684845, "learning_rate": 9.981600724125189e-05, "loss": 0.0555, "step": 1520 }, { "grad_norm": 0.6065496206283569, "learning_rate": 9.980885321896685e-05, "loss": 0.0509, "step": 1530 }, { "grad_norm": 0.6572223901748657, "learning_rate": 9.980156302117905e-05, "loss": 0.044, "step": 1540 }, { "grad_norm": 0.6978927254676819, "learning_rate": 9.979413666781963e-05, "loss": 0.0465, "step": 1550 }, { "grad_norm": 0.5508580803871155, "learning_rate": 9.978657417919193e-05, "loss": 0.0452, "step": 1560 }, { "grad_norm": 0.5769541263580322, "learning_rate": 9.977887557597153e-05, "loss": 0.0475, "step": 1570 }, { "grad_norm": 0.5610742568969727, "learning_rate": 9.97710408792061e-05, "loss": 0.0469, "step": 1580 }, { "grad_norm": 0.5692776441574097, "learning_rate": 9.976307011031542e-05, "loss": 0.0449, "step": 1590 }, { "grad_norm": 0.5226185321807861, "learning_rate": 9.975496329109126e-05, "loss": 0.0476, "step": 1600 }, { "grad_norm": 0.7111744284629822, "learning_rate": 9.974672044369732e-05, "loss": 0.047, "step": 1610 }, { "grad_norm": 0.514858067035675, "learning_rate": 9.97383415906693e-05, "loss": 0.043, "step": 1620 }, { "grad_norm": 0.5856963396072388, "learning_rate": 9.97298267549146e-05, "loss": 0.0471, "step": 1630 }, { "grad_norm": 0.6191436052322388, "learning_rate": 9.972117595971249e-05, "loss": 0.0422, "step": 1640 }, { "grad_norm": 0.5670982599258423, "learning_rate": 9.971238922871391e-05, "loss": 0.0419, "step": 1650 }, { "grad_norm": 0.7190003991127014, "learning_rate": 9.970346658594142e-05, "loss": 0.0453, "step": 1660 }, { "grad_norm": 0.6552428007125854, "learning_rate": 9.969440805578923e-05, "loss": 0.046, "step": 1670 }, { "grad_norm": 0.578118622303009, "learning_rate": 9.968521366302298e-05, "loss": 0.0392, "step": 1680 }, { "grad_norm": 0.7054030895233154, "learning_rate": 9.967588343277981e-05, "loss": 0.0455, "step": 1690 }, { "grad_norm": 0.6531293392181396, "learning_rate": 9.966641739056818e-05, "loss": 0.0421, "step": 1700 }, { "grad_norm": 0.6111751198768616, "learning_rate": 9.965681556226793e-05, "loss": 0.0517, "step": 1710 }, { "grad_norm": 0.4928556978702545, "learning_rate": 9.964707797413006e-05, "loss": 0.044, "step": 1720 }, { "grad_norm": 0.6597058773040771, "learning_rate": 9.963720465277679e-05, "loss": 0.047, "step": 1730 }, { "grad_norm": 0.6202155351638794, "learning_rate": 9.96271956252014e-05, "loss": 0.0384, "step": 1740 }, { "grad_norm": 0.5262959599494934, "learning_rate": 9.961705091876816e-05, "loss": 0.0425, "step": 1750 }, { "grad_norm": 0.6935763955116272, "learning_rate": 9.960677056121235e-05, "loss": 0.0409, "step": 1760 }, { "grad_norm": 0.6149827837944031, "learning_rate": 9.959635458064005e-05, "loss": 0.0383, "step": 1770 }, { "grad_norm": 0.5901826024055481, "learning_rate": 9.958580300552815e-05, "loss": 0.0426, "step": 1780 }, { "grad_norm": 0.5597098469734192, "learning_rate": 9.957511586472426e-05, "loss": 0.0352, "step": 1790 }, { "grad_norm": 0.5581690073013306, "learning_rate": 9.956429318744662e-05, "loss": 0.0366, "step": 1800 }, { "grad_norm": 0.5969916582107544, "learning_rate": 9.955333500328404e-05, "loss": 0.0355, "step": 1810 }, { "grad_norm": 0.5474916696548462, "learning_rate": 9.95422413421957e-05, "loss": 0.0376, "step": 1820 }, { "grad_norm": 0.5651562809944153, "learning_rate": 9.953101223451133e-05, "loss": 0.0359, "step": 1830 }, { "grad_norm": 0.6243921518325806, "learning_rate": 9.951964771093085e-05, "loss": 0.0373, "step": 1840 }, { "grad_norm": 0.4624647796154022, "learning_rate": 9.950814780252442e-05, "loss": 0.0347, "step": 1850 }, { "grad_norm": 0.5893751382827759, "learning_rate": 9.949651254073236e-05, "loss": 0.0408, "step": 1860 }, { "grad_norm": 0.526287317276001, "learning_rate": 9.948474195736504e-05, "loss": 0.0388, "step": 1870 }, { "grad_norm": 0.6111840605735779, "learning_rate": 9.947283608460277e-05, "loss": 0.0346, "step": 1880 }, { "grad_norm": 0.46461328864097595, "learning_rate": 9.946079495499577e-05, "loss": 0.0411, "step": 1890 }, { "grad_norm": 0.610548734664917, "learning_rate": 9.944861860146401e-05, "loss": 0.0407, "step": 1900 }, { "grad_norm": 0.5339504480361938, "learning_rate": 9.943630705729719e-05, "loss": 0.0398, "step": 1910 }, { "grad_norm": 0.46559029817581177, "learning_rate": 9.942386035615459e-05, "loss": 0.039, "step": 1920 }, { "grad_norm": 0.7745798826217651, "learning_rate": 9.941127853206503e-05, "loss": 0.04, "step": 1930 }, { "grad_norm": 0.5811882019042969, "learning_rate": 9.939856161942673e-05, "loss": 0.0425, "step": 1940 }, { "grad_norm": 0.4856541156768799, "learning_rate": 9.938570965300724e-05, "loss": 0.0363, "step": 1950 }, { "grad_norm": 0.5952467918395996, "learning_rate": 9.937272266794335e-05, "loss": 0.0439, "step": 1960 }, { "grad_norm": 0.5669976472854614, "learning_rate": 9.935960069974096e-05, "loss": 0.05, "step": 1970 }, { "grad_norm": 0.5959198474884033, "learning_rate": 9.934634378427506e-05, "loss": 0.0382, "step": 1980 }, { "grad_norm": 0.520875096321106, "learning_rate": 9.933295195778954e-05, "loss": 0.0386, "step": 1990 }, { "grad_norm": 0.4351758360862732, "learning_rate": 9.931942525689715e-05, "loss": 0.0488, "step": 2000 }, { "grad_norm": 0.6345981359481812, "learning_rate": 9.930576371857936e-05, "loss": 0.0391, "step": 2010 }, { "grad_norm": 0.6230748295783997, "learning_rate": 9.929196738018629e-05, "loss": 0.0388, "step": 2020 }, { "grad_norm": 0.5425089001655579, "learning_rate": 9.927803627943662e-05, "loss": 0.0395, "step": 2030 }, { "grad_norm": 0.49332770705223083, "learning_rate": 9.926397045441744e-05, "loss": 0.039, "step": 2040 }, { "grad_norm": 0.6731558442115784, "learning_rate": 9.924976994358417e-05, "loss": 0.0427, "step": 2050 }, { "grad_norm": 0.5310463309288025, "learning_rate": 9.923543478576048e-05, "loss": 0.0474, "step": 2060 }, { "grad_norm": 0.548930823802948, "learning_rate": 9.922096502013813e-05, "loss": 0.0423, "step": 2070 }, { "grad_norm": 0.5744786262512207, "learning_rate": 9.92063606862769e-05, "loss": 0.0372, "step": 2080 }, { "grad_norm": 0.6390929222106934, "learning_rate": 9.919162182410453e-05, "loss": 0.0368, "step": 2090 }, { "grad_norm": 0.5252511501312256, "learning_rate": 9.917674847391645e-05, "loss": 0.038, "step": 2100 }, { "grad_norm": 0.5656434297561646, "learning_rate": 9.916174067637584e-05, "loss": 0.0333, "step": 2110 }, { "grad_norm": 0.5288258790969849, "learning_rate": 9.914659847251348e-05, "loss": 0.0406, "step": 2120 }, { "grad_norm": 0.5040147304534912, "learning_rate": 9.913132190372753e-05, "loss": 0.0369, "step": 2130 }, { "grad_norm": 0.5128138661384583, "learning_rate": 9.911591101178359e-05, "loss": 0.0368, "step": 2140 }, { "grad_norm": 0.4942684769630432, "learning_rate": 9.910036583881443e-05, "loss": 0.0334, "step": 2150 }, { "grad_norm": 0.5318565368652344, "learning_rate": 9.908468642731995e-05, "loss": 0.0325, "step": 2160 }, { "grad_norm": 0.5772367715835571, "learning_rate": 9.906887282016707e-05, "loss": 0.0344, "step": 2170 }, { "grad_norm": 0.5957911014556885, "learning_rate": 9.90529250605896e-05, "loss": 0.0368, "step": 2180 }, { "grad_norm": 0.6259480714797974, "learning_rate": 9.903684319218809e-05, "loss": 0.0375, "step": 2190 }, { "grad_norm": 0.691277801990509, "learning_rate": 9.902062725892976e-05, "loss": 0.0402, "step": 2200 }, { "grad_norm": 0.624859094619751, "learning_rate": 9.900427730514834e-05, "loss": 0.0316, "step": 2210 }, { "grad_norm": 0.46915674209594727, "learning_rate": 9.8987793375544e-05, "loss": 0.0352, "step": 2220 }, { "grad_norm": 0.5559591054916382, "learning_rate": 9.897117551518318e-05, "loss": 0.0353, "step": 2230 }, { "grad_norm": 0.47577548027038574, "learning_rate": 9.895442376949844e-05, "loss": 0.0395, "step": 2240 }, { "grad_norm": 0.7231595516204834, "learning_rate": 9.893753818428845e-05, "loss": 0.0442, "step": 2250 }, { "grad_norm": 0.4607575535774231, "learning_rate": 9.892051880571773e-05, "loss": 0.037, "step": 2260 }, { "grad_norm": 0.4901242256164551, "learning_rate": 9.890336568031663e-05, "loss": 0.0342, "step": 2270 }, { "grad_norm": 0.46413323283195496, "learning_rate": 9.888607885498113e-05, "loss": 0.0386, "step": 2280 }, { "grad_norm": 0.5028432607650757, "learning_rate": 9.886865837697275e-05, "loss": 0.0384, "step": 2290 }, { "grad_norm": 0.6079827547073364, "learning_rate": 9.88511042939184e-05, "loss": 0.0416, "step": 2300 }, { "grad_norm": 0.6189248561859131, "learning_rate": 9.883341665381028e-05, "loss": 0.0372, "step": 2310 }, { "grad_norm": 0.569456160068512, "learning_rate": 9.881559550500575e-05, "loss": 0.0317, "step": 2320 }, { "grad_norm": 0.5782006978988647, "learning_rate": 9.879764089622712e-05, "loss": 0.0363, "step": 2330 }, { "grad_norm": 0.6612024307250977, "learning_rate": 9.87795528765616e-05, "loss": 0.0386, "step": 2340 }, { "grad_norm": 0.45619797706604004, "learning_rate": 9.876133149546118e-05, "loss": 0.0385, "step": 2350 }, { "grad_norm": 0.4743977189064026, "learning_rate": 9.874297680274238e-05, "loss": 0.0384, "step": 2360 }, { "grad_norm": 0.5303918719291687, "learning_rate": 9.872448884858624e-05, "loss": 0.0364, "step": 2370 }, { "grad_norm": 0.5923212766647339, "learning_rate": 9.870586768353815e-05, "loss": 0.0366, "step": 2380 }, { "grad_norm": 0.5156052112579346, "learning_rate": 9.868711335850764e-05, "loss": 0.0412, "step": 2390 }, { "grad_norm": 0.4702778458595276, "learning_rate": 9.866822592476833e-05, "loss": 0.0353, "step": 2400 }, { "grad_norm": 0.4955006241798401, "learning_rate": 9.86492054339577e-05, "loss": 0.0356, "step": 2410 }, { "grad_norm": 0.4722374677658081, "learning_rate": 9.863005193807711e-05, "loss": 0.0328, "step": 2420 }, { "grad_norm": 0.5261074900627136, "learning_rate": 9.861076548949143e-05, "loss": 0.0314, "step": 2430 }, { "grad_norm": 0.43109720945358276, "learning_rate": 9.859134614092912e-05, "loss": 0.0306, "step": 2440 }, { "grad_norm": 0.5150691270828247, "learning_rate": 9.857179394548191e-05, "loss": 0.0331, "step": 2450 }, { "grad_norm": 0.413881778717041, "learning_rate": 9.855210895660477e-05, "loss": 0.0313, "step": 2460 }, { "grad_norm": 0.5778813362121582, "learning_rate": 9.853229122811568e-05, "loss": 0.0327, "step": 2470 }, { "grad_norm": 0.5499809980392456, "learning_rate": 9.851234081419559e-05, "loss": 0.0371, "step": 2480 }, { "grad_norm": 0.533755898475647, "learning_rate": 9.849225776938814e-05, "loss": 0.0347, "step": 2490 }, { "grad_norm": 0.5036794543266296, "learning_rate": 9.847204214859964e-05, "loss": 0.0365, "step": 2500 }, { "grad_norm": 0.4547636806964874, "learning_rate": 9.845169400709879e-05, "loss": 0.0284, "step": 2510 }, { "grad_norm": 0.4148177206516266, "learning_rate": 9.843121340051664e-05, "loss": 0.0338, "step": 2520 }, { "grad_norm": 0.4307814836502075, "learning_rate": 9.841060038484641e-05, "loss": 0.0401, "step": 2530 }, { "grad_norm": 0.5055217146873474, "learning_rate": 9.838985501644328e-05, "loss": 0.0413, "step": 2540 }, { "grad_norm": 0.5252987742424011, "learning_rate": 9.83689773520243e-05, "loss": 0.0334, "step": 2550 }, { "grad_norm": 0.5325053334236145, "learning_rate": 9.834796744866819e-05, "loss": 0.0339, "step": 2560 }, { "grad_norm": 0.5485632419586182, "learning_rate": 9.832682536381525e-05, "loss": 0.0354, "step": 2570 }, { "grad_norm": 0.5406777262687683, "learning_rate": 9.830555115526711e-05, "loss": 0.0368, "step": 2580 }, { "grad_norm": 0.37698280811309814, "learning_rate": 9.828414488118667e-05, "loss": 0.0336, "step": 2590 }, { "grad_norm": 0.5253736972808838, "learning_rate": 9.826260660009785e-05, "loss": 0.0337, "step": 2600 }, { "grad_norm": 0.482319176197052, "learning_rate": 9.824093637088547e-05, "loss": 0.0299, "step": 2610 }, { "grad_norm": 0.43845584988594055, "learning_rate": 9.821913425279514e-05, "loss": 0.032, "step": 2620 }, { "grad_norm": 0.4526597559452057, "learning_rate": 9.8197200305433e-05, "loss": 0.034, "step": 2630 }, { "grad_norm": 0.45589521527290344, "learning_rate": 9.817513458876564e-05, "loss": 0.0464, "step": 2640 }, { "grad_norm": 0.5381149649620056, "learning_rate": 9.815293716311987e-05, "loss": 0.0334, "step": 2650 }, { "grad_norm": 0.5279123187065125, "learning_rate": 9.813060808918262e-05, "loss": 0.0318, "step": 2660 }, { "grad_norm": 0.3532435894012451, "learning_rate": 9.810814742800069e-05, "loss": 0.0285, "step": 2670 }, { "grad_norm": 0.3765302896499634, "learning_rate": 9.808555524098074e-05, "loss": 0.0289, "step": 2680 }, { "grad_norm": 0.46037837862968445, "learning_rate": 9.806283158988887e-05, "loss": 0.0291, "step": 2690 }, { "grad_norm": 0.483735591173172, "learning_rate": 9.803997653685072e-05, "loss": 0.0392, "step": 2700 }, { "grad_norm": 0.45865148305892944, "learning_rate": 9.801699014435112e-05, "loss": 0.0393, "step": 2710 }, { "grad_norm": 0.4620376229286194, "learning_rate": 9.799387247523398e-05, "loss": 0.0352, "step": 2720 }, { "grad_norm": 0.41832435131073, "learning_rate": 9.797062359270215e-05, "loss": 0.0319, "step": 2730 }, { "grad_norm": 0.4439375400543213, "learning_rate": 9.794724356031715e-05, "loss": 0.0307, "step": 2740 }, { "grad_norm": 0.5037664771080017, "learning_rate": 9.792373244199913e-05, "loss": 0.0306, "step": 2750 }, { "grad_norm": 0.378164678812027, "learning_rate": 9.790009030202658e-05, "loss": 0.0313, "step": 2760 }, { "grad_norm": 0.5053073763847351, "learning_rate": 9.78763172050362e-05, "loss": 0.0295, "step": 2770 }, { "grad_norm": 0.4680381119251251, "learning_rate": 9.785241321602274e-05, "loss": 0.0277, "step": 2780 }, { "grad_norm": 0.4624013304710388, "learning_rate": 9.782837840033879e-05, "loss": 0.0288, "step": 2790 }, { "grad_norm": 0.5074241757392883, "learning_rate": 9.780421282369461e-05, "loss": 0.0292, "step": 2800 }, { "grad_norm": 0.4835506081581116, "learning_rate": 9.777991655215797e-05, "loss": 0.0294, "step": 2810 }, { "grad_norm": 0.5738292336463928, "learning_rate": 9.775548965215394e-05, "loss": 0.0295, "step": 2820 }, { "grad_norm": 0.5334445238113403, "learning_rate": 9.773093219046474e-05, "loss": 0.0293, "step": 2830 }, { "grad_norm": 0.4011390507221222, "learning_rate": 9.770624423422954e-05, "loss": 0.0291, "step": 2840 }, { "grad_norm": 0.41171419620513916, "learning_rate": 9.768142585094426e-05, "loss": 0.0302, "step": 2850 }, { "grad_norm": 0.46391263604164124, "learning_rate": 9.765647710846142e-05, "loss": 0.0405, "step": 2860 }, { "grad_norm": 0.5071845650672913, "learning_rate": 9.763139807498991e-05, "loss": 0.0285, "step": 2870 }, { "grad_norm": 0.4814237058162689, "learning_rate": 9.760618881909487e-05, "loss": 0.0317, "step": 2880 }, { "grad_norm": 0.5396919846534729, "learning_rate": 9.758084940969744e-05, "loss": 0.0316, "step": 2890 }, { "grad_norm": 0.5363779664039612, "learning_rate": 9.755537991607459e-05, "loss": 0.027, "step": 2900 }, { "grad_norm": 0.505138099193573, "learning_rate": 9.752978040785895e-05, "loss": 0.0354, "step": 2910 }, { "grad_norm": 0.5476271510124207, "learning_rate": 9.750405095503859e-05, "loss": 0.0299, "step": 2920 }, { "grad_norm": 0.5189036130905151, "learning_rate": 9.747819162795686e-05, "loss": 0.0331, "step": 2930 }, { "grad_norm": 0.45717042684555054, "learning_rate": 9.745220249731217e-05, "loss": 0.026, "step": 2940 }, { "grad_norm": 0.4337165355682373, "learning_rate": 9.742608363415781e-05, "loss": 0.0272, "step": 2950 }, { "grad_norm": 0.4811023771762848, "learning_rate": 9.739983510990176e-05, "loss": 0.0288, "step": 2960 }, { "grad_norm": 0.3455168902873993, "learning_rate": 9.737345699630647e-05, "loss": 0.0298, "step": 2970 }, { "grad_norm": 0.5057815313339233, "learning_rate": 9.734694936548869e-05, "loss": 0.0332, "step": 2980 }, { "grad_norm": 0.38619765639305115, "learning_rate": 9.732031228991932e-05, "loss": 0.0256, "step": 2990 }, { "grad_norm": 0.3297816514968872, "learning_rate": 9.729354584242302e-05, "loss": 0.0355, "step": 3000 }, { "grad_norm": 0.5174765586853027, "learning_rate": 9.726665009617832e-05, "loss": 0.0309, "step": 3010 }, { "grad_norm": 0.43245866894721985, "learning_rate": 9.723962512471714e-05, "loss": 0.033, "step": 3020 }, { "grad_norm": 0.516598105430603, "learning_rate": 9.72124710019247e-05, "loss": 0.03, "step": 3030 }, { "grad_norm": 0.48712822794914246, "learning_rate": 9.718518780203934e-05, "loss": 0.0322, "step": 3040 }, { "grad_norm": 0.3674415946006775, "learning_rate": 9.715777559965228e-05, "loss": 0.0319, "step": 3050 }, { "grad_norm": 0.4218079149723053, "learning_rate": 9.713023446970746e-05, "loss": 0.0255, "step": 3060 }, { "grad_norm": 0.4967867136001587, "learning_rate": 9.710256448750126e-05, "loss": 0.0311, "step": 3070 }, { "grad_norm": 0.497653067111969, "learning_rate": 9.707476572868235e-05, "loss": 0.0341, "step": 3080 }, { "grad_norm": 0.4222137928009033, "learning_rate": 9.704683826925149e-05, "loss": 0.0273, "step": 3090 }, { "grad_norm": 0.37705838680267334, "learning_rate": 9.701878218556129e-05, "loss": 0.036, "step": 3100 }, { "grad_norm": 0.5626199841499329, "learning_rate": 9.699059755431598e-05, "loss": 0.0331, "step": 3110 }, { "grad_norm": 0.46293774247169495, "learning_rate": 9.696228445257132e-05, "loss": 0.0277, "step": 3120 }, { "grad_norm": 0.42764750123023987, "learning_rate": 9.693384295773419e-05, "loss": 0.0327, "step": 3130 }, { "grad_norm": 0.4717363715171814, "learning_rate": 9.690527314756259e-05, "loss": 0.0339, "step": 3140 }, { "grad_norm": 0.458967387676239, "learning_rate": 9.687657510016527e-05, "loss": 0.0261, "step": 3150 }, { "grad_norm": 0.45871081948280334, "learning_rate": 9.684774889400161e-05, "loss": 0.0309, "step": 3160 }, { "grad_norm": 0.5132860541343689, "learning_rate": 9.681879460788135e-05, "loss": 0.0264, "step": 3170 }, { "grad_norm": 0.4729975461959839, "learning_rate": 9.67897123209644e-05, "loss": 0.0315, "step": 3180 }, { "grad_norm": 0.4921012818813324, "learning_rate": 9.676050211276062e-05, "loss": 0.035, "step": 3190 }, { "grad_norm": 0.4574073255062103, "learning_rate": 9.673116406312962e-05, "loss": 0.0284, "step": 3200 }, { "grad_norm": 0.48541590571403503, "learning_rate": 9.67016982522805e-05, "loss": 0.028, "step": 3210 }, { "grad_norm": 0.4924331307411194, "learning_rate": 9.667210476077164e-05, "loss": 0.028, "step": 3220 }, { "grad_norm": 0.5730510950088501, "learning_rate": 9.664238366951055e-05, "loss": 0.0288, "step": 3230 }, { "grad_norm": 0.5551027059555054, "learning_rate": 9.661253505975355e-05, "loss": 0.0269, "step": 3240 }, { "grad_norm": 0.4366356134414673, "learning_rate": 9.658255901310557e-05, "loss": 0.0301, "step": 3250 }, { "grad_norm": 0.5327138304710388, "learning_rate": 9.655245561152e-05, "loss": 0.0278, "step": 3260 }, { "grad_norm": 0.4516207277774811, "learning_rate": 9.65222249372984e-05, "loss": 0.0266, "step": 3270 }, { "grad_norm": 0.4709407687187195, "learning_rate": 9.649186707309026e-05, "loss": 0.0325, "step": 3280 }, { "grad_norm": 0.36673372983932495, "learning_rate": 9.646138210189283e-05, "loss": 0.0285, "step": 3290 }, { "grad_norm": 0.5308244824409485, "learning_rate": 9.643077010705087e-05, "loss": 0.0281, "step": 3300 }, { "grad_norm": 0.45568153262138367, "learning_rate": 9.640003117225637e-05, "loss": 0.0286, "step": 3310 }, { "grad_norm": 0.4082559049129486, "learning_rate": 9.636916538154846e-05, "loss": 0.0241, "step": 3320 }, { "grad_norm": 0.48012563586235046, "learning_rate": 9.633817281931296e-05, "loss": 0.0297, "step": 3330 }, { "grad_norm": 0.4177444875240326, "learning_rate": 9.630705357028242e-05, "loss": 0.032, "step": 3340 }, { "grad_norm": 0.48793429136276245, "learning_rate": 9.627580771953563e-05, "loss": 0.0285, "step": 3350 }, { "grad_norm": 0.4371464252471924, "learning_rate": 9.624443535249759e-05, "loss": 0.0275, "step": 3360 }, { "grad_norm": 0.4983312487602234, "learning_rate": 9.621293655493913e-05, "loss": 0.0254, "step": 3370 }, { "grad_norm": 0.5624396204948425, "learning_rate": 9.618131141297675e-05, "loss": 0.027, "step": 3380 }, { "grad_norm": 0.43570947647094727, "learning_rate": 9.614956001307242e-05, "loss": 0.0301, "step": 3390 }, { "grad_norm": 0.4448493719100952, "learning_rate": 9.611768244203321e-05, "loss": 0.0351, "step": 3400 }, { "grad_norm": 0.4213621914386749, "learning_rate": 9.60856787870112e-05, "loss": 0.0292, "step": 3410 }, { "grad_norm": 0.4154338836669922, "learning_rate": 9.605354913550318e-05, "loss": 0.0262, "step": 3420 }, { "grad_norm": 0.45102718472480774, "learning_rate": 9.602129357535037e-05, "loss": 0.0313, "step": 3430 }, { "grad_norm": 0.38145503401756287, "learning_rate": 9.598891219473825e-05, "loss": 0.027, "step": 3440 }, { "grad_norm": 0.41790488362312317, "learning_rate": 9.595640508219625e-05, "loss": 0.0291, "step": 3450 }, { "grad_norm": 0.4644753336906433, "learning_rate": 9.592377232659761e-05, "loss": 0.0249, "step": 3460 }, { "grad_norm": 0.4731713533401489, "learning_rate": 9.589101401715904e-05, "loss": 0.0263, "step": 3470 }, { "grad_norm": 0.42398542165756226, "learning_rate": 9.585813024344045e-05, "loss": 0.026, "step": 3480 }, { "grad_norm": 0.5419644117355347, "learning_rate": 9.58251210953449e-05, "loss": 0.0296, "step": 3490 }, { "grad_norm": 0.463670939207077, "learning_rate": 9.579198666311809e-05, "loss": 0.0238, "step": 3500 }, { "grad_norm": 0.39643239974975586, "learning_rate": 9.575872703734832e-05, "loss": 0.0292, "step": 3510 }, { "grad_norm": 0.3542700409889221, "learning_rate": 9.572534230896611e-05, "loss": 0.0231, "step": 3520 }, { "grad_norm": 0.43060752749443054, "learning_rate": 9.569183256924403e-05, "loss": 0.025, "step": 3530 }, { "grad_norm": 0.40233463048934937, "learning_rate": 9.565819790979646e-05, "loss": 0.0422, "step": 3540 }, { "grad_norm": 0.4497774839401245, "learning_rate": 9.562443842257925e-05, "loss": 0.029, "step": 3550 }, { "grad_norm": 0.5018470287322998, "learning_rate": 9.559055419988956e-05, "loss": 0.0283, "step": 3560 }, { "grad_norm": 0.47868454456329346, "learning_rate": 9.555654533436557e-05, "loss": 0.0349, "step": 3570 }, { "grad_norm": 0.4413691759109497, "learning_rate": 9.552241191898621e-05, "loss": 0.0238, "step": 3580 }, { "grad_norm": 0.40998080372810364, "learning_rate": 9.548815404707092e-05, "loss": 0.03, "step": 3590 }, { "grad_norm": 0.43824273347854614, "learning_rate": 9.545377181227942e-05, "loss": 0.0284, "step": 3600 }, { "grad_norm": 0.4570449888706207, "learning_rate": 9.541926530861145e-05, "loss": 0.0266, "step": 3610 }, { "grad_norm": 0.44766074419021606, "learning_rate": 9.538463463040645e-05, "loss": 0.0278, "step": 3620 }, { "grad_norm": 0.481611967086792, "learning_rate": 9.534987987234337e-05, "loss": 0.0277, "step": 3630 }, { "grad_norm": 0.4858357608318329, "learning_rate": 9.53150011294404e-05, "loss": 0.0265, "step": 3640 }, { "grad_norm": 0.40574368834495544, "learning_rate": 9.527999849705471e-05, "loss": 0.0297, "step": 3650 }, { "grad_norm": 0.4581122100353241, "learning_rate": 9.524487207088213e-05, "loss": 0.0224, "step": 3660 }, { "grad_norm": 0.4100882411003113, "learning_rate": 9.520962194695698e-05, "loss": 0.0239, "step": 3670 }, { "grad_norm": 0.40333643555641174, "learning_rate": 9.517424822165175e-05, "loss": 0.0238, "step": 3680 }, { "grad_norm": 0.5596145987510681, "learning_rate": 9.513875099167685e-05, "loss": 0.0245, "step": 3690 }, { "grad_norm": 0.5230712890625, "learning_rate": 9.510313035408035e-05, "loss": 0.0262, "step": 3700 }, { "grad_norm": 0.39155617356300354, "learning_rate": 9.506738640624775e-05, "loss": 0.0264, "step": 3710 }, { "grad_norm": 0.4129464328289032, "learning_rate": 9.50315192459016e-05, "loss": 0.0208, "step": 3720 }, { "grad_norm": 0.5159543752670288, "learning_rate": 9.499552897110136e-05, "loss": 0.0239, "step": 3730 }, { "grad_norm": 0.5178094506263733, "learning_rate": 9.495941568024304e-05, "loss": 0.0253, "step": 3740 }, { "grad_norm": 0.43580612540245056, "learning_rate": 9.492317947205904e-05, "loss": 0.0268, "step": 3750 }, { "grad_norm": 0.4596274495124817, "learning_rate": 9.488682044561775e-05, "loss": 0.0256, "step": 3760 }, { "grad_norm": 0.41573286056518555, "learning_rate": 9.485033870032335e-05, "loss": 0.0243, "step": 3770 }, { "grad_norm": 0.47876912355422974, "learning_rate": 9.481373433591556e-05, "loss": 0.0215, "step": 3780 }, { "grad_norm": 0.4741547703742981, "learning_rate": 9.47770074524693e-05, "loss": 0.027, "step": 3790 }, { "grad_norm": 0.4306631088256836, "learning_rate": 9.474015815039446e-05, "loss": 0.0277, "step": 3800 }, { "grad_norm": 0.46127429604530334, "learning_rate": 9.470318653043565e-05, "loss": 0.0273, "step": 3810 }, { "grad_norm": 0.5021414160728455, "learning_rate": 9.466609269367185e-05, "loss": 0.0263, "step": 3820 }, { "grad_norm": 0.5333779454231262, "learning_rate": 9.46288767415162e-05, "loss": 0.0234, "step": 3830 }, { "grad_norm": 0.4366990625858307, "learning_rate": 9.459153877571567e-05, "loss": 0.0225, "step": 3840 }, { "grad_norm": 0.4819251298904419, "learning_rate": 9.455407889835087e-05, "loss": 0.0238, "step": 3850 }, { "grad_norm": 0.3999616503715515, "learning_rate": 9.451649721183564e-05, "loss": 0.0234, "step": 3860 }, { "grad_norm": 0.37807697057724, "learning_rate": 9.447879381891692e-05, "loss": 0.0258, "step": 3870 }, { "grad_norm": 0.5266739130020142, "learning_rate": 9.444096882267428e-05, "loss": 0.0329, "step": 3880 }, { "grad_norm": 0.3961910903453827, "learning_rate": 9.440302232651988e-05, "loss": 0.0226, "step": 3890 }, { "grad_norm": 0.3786242604255676, "learning_rate": 9.436495443419795e-05, "loss": 0.024, "step": 3900 }, { "grad_norm": 0.4175941050052643, "learning_rate": 9.432676524978466e-05, "loss": 0.0219, "step": 3910 }, { "grad_norm": 0.44096827507019043, "learning_rate": 9.42884548776878e-05, "loss": 0.0253, "step": 3920 }, { "grad_norm": 0.41201087832450867, "learning_rate": 9.425002342264646e-05, "loss": 0.0223, "step": 3930 }, { "grad_norm": 0.5009353160858154, "learning_rate": 9.421147098973077e-05, "loss": 0.0266, "step": 3940 }, { "grad_norm": 0.5505723357200623, "learning_rate": 9.41727976843416e-05, "loss": 0.0258, "step": 3950 }, { "grad_norm": 0.45981982350349426, "learning_rate": 9.413400361221029e-05, "loss": 0.0279, "step": 3960 }, { "grad_norm": 0.4804719388484955, "learning_rate": 9.409508887939835e-05, "loss": 0.022, "step": 3970 }, { "grad_norm": 0.4238436222076416, "learning_rate": 9.40560535922972e-05, "loss": 0.0212, "step": 3980 }, { "grad_norm": 0.403974324464798, "learning_rate": 9.40168978576278e-05, "loss": 0.0189, "step": 3990 }, { "grad_norm": 0.48837044835090637, "learning_rate": 9.397762178244043e-05, "loss": 0.0244, "step": 4000 }, { "grad_norm": 0.48128196597099304, "learning_rate": 9.393822547411439e-05, "loss": 0.0217, "step": 4010 }, { "grad_norm": 0.3272818624973297, "learning_rate": 9.389870904035769e-05, "loss": 0.0242, "step": 4020 }, { "grad_norm": 0.36953118443489075, "learning_rate": 9.385907258920672e-05, "loss": 0.0246, "step": 4030 }, { "grad_norm": 0.41161492466926575, "learning_rate": 9.381931622902607e-05, "loss": 0.021, "step": 4040 }, { "grad_norm": 0.4544064998626709, "learning_rate": 9.377944006850807e-05, "loss": 0.0193, "step": 4050 }, { "grad_norm": 0.47396498918533325, "learning_rate": 9.373944421667265e-05, "loss": 0.0213, "step": 4060 }, { "grad_norm": 0.4621795117855072, "learning_rate": 9.369932878286691e-05, "loss": 0.0266, "step": 4070 }, { "grad_norm": 0.5184421539306641, "learning_rate": 9.365909387676494e-05, "loss": 0.0196, "step": 4080 }, { "grad_norm": 0.4004800319671631, "learning_rate": 9.361873960836744e-05, "loss": 0.0263, "step": 4090 }, { "grad_norm": 0.3737598657608032, "learning_rate": 9.357826608800142e-05, "loss": 0.0196, "step": 4100 }, { "grad_norm": 0.4000731110572815, "learning_rate": 9.353767342631994e-05, "loss": 0.0203, "step": 4110 }, { "grad_norm": 0.3826330006122589, "learning_rate": 9.34969617343018e-05, "loss": 0.0219, "step": 4120 }, { "grad_norm": 0.5988262891769409, "learning_rate": 9.345613112325122e-05, "loss": 0.0204, "step": 4130 }, { "grad_norm": 0.4280189275741577, "learning_rate": 9.34151817047975e-05, "loss": 0.0224, "step": 4140 }, { "grad_norm": 0.3716961145401001, "learning_rate": 9.33741135908948e-05, "loss": 0.0262, "step": 4150 }, { "grad_norm": 0.4295980930328369, "learning_rate": 9.33329268938218e-05, "loss": 0.0207, "step": 4160 }, { "grad_norm": 0.425942063331604, "learning_rate": 9.329162172618132e-05, "loss": 0.0238, "step": 4170 }, { "grad_norm": 0.416522741317749, "learning_rate": 9.325019820090013e-05, "loss": 0.0226, "step": 4180 }, { "grad_norm": 0.5610533952713013, "learning_rate": 9.320865643122855e-05, "loss": 0.0208, "step": 4190 }, { "grad_norm": 0.379802942276001, "learning_rate": 9.316699653074023e-05, "loss": 0.022, "step": 4200 }, { "grad_norm": 0.4576219618320465, "learning_rate": 9.312521861333172e-05, "loss": 0.0166, "step": 4210 }, { "grad_norm": 0.45310190320014954, "learning_rate": 9.308332279322224e-05, "loss": 0.0242, "step": 4220 }, { "grad_norm": 0.4080248177051544, "learning_rate": 9.304130918495338e-05, "loss": 0.0224, "step": 4230 }, { "grad_norm": 0.33399489521980286, "learning_rate": 9.299917790338874e-05, "loss": 0.0187, "step": 4240 }, { "grad_norm": 0.356057733297348, "learning_rate": 9.295692906371363e-05, "loss": 0.0173, "step": 4250 }, { "grad_norm": 0.42619287967681885, "learning_rate": 9.291456278143476e-05, "loss": 0.0264, "step": 4260 }, { "grad_norm": 0.3479536175727844, "learning_rate": 9.287207917237994e-05, "loss": 0.0213, "step": 4270 }, { "grad_norm": 0.3362795114517212, "learning_rate": 9.282947835269773e-05, "loss": 0.0206, "step": 4280 }, { "grad_norm": 0.43236204981803894, "learning_rate": 9.278676043885715e-05, "loss": 0.0191, "step": 4290 }, { "grad_norm": 0.32585880160331726, "learning_rate": 9.274392554764733e-05, "loss": 0.0194, "step": 4300 }, { "grad_norm": 0.4723697900772095, "learning_rate": 9.270097379617723e-05, "loss": 0.016, "step": 4310 }, { "grad_norm": 0.42713454365730286, "learning_rate": 9.26579053018753e-05, "loss": 0.0154, "step": 4320 }, { "grad_norm": 0.33830246329307556, "learning_rate": 9.261472018248918e-05, "loss": 0.0146, "step": 4330 }, { "grad_norm": 0.4066753387451172, "learning_rate": 9.25714185560853e-05, "loss": 0.0259, "step": 4340 }, { "grad_norm": 0.448772668838501, "learning_rate": 9.252800054104868e-05, "loss": 0.0187, "step": 4350 }, { "grad_norm": 0.4219300448894501, "learning_rate": 9.248446625608252e-05, "loss": 0.0208, "step": 4360 }, { "grad_norm": 0.39920371770858765, "learning_rate": 9.244081582020789e-05, "loss": 0.0175, "step": 4370 }, { "grad_norm": 0.42131638526916504, "learning_rate": 9.239704935276339e-05, "loss": 0.0182, "step": 4380 }, { "grad_norm": 0.45648935437202454, "learning_rate": 9.235316697340489e-05, "loss": 0.0158, "step": 4390 }, { "grad_norm": 0.42188429832458496, "learning_rate": 9.230916880210512e-05, "loss": 0.0183, "step": 4400 }, { "grad_norm": 0.36581969261169434, "learning_rate": 9.226505495915342e-05, "loss": 0.0147, "step": 4410 }, { "grad_norm": 0.42502549290657043, "learning_rate": 9.222082556515536e-05, "loss": 0.0198, "step": 4420 }, { "grad_norm": 0.35229989886283875, "learning_rate": 9.217648074103242e-05, "loss": 0.0153, "step": 4430 }, { "grad_norm": 0.4085313379764557, "learning_rate": 9.213202060802161e-05, "loss": 0.0192, "step": 4440 }, { "grad_norm": 0.4650028645992279, "learning_rate": 9.208744528767528e-05, "loss": 0.0173, "step": 4450 }, { "grad_norm": 0.4048616886138916, "learning_rate": 9.204275490186064e-05, "loss": 0.0204, "step": 4460 }, { "grad_norm": 0.4178619980812073, "learning_rate": 9.199794957275949e-05, "loss": 0.0204, "step": 4470 }, { "grad_norm": 0.46256691217422485, "learning_rate": 9.19530294228679e-05, "loss": 0.0177, "step": 4480 }, { "grad_norm": 0.35352519154548645, "learning_rate": 9.190799457499583e-05, "loss": 0.028, "step": 4490 }, { "grad_norm": 0.4470050632953644, "learning_rate": 9.186284515226686e-05, "loss": 0.0194, "step": 4500 }, { "grad_norm": 0.3508913815021515, "learning_rate": 9.181758127811777e-05, "loss": 0.0241, "step": 4510 }, { "grad_norm": 0.411702424287796, "learning_rate": 9.177220307629825e-05, "loss": 0.0204, "step": 4520 }, { "grad_norm": 0.4468960762023926, "learning_rate": 9.172671067087059e-05, "loss": 0.0194, "step": 4530 }, { "grad_norm": 0.4807928204536438, "learning_rate": 9.16811041862093e-05, "loss": 0.0256, "step": 4540 }, { "grad_norm": 0.39205247163772583, "learning_rate": 9.163538374700076e-05, "loss": 0.0185, "step": 4550 }, { "grad_norm": 0.44329723715782166, "learning_rate": 9.158954947824287e-05, "loss": 0.0178, "step": 4560 }, { "grad_norm": 0.47283023595809937, "learning_rate": 9.154360150524482e-05, "loss": 0.0174, "step": 4570 }, { "grad_norm": 0.38849857449531555, "learning_rate": 9.14975399536266e-05, "loss": 0.0143, "step": 4580 }, { "grad_norm": 0.3656264543533325, "learning_rate": 9.14513649493187e-05, "loss": 0.0212, "step": 4590 }, { "grad_norm": 0.4674840271472931, "learning_rate": 9.140507661856187e-05, "loss": 0.0153, "step": 4600 }, { "grad_norm": 0.4313472509384155, "learning_rate": 9.135867508790661e-05, "loss": 0.0214, "step": 4610 }, { "grad_norm": 0.3471619486808777, "learning_rate": 9.131216048421291e-05, "loss": 0.0165, "step": 4620 }, { "grad_norm": 0.4542539715766907, "learning_rate": 9.126553293464998e-05, "loss": 0.0189, "step": 4630 }, { "grad_norm": 0.47608688473701477, "learning_rate": 9.121879256669572e-05, "loss": 0.017, "step": 4640 }, { "grad_norm": 0.3959465026855469, "learning_rate": 9.117193950813652e-05, "loss": 0.0164, "step": 4650 }, { "grad_norm": 0.408431738615036, "learning_rate": 9.112497388706685e-05, "loss": 0.0255, "step": 4660 }, { "grad_norm": 0.4116475582122803, "learning_rate": 9.10778958318889e-05, "loss": 0.0174, "step": 4670 }, { "grad_norm": 0.3917919993400574, "learning_rate": 9.103070547131232e-05, "loss": 0.0199, "step": 4680 }, { "grad_norm": 0.3482106029987335, "learning_rate": 9.098340293435375e-05, "loss": 0.0179, "step": 4690 }, { "grad_norm": 0.34646838903427124, "learning_rate": 9.093598835033649e-05, "loss": 0.0174, "step": 4700 }, { "grad_norm": 0.39419376850128174, "learning_rate": 9.088846184889021e-05, "loss": 0.0191, "step": 4710 }, { "grad_norm": 0.4543268084526062, "learning_rate": 9.084082355995057e-05, "loss": 0.0213, "step": 4720 }, { "grad_norm": 0.4212946891784668, "learning_rate": 9.079307361375882e-05, "loss": 0.0181, "step": 4730 }, { "grad_norm": 0.3014923334121704, "learning_rate": 9.074521214086149e-05, "loss": 0.019, "step": 4740 }, { "grad_norm": 0.36527299880981445, "learning_rate": 9.069723927211001e-05, "loss": 0.0179, "step": 4750 }, { "grad_norm": 0.3752840757369995, "learning_rate": 9.064915513866037e-05, "loss": 0.0183, "step": 4760 }, { "grad_norm": 0.42201003432273865, "learning_rate": 9.060095987197279e-05, "loss": 0.0162, "step": 4770 }, { "grad_norm": 0.3307137191295624, "learning_rate": 9.055265360381126e-05, "loss": 0.0206, "step": 4780 }, { "grad_norm": 0.33322593569755554, "learning_rate": 9.050423646624326e-05, "loss": 0.016, "step": 4790 }, { "grad_norm": 0.35324618220329285, "learning_rate": 9.045570859163943e-05, "loss": 0.0194, "step": 4800 }, { "grad_norm": 0.427572637796402, "learning_rate": 9.04070701126731e-05, "loss": 0.015, "step": 4810 }, { "grad_norm": 0.3561609983444214, "learning_rate": 9.035832116232001e-05, "loss": 0.0145, "step": 4820 }, { "grad_norm": 0.37716561555862427, "learning_rate": 9.030946187385796e-05, "loss": 0.016, "step": 4830 }, { "grad_norm": 0.39859738945961, "learning_rate": 9.026049238086635e-05, "loss": 0.0178, "step": 4840 }, { "grad_norm": 0.4500395655632019, "learning_rate": 9.021141281722591e-05, "loss": 0.0202, "step": 4850 }, { "grad_norm": 0.34830138087272644, "learning_rate": 9.01622233171183e-05, "loss": 0.0169, "step": 4860 }, { "grad_norm": 0.3729107677936554, "learning_rate": 9.011292401502574e-05, "loss": 0.0212, "step": 4870 }, { "grad_norm": 0.3912448585033417, "learning_rate": 9.006351504573063e-05, "loss": 0.0146, "step": 4880 }, { "grad_norm": 0.4137353003025055, "learning_rate": 9.001399654431519e-05, "loss": 0.0171, "step": 4890 }, { "grad_norm": 0.4444160759449005, "learning_rate": 8.996436864616116e-05, "loss": 0.0162, "step": 4900 }, { "grad_norm": 0.3148241639137268, "learning_rate": 8.991463148694925e-05, "loss": 0.0191, "step": 4910 }, { "grad_norm": 0.4391416907310486, "learning_rate": 8.986478520265902e-05, "loss": 0.0187, "step": 4920 }, { "grad_norm": 0.4296688139438629, "learning_rate": 8.981482992956827e-05, "loss": 0.0143, "step": 4930 }, { "grad_norm": 0.29728299379348755, "learning_rate": 8.976476580425282e-05, "loss": 0.0148, "step": 4940 }, { "grad_norm": 0.4356195032596588, "learning_rate": 8.971459296358606e-05, "loss": 0.0287, "step": 4950 }, { "grad_norm": 0.4179481565952301, "learning_rate": 8.966431154473864e-05, "loss": 0.0157, "step": 4960 }, { "grad_norm": 0.3610477149486542, "learning_rate": 8.961392168517803e-05, "loss": 0.0159, "step": 4970 }, { "grad_norm": 0.34345686435699463, "learning_rate": 8.956342352266821e-05, "loss": 0.016, "step": 4980 }, { "grad_norm": 0.3698787987232208, "learning_rate": 8.95128171952692e-05, "loss": 0.0214, "step": 4990 }, { "grad_norm": 0.327648788690567, "learning_rate": 8.946210284133676e-05, "loss": 0.0173, "step": 5000 }, { "grad_norm": 0.2809329330921173, "learning_rate": 8.941128059952201e-05, "loss": 0.0132, "step": 5010 }, { "grad_norm": 0.31239569187164307, "learning_rate": 8.936035060877102e-05, "loss": 0.0244, "step": 5020 }, { "grad_norm": 0.40824398398399353, "learning_rate": 8.930931300832443e-05, "loss": 0.0181, "step": 5030 }, { "grad_norm": 0.36586353182792664, "learning_rate": 8.925816793771711e-05, "loss": 0.0148, "step": 5040 }, { "grad_norm": 0.2970711886882782, "learning_rate": 8.92069155367777e-05, "loss": 0.0139, "step": 5050 }, { "grad_norm": 0.3478129506111145, "learning_rate": 8.915555594562834e-05, "loss": 0.0172, "step": 5060 }, { "grad_norm": 0.3716062903404236, "learning_rate": 8.910408930468416e-05, "loss": 0.016, "step": 5070 }, { "grad_norm": 0.35524141788482666, "learning_rate": 8.905251575465303e-05, "loss": 0.0128, "step": 5080 }, { "grad_norm": 0.46534451842308044, "learning_rate": 8.900083543653502e-05, "loss": 0.0192, "step": 5090 }, { "grad_norm": 0.4329080283641815, "learning_rate": 8.894904849162218e-05, "loss": 0.0176, "step": 5100 }, { "grad_norm": 0.33941879868507385, "learning_rate": 8.889715506149802e-05, "loss": 0.0161, "step": 5110 }, { "grad_norm": 0.37832191586494446, "learning_rate": 8.884515528803722e-05, "loss": 0.0157, "step": 5120 }, { "grad_norm": 0.36182844638824463, "learning_rate": 8.879304931340517e-05, "loss": 0.0148, "step": 5130 }, { "grad_norm": 0.3270893096923828, "learning_rate": 8.874083728005759e-05, "loss": 0.014, "step": 5140 }, { "grad_norm": 0.3815666437149048, "learning_rate": 8.868851933074021e-05, "loss": 0.0207, "step": 5150 }, { "grad_norm": 0.4211690425872803, "learning_rate": 8.863609560848829e-05, "loss": 0.0147, "step": 5160 }, { "grad_norm": 0.38520973920822144, "learning_rate": 8.85835662566263e-05, "loss": 0.0167, "step": 5170 }, { "grad_norm": 0.4238542914390564, "learning_rate": 8.853093141876747e-05, "loss": 0.0123, "step": 5180 }, { "grad_norm": 0.4630364179611206, "learning_rate": 8.847819123881343e-05, "loss": 0.0157, "step": 5190 }, { "grad_norm": 0.3982260227203369, "learning_rate": 8.842534586095383e-05, "loss": 0.0156, "step": 5200 }, { "grad_norm": 0.418029248714447, "learning_rate": 8.837239542966593e-05, "loss": 0.0147, "step": 5210 }, { "grad_norm": 0.42096608877182007, "learning_rate": 8.831934008971417e-05, "loss": 0.0155, "step": 5220 }, { "grad_norm": 0.40651246905326843, "learning_rate": 8.826617998614982e-05, "loss": 0.0134, "step": 5230 }, { "grad_norm": 0.35442405939102173, "learning_rate": 8.821291526431056e-05, "loss": 0.0149, "step": 5240 }, { "grad_norm": 0.4477267563343048, "learning_rate": 8.815954606982015e-05, "loss": 0.017, "step": 5250 }, { "grad_norm": 0.42138704657554626, "learning_rate": 8.810607254858789e-05, "loss": 0.0161, "step": 5260 }, { "grad_norm": 0.3887590169906616, "learning_rate": 8.805249484680838e-05, "loss": 0.0177, "step": 5270 }, { "grad_norm": 0.3340283930301666, "learning_rate": 8.799881311096096e-05, "loss": 0.0132, "step": 5280 }, { "grad_norm": 0.37869396805763245, "learning_rate": 8.794502748780949e-05, "loss": 0.0158, "step": 5290 }, { "grad_norm": 0.38290390372276306, "learning_rate": 8.78911381244018e-05, "loss": 0.0142, "step": 5300 }, { "grad_norm": 0.3591060936450958, "learning_rate": 8.783714516806933e-05, "loss": 0.0151, "step": 5310 }, { "grad_norm": 0.4037642478942871, "learning_rate": 8.77830487664268e-05, "loss": 0.0128, "step": 5320 }, { "grad_norm": 0.25061705708503723, "learning_rate": 8.772884906737167e-05, "loss": 0.0162, "step": 5330 }, { "grad_norm": 0.3075624108314514, "learning_rate": 8.767454621908387e-05, "loss": 0.0136, "step": 5340 }, { "grad_norm": 0.38026294112205505, "learning_rate": 8.76201403700253e-05, "loss": 0.0136, "step": 5350 }, { "grad_norm": 0.48398804664611816, "learning_rate": 8.756563166893949e-05, "loss": 0.0132, "step": 5360 }, { "grad_norm": 0.3468417823314667, "learning_rate": 8.751102026485113e-05, "loss": 0.0151, "step": 5370 }, { "grad_norm": 0.3630031943321228, "learning_rate": 8.745630630706571e-05, "loss": 0.0158, "step": 5380 }, { "grad_norm": 0.5174519419670105, "learning_rate": 8.740148994516912e-05, "loss": 0.0165, "step": 5390 }, { "grad_norm": 0.3353045582771301, "learning_rate": 8.73465713290272e-05, "loss": 0.0133, "step": 5400 }, { "grad_norm": 0.3345337212085724, "learning_rate": 8.729155060878533e-05, "loss": 0.0186, "step": 5410 }, { "grad_norm": 0.3502795100212097, "learning_rate": 8.723642793486809e-05, "loss": 0.017, "step": 5420 }, { "grad_norm": 0.33117562532424927, "learning_rate": 8.718120345797873e-05, "loss": 0.0156, "step": 5430 }, { "grad_norm": 0.3348385989665985, "learning_rate": 8.712587732909889e-05, "loss": 0.0131, "step": 5440 }, { "grad_norm": 0.31218665838241577, "learning_rate": 8.707044969948806e-05, "loss": 0.0139, "step": 5450 }, { "grad_norm": 0.36143720149993896, "learning_rate": 8.701492072068329e-05, "loss": 0.0162, "step": 5460 }, { "grad_norm": 0.397625207901001, "learning_rate": 8.695929054449869e-05, "loss": 0.0162, "step": 5470 }, { "grad_norm": 0.35156044363975525, "learning_rate": 8.690355932302501e-05, "loss": 0.0149, "step": 5480 }, { "grad_norm": 0.3862064778804779, "learning_rate": 8.684772720862931e-05, "loss": 0.0134, "step": 5490 }, { "grad_norm": 0.33415740728378296, "learning_rate": 8.679179435395446e-05, "loss": 0.0156, "step": 5500 }, { "grad_norm": 0.33985161781311035, "learning_rate": 8.673576091191874e-05, "loss": 0.0152, "step": 5510 }, { "grad_norm": 0.43412765860557556, "learning_rate": 8.667962703571541e-05, "loss": 0.0153, "step": 5520 }, { "grad_norm": 0.29241663217544556, "learning_rate": 8.662339287881238e-05, "loss": 0.0145, "step": 5530 }, { "grad_norm": 0.26985955238342285, "learning_rate": 8.656705859495169e-05, "loss": 0.0095, "step": 5540 }, { "grad_norm": 0.3288934528827667, "learning_rate": 8.651062433814912e-05, "loss": 0.0148, "step": 5550 }, { "grad_norm": 0.32042691111564636, "learning_rate": 8.645409026269375e-05, "loss": 0.0178, "step": 5560 }, { "grad_norm": 0.29201775789260864, "learning_rate": 8.639745652314759e-05, "loss": 0.0136, "step": 5570 }, { "grad_norm": 0.33705347776412964, "learning_rate": 8.634072327434515e-05, "loss": 0.0199, "step": 5580 }, { "grad_norm": 0.43964189291000366, "learning_rate": 8.628389067139294e-05, "loss": 0.0153, "step": 5590 }, { "grad_norm": 0.3852575421333313, "learning_rate": 8.622695886966911e-05, "loss": 0.0124, "step": 5600 }, { "grad_norm": 0.3601333200931549, "learning_rate": 8.616992802482308e-05, "loss": 0.0115, "step": 5610 }, { "grad_norm": 0.3712993562221527, "learning_rate": 8.611279829277496e-05, "loss": 0.0129, "step": 5620 }, { "grad_norm": 0.3430801033973694, "learning_rate": 8.605556982971528e-05, "loss": 0.0119, "step": 5630 }, { "grad_norm": 0.2783951163291931, "learning_rate": 8.599824279210447e-05, "loss": 0.0113, "step": 5640 }, { "grad_norm": 0.3604603111743927, "learning_rate": 8.594081733667243e-05, "loss": 0.016, "step": 5650 }, { "grad_norm": 0.4052552282810211, "learning_rate": 8.58832936204182e-05, "loss": 0.0141, "step": 5660 }, { "grad_norm": 0.2946913242340088, "learning_rate": 8.582567180060942e-05, "loss": 0.0188, "step": 5670 }, { "grad_norm": 0.29554295539855957, "learning_rate": 8.576795203478194e-05, "loss": 0.0147, "step": 5680 }, { "grad_norm": 0.30913200974464417, "learning_rate": 8.571013448073939e-05, "loss": 0.0155, "step": 5690 }, { "grad_norm": 0.3153333067893982, "learning_rate": 8.565221929655275e-05, "loss": 0.0116, "step": 5700 }, { "grad_norm": 0.26914530992507935, "learning_rate": 8.559420664055992e-05, "loss": 0.0125, "step": 5710 }, { "grad_norm": 0.3266845941543579, "learning_rate": 8.553609667136532e-05, "loss": 0.0127, "step": 5720 }, { "grad_norm": 0.36770594120025635, "learning_rate": 8.547788954783936e-05, "loss": 0.0132, "step": 5730 }, { "grad_norm": 0.3868075907230377, "learning_rate": 8.541958542911808e-05, "loss": 0.0137, "step": 5740 }, { "grad_norm": 0.3873762786388397, "learning_rate": 8.536118447460275e-05, "loss": 0.016, "step": 5750 }, { "grad_norm": 0.34997740387916565, "learning_rate": 8.530268684395932e-05, "loss": 0.012, "step": 5760 }, { "grad_norm": 0.36314913630485535, "learning_rate": 8.524409269711807e-05, "loss": 0.014, "step": 5770 }, { "grad_norm": 0.2800992727279663, "learning_rate": 8.51854021942732e-05, "loss": 0.0111, "step": 5780 }, { "grad_norm": 0.3715326488018036, "learning_rate": 8.512661549588227e-05, "loss": 0.0128, "step": 5790 }, { "grad_norm": 0.3508760631084442, "learning_rate": 8.506773276266588e-05, "loss": 0.0123, "step": 5800 }, { "grad_norm": 0.31156125664711, "learning_rate": 8.500875415560721e-05, "loss": 0.0104, "step": 5810 }, { "grad_norm": 0.28672730922698975, "learning_rate": 8.494967983595144e-05, "loss": 0.0138, "step": 5820 }, { "grad_norm": 0.2949328124523163, "learning_rate": 8.489050996520558e-05, "loss": 0.0111, "step": 5830 }, { "grad_norm": 0.3339660167694092, "learning_rate": 8.483124470513775e-05, "loss": 0.0125, "step": 5840 }, { "grad_norm": 0.37675192952156067, "learning_rate": 8.477188421777692e-05, "loss": 0.013, "step": 5850 }, { "grad_norm": 0.4156615436077118, "learning_rate": 8.47124286654124e-05, "loss": 0.0188, "step": 5860 }, { "grad_norm": 0.3377411365509033, "learning_rate": 8.465287821059341e-05, "loss": 0.0197, "step": 5870 }, { "grad_norm": 0.31554165482521057, "learning_rate": 8.45932330161286e-05, "loss": 0.0125, "step": 5880 }, { "grad_norm": 0.3891998529434204, "learning_rate": 8.453349324508567e-05, "loss": 0.0169, "step": 5890 }, { "grad_norm": 0.2835284173488617, "learning_rate": 8.447365906079088e-05, "loss": 0.0172, "step": 5900 }, { "grad_norm": 0.3825901746749878, "learning_rate": 8.441373062682856e-05, "loss": 0.0146, "step": 5910 }, { "grad_norm": 0.3294428884983063, "learning_rate": 8.43537081070408e-05, "loss": 0.0218, "step": 5920 }, { "grad_norm": 0.3541003167629242, "learning_rate": 8.429359166552689e-05, "loss": 0.0132, "step": 5930 }, { "grad_norm": 0.3192877173423767, "learning_rate": 8.423338146664284e-05, "loss": 0.014, "step": 5940 }, { "grad_norm": 0.3500727713108063, "learning_rate": 8.417307767500107e-05, "loss": 0.0115, "step": 5950 }, { "grad_norm": 0.3229285478591919, "learning_rate": 8.411268045546983e-05, "loss": 0.0121, "step": 5960 }, { "grad_norm": 0.4392866790294647, "learning_rate": 8.405218997317281e-05, "loss": 0.0125, "step": 5970 }, { "grad_norm": 0.3409421145915985, "learning_rate": 8.399160639348869e-05, "loss": 0.0115, "step": 5980 }, { "grad_norm": 0.3397701680660248, "learning_rate": 8.393092988205065e-05, "loss": 0.0144, "step": 5990 }, { "grad_norm": 0.2932409346103668, "learning_rate": 8.387016060474597e-05, "loss": 0.0155, "step": 6000 }, { "grad_norm": 0.30588042736053467, "learning_rate": 8.380929872771551e-05, "loss": 0.018, "step": 6010 }, { "grad_norm": 0.2569223642349243, "learning_rate": 8.374834441735335e-05, "loss": 0.0139, "step": 6020 }, { "grad_norm": 0.37832796573638916, "learning_rate": 8.368729784030622e-05, "loss": 0.012, "step": 6030 }, { "grad_norm": 0.3126446604728699, "learning_rate": 8.362615916347315e-05, "loss": 0.0138, "step": 6040 }, { "grad_norm": 0.2546840310096741, "learning_rate": 8.356492855400493e-05, "loss": 0.0122, "step": 6050 }, { "grad_norm": 0.3276226818561554, "learning_rate": 8.350360617930371e-05, "loss": 0.0118, "step": 6060 }, { "grad_norm": 0.3676457405090332, "learning_rate": 8.344219220702255e-05, "loss": 0.0117, "step": 6070 }, { "grad_norm": 0.35417577624320984, "learning_rate": 8.338068680506485e-05, "loss": 0.0104, "step": 6080 }, { "grad_norm": 0.26148155331611633, "learning_rate": 8.33190901415841e-05, "loss": 0.0142, "step": 6090 }, { "grad_norm": 0.29308485984802246, "learning_rate": 8.325740238498317e-05, "loss": 0.0111, "step": 6100 }, { "grad_norm": 0.41223078966140747, "learning_rate": 8.319562370391406e-05, "loss": 0.0135, "step": 6110 }, { "grad_norm": 0.38607892394065857, "learning_rate": 8.31337542672773e-05, "loss": 0.0131, "step": 6120 }, { "grad_norm": 0.3013926148414612, "learning_rate": 8.307179424422158e-05, "loss": 0.0115, "step": 6130 }, { "grad_norm": 0.34792882204055786, "learning_rate": 8.300974380414327e-05, "loss": 0.0165, "step": 6140 }, { "grad_norm": 0.3160726726055145, "learning_rate": 8.294760311668586e-05, "loss": 0.0133, "step": 6150 }, { "grad_norm": 0.3607368469238281, "learning_rate": 8.288537235173961e-05, "loss": 0.0152, "step": 6160 }, { "grad_norm": 0.2962538003921509, "learning_rate": 8.282305167944108e-05, "loss": 0.0146, "step": 6170 }, { "grad_norm": 0.3576897978782654, "learning_rate": 8.276064127017262e-05, "loss": 0.0132, "step": 6180 }, { "grad_norm": 0.3158738911151886, "learning_rate": 8.269814129456189e-05, "loss": 0.0165, "step": 6190 }, { "grad_norm": 0.3886716663837433, "learning_rate": 8.263555192348143e-05, "loss": 0.0164, "step": 6200 }, { "grad_norm": 0.33075031638145447, "learning_rate": 8.257287332804819e-05, "loss": 0.024, "step": 6210 }, { "grad_norm": 0.40429654717445374, "learning_rate": 8.251010567962307e-05, "loss": 0.0142, "step": 6220 }, { "grad_norm": 0.26305797696113586, "learning_rate": 8.244724914981041e-05, "loss": 0.0127, "step": 6230 }, { "grad_norm": 0.26651525497436523, "learning_rate": 8.238430391045757e-05, "loss": 0.0099, "step": 6240 }, { "grad_norm": 0.22140610218048096, "learning_rate": 8.232127013365445e-05, "loss": 0.0162, "step": 6250 }, { "grad_norm": 0.3048286736011505, "learning_rate": 8.225814799173295e-05, "loss": 0.0219, "step": 6260 }, { "grad_norm": 0.27734512090682983, "learning_rate": 8.219493765726663e-05, "loss": 0.012, "step": 6270 }, { "grad_norm": 0.3474031090736389, "learning_rate": 8.21316393030701e-05, "loss": 0.0122, "step": 6280 }, { "grad_norm": 0.3461661636829376, "learning_rate": 8.206825310219865e-05, "loss": 0.0142, "step": 6290 }, { "grad_norm": 0.3480895161628723, "learning_rate": 8.200477922794776e-05, "loss": 0.0155, "step": 6300 }, { "grad_norm": 0.29770898818969727, "learning_rate": 8.194121785385256e-05, "loss": 0.0122, "step": 6310 }, { "grad_norm": 0.27077922224998474, "learning_rate": 8.187756915368741e-05, "loss": 0.0114, "step": 6320 }, { "grad_norm": 0.37504443526268005, "learning_rate": 8.181383330146544e-05, "loss": 0.0123, "step": 6330 }, { "grad_norm": 0.27733808755874634, "learning_rate": 8.175001047143804e-05, "loss": 0.0152, "step": 6340 }, { "grad_norm": 0.2905326187610626, "learning_rate": 8.168610083809438e-05, "loss": 0.014, "step": 6350 }, { "grad_norm": 0.3680465519428253, "learning_rate": 8.162210457616095e-05, "loss": 0.0128, "step": 6360 }, { "grad_norm": 0.41437506675720215, "learning_rate": 8.155802186060109e-05, "loss": 0.0168, "step": 6370 }, { "grad_norm": 0.2748274505138397, "learning_rate": 8.149385286661453e-05, "loss": 0.014, "step": 6380 }, { "grad_norm": 0.27940356731414795, "learning_rate": 8.14295977696368e-05, "loss": 0.0106, "step": 6390 }, { "grad_norm": 0.3001856505870819, "learning_rate": 8.13652567453389e-05, "loss": 0.0161, "step": 6400 }, { "grad_norm": 0.3228931725025177, "learning_rate": 8.130082996962676e-05, "loss": 0.0108, "step": 6410 }, { "grad_norm": 0.37547504901885986, "learning_rate": 8.123631761864068e-05, "loss": 0.0108, "step": 6420 }, { "grad_norm": 0.3306344449520111, "learning_rate": 8.1171719868755e-05, "loss": 0.012, "step": 6430 }, { "grad_norm": 0.30303868651390076, "learning_rate": 8.110703689657748e-05, "loss": 0.0132, "step": 6440 }, { "grad_norm": 0.3219710886478424, "learning_rate": 8.104226887894892e-05, "loss": 0.0247, "step": 6450 }, { "grad_norm": 0.25483033061027527, "learning_rate": 8.097741599294257e-05, "loss": 0.0137, "step": 6460 }, { "grad_norm": 0.34695756435394287, "learning_rate": 8.091247841586378e-05, "loss": 0.0118, "step": 6470 }, { "grad_norm": 0.31917816400527954, "learning_rate": 8.084745632524939e-05, "loss": 0.0153, "step": 6480 }, { "grad_norm": 0.298793762922287, "learning_rate": 8.07823498988673e-05, "loss": 0.013, "step": 6490 }, { "grad_norm": 0.29808494448661804, "learning_rate": 8.071715931471602e-05, "loss": 0.0141, "step": 6500 }, { "grad_norm": 0.2866530120372772, "learning_rate": 8.06518847510241e-05, "loss": 0.013, "step": 6510 }, { "grad_norm": 0.2647181451320648, "learning_rate": 8.058652638624971e-05, "loss": 0.013, "step": 6520 }, { "grad_norm": 0.2521056830883026, "learning_rate": 8.052108439908013e-05, "loss": 0.0117, "step": 6530 }, { "grad_norm": 0.30903002619743347, "learning_rate": 8.045555896843125e-05, "loss": 0.0127, "step": 6540 }, { "grad_norm": 0.3295891284942627, "learning_rate": 8.03899502734471e-05, "loss": 0.0116, "step": 6550 }, { "grad_norm": 0.3196203410625458, "learning_rate": 8.032425849349931e-05, "loss": 0.0096, "step": 6560 }, { "grad_norm": 0.2777307331562042, "learning_rate": 8.025848380818674e-05, "loss": 0.0137, "step": 6570 }, { "grad_norm": 0.2528461217880249, "learning_rate": 8.019262639733487e-05, "loss": 0.012, "step": 6580 }, { "grad_norm": 0.29351767897605896, "learning_rate": 8.012668644099531e-05, "loss": 0.0116, "step": 6590 }, { "grad_norm": 0.3499806821346283, "learning_rate": 8.006066411944542e-05, "loss": 0.0145, "step": 6600 }, { "grad_norm": 0.3001391589641571, "learning_rate": 7.999455961318769e-05, "loss": 0.0107, "step": 6610 }, { "grad_norm": 0.34684231877326965, "learning_rate": 7.992837310294932e-05, "loss": 0.0174, "step": 6620 }, { "grad_norm": 0.29393184185028076, "learning_rate": 7.986210476968167e-05, "loss": 0.0139, "step": 6630 }, { "grad_norm": 0.29339519143104553, "learning_rate": 7.97957547945599e-05, "loss": 0.0174, "step": 6640 }, { "grad_norm": 0.4209054112434387, "learning_rate": 7.972932335898226e-05, "loss": 0.0115, "step": 6650 }, { "grad_norm": 0.28844037652015686, "learning_rate": 7.966281064456975e-05, "loss": 0.0121, "step": 6660 }, { "grad_norm": 0.3526553511619568, "learning_rate": 7.959621683316563e-05, "loss": 0.0119, "step": 6670 }, { "grad_norm": 0.42336228489875793, "learning_rate": 7.952954210683481e-05, "loss": 0.016, "step": 6680 }, { "grad_norm": 0.3053176701068878, "learning_rate": 7.946278664786345e-05, "loss": 0.0143, "step": 6690 }, { "grad_norm": 0.25422704219818115, "learning_rate": 7.939595063875842e-05, "loss": 0.0129, "step": 6700 }, { "grad_norm": 0.2565918266773224, "learning_rate": 7.932903426224683e-05, "loss": 0.0118, "step": 6710 }, { "grad_norm": 0.29806122183799744, "learning_rate": 7.926203770127552e-05, "loss": 0.0112, "step": 6720 }, { "grad_norm": 0.29177016019821167, "learning_rate": 7.919496113901046e-05, "loss": 0.0138, "step": 6730 }, { "grad_norm": 0.24790076911449432, "learning_rate": 7.912780475883649e-05, "loss": 0.0104, "step": 6740 }, { "grad_norm": 0.3190149664878845, "learning_rate": 7.906056874435652e-05, "loss": 0.0127, "step": 6750 }, { "grad_norm": 0.28794440627098083, "learning_rate": 7.899325327939131e-05, "loss": 0.0094, "step": 6760 }, { "grad_norm": 0.24909764528274536, "learning_rate": 7.892585854797872e-05, "loss": 0.0135, "step": 6770 }, { "grad_norm": 0.358511745929718, "learning_rate": 7.88583847343734e-05, "loss": 0.0119, "step": 6780 }, { "grad_norm": 0.2664150893688202, "learning_rate": 7.879083202304616e-05, "loss": 0.0137, "step": 6790 }, { "grad_norm": 0.24875850975513458, "learning_rate": 7.872320059868355e-05, "loss": 0.0144, "step": 6800 }, { "grad_norm": 0.2646051347255707, "learning_rate": 7.865549064618729e-05, "loss": 0.0101, "step": 6810 }, { "grad_norm": 0.3005359172821045, "learning_rate": 7.858770235067381e-05, "loss": 0.0189, "step": 6820 }, { "grad_norm": 0.30753788352012634, "learning_rate": 7.851983589747374e-05, "loss": 0.0128, "step": 6830 }, { "grad_norm": 0.3018791675567627, "learning_rate": 7.845189147213133e-05, "loss": 0.0103, "step": 6840 }, { "grad_norm": 0.27601730823516846, "learning_rate": 7.838386926040407e-05, "loss": 0.012, "step": 6850 }, { "grad_norm": 0.18579219281673431, "learning_rate": 7.83157694482621e-05, "loss": 0.0088, "step": 6860 }, { "grad_norm": 0.3077498972415924, "learning_rate": 7.824759222188768e-05, "loss": 0.0135, "step": 6870 }, { "grad_norm": 0.3342336118221283, "learning_rate": 7.817933776767478e-05, "loss": 0.0107, "step": 6880 }, { "grad_norm": 0.2645319998264313, "learning_rate": 7.811100627222842e-05, "loss": 0.0079, "step": 6890 }, { "grad_norm": 0.34547173976898193, "learning_rate": 7.804259792236435e-05, "loss": 0.0106, "step": 6900 }, { "grad_norm": 0.3139336407184601, "learning_rate": 7.797411290510835e-05, "loss": 0.0143, "step": 6910 }, { "grad_norm": 0.32347217202186584, "learning_rate": 7.790555140769586e-05, "loss": 0.0116, "step": 6920 }, { "grad_norm": 0.2938658595085144, "learning_rate": 7.78369136175714e-05, "loss": 0.0151, "step": 6930 }, { "grad_norm": 0.3143678605556488, "learning_rate": 7.776819972238806e-05, "loss": 0.0157, "step": 6940 }, { "grad_norm": 0.3636862337589264, "learning_rate": 7.7699409910007e-05, "loss": 0.012, "step": 6950 }, { "grad_norm": 0.2957841157913208, "learning_rate": 7.763054436849694e-05, "loss": 0.0137, "step": 6960 }, { "grad_norm": 0.3176383078098297, "learning_rate": 7.756160328613364e-05, "loss": 0.0138, "step": 6970 }, { "grad_norm": 0.38332003355026245, "learning_rate": 7.749258685139942e-05, "loss": 0.0103, "step": 6980 }, { "grad_norm": 0.3329831659793854, "learning_rate": 7.742349525298253e-05, "loss": 0.0118, "step": 6990 }, { "grad_norm": 0.373206228017807, "learning_rate": 7.735432867977679e-05, "loss": 0.0096, "step": 7000 }, { "grad_norm": 0.22941261529922485, "learning_rate": 7.728508732088096e-05, "loss": 0.0141, "step": 7010 }, { "grad_norm": 0.3203655183315277, "learning_rate": 7.721577136559825e-05, "loss": 0.0142, "step": 7020 }, { "grad_norm": 0.2580220103263855, "learning_rate": 7.714638100343588e-05, "loss": 0.0119, "step": 7030 }, { "grad_norm": 0.23703204095363617, "learning_rate": 7.707691642410444e-05, "loss": 0.0114, "step": 7040 }, { "grad_norm": 0.2933865189552307, "learning_rate": 7.70073778175174e-05, "loss": 0.0125, "step": 7050 }, { "grad_norm": 0.3587990403175354, "learning_rate": 7.69377653737907e-05, "loss": 0.0103, "step": 7060 }, { "grad_norm": 0.3031073212623596, "learning_rate": 7.686807928324209e-05, "loss": 0.0119, "step": 7070 }, { "grad_norm": 0.24994587898254395, "learning_rate": 7.679831973639065e-05, "loss": 0.0108, "step": 7080 }, { "grad_norm": 0.2641933262348175, "learning_rate": 7.672848692395637e-05, "loss": 0.0124, "step": 7090 }, { "grad_norm": 0.3555925488471985, "learning_rate": 7.665858103685944e-05, "loss": 0.0106, "step": 7100 }, { "grad_norm": 0.32579174637794495, "learning_rate": 7.658860226621991e-05, "loss": 0.0129, "step": 7110 }, { "grad_norm": 0.2926287353038788, "learning_rate": 7.651855080335708e-05, "loss": 0.0127, "step": 7120 }, { "grad_norm": 0.32705703377723694, "learning_rate": 7.644842683978896e-05, "loss": 0.0103, "step": 7130 }, { "grad_norm": 0.3596729636192322, "learning_rate": 7.63782305672318e-05, "loss": 0.0153, "step": 7140 }, { "grad_norm": 0.3586880564689636, "learning_rate": 7.63079621775995e-05, "loss": 0.0132, "step": 7150 }, { "grad_norm": 0.23977535963058472, "learning_rate": 7.623762186300319e-05, "loss": 0.0125, "step": 7160 }, { "grad_norm": 0.3540763854980469, "learning_rate": 7.616720981575057e-05, "loss": 0.0133, "step": 7170 }, { "grad_norm": 0.33983471989631653, "learning_rate": 7.609672622834552e-05, "loss": 0.0102, "step": 7180 }, { "grad_norm": 0.30762892961502075, "learning_rate": 7.602617129348747e-05, "loss": 0.0108, "step": 7190 }, { "grad_norm": 0.3010900020599365, "learning_rate": 7.595554520407088e-05, "loss": 0.0111, "step": 7200 }, { "grad_norm": 0.2769547402858734, "learning_rate": 7.588484815318484e-05, "loss": 0.0126, "step": 7210 }, { "grad_norm": 0.28370919823646545, "learning_rate": 7.581408033411234e-05, "loss": 0.0101, "step": 7220 }, { "grad_norm": 0.32081568241119385, "learning_rate": 7.574324194032995e-05, "loss": 0.009, "step": 7230 }, { "grad_norm": 0.2977130711078644, "learning_rate": 7.567233316550705e-05, "loss": 0.012, "step": 7240 }, { "grad_norm": 0.22645479440689087, "learning_rate": 7.560135420350562e-05, "loss": 0.0087, "step": 7250 }, { "grad_norm": 0.3486950397491455, "learning_rate": 7.553030524837935e-05, "loss": 0.0184, "step": 7260 }, { "grad_norm": 0.3366019129753113, "learning_rate": 7.545918649437341e-05, "loss": 0.0109, "step": 7270 }, { "grad_norm": 0.3164430856704712, "learning_rate": 7.538799813592377e-05, "loss": 0.0121, "step": 7280 }, { "grad_norm": 0.30747735500335693, "learning_rate": 7.531674036765662e-05, "loss": 0.0127, "step": 7290 }, { "grad_norm": 0.2376401573419571, "learning_rate": 7.524541338438807e-05, "loss": 0.0085, "step": 7300 }, { "grad_norm": 0.263528436422348, "learning_rate": 7.517401738112328e-05, "loss": 0.0116, "step": 7310 }, { "grad_norm": 0.35192421078681946, "learning_rate": 7.510255255305628e-05, "loss": 0.0097, "step": 7320 }, { "grad_norm": 0.3433808386325836, "learning_rate": 7.503101909556911e-05, "loss": 0.0106, "step": 7330 }, { "grad_norm": 0.2994978129863739, "learning_rate": 7.495941720423154e-05, "loss": 0.0101, "step": 7340 }, { "grad_norm": 0.2600700557231903, "learning_rate": 7.488774707480042e-05, "loss": 0.0116, "step": 7350 }, { "grad_norm": 0.3085675537586212, "learning_rate": 7.481600890321911e-05, "loss": 0.0088, "step": 7360 }, { "grad_norm": 0.2523610591888428, "learning_rate": 7.474420288561708e-05, "loss": 0.0149, "step": 7370 }, { "grad_norm": 0.3608851432800293, "learning_rate": 7.467232921830921e-05, "loss": 0.015, "step": 7380 }, { "grad_norm": 0.27163684368133545, "learning_rate": 7.460038809779537e-05, "loss": 0.0092, "step": 7390 }, { "grad_norm": 0.3165534436702728, "learning_rate": 7.452837972075983e-05, "loss": 0.0094, "step": 7400 }, { "grad_norm": 0.2729543149471283, "learning_rate": 7.445630428407074e-05, "loss": 0.0093, "step": 7410 }, { "grad_norm": 0.35123759508132935, "learning_rate": 7.43841619847796e-05, "loss": 0.0106, "step": 7420 }, { "grad_norm": 0.2810732424259186, "learning_rate": 7.431195302012072e-05, "loss": 0.0097, "step": 7430 }, { "grad_norm": 0.3775671720504761, "learning_rate": 7.423967758751061e-05, "loss": 0.0119, "step": 7440 }, { "grad_norm": 0.2957786023616791, "learning_rate": 7.416733588454758e-05, "loss": 0.0129, "step": 7450 }, { "grad_norm": 0.3434288799762726, "learning_rate": 7.409492810901106e-05, "loss": 0.0132, "step": 7460 }, { "grad_norm": 0.35347649455070496, "learning_rate": 7.402245445886116e-05, "loss": 0.0147, "step": 7470 }, { "grad_norm": 0.3285079598426819, "learning_rate": 7.394991513223806e-05, "loss": 0.011, "step": 7480 }, { "grad_norm": 0.29947003722190857, "learning_rate": 7.38773103274615e-05, "loss": 0.0132, "step": 7490 }, { "grad_norm": 0.28526970744132996, "learning_rate": 7.380464024303028e-05, "loss": 0.0105, "step": 7500 }, { "grad_norm": 0.2930798828601837, "learning_rate": 7.373190507762162e-05, "loss": 0.0127, "step": 7510 }, { "grad_norm": 0.2921172082424164, "learning_rate": 7.365910503009066e-05, "loss": 0.0156, "step": 7520 }, { "grad_norm": 0.3323417007923126, "learning_rate": 7.358624029946996e-05, "loss": 0.0088, "step": 7530 }, { "grad_norm": 0.29670819640159607, "learning_rate": 7.351331108496893e-05, "loss": 0.0095, "step": 7540 }, { "grad_norm": 0.32244303822517395, "learning_rate": 7.344031758597325e-05, "loss": 0.0137, "step": 7550 }, { "grad_norm": 0.25546425580978394, "learning_rate": 7.336726000204435e-05, "loss": 0.0101, "step": 7560 }, { "grad_norm": 0.30756881833076477, "learning_rate": 7.32941385329189e-05, "loss": 0.0129, "step": 7570 }, { "grad_norm": 0.2749859690666199, "learning_rate": 7.322095337850816e-05, "loss": 0.0095, "step": 7580 }, { "grad_norm": 0.41934898495674133, "learning_rate": 7.314770473889758e-05, "loss": 0.0154, "step": 7590 }, { "grad_norm": 0.2750692665576935, "learning_rate": 7.307439281434615e-05, "loss": 0.0089, "step": 7600 }, { "grad_norm": 0.28263887763023376, "learning_rate": 7.300101780528585e-05, "loss": 0.0127, "step": 7610 }, { "grad_norm": 0.2647198438644409, "learning_rate": 7.292757991232117e-05, "loss": 0.0155, "step": 7620 }, { "grad_norm": 0.30357304215431213, "learning_rate": 7.285407933622848e-05, "loss": 0.0122, "step": 7630 }, { "grad_norm": 0.2601131796836853, "learning_rate": 7.278051627795557e-05, "loss": 0.0173, "step": 7640 }, { "grad_norm": 0.2693704664707184, "learning_rate": 7.270689093862105e-05, "loss": 0.0123, "step": 7650 }, { "grad_norm": 0.3310806453227997, "learning_rate": 7.263320351951374e-05, "loss": 0.009, "step": 7660 }, { "grad_norm": 0.2853841483592987, "learning_rate": 7.255945422209227e-05, "loss": 0.0104, "step": 7670 }, { "grad_norm": 0.19990304112434387, "learning_rate": 7.248564324798437e-05, "loss": 0.0105, "step": 7680 }, { "grad_norm": 0.20870745182037354, "learning_rate": 7.241177079898644e-05, "loss": 0.0126, "step": 7690 }, { "grad_norm": 0.29264724254608154, "learning_rate": 7.233783707706295e-05, "loss": 0.0108, "step": 7700 }, { "grad_norm": 0.26418036222457886, "learning_rate": 7.226384228434586e-05, "loss": 0.0121, "step": 7710 }, { "grad_norm": 0.19485041499137878, "learning_rate": 7.21897866231341e-05, "loss": 0.0136, "step": 7720 }, { "grad_norm": 0.21631906926631927, "learning_rate": 7.211567029589303e-05, "loss": 0.0103, "step": 7730 }, { "grad_norm": 0.2685507833957672, "learning_rate": 7.204149350525387e-05, "loss": 0.0088, "step": 7740 }, { "grad_norm": 0.21937017142772675, "learning_rate": 7.196725645401309e-05, "loss": 0.0118, "step": 7750 }, { "grad_norm": 0.40596023201942444, "learning_rate": 7.1892959345132e-05, "loss": 0.0106, "step": 7760 }, { "grad_norm": 0.3677843511104584, "learning_rate": 7.181860238173605e-05, "loss": 0.0155, "step": 7770 }, { "grad_norm": 0.3440069854259491, "learning_rate": 7.174418576711432e-05, "loss": 0.012, "step": 7780 }, { "grad_norm": 0.3115421533584595, "learning_rate": 7.1669709704719e-05, "loss": 0.0129, "step": 7790 }, { "grad_norm": 0.2567780315876007, "learning_rate": 7.159517439816481e-05, "loss": 0.0186, "step": 7800 }, { "grad_norm": 0.26565659046173096, "learning_rate": 7.152058005122842e-05, "loss": 0.0117, "step": 7810 }, { "grad_norm": 0.2598077654838562, "learning_rate": 7.144592686784793e-05, "loss": 0.0105, "step": 7820 }, { "grad_norm": 0.3031388819217682, "learning_rate": 7.137121505212229e-05, "loss": 0.0103, "step": 7830 }, { "grad_norm": 0.334942102432251, "learning_rate": 7.129644480831077e-05, "loss": 0.0198, "step": 7840 }, { "grad_norm": 0.28533506393432617, "learning_rate": 7.122161634083234e-05, "loss": 0.0103, "step": 7850 }, { "grad_norm": 0.27979883551597595, "learning_rate": 7.114672985426516e-05, "loss": 0.0097, "step": 7860 }, { "grad_norm": 0.21115346252918243, "learning_rate": 7.107178555334606e-05, "loss": 0.0151, "step": 7870 }, { "grad_norm": 0.36340436339378357, "learning_rate": 7.099678364296989e-05, "loss": 0.0115, "step": 7880 }, { "grad_norm": 0.21899573504924774, "learning_rate": 7.0921724328189e-05, "loss": 0.0102, "step": 7890 }, { "grad_norm": 0.3320227861404419, "learning_rate": 7.084660781421268e-05, "loss": 0.0132, "step": 7900 }, { "grad_norm": 0.27939140796661377, "learning_rate": 7.077143430640662e-05, "loss": 0.0119, "step": 7910 }, { "grad_norm": 0.36499500274658203, "learning_rate": 7.069620401029232e-05, "loss": 0.0099, "step": 7920 }, { "grad_norm": 0.3523150682449341, "learning_rate": 7.062091713154655e-05, "loss": 0.0093, "step": 7930 }, { "grad_norm": 0.2645350396633148, "learning_rate": 7.054557387600075e-05, "loss": 0.0097, "step": 7940 }, { "grad_norm": 0.20298346877098083, "learning_rate": 7.04701744496405e-05, "loss": 0.0086, "step": 7950 }, { "grad_norm": 0.20196221768856049, "learning_rate": 7.039471905860495e-05, "loss": 0.0096, "step": 7960 }, { "grad_norm": 0.25855553150177, "learning_rate": 7.031920790918628e-05, "loss": 0.0097, "step": 7970 }, { "grad_norm": 0.28350019454956055, "learning_rate": 7.024364120782906e-05, "loss": 0.0105, "step": 7980 }, { "grad_norm": 0.28542742133140564, "learning_rate": 7.016801916112978e-05, "loss": 0.0137, "step": 7990 }, { "grad_norm": 0.24646037817001343, "learning_rate": 7.009234197583623e-05, "loss": 0.0103, "step": 8000 }, { "grad_norm": 0.26659995317459106, "learning_rate": 7.001660985884692e-05, "loss": 0.0088, "step": 8010 }, { "grad_norm": 0.3342621624469757, "learning_rate": 6.994082301721063e-05, "loss": 0.0114, "step": 8020 }, { "grad_norm": 0.2519735097885132, "learning_rate": 6.986498165812563e-05, "loss": 0.009, "step": 8030 }, { "grad_norm": 0.25942492485046387, "learning_rate": 6.978908598893932e-05, "loss": 0.0089, "step": 8040 }, { "grad_norm": 0.2715606391429901, "learning_rate": 6.971313621714756e-05, "loss": 0.0129, "step": 8050 }, { "grad_norm": 0.28495460748672485, "learning_rate": 6.96371325503941e-05, "loss": 0.0086, "step": 8060 }, { "grad_norm": 0.2748803198337555, "learning_rate": 6.956107519647014e-05, "loss": 0.0079, "step": 8070 }, { "grad_norm": 0.19981186091899872, "learning_rate": 6.94849643633135e-05, "loss": 0.011, "step": 8080 }, { "grad_norm": 0.25709986686706543, "learning_rate": 6.940880025900834e-05, "loss": 0.0091, "step": 8090 }, { "grad_norm": 0.31318721175193787, "learning_rate": 6.933258309178438e-05, "loss": 0.0128, "step": 8100 }, { "grad_norm": 0.2338728904724121, "learning_rate": 6.925631307001646e-05, "loss": 0.01, "step": 8110 }, { "grad_norm": 0.2807973027229309, "learning_rate": 6.91799904022239e-05, "loss": 0.0117, "step": 8120 }, { "grad_norm": 0.16702820360660553, "learning_rate": 6.910361529706997e-05, "loss": 0.0081, "step": 8130 }, { "grad_norm": 0.2894277572631836, "learning_rate": 6.902718796336131e-05, "loss": 0.0086, "step": 8140 }, { "grad_norm": 0.3167431950569153, "learning_rate": 6.895070861004729e-05, "loss": 0.0099, "step": 8150 }, { "grad_norm": 0.35957372188568115, "learning_rate": 6.887417744621956e-05, "loss": 0.0103, "step": 8160 }, { "grad_norm": 0.2503855526447296, "learning_rate": 6.87975946811114e-05, "loss": 0.0088, "step": 8170 }, { "grad_norm": 0.24923115968704224, "learning_rate": 6.872096052409718e-05, "loss": 0.0101, "step": 8180 }, { "grad_norm": 0.27979594469070435, "learning_rate": 6.864427518469174e-05, "loss": 0.0088, "step": 8190 }, { "grad_norm": 0.22051957249641418, "learning_rate": 6.856753887254986e-05, "loss": 0.009, "step": 8200 }, { "grad_norm": 0.2485746592283249, "learning_rate": 6.849075179746572e-05, "loss": 0.0132, "step": 8210 }, { "grad_norm": 0.29662564396858215, "learning_rate": 6.841391416937221e-05, "loss": 0.0109, "step": 8220 }, { "grad_norm": 0.22791483998298645, "learning_rate": 6.833702619834053e-05, "loss": 0.0119, "step": 8230 }, { "grad_norm": 0.2398047000169754, "learning_rate": 6.82600880945794e-05, "loss": 0.0114, "step": 8240 }, { "grad_norm": 0.28467315435409546, "learning_rate": 6.818310006843468e-05, "loss": 0.0109, "step": 8250 }, { "grad_norm": 0.39719972014427185, "learning_rate": 6.810606233038868e-05, "loss": 0.0125, "step": 8260 }, { "grad_norm": 0.367841899394989, "learning_rate": 6.802897509105966e-05, "loss": 0.0093, "step": 8270 }, { "grad_norm": 0.28418073058128357, "learning_rate": 6.79518385612012e-05, "loss": 0.011, "step": 8280 }, { "grad_norm": 0.33649322390556335, "learning_rate": 6.787465295170157e-05, "loss": 0.0111, "step": 8290 }, { "grad_norm": 0.20563358068466187, "learning_rate": 6.779741847358332e-05, "loss": 0.0101, "step": 8300 }, { "grad_norm": 0.2554636299610138, "learning_rate": 6.772013533800256e-05, "loss": 0.0078, "step": 8310 }, { "grad_norm": 0.35546278953552246, "learning_rate": 6.764280375624843e-05, "loss": 0.0112, "step": 8320 }, { "grad_norm": 0.33014950156211853, "learning_rate": 6.756542393974252e-05, "loss": 0.0136, "step": 8330 }, { "grad_norm": 0.34563302993774414, "learning_rate": 6.748799610003828e-05, "loss": 0.0111, "step": 8340 }, { "grad_norm": 0.28476187586784363, "learning_rate": 6.741052044882048e-05, "loss": 0.0133, "step": 8350 }, { "grad_norm": 0.21680086851119995, "learning_rate": 6.73329971979046e-05, "loss": 0.01, "step": 8360 }, { "grad_norm": 0.24411574006080627, "learning_rate": 6.725542655923625e-05, "loss": 0.0101, "step": 8370 }, { "grad_norm": 0.2673936188220978, "learning_rate": 6.717780874489057e-05, "loss": 0.0142, "step": 8380 }, { "grad_norm": 0.22214150428771973, "learning_rate": 6.710014396707172e-05, "loss": 0.0081, "step": 8390 }, { "grad_norm": 0.21424426138401031, "learning_rate": 6.702243243811221e-05, "loss": 0.0094, "step": 8400 }, { "grad_norm": 0.326800137758255, "learning_rate": 6.694467437047244e-05, "loss": 0.0109, "step": 8410 }, { "grad_norm": 0.33343732357025146, "learning_rate": 6.686686997673997e-05, "loss": 0.0073, "step": 8420 }, { "grad_norm": 0.2529543936252594, "learning_rate": 6.678901946962903e-05, "loss": 0.0091, "step": 8430 }, { "grad_norm": 0.33415693044662476, "learning_rate": 6.671112306197996e-05, "loss": 0.0093, "step": 8440 }, { "grad_norm": 0.21618202328681946, "learning_rate": 6.663318096675854e-05, "loss": 0.0109, "step": 8450 }, { "grad_norm": 0.17679205536842346, "learning_rate": 6.655519339705552e-05, "loss": 0.0081, "step": 8460 }, { "grad_norm": 0.31180447340011597, "learning_rate": 6.647716056608588e-05, "loss": 0.0075, "step": 8470 }, { "grad_norm": 0.3059850335121155, "learning_rate": 6.639908268718843e-05, "loss": 0.0104, "step": 8480 }, { "grad_norm": 0.2953120768070221, "learning_rate": 6.632095997382514e-05, "loss": 0.0136, "step": 8490 }, { "grad_norm": 0.26256608963012695, "learning_rate": 6.624279263958047e-05, "loss": 0.0093, "step": 8500 }, { "grad_norm": 0.31494539976119995, "learning_rate": 6.616458089816097e-05, "loss": 0.0106, "step": 8510 }, { "grad_norm": 0.376949280500412, "learning_rate": 6.608632496339454e-05, "loss": 0.0099, "step": 8520 }, { "grad_norm": 0.2433367669582367, "learning_rate": 6.600802504922988e-05, "loss": 0.0102, "step": 8530 }, { "grad_norm": 0.37388041615486145, "learning_rate": 6.592968136973604e-05, "loss": 0.0105, "step": 8540 }, { "grad_norm": 0.28184112906455994, "learning_rate": 6.585129413910159e-05, "loss": 0.0097, "step": 8550 }, { "grad_norm": 0.2922300696372986, "learning_rate": 6.577286357163424e-05, "loss": 0.01, "step": 8560 }, { "grad_norm": 0.2782540023326874, "learning_rate": 6.569438988176018e-05, "loss": 0.0082, "step": 8570 }, { "grad_norm": 0.27248328924179077, "learning_rate": 6.561587328402347e-05, "loss": 0.0088, "step": 8580 }, { "grad_norm": 0.24880534410476685, "learning_rate": 6.553731399308549e-05, "loss": 0.0079, "step": 8590 }, { "grad_norm": 0.2515351176261902, "learning_rate": 6.545871222372436e-05, "loss": 0.0073, "step": 8600 }, { "grad_norm": 0.18304027616977692, "learning_rate": 6.538006819083426e-05, "loss": 0.0071, "step": 8610 }, { "grad_norm": 0.3344535231590271, "learning_rate": 6.530138210942505e-05, "loss": 0.0101, "step": 8620 }, { "grad_norm": 0.2146245390176773, "learning_rate": 6.522265419462141e-05, "loss": 0.0121, "step": 8630 }, { "grad_norm": 0.24218611419200897, "learning_rate": 6.514388466166248e-05, "loss": 0.0096, "step": 8640 }, { "grad_norm": 0.16334691643714905, "learning_rate": 6.506507372590119e-05, "loss": 0.0079, "step": 8650 }, { "grad_norm": 0.33363232016563416, "learning_rate": 6.498622160280355e-05, "loss": 0.0098, "step": 8660 }, { "grad_norm": 0.35225972533226013, "learning_rate": 6.490732850794832e-05, "loss": 0.0103, "step": 8670 }, { "grad_norm": 0.3227727711200714, "learning_rate": 6.482839465702616e-05, "loss": 0.0107, "step": 8680 }, { "grad_norm": 0.2620507776737213, "learning_rate": 6.474942026583923e-05, "loss": 0.0104, "step": 8690 }, { "grad_norm": 0.2854481041431427, "learning_rate": 6.467040555030052e-05, "loss": 0.0153, "step": 8700 }, { "grad_norm": 0.24487437307834625, "learning_rate": 6.459135072643321e-05, "loss": 0.0099, "step": 8710 }, { "grad_norm": 0.23461481928825378, "learning_rate": 6.451225601037019e-05, "loss": 0.0109, "step": 8720 }, { "grad_norm": 0.2263409048318863, "learning_rate": 6.443312161835338e-05, "loss": 0.0105, "step": 8730 }, { "grad_norm": 0.3112694025039673, "learning_rate": 6.43539477667332e-05, "loss": 0.0116, "step": 8740 }, { "grad_norm": 0.2899706959724426, "learning_rate": 6.427473467196793e-05, "loss": 0.0109, "step": 8750 }, { "grad_norm": 0.319865882396698, "learning_rate": 6.419548255062315e-05, "loss": 0.0142, "step": 8760 }, { "grad_norm": 0.2569391429424286, "learning_rate": 6.411619161937112e-05, "loss": 0.0185, "step": 8770 }, { "grad_norm": 0.2740200161933899, "learning_rate": 6.403686209499022e-05, "loss": 0.0101, "step": 8780 }, { "grad_norm": 0.29832252860069275, "learning_rate": 6.395749419436437e-05, "loss": 0.0094, "step": 8790 }, { "grad_norm": 0.275097519159317, "learning_rate": 6.387808813448234e-05, "loss": 0.0106, "step": 8800 }, { "grad_norm": 0.32685816287994385, "learning_rate": 6.37986441324373e-05, "loss": 0.0086, "step": 8810 }, { "grad_norm": 0.28648194670677185, "learning_rate": 6.37191624054261e-05, "loss": 0.0117, "step": 8820 }, { "grad_norm": 0.2401561737060547, "learning_rate": 6.363964317074872e-05, "loss": 0.0081, "step": 8830 }, { "grad_norm": 0.2832534909248352, "learning_rate": 6.356008664580776e-05, "loss": 0.0136, "step": 8840 }, { "grad_norm": 0.2052382379770279, "learning_rate": 6.348049304810771e-05, "loss": 0.0097, "step": 8850 }, { "grad_norm": 0.3278440833091736, "learning_rate": 6.340086259525442e-05, "loss": 0.0083, "step": 8860 }, { "grad_norm": 0.34554144740104675, "learning_rate": 6.332119550495448e-05, "loss": 0.0098, "step": 8870 }, { "grad_norm": 0.2610031068325043, "learning_rate": 6.324149199501473e-05, "loss": 0.01, "step": 8880 }, { "grad_norm": 0.22511707246303558, "learning_rate": 6.316175228334146e-05, "loss": 0.0092, "step": 8890 }, { "grad_norm": 0.2637081742286682, "learning_rate": 6.308197658794003e-05, "loss": 0.0128, "step": 8900 }, { "grad_norm": 0.31135818362236023, "learning_rate": 6.300216512691417e-05, "loss": 0.008, "step": 8910 }, { "grad_norm": 0.23880526423454285, "learning_rate": 6.292231811846532e-05, "loss": 0.0101, "step": 8920 }, { "grad_norm": 0.25867730379104614, "learning_rate": 6.284243578089217e-05, "loss": 0.0088, "step": 8930 }, { "grad_norm": 0.26295626163482666, "learning_rate": 6.276251833258999e-05, "loss": 0.0081, "step": 8940 }, { "grad_norm": 0.28615128993988037, "learning_rate": 6.268256599205003e-05, "loss": 0.0104, "step": 8950 }, { "grad_norm": 0.292758971452713, "learning_rate": 6.260257897785892e-05, "loss": 0.009, "step": 8960 }, { "grad_norm": 0.2305050492286682, "learning_rate": 6.252255750869811e-05, "loss": 0.01, "step": 8970 }, { "grad_norm": 0.2941057085990906, "learning_rate": 6.244250180334325e-05, "loss": 0.0136, "step": 8980 }, { "grad_norm": 0.3271690607070923, "learning_rate": 6.236241208066356e-05, "loss": 0.0111, "step": 8990 }, { "grad_norm": 0.2185642123222351, "learning_rate": 6.228228855962133e-05, "loss": 0.0074, "step": 9000 }, { "grad_norm": 0.26744088530540466, "learning_rate": 6.220213145927115e-05, "loss": 0.0067, "step": 9010 }, { "grad_norm": 0.20686663687229156, "learning_rate": 6.212194099875951e-05, "loss": 0.0094, "step": 9020 }, { "grad_norm": 0.2725589871406555, "learning_rate": 6.204171739732405e-05, "loss": 0.0083, "step": 9030 }, { "grad_norm": 0.3449211120605469, "learning_rate": 6.196146087429303e-05, "loss": 0.0085, "step": 9040 }, { "grad_norm": 0.24136734008789062, "learning_rate": 6.188117164908474e-05, "loss": 0.0094, "step": 9050 }, { "grad_norm": 0.2191270887851715, "learning_rate": 6.180084994120684e-05, "loss": 0.011, "step": 9060 }, { "grad_norm": 0.4127255082130432, "learning_rate": 6.17204959702558e-05, "loss": 0.0115, "step": 9070 }, { "grad_norm": 0.25778821110725403, "learning_rate": 6.164010995591635e-05, "loss": 0.0073, "step": 9080 }, { "grad_norm": 0.19585411250591278, "learning_rate": 6.155969211796076e-05, "loss": 0.0093, "step": 9090 }, { "grad_norm": 0.1971423476934433, "learning_rate": 6.147924267624829e-05, "loss": 0.0065, "step": 9100 }, { "grad_norm": 0.18513402342796326, "learning_rate": 6.13987618507247e-05, "loss": 0.0082, "step": 9110 }, { "grad_norm": 0.30468112230300903, "learning_rate": 6.131824986142147e-05, "loss": 0.0108, "step": 9120 }, { "grad_norm": 0.2643079161643982, "learning_rate": 6.123770692845529e-05, "loss": 0.0081, "step": 9130 }, { "grad_norm": 0.35840684175491333, "learning_rate": 6.11571332720275e-05, "loss": 0.0081, "step": 9140 }, { "grad_norm": 0.34239524602890015, "learning_rate": 6.107652911242336e-05, "loss": 0.0118, "step": 9150 }, { "grad_norm": 0.3136473596096039, "learning_rate": 6.0995894670011586e-05, "loss": 0.0128, "step": 9160 }, { "grad_norm": 0.3185141980648041, "learning_rate": 6.091523016524368e-05, "loss": 0.0133, "step": 9170 }, { "grad_norm": 0.2437521517276764, "learning_rate": 6.083453581865328e-05, "loss": 0.0116, "step": 9180 }, { "grad_norm": 0.24852575361728668, "learning_rate": 6.075381185085568e-05, "loss": 0.0101, "step": 9190 }, { "grad_norm": 0.21035079658031464, "learning_rate": 6.067305848254709e-05, "loss": 0.0089, "step": 9200 }, { "grad_norm": 0.19124074280261993, "learning_rate": 6.059227593450418e-05, "loss": 0.01, "step": 9210 }, { "grad_norm": 0.22028234601020813, "learning_rate": 6.051146442758333e-05, "loss": 0.0128, "step": 9220 }, { "grad_norm": 0.2855907082557678, "learning_rate": 6.043062418272012e-05, "loss": 0.0103, "step": 9230 }, { "grad_norm": 0.23253290355205536, "learning_rate": 6.0349755420928666e-05, "loss": 0.0075, "step": 9240 }, { "grad_norm": 0.22832125425338745, "learning_rate": 6.0268858363301105e-05, "loss": 0.0074, "step": 9250 }, { "grad_norm": 0.22071580588817596, "learning_rate": 6.018793323100689e-05, "loss": 0.0106, "step": 9260 }, { "grad_norm": 0.3454406261444092, "learning_rate": 6.0106980245292255e-05, "loss": 0.011, "step": 9270 }, { "grad_norm": 0.3467009663581848, "learning_rate": 6.002599962747957e-05, "loss": 0.0087, "step": 9280 }, { "grad_norm": 0.2289619743824005, "learning_rate": 5.994499159896673e-05, "loss": 0.0068, "step": 9290 }, { "grad_norm": 0.2502879202365875, "learning_rate": 5.9863956381226607e-05, "loss": 0.0138, "step": 9300 }, { "grad_norm": 0.23016954958438873, "learning_rate": 5.9782894195806394e-05, "loss": 0.0088, "step": 9310 }, { "grad_norm": 0.3265341520309448, "learning_rate": 5.9701805264327004e-05, "loss": 0.0122, "step": 9320 }, { "grad_norm": 0.2779223322868347, "learning_rate": 5.96206898084825e-05, "loss": 0.0073, "step": 9330 }, { "grad_norm": 0.20430560410022736, "learning_rate": 5.953954805003942e-05, "loss": 0.0106, "step": 9340 }, { "grad_norm": 0.22642415761947632, "learning_rate": 5.945838021083623e-05, "loss": 0.0083, "step": 9350 }, { "grad_norm": 0.22150662541389465, "learning_rate": 5.9377186512782714e-05, "loss": 0.0065, "step": 9360 }, { "grad_norm": 0.364218145608902, "learning_rate": 5.929596717785935e-05, "loss": 0.0089, "step": 9370 }, { "grad_norm": 0.24294275045394897, "learning_rate": 5.921472242811668e-05, "loss": 0.0079, "step": 9380 }, { "grad_norm": 0.23857471346855164, "learning_rate": 5.913345248567475e-05, "loss": 0.0125, "step": 9390 }, { "grad_norm": 0.17391999065876007, "learning_rate": 5.905215757272248e-05, "loss": 0.0162, "step": 9400 }, { "grad_norm": 0.2742446959018707, "learning_rate": 5.897083791151706e-05, "loss": 0.009, "step": 9410 }, { "grad_norm": 0.27407529950141907, "learning_rate": 5.888949372438336e-05, "loss": 0.0104, "step": 9420 }, { "grad_norm": 0.1971682459115982, "learning_rate": 5.8808125233713255e-05, "loss": 0.008, "step": 9430 }, { "grad_norm": 0.17385496199131012, "learning_rate": 5.872673266196509e-05, "loss": 0.007, "step": 9440 }, { "grad_norm": 0.2608735263347626, "learning_rate": 5.864531623166305e-05, "loss": 0.0083, "step": 9450 }, { "grad_norm": 0.2396305799484253, "learning_rate": 5.856387616539656e-05, "loss": 0.0082, "step": 9460 }, { "grad_norm": 0.26889148354530334, "learning_rate": 5.848241268581967e-05, "loss": 0.0085, "step": 9470 }, { "grad_norm": 0.26665395498275757, "learning_rate": 5.840092601565037e-05, "loss": 0.0094, "step": 9480 }, { "grad_norm": 0.23885580897331238, "learning_rate": 5.8319416377670144e-05, "loss": 0.008, "step": 9490 }, { "grad_norm": 0.2632520794868469, "learning_rate": 5.82378839947232e-05, "loss": 0.0098, "step": 9500 }, { "grad_norm": 0.3209339678287506, "learning_rate": 5.815632908971599e-05, "loss": 0.0106, "step": 9510 }, { "grad_norm": 0.282398521900177, "learning_rate": 5.80747518856165e-05, "loss": 0.0099, "step": 9520 }, { "grad_norm": 0.3100825250148773, "learning_rate": 5.799315260545367e-05, "loss": 0.0134, "step": 9530 }, { "grad_norm": 0.2550257444381714, "learning_rate": 5.791153147231686e-05, "loss": 0.0135, "step": 9540 }, { "grad_norm": 0.3137185275554657, "learning_rate": 5.782988870935509e-05, "loss": 0.008, "step": 9550 }, { "grad_norm": 0.23910042643547058, "learning_rate": 5.774822453977657e-05, "loss": 0.0087, "step": 9560 }, { "grad_norm": 0.21105986833572388, "learning_rate": 5.7666539186848036e-05, "loss": 0.009, "step": 9570 }, { "grad_norm": 0.2725152373313904, "learning_rate": 5.758483287389411e-05, "loss": 0.0143, "step": 9580 }, { "grad_norm": 0.2424250990152359, "learning_rate": 5.7503105824296735e-05, "loss": 0.0122, "step": 9590 }, { "grad_norm": 0.20699156820774078, "learning_rate": 5.742135826149453e-05, "loss": 0.0092, "step": 9600 }, { "grad_norm": 0.19423116743564606, "learning_rate": 5.7339590408982223e-05, "loss": 0.0065, "step": 9610 }, { "grad_norm": 0.2932196259498596, "learning_rate": 5.725780249031e-05, "loss": 0.0091, "step": 9620 }, { "grad_norm": 0.3803527057170868, "learning_rate": 5.717599472908292e-05, "loss": 0.0109, "step": 9630 }, { "grad_norm": 0.3079898953437805, "learning_rate": 5.7094167348960237e-05, "loss": 0.0084, "step": 9640 }, { "grad_norm": 0.2345152646303177, "learning_rate": 5.7012320573654945e-05, "loss": 0.0093, "step": 9650 }, { "grad_norm": 0.19556953012943268, "learning_rate": 5.693045462693295e-05, "loss": 0.008, "step": 9660 }, { "grad_norm": 0.22584684193134308, "learning_rate": 5.684856973261266e-05, "loss": 0.0073, "step": 9670 }, { "grad_norm": 0.2571251094341278, "learning_rate": 5.6766666114564215e-05, "loss": 0.0099, "step": 9680 }, { "grad_norm": 0.3253817856311798, "learning_rate": 5.668474399670899e-05, "loss": 0.0089, "step": 9690 }, { "grad_norm": 0.23601661622524261, "learning_rate": 5.660280360301896e-05, "loss": 0.0085, "step": 9700 }, { "grad_norm": 0.18850122392177582, "learning_rate": 5.652084515751599e-05, "loss": 0.0063, "step": 9710 }, { "grad_norm": 0.2345300316810608, "learning_rate": 5.643886888427137e-05, "loss": 0.009, "step": 9720 }, { "grad_norm": 0.21616721153259277, "learning_rate": 5.6356875007405074e-05, "loss": 0.0105, "step": 9730 }, { "grad_norm": 0.23095466196537018, "learning_rate": 5.627486375108525e-05, "loss": 0.0113, "step": 9740 }, { "grad_norm": 0.20307700335979462, "learning_rate": 5.619283533952754e-05, "loss": 0.0098, "step": 9750 }, { "grad_norm": 0.3022615313529968, "learning_rate": 5.6110789996994474e-05, "loss": 0.0087, "step": 9760 }, { "grad_norm": 0.2918921411037445, "learning_rate": 5.602872794779491e-05, "loss": 0.0115, "step": 9770 }, { "grad_norm": 0.3074280023574829, "learning_rate": 5.594664941628334e-05, "loss": 0.0087, "step": 9780 }, { "grad_norm": 0.2652987241744995, "learning_rate": 5.5864554626859324e-05, "loss": 0.0102, "step": 9790 }, { "grad_norm": 0.2364000827074051, "learning_rate": 5.578244380396691e-05, "loss": 0.0071, "step": 9800 }, { "grad_norm": 0.25912168622016907, "learning_rate": 5.570031717209394e-05, "loss": 0.0073, "step": 9810 }, { "grad_norm": 0.2132926732301712, "learning_rate": 5.561817495577147e-05, "loss": 0.0081, "step": 9820 }, { "grad_norm": 0.3217675983905792, "learning_rate": 5.5536017379573215e-05, "loss": 0.0095, "step": 9830 }, { "grad_norm": 0.21416251361370087, "learning_rate": 5.545384466811483e-05, "loss": 0.0088, "step": 9840 }, { "grad_norm": 0.30158206820487976, "learning_rate": 5.5371657046053384e-05, "loss": 0.0127, "step": 9850 }, { "grad_norm": 0.17979058623313904, "learning_rate": 5.528945473808669e-05, "loss": 0.0083, "step": 9860 }, { "grad_norm": 0.1995510756969452, "learning_rate": 5.520723796895272e-05, "loss": 0.0063, "step": 9870 }, { "grad_norm": 0.21220991015434265, "learning_rate": 5.512500696342897e-05, "loss": 0.0077, "step": 9880 }, { "grad_norm": 0.3290112316608429, "learning_rate": 5.504276194633188e-05, "loss": 0.0107, "step": 9890 }, { "grad_norm": 0.28488659858703613, "learning_rate": 5.49605031425162e-05, "loss": 0.0081, "step": 9900 }, { "grad_norm": 0.32763949036598206, "learning_rate": 5.487823077687434e-05, "loss": 0.0135, "step": 9910 }, { "grad_norm": 0.23580661416053772, "learning_rate": 5.4795945074335806e-05, "loss": 0.0094, "step": 9920 }, { "grad_norm": 0.22253672778606415, "learning_rate": 5.471364625986657e-05, "loss": 0.008, "step": 9930 }, { "grad_norm": 0.20649607479572296, "learning_rate": 5.463133455846845e-05, "loss": 0.0064, "step": 9940 }, { "grad_norm": 0.21485736966133118, "learning_rate": 5.4549010195178505e-05, "loss": 0.0142, "step": 9950 }, { "grad_norm": 0.22330300509929657, "learning_rate": 5.446667339506838e-05, "loss": 0.008, "step": 9960 }, { "grad_norm": 0.2618495523929596, "learning_rate": 5.4384324383243756e-05, "loss": 0.0099, "step": 9970 }, { "grad_norm": 0.28015658259391785, "learning_rate": 5.430196338484368e-05, "loss": 0.011, "step": 9980 }, { "grad_norm": 0.20648691058158875, "learning_rate": 5.4219590625039975e-05, "loss": 0.0083, "step": 9990 }, { "grad_norm": 0.22049671411514282, "learning_rate": 5.413720632903664e-05, "loss": 0.008, "step": 10000 }, { "grad_norm": 0.26092201471328735, "learning_rate": 5.405481072206917e-05, "loss": 0.0068, "step": 10010 }, { "grad_norm": 0.2266596406698227, "learning_rate": 5.397240402940402e-05, "loss": 0.0107, "step": 10020 }, { "grad_norm": 0.2024102360010147, "learning_rate": 5.388998647633794e-05, "loss": 0.0081, "step": 10030 }, { "grad_norm": 0.29881739616394043, "learning_rate": 5.380755828819737e-05, "loss": 0.0109, "step": 10040 }, { "grad_norm": 0.31106695532798767, "learning_rate": 5.3725119690337846e-05, "loss": 0.0079, "step": 10050 }, { "grad_norm": 0.22892136871814728, "learning_rate": 5.3642670908143324e-05, "loss": 0.007, "step": 10060 }, { "grad_norm": 0.30826979875564575, "learning_rate": 5.356021216702562e-05, "loss": 0.0073, "step": 10070 }, { "grad_norm": 0.27872055768966675, "learning_rate": 5.347774369242381e-05, "loss": 0.0075, "step": 10080 }, { "grad_norm": 0.3187792897224426, "learning_rate": 5.3395265709803545e-05, "loss": 0.0072, "step": 10090 }, { "grad_norm": 0.24945269525051117, "learning_rate": 5.331277844465647e-05, "loss": 0.0121, "step": 10100 }, { "grad_norm": 0.21182595193386078, "learning_rate": 5.323028212249963e-05, "loss": 0.012, "step": 10110 }, { "grad_norm": 0.28724876046180725, "learning_rate": 5.314777696887481e-05, "loss": 0.0088, "step": 10120 }, { "grad_norm": 0.2590867877006531, "learning_rate": 5.306526320934796e-05, "loss": 0.0099, "step": 10130 }, { "grad_norm": 0.2745634615421295, "learning_rate": 5.298274106950854e-05, "loss": 0.007, "step": 10140 }, { "grad_norm": 0.22778289020061493, "learning_rate": 5.290021077496893e-05, "loss": 0.0081, "step": 10150 }, { "grad_norm": 0.2155156284570694, "learning_rate": 5.2817672551363816e-05, "loss": 0.0073, "step": 10160 }, { "grad_norm": 0.2104647010564804, "learning_rate": 5.273512662434952e-05, "loss": 0.0085, "step": 10170 }, { "grad_norm": 0.2083871066570282, "learning_rate": 5.265257321960349e-05, "loss": 0.0093, "step": 10180 }, { "grad_norm": 0.25575950741767883, "learning_rate": 5.257001256282357e-05, "loss": 0.0086, "step": 10190 }, { "grad_norm": 0.2434515506029129, "learning_rate": 5.248744487972742e-05, "loss": 0.0094, "step": 10200 }, { "grad_norm": 0.28249382972717285, "learning_rate": 5.240487039605196e-05, "loss": 0.0087, "step": 10210 }, { "grad_norm": 0.20960280299186707, "learning_rate": 5.232228933755267e-05, "loss": 0.0097, "step": 10220 }, { "grad_norm": 0.17936058342456818, "learning_rate": 5.2239701930003006e-05, "loss": 0.0065, "step": 10230 }, { "grad_norm": 0.21344617009162903, "learning_rate": 5.215710839919379e-05, "loss": 0.0077, "step": 10240 }, { "grad_norm": 0.2035842388868332, "learning_rate": 5.207450897093257e-05, "loss": 0.0086, "step": 10250 }, { "grad_norm": 0.21530857682228088, "learning_rate": 5.1991903871043046e-05, "loss": 0.0071, "step": 10260 }, { "grad_norm": 0.21515516936779022, "learning_rate": 5.190929332536439e-05, "loss": 0.0084, "step": 10270 }, { "grad_norm": 0.30118030309677124, "learning_rate": 5.182667755975071e-05, "loss": 0.0092, "step": 10280 }, { "grad_norm": 0.2057802826166153, "learning_rate": 5.1744056800070315e-05, "loss": 0.0078, "step": 10290 }, { "grad_norm": 0.26955923438072205, "learning_rate": 5.166143127220524e-05, "loss": 0.0097, "step": 10300 }, { "grad_norm": 0.2716047763824463, "learning_rate": 5.1578801202050485e-05, "loss": 0.0088, "step": 10310 }, { "grad_norm": 0.2820281386375427, "learning_rate": 5.149616681551355e-05, "loss": 0.0088, "step": 10320 }, { "grad_norm": 0.25352105498313904, "learning_rate": 5.141352833851367e-05, "loss": 0.0073, "step": 10330 }, { "grad_norm": 0.15962743759155273, "learning_rate": 5.1330885996981285e-05, "loss": 0.0078, "step": 10340 }, { "grad_norm": 0.22716277837753296, "learning_rate": 5.124824001685741e-05, "loss": 0.0075, "step": 10350 }, { "grad_norm": 0.20729690790176392, "learning_rate": 5.116559062409298e-05, "loss": 0.0059, "step": 10360 }, { "grad_norm": 0.2034323364496231, "learning_rate": 5.10829380446483e-05, "loss": 0.0094, "step": 10370 }, { "grad_norm": 0.21604637801647186, "learning_rate": 5.100028250449235e-05, "loss": 0.008, "step": 10380 }, { "grad_norm": 0.2015807330608368, "learning_rate": 5.0917624229602234e-05, "loss": 0.0065, "step": 10390 }, { "grad_norm": 0.18266531825065613, "learning_rate": 5.0834963445962524e-05, "loss": 0.0089, "step": 10400 }, { "grad_norm": 0.252404123544693, "learning_rate": 5.075230037956461e-05, "loss": 0.0083, "step": 10410 }, { "grad_norm": 0.22738321125507355, "learning_rate": 5.0669635256406213e-05, "loss": 0.0145, "step": 10420 }, { "grad_norm": 0.3449110984802246, "learning_rate": 5.058696830249058e-05, "loss": 0.01, "step": 10430 }, { "grad_norm": 0.30087509751319885, "learning_rate": 5.050429974382602e-05, "loss": 0.0156, "step": 10440 }, { "grad_norm": 0.27621498703956604, "learning_rate": 5.042162980642523e-05, "loss": 0.0075, "step": 10450 }, { "grad_norm": 0.2740698754787445, "learning_rate": 5.033895871630462e-05, "loss": 0.011, "step": 10460 }, { "grad_norm": 0.23271113634109497, "learning_rate": 5.025628669948386e-05, "loss": 0.007, "step": 10470 }, { "grad_norm": 0.25937408208847046, "learning_rate": 5.017361398198502e-05, "loss": 0.0065, "step": 10480 }, { "grad_norm": 0.2318635731935501, "learning_rate": 5.009094078983221e-05, "loss": 0.0069, "step": 10490 }, { "grad_norm": 0.20766286551952362, "learning_rate": 5.000826734905073e-05, "loss": 0.0096, "step": 10500 }, { "grad_norm": 0.2584581971168518, "learning_rate": 4.9925593885666645e-05, "loss": 0.0086, "step": 10510 }, { "grad_norm": 0.2603772282600403, "learning_rate": 4.984292062570602e-05, "loss": 0.0063, "step": 10520 }, { "grad_norm": 0.23702503740787506, "learning_rate": 4.976024779519442e-05, "loss": 0.0079, "step": 10530 }, { "grad_norm": 0.2646651268005371, "learning_rate": 4.9677575620156194e-05, "loss": 0.0092, "step": 10540 }, { "grad_norm": 0.1895063817501068, "learning_rate": 4.959490432661391e-05, "loss": 0.0093, "step": 10550 }, { "grad_norm": 0.21291036903858185, "learning_rate": 4.9512234140587726e-05, "loss": 0.0075, "step": 10560 }, { "grad_norm": 0.20933032035827637, "learning_rate": 4.942956528809477e-05, "loss": 0.0077, "step": 10570 }, { "grad_norm": 0.20019420981407166, "learning_rate": 4.934689799514854e-05, "loss": 0.0097, "step": 10580 }, { "grad_norm": 0.26344871520996094, "learning_rate": 4.926423248775827e-05, "loss": 0.0096, "step": 10590 }, { "grad_norm": 0.20692236721515656, "learning_rate": 4.918156899192826e-05, "loss": 0.0131, "step": 10600 }, { "grad_norm": 0.23554596304893494, "learning_rate": 4.909890773365738e-05, "loss": 0.0137, "step": 10610 }, { "grad_norm": 0.20528851449489594, "learning_rate": 4.9016248938938344e-05, "loss": 0.0105, "step": 10620 }, { "grad_norm": 0.26955676078796387, "learning_rate": 4.8933592833757156e-05, "loss": 0.0132, "step": 10630 }, { "grad_norm": 0.24838532507419586, "learning_rate": 4.8850939644092435e-05, "loss": 0.0136, "step": 10640 }, { "grad_norm": 0.2874491214752197, "learning_rate": 4.876828959591485e-05, "loss": 0.0079, "step": 10650 }, { "grad_norm": 0.22860568761825562, "learning_rate": 4.8685642915186474e-05, "loss": 0.0089, "step": 10660 }, { "grad_norm": 0.23194189369678497, "learning_rate": 4.860299982786018e-05, "loss": 0.0108, "step": 10670 }, { "grad_norm": 0.23060914874076843, "learning_rate": 4.852036055987901e-05, "loss": 0.0097, "step": 10680 }, { "grad_norm": 0.385558545589447, "learning_rate": 4.843772533717558e-05, "loss": 0.0093, "step": 10690 }, { "grad_norm": 0.24155037105083466, "learning_rate": 4.835509438567142e-05, "loss": 0.009, "step": 10700 }, { "grad_norm": 0.25894564390182495, "learning_rate": 4.827246793127639e-05, "loss": 0.0079, "step": 10710 }, { "grad_norm": 0.22689871490001678, "learning_rate": 4.818984619988807e-05, "loss": 0.0059, "step": 10720 }, { "grad_norm": 0.17574647068977356, "learning_rate": 4.810722941739115e-05, "loss": 0.0056, "step": 10730 }, { "grad_norm": 0.2433113157749176, "learning_rate": 4.8024617809656684e-05, "loss": 0.0084, "step": 10740 }, { "grad_norm": 0.28003692626953125, "learning_rate": 4.794201160254171e-05, "loss": 0.0094, "step": 10750 }, { "grad_norm": 0.23488906025886536, "learning_rate": 4.785941102188844e-05, "loss": 0.006, "step": 10760 }, { "grad_norm": 0.2274412214756012, "learning_rate": 4.7776816293523686e-05, "loss": 0.0084, "step": 10770 }, { "grad_norm": 0.19838649034500122, "learning_rate": 4.769422764325832e-05, "loss": 0.0064, "step": 10780 }, { "grad_norm": 0.18760640919208527, "learning_rate": 4.76116452968865e-05, "loss": 0.0062, "step": 10790 }, { "grad_norm": 0.2381298542022705, "learning_rate": 4.752906948018525e-05, "loss": 0.0071, "step": 10800 }, { "grad_norm": 0.2588251829147339, "learning_rate": 4.7446500418913684e-05, "loss": 0.0095, "step": 10810 }, { "grad_norm": 0.2380070835351944, "learning_rate": 4.736393833881247e-05, "loss": 0.0068, "step": 10820 }, { "grad_norm": 0.23917271196842194, "learning_rate": 4.7281383465603194e-05, "loss": 0.0097, "step": 10830 }, { "grad_norm": 0.18200017511844635, "learning_rate": 4.71988360249877e-05, "loss": 0.006, "step": 10840 }, { "grad_norm": 0.23980382084846497, "learning_rate": 4.7116296242647554e-05, "loss": 0.0085, "step": 10850 }, { "grad_norm": 0.24523408710956573, "learning_rate": 4.703376434424336e-05, "loss": 0.0086, "step": 10860 }, { "grad_norm": 0.18079398572444916, "learning_rate": 4.695124055541421e-05, "loss": 0.0086, "step": 10870 }, { "grad_norm": 0.21983122825622559, "learning_rate": 4.6868725101776934e-05, "loss": 0.0076, "step": 10880 }, { "grad_norm": 0.22801896929740906, "learning_rate": 4.678621820892567e-05, "loss": 0.0077, "step": 10890 }, { "grad_norm": 0.23620279133319855, "learning_rate": 4.670372010243111e-05, "loss": 0.0085, "step": 10900 }, { "grad_norm": 0.1909436583518982, "learning_rate": 4.662123100783992e-05, "loss": 0.0094, "step": 10910 }, { "grad_norm": 0.22192321717739105, "learning_rate": 4.653875115067415e-05, "loss": 0.0081, "step": 10920 }, { "grad_norm": 0.17492982745170593, "learning_rate": 4.6456280756430545e-05, "loss": 0.0076, "step": 10930 }, { "grad_norm": 0.17999714612960815, "learning_rate": 4.637382005058004e-05, "loss": 0.0072, "step": 10940 }, { "grad_norm": 0.19913795590400696, "learning_rate": 4.629136925856705e-05, "loss": 0.0108, "step": 10950 }, { "grad_norm": 0.1705617606639862, "learning_rate": 4.6208928605808895e-05, "loss": 0.0086, "step": 10960 }, { "grad_norm": 0.2795408368110657, "learning_rate": 4.612649831769519e-05, "loss": 0.0093, "step": 10970 }, { "grad_norm": 0.2092956155538559, "learning_rate": 4.604407861958715e-05, "loss": 0.0077, "step": 10980 }, { "grad_norm": 0.2425389438867569, "learning_rate": 4.5961669736817114e-05, "loss": 0.0069, "step": 10990 }, { "grad_norm": 0.23790328204631805, "learning_rate": 4.5879271894687814e-05, "loss": 0.0064, "step": 11000 }, { "grad_norm": 0.20760825276374817, "learning_rate": 4.5796885318471826e-05, "loss": 0.008, "step": 11010 }, { "grad_norm": 0.24325376749038696, "learning_rate": 4.571451023341086e-05, "loss": 0.0097, "step": 11020 }, { "grad_norm": 0.24800147116184235, "learning_rate": 4.563214686471527e-05, "loss": 0.0113, "step": 11030 }, { "grad_norm": 0.31807875633239746, "learning_rate": 4.5549795437563365e-05, "loss": 0.0088, "step": 11040 }, { "grad_norm": 0.24229101836681366, "learning_rate": 4.546745617710081e-05, "loss": 0.009, "step": 11050 }, { "grad_norm": 0.22642278671264648, "learning_rate": 4.5385129308440014e-05, "loss": 0.0072, "step": 11060 }, { "grad_norm": 0.2451431155204773, "learning_rate": 4.530281505665944e-05, "loss": 0.0062, "step": 11070 }, { "grad_norm": 0.1859740912914276, "learning_rate": 4.5220513646803134e-05, "loss": 0.0119, "step": 11080 }, { "grad_norm": 0.24542033672332764, "learning_rate": 4.513822530388003e-05, "loss": 0.0085, "step": 11090 }, { "grad_norm": 0.28155162930488586, "learning_rate": 4.5055950252863296e-05, "loss": 0.0074, "step": 11100 }, { "grad_norm": 0.2138509750366211, "learning_rate": 4.4973688718689803e-05, "loss": 0.0079, "step": 11110 }, { "grad_norm": 0.2261638045310974, "learning_rate": 4.4891440926259406e-05, "loss": 0.0071, "step": 11120 }, { "grad_norm": 0.2542162835597992, "learning_rate": 4.480920710043443e-05, "loss": 0.0097, "step": 11130 }, { "grad_norm": 0.15774618089199066, "learning_rate": 4.4726987466039044e-05, "loss": 0.0062, "step": 11140 }, { "grad_norm": 0.23213805258274078, "learning_rate": 4.46447822478586e-05, "loss": 0.0063, "step": 11150 }, { "grad_norm": 0.25310662388801575, "learning_rate": 4.4562591670638974e-05, "loss": 0.0057, "step": 11160 }, { "grad_norm": 0.2297695130109787, "learning_rate": 4.4480415959086105e-05, "loss": 0.0093, "step": 11170 }, { "grad_norm": 0.2621958255767822, "learning_rate": 4.439825533786522e-05, "loss": 0.0071, "step": 11180 }, { "grad_norm": 0.22085212171077728, "learning_rate": 4.431611003160035e-05, "loss": 0.0098, "step": 11190 }, { "grad_norm": 0.18988831341266632, "learning_rate": 4.4233980264873636e-05, "loss": 0.0051, "step": 11200 }, { "grad_norm": 0.2241099625825882, "learning_rate": 4.4151866262224684e-05, "loss": 0.0067, "step": 11210 }, { "grad_norm": 0.27083107829093933, "learning_rate": 4.406976824815006e-05, "loss": 0.0068, "step": 11220 }, { "grad_norm": 0.26719561219215393, "learning_rate": 4.3987686447102595e-05, "loss": 0.0074, "step": 11230 }, { "grad_norm": 0.20571677386760712, "learning_rate": 4.3905621083490804e-05, "loss": 0.0068, "step": 11240 }, { "grad_norm": 0.15599504113197327, "learning_rate": 4.3823572381678286e-05, "loss": 0.006, "step": 11250 }, { "grad_norm": 0.21513454616069794, "learning_rate": 4.374154056598301e-05, "loss": 0.0073, "step": 11260 }, { "grad_norm": 0.22201840579509735, "learning_rate": 4.3659525860676845e-05, "loss": 0.0064, "step": 11270 }, { "grad_norm": 0.22313819825649261, "learning_rate": 4.3577528489984854e-05, "loss": 0.006, "step": 11280 }, { "grad_norm": 0.23398296535015106, "learning_rate": 4.349554867808476e-05, "loss": 0.0061, "step": 11290 }, { "grad_norm": 0.21648810803890228, "learning_rate": 4.34135866491062e-05, "loss": 0.0068, "step": 11300 }, { "grad_norm": 0.2085336595773697, "learning_rate": 4.333164262713022e-05, "loss": 0.0142, "step": 11310 }, { "grad_norm": 0.27589038014411926, "learning_rate": 4.324971683618868e-05, "loss": 0.0101, "step": 11320 }, { "grad_norm": 0.21786828339099884, "learning_rate": 4.316780950026354e-05, "loss": 0.0053, "step": 11330 }, { "grad_norm": 0.2073872834444046, "learning_rate": 4.308592084328637e-05, "loss": 0.009, "step": 11340 }, { "grad_norm": 0.22114667296409607, "learning_rate": 4.3004051089137576e-05, "loss": 0.0079, "step": 11350 }, { "grad_norm": 0.22930344939231873, "learning_rate": 4.292220046164597e-05, "loss": 0.0069, "step": 11360 }, { "grad_norm": 0.22985005378723145, "learning_rate": 4.2840369184588035e-05, "loss": 0.0063, "step": 11370 }, { "grad_norm": 0.18045322597026825, "learning_rate": 4.2758557481687345e-05, "loss": 0.0055, "step": 11380 }, { "grad_norm": 0.2663944661617279, "learning_rate": 4.267676557661403e-05, "loss": 0.0079, "step": 11390 }, { "grad_norm": 0.24303773045539856, "learning_rate": 4.2594993692983955e-05, "loss": 0.0065, "step": 11400 }, { "grad_norm": 0.15300486981868744, "learning_rate": 4.251324205435837e-05, "loss": 0.0067, "step": 11410 }, { "grad_norm": 0.22211404144763947, "learning_rate": 4.243151088424312e-05, "loss": 0.0069, "step": 11420 }, { "grad_norm": 0.21946389973163605, "learning_rate": 4.234980040608813e-05, "loss": 0.0084, "step": 11430 }, { "grad_norm": 0.21228325366973877, "learning_rate": 4.22681108432867e-05, "loss": 0.0063, "step": 11440 }, { "grad_norm": 0.23013344407081604, "learning_rate": 4.2186442419174984e-05, "loss": 0.0063, "step": 11450 }, { "grad_norm": 0.20971117913722992, "learning_rate": 4.210479535703133e-05, "loss": 0.0083, "step": 11460 }, { "grad_norm": 0.22944243252277374, "learning_rate": 4.202316988007567e-05, "loss": 0.0081, "step": 11470 }, { "grad_norm": 0.2118869572877884, "learning_rate": 4.194156621146901e-05, "loss": 0.0087, "step": 11480 }, { "grad_norm": 0.26815682649612427, "learning_rate": 4.1859984574312596e-05, "loss": 0.0093, "step": 11490 }, { "grad_norm": 0.21731431782245636, "learning_rate": 4.177842519164752e-05, "loss": 0.0065, "step": 11500 }, { "grad_norm": 0.2921445965766907, "learning_rate": 4.169688828645404e-05, "loss": 0.0073, "step": 11510 }, { "grad_norm": 0.1848747879266739, "learning_rate": 4.161537408165092e-05, "loss": 0.0056, "step": 11520 }, { "grad_norm": 0.18249960243701935, "learning_rate": 4.1533882800094924e-05, "loss": 0.0082, "step": 11530 }, { "grad_norm": 0.15499596297740936, "learning_rate": 4.145241466458005e-05, "loss": 0.0094, "step": 11540 }, { "grad_norm": 0.19080030918121338, "learning_rate": 4.13709698978371e-05, "loss": 0.0077, "step": 11550 }, { "grad_norm": 0.1779751032590866, "learning_rate": 4.1289548722532944e-05, "loss": 0.007, "step": 11560 }, { "grad_norm": 0.22897697985172272, "learning_rate": 4.120815136126999e-05, "loss": 0.0075, "step": 11570 }, { "grad_norm": 0.2294849157333374, "learning_rate": 4.112677803658548e-05, "loss": 0.0095, "step": 11580 }, { "grad_norm": 0.20816902816295624, "learning_rate": 4.1045428970951e-05, "loss": 0.0066, "step": 11590 }, { "grad_norm": 0.22296547889709473, "learning_rate": 4.0964104386771785e-05, "loss": 0.0088, "step": 11600 }, { "grad_norm": 0.15643584728240967, "learning_rate": 4.0882804506386144e-05, "loss": 0.0053, "step": 11610 }, { "grad_norm": 0.28590163588523865, "learning_rate": 4.080152955206485e-05, "loss": 0.0062, "step": 11620 }, { "grad_norm": 0.18030160665512085, "learning_rate": 4.0720279746010505e-05, "loss": 0.0082, "step": 11630 }, { "grad_norm": 0.19158947467803955, "learning_rate": 4.063905531035699e-05, "loss": 0.0068, "step": 11640 }, { "grad_norm": 0.2148633450269699, "learning_rate": 4.055785646716882e-05, "loss": 0.0067, "step": 11650 }, { "grad_norm": 0.2393154799938202, "learning_rate": 4.047668343844051e-05, "loss": 0.0101, "step": 11660 }, { "grad_norm": 0.31594404578208923, "learning_rate": 4.039553644609604e-05, "loss": 0.0075, "step": 11670 }, { "grad_norm": 0.3099832236766815, "learning_rate": 4.0314415711988176e-05, "loss": 0.0084, "step": 11680 }, { "grad_norm": 0.2928639054298401, "learning_rate": 4.023332145789792e-05, "loss": 0.0064, "step": 11690 }, { "grad_norm": 0.2324325442314148, "learning_rate": 4.015225390553385e-05, "loss": 0.0101, "step": 11700 }, { "grad_norm": 0.1799289435148239, "learning_rate": 4.007121327653158e-05, "loss": 0.0067, "step": 11710 }, { "grad_norm": 0.2345605045557022, "learning_rate": 3.9990199792453064e-05, "loss": 0.0104, "step": 11720 }, { "grad_norm": 0.22300177812576294, "learning_rate": 3.9909213674786103e-05, "loss": 0.0071, "step": 11730 }, { "grad_norm": 0.18316997587680817, "learning_rate": 3.982825514494363e-05, "loss": 0.0116, "step": 11740 }, { "grad_norm": 0.1820177286863327, "learning_rate": 3.974732442426319e-05, "loss": 0.0065, "step": 11750 }, { "grad_norm": 0.2372979074716568, "learning_rate": 3.966642173400629e-05, "loss": 0.0092, "step": 11760 }, { "grad_norm": 0.22073568403720856, "learning_rate": 3.9585547295357764e-05, "loss": 0.0073, "step": 11770 }, { "grad_norm": 0.18935738503932953, "learning_rate": 3.950470132942526e-05, "loss": 0.0062, "step": 11780 }, { "grad_norm": 0.1814570277929306, "learning_rate": 3.942388405723856e-05, "loss": 0.006, "step": 11790 }, { "grad_norm": 0.2537878751754761, "learning_rate": 3.9343095699749e-05, "loss": 0.0092, "step": 11800 }, { "grad_norm": 0.20250701904296875, "learning_rate": 3.9262336477828874e-05, "loss": 0.0086, "step": 11810 }, { "grad_norm": 0.20579147338867188, "learning_rate": 3.9181606612270794e-05, "loss": 0.0068, "step": 11820 }, { "grad_norm": 0.15340419113636017, "learning_rate": 3.910090632378713e-05, "loss": 0.0047, "step": 11830 }, { "grad_norm": 0.21056094765663147, "learning_rate": 3.90202358330094e-05, "loss": 0.0087, "step": 11840 }, { "grad_norm": 0.1647689789533615, "learning_rate": 3.8939595360487656e-05, "loss": 0.0061, "step": 11850 }, { "grad_norm": 0.17378567159175873, "learning_rate": 3.885898512668984e-05, "loss": 0.0055, "step": 11860 }, { "grad_norm": 0.179255411028862, "learning_rate": 3.877840535200127e-05, "loss": 0.0112, "step": 11870 }, { "grad_norm": 0.2528724670410156, "learning_rate": 3.869785625672397e-05, "loss": 0.0057, "step": 11880 }, { "grad_norm": 0.30834850668907166, "learning_rate": 3.8617338061076094e-05, "loss": 0.0065, "step": 11890 }, { "grad_norm": 0.21379484236240387, "learning_rate": 3.853685098519132e-05, "loss": 0.0061, "step": 11900 }, { "grad_norm": 0.21671782433986664, "learning_rate": 3.845639524911823e-05, "loss": 0.0058, "step": 11910 }, { "grad_norm": 0.1984700709581375, "learning_rate": 3.837597107281974e-05, "loss": 0.0061, "step": 11920 }, { "grad_norm": 0.13126368820667267, "learning_rate": 3.829557867617247e-05, "loss": 0.0076, "step": 11930 }, { "grad_norm": 0.22941109538078308, "learning_rate": 3.821521827896618e-05, "loss": 0.0089, "step": 11940 }, { "grad_norm": 0.20077037811279297, "learning_rate": 3.81348901009031e-05, "loss": 0.0079, "step": 11950 }, { "grad_norm": 0.27187028527259827, "learning_rate": 3.805459436159741e-05, "loss": 0.0054, "step": 11960 }, { "grad_norm": 0.250042200088501, "learning_rate": 3.797433128057461e-05, "loss": 0.0098, "step": 11970 }, { "grad_norm": 0.1824905276298523, "learning_rate": 3.789410107727089e-05, "loss": 0.008, "step": 11980 }, { "grad_norm": 0.1979016214609146, "learning_rate": 3.781390397103257e-05, "loss": 0.0078, "step": 11990 }, { "grad_norm": 0.18383477628231049, "learning_rate": 3.7733740181115455e-05, "loss": 0.0097, "step": 12000 }, { "grad_norm": 0.14712467789649963, "learning_rate": 3.7653609926684306e-05, "loss": 0.0064, "step": 12010 }, { "grad_norm": 0.2039942592382431, "learning_rate": 3.757351342681217e-05, "loss": 0.0064, "step": 12020 }, { "grad_norm": 0.19466397166252136, "learning_rate": 3.749345090047982e-05, "loss": 0.0064, "step": 12030 }, { "grad_norm": 0.24284909665584564, "learning_rate": 3.741342256657515e-05, "loss": 0.0065, "step": 12040 }, { "grad_norm": 0.21631167829036713, "learning_rate": 3.7333428643892567e-05, "loss": 0.0053, "step": 12050 }, { "grad_norm": 0.17816664278507233, "learning_rate": 3.725346935113239e-05, "loss": 0.0058, "step": 12060 }, { "grad_norm": 0.17582198977470398, "learning_rate": 3.717354490690029e-05, "loss": 0.0087, "step": 12070 }, { "grad_norm": 0.17063601315021515, "learning_rate": 3.709365552970664e-05, "loss": 0.0066, "step": 12080 }, { "grad_norm": 0.19023405015468597, "learning_rate": 3.7013801437965945e-05, "loss": 0.0062, "step": 12090 }, { "grad_norm": 0.17706483602523804, "learning_rate": 3.693398284999623e-05, "loss": 0.0084, "step": 12100 }, { "grad_norm": 0.20245179533958435, "learning_rate": 3.6854199984018484e-05, "loss": 0.0058, "step": 12110 }, { "grad_norm": 0.20159313082695007, "learning_rate": 3.677445305815601e-05, "loss": 0.0064, "step": 12120 }, { "grad_norm": 0.21612194180488586, "learning_rate": 3.669474229043387e-05, "loss": 0.0077, "step": 12130 }, { "grad_norm": 0.18706907331943512, "learning_rate": 3.6615067898778235e-05, "loss": 0.0077, "step": 12140 }, { "grad_norm": 0.24379310011863708, "learning_rate": 3.6535430101015866e-05, "loss": 0.0088, "step": 12150 }, { "grad_norm": 0.3636125922203064, "learning_rate": 3.645582911487345e-05, "loss": 0.0069, "step": 12160 }, { "grad_norm": 0.20372240245342255, "learning_rate": 3.637626515797706e-05, "loss": 0.0049, "step": 12170 }, { "grad_norm": 0.2063855528831482, "learning_rate": 3.629673844785152e-05, "loss": 0.0058, "step": 12180 }, { "grad_norm": 0.1806401014328003, "learning_rate": 3.621724920191979e-05, "loss": 0.0067, "step": 12190 }, { "grad_norm": 0.21033532917499542, "learning_rate": 3.6137797637502444e-05, "loss": 0.0111, "step": 12200 }, { "grad_norm": 0.24749749898910522, "learning_rate": 3.6058383971817035e-05, "loss": 0.0071, "step": 12210 }, { "grad_norm": 0.22450430691242218, "learning_rate": 3.59790084219775e-05, "loss": 0.0068, "step": 12220 }, { "grad_norm": 0.36298665404319763, "learning_rate": 3.589967120499353e-05, "loss": 0.0114, "step": 12230 }, { "grad_norm": 0.32570943236351013, "learning_rate": 3.5820372537770075e-05, "loss": 0.0074, "step": 12240 }, { "grad_norm": 0.2093043178319931, "learning_rate": 3.5741112637106655e-05, "loss": 0.0057, "step": 12250 }, { "grad_norm": 0.19811396300792694, "learning_rate": 3.5661891719696804e-05, "loss": 0.0094, "step": 12260 }, { "grad_norm": 0.15150699019432068, "learning_rate": 3.5582710002127504e-05, "loss": 0.0057, "step": 12270 }, { "grad_norm": 0.18276816606521606, "learning_rate": 3.550356770087853e-05, "loss": 0.0069, "step": 12280 }, { "grad_norm": 0.1798425316810608, "learning_rate": 3.5424465032321914e-05, "loss": 0.0088, "step": 12290 }, { "grad_norm": 0.22988200187683105, "learning_rate": 3.5345402212721335e-05, "loss": 0.0098, "step": 12300 }, { "grad_norm": 0.1798902451992035, "learning_rate": 3.526637945823152e-05, "loss": 0.0072, "step": 12310 }, { "grad_norm": 0.18393878638744354, "learning_rate": 3.518739698489767e-05, "loss": 0.0066, "step": 12320 }, { "grad_norm": 0.2207607924938202, "learning_rate": 3.510845500865485e-05, "loss": 0.0097, "step": 12330 }, { "grad_norm": 0.24832548201084137, "learning_rate": 3.502955374532739e-05, "loss": 0.0071, "step": 12340 }, { "grad_norm": 0.25870540738105774, "learning_rate": 3.495069341062836e-05, "loss": 0.0075, "step": 12350 }, { "grad_norm": 0.18368877470493317, "learning_rate": 3.4871874220158896e-05, "loss": 0.0068, "step": 12360 }, { "grad_norm": 0.18090079724788666, "learning_rate": 3.479309638940762e-05, "loss": 0.0052, "step": 12370 }, { "grad_norm": 0.16779513657093048, "learning_rate": 3.4714360133750146e-05, "loss": 0.0073, "step": 12380 }, { "grad_norm": 0.15940751135349274, "learning_rate": 3.463566566844839e-05, "loss": 0.0051, "step": 12390 }, { "grad_norm": 0.24262839555740356, "learning_rate": 3.4557013208650016e-05, "loss": 0.0094, "step": 12400 }, { "grad_norm": 0.24909386038780212, "learning_rate": 3.4478402969387857e-05, "loss": 0.0079, "step": 12410 }, { "grad_norm": 0.19093437492847443, "learning_rate": 3.4399835165579266e-05, "loss": 0.0071, "step": 12420 }, { "grad_norm": 0.25968360900878906, "learning_rate": 3.4321310012025645e-05, "loss": 0.0081, "step": 12430 }, { "grad_norm": 0.2364204376935959, "learning_rate": 3.424282772341176e-05, "loss": 0.0084, "step": 12440 }, { "grad_norm": 0.23931537568569183, "learning_rate": 3.416438851430519e-05, "loss": 0.0071, "step": 12450 }, { "grad_norm": 0.21466895937919617, "learning_rate": 3.408599259915577e-05, "loss": 0.0067, "step": 12460 }, { "grad_norm": 0.2476363629102707, "learning_rate": 3.400764019229487e-05, "loss": 0.0056, "step": 12470 }, { "grad_norm": 0.2308126986026764, "learning_rate": 3.3929331507935035e-05, "loss": 0.0091, "step": 12480 }, { "grad_norm": 0.222612202167511, "learning_rate": 3.3851066760169196e-05, "loss": 0.012, "step": 12490 }, { "grad_norm": 0.18202485144138336, "learning_rate": 3.377284616297021e-05, "loss": 0.0065, "step": 12500 }, { "grad_norm": 0.18596436083316803, "learning_rate": 3.3694669930190166e-05, "loss": 0.009, "step": 12510 }, { "grad_norm": 0.21673136949539185, "learning_rate": 3.36165382755599e-05, "loss": 0.0085, "step": 12520 }, { "grad_norm": 0.18619108200073242, "learning_rate": 3.35384514126884e-05, "loss": 0.0049, "step": 12530 }, { "grad_norm": 0.14304818212985992, "learning_rate": 3.3460409555062154e-05, "loss": 0.0058, "step": 12540 }, { "grad_norm": 0.188069149851799, "learning_rate": 3.3382412916044645e-05, "loss": 0.0062, "step": 12550 }, { "grad_norm": 0.2641826570034027, "learning_rate": 3.330446170887566e-05, "loss": 0.0083, "step": 12560 }, { "grad_norm": 0.20169927179813385, "learning_rate": 3.3226556146670834e-05, "loss": 0.0054, "step": 12570 }, { "grad_norm": 0.2031773030757904, "learning_rate": 3.314869644242102e-05, "loss": 0.0058, "step": 12580 }, { "grad_norm": 0.2010979801416397, "learning_rate": 3.3070882808991674e-05, "loss": 0.0126, "step": 12590 }, { "grad_norm": 0.2006341814994812, "learning_rate": 3.2993115459122305e-05, "loss": 0.0056, "step": 12600 }, { "grad_norm": 0.21729214489459991, "learning_rate": 3.2915394605425835e-05, "loss": 0.0081, "step": 12610 }, { "grad_norm": 0.18181125819683075, "learning_rate": 3.283772046038816e-05, "loss": 0.0066, "step": 12620 }, { "grad_norm": 0.1822119951248169, "learning_rate": 3.276009323636739e-05, "loss": 0.0072, "step": 12630 }, { "grad_norm": 0.2782512605190277, "learning_rate": 3.268251314559344e-05, "loss": 0.0063, "step": 12640 }, { "grad_norm": 0.3128264844417572, "learning_rate": 3.2604980400167254e-05, "loss": 0.0083, "step": 12650 }, { "grad_norm": 0.18374745547771454, "learning_rate": 3.252749521206042e-05, "loss": 0.0064, "step": 12660 }, { "grad_norm": 0.2119012176990509, "learning_rate": 3.2450057793114494e-05, "loss": 0.0134, "step": 12670 }, { "grad_norm": 0.1813078075647354, "learning_rate": 3.2372668355040435e-05, "loss": 0.0055, "step": 12680 }, { "grad_norm": 0.3598603308200836, "learning_rate": 3.2295327109418005e-05, "loss": 0.0101, "step": 12690 }, { "grad_norm": 0.2781860828399658, "learning_rate": 3.221803426769518e-05, "loss": 0.0087, "step": 12700 }, { "grad_norm": 0.19660301506519318, "learning_rate": 3.214079004118768e-05, "loss": 0.0088, "step": 12710 }, { "grad_norm": 0.18734294176101685, "learning_rate": 3.2063594641078234e-05, "loss": 0.0093, "step": 12720 }, { "grad_norm": 0.19372877478599548, "learning_rate": 3.198644827841616e-05, "loss": 0.0097, "step": 12730 }, { "grad_norm": 0.1916368454694748, "learning_rate": 3.1909351164116654e-05, "loss": 0.0062, "step": 12740 }, { "grad_norm": 0.1832991987466812, "learning_rate": 3.183230350896026e-05, "loss": 0.0084, "step": 12750 }, { "grad_norm": 0.2088966965675354, "learning_rate": 3.1755305523592337e-05, "loss": 0.0081, "step": 12760 }, { "grad_norm": 0.25852876901626587, "learning_rate": 3.167835741852245e-05, "loss": 0.0076, "step": 12770 }, { "grad_norm": 0.21350952982902527, "learning_rate": 3.160145940412378e-05, "loss": 0.0081, "step": 12780 }, { "grad_norm": 0.22510845959186554, "learning_rate": 3.1524611690632545e-05, "loss": 0.0044, "step": 12790 }, { "grad_norm": 0.1736249029636383, "learning_rate": 3.144781448814746e-05, "loss": 0.0083, "step": 12800 }, { "grad_norm": 0.21229802072048187, "learning_rate": 3.1371068006629145e-05, "loss": 0.0095, "step": 12810 }, { "grad_norm": 0.194345623254776, "learning_rate": 3.129437245589956e-05, "loss": 0.0084, "step": 12820 }, { "grad_norm": 0.2444475144147873, "learning_rate": 3.121772804564143e-05, "loss": 0.0103, "step": 12830 }, { "grad_norm": 0.23777633905410767, "learning_rate": 3.11411349853976e-05, "loss": 0.0071, "step": 12840 }, { "grad_norm": 0.19102996587753296, "learning_rate": 3.10645934845706e-05, "loss": 0.006, "step": 12850 }, { "grad_norm": 0.16968311369419098, "learning_rate": 3.098810375242196e-05, "loss": 0.0068, "step": 12860 }, { "grad_norm": 0.21856744587421417, "learning_rate": 3.0911665998071704e-05, "loss": 0.0082, "step": 12870 }, { "grad_norm": 0.14362917840480804, "learning_rate": 3.083528043049774e-05, "loss": 0.0067, "step": 12880 }, { "grad_norm": 0.18467217683792114, "learning_rate": 3.0758947258535255e-05, "loss": 0.0061, "step": 12890 }, { "grad_norm": 0.18513968586921692, "learning_rate": 3.068266669087625e-05, "loss": 0.0078, "step": 12900 }, { "grad_norm": 0.20277082920074463, "learning_rate": 3.060643893606887e-05, "loss": 0.0074, "step": 12910 }, { "grad_norm": 0.1835511475801468, "learning_rate": 3.053026420251693e-05, "loss": 0.0053, "step": 12920 }, { "grad_norm": 0.16996018588542938, "learning_rate": 3.0454142698479183e-05, "loss": 0.006, "step": 12930 }, { "grad_norm": 0.24543605744838715, "learning_rate": 3.0378074632068954e-05, "loss": 0.0066, "step": 12940 }, { "grad_norm": 0.1390630453824997, "learning_rate": 3.0302060211253408e-05, "loss": 0.0106, "step": 12950 }, { "grad_norm": 0.23156525194644928, "learning_rate": 3.0226099643853073e-05, "loss": 0.0066, "step": 12960 }, { "grad_norm": 0.150277242064476, "learning_rate": 3.0150193137541283e-05, "loss": 0.0075, "step": 12970 }, { "grad_norm": 0.1806921511888504, "learning_rate": 3.0074340899843467e-05, "loss": 0.0078, "step": 12980 }, { "grad_norm": 0.1816614270210266, "learning_rate": 2.999854313813677e-05, "loss": 0.0049, "step": 12990 }, { "grad_norm": 0.18437013030052185, "learning_rate": 2.9922800059649382e-05, "loss": 0.0071, "step": 13000 }, { "grad_norm": 0.20052628219127655, "learning_rate": 2.9847111871459976e-05, "loss": 0.0053, "step": 13010 }, { "grad_norm": 0.1503404974937439, "learning_rate": 2.977147878049721e-05, "loss": 0.0061, "step": 13020 }, { "grad_norm": 0.17255957424640656, "learning_rate": 2.9695900993539006e-05, "loss": 0.0054, "step": 13030 }, { "grad_norm": 0.17958895862102509, "learning_rate": 2.9620378717212183e-05, "loss": 0.0062, "step": 13040 }, { "grad_norm": 0.1710553616285324, "learning_rate": 2.9544912157991745e-05, "loss": 0.0088, "step": 13050 }, { "grad_norm": 0.2009718269109726, "learning_rate": 2.9469501522200405e-05, "loss": 0.0049, "step": 13060 }, { "grad_norm": 0.23603618144989014, "learning_rate": 2.9394147016007946e-05, "loss": 0.0062, "step": 13070 }, { "grad_norm": 0.18876326084136963, "learning_rate": 2.9318848845430702e-05, "loss": 0.0067, "step": 13080 }, { "grad_norm": 0.1263684630393982, "learning_rate": 2.9243607216331013e-05, "loss": 0.004, "step": 13090 }, { "grad_norm": 0.18033087253570557, "learning_rate": 2.916842233441661e-05, "loss": 0.0103, "step": 13100 }, { "grad_norm": 0.17048488557338715, "learning_rate": 2.90932944052401e-05, "loss": 0.0065, "step": 13110 }, { "grad_norm": 0.20619574189186096, "learning_rate": 2.9018223634198354e-05, "loss": 0.0088, "step": 13120 }, { "grad_norm": 0.32705435156822205, "learning_rate": 2.8943210226532025e-05, "loss": 0.0072, "step": 13130 }, { "grad_norm": 0.25954321026802063, "learning_rate": 2.8868254387324857e-05, "loss": 0.0123, "step": 13140 }, { "grad_norm": 0.12593084573745728, "learning_rate": 2.8793356321503306e-05, "loss": 0.0061, "step": 13150 }, { "grad_norm": 0.15029622614383698, "learning_rate": 2.87185162338358e-05, "loss": 0.0061, "step": 13160 }, { "grad_norm": 0.1333160251379013, "learning_rate": 2.8643734328932253e-05, "loss": 0.0081, "step": 13170 }, { "grad_norm": 0.14292830228805542, "learning_rate": 2.856901081124359e-05, "loss": 0.0063, "step": 13180 }, { "grad_norm": 0.17720255255699158, "learning_rate": 2.8494345885061002e-05, "loss": 0.0099, "step": 13190 }, { "grad_norm": 0.2270614057779312, "learning_rate": 2.8419739754515616e-05, "loss": 0.0094, "step": 13200 }, { "grad_norm": 0.20722974836826324, "learning_rate": 2.8345192623577666e-05, "loss": 0.0063, "step": 13210 }, { "grad_norm": 0.19621099531650543, "learning_rate": 2.8270704696056193e-05, "loss": 0.0136, "step": 13220 }, { "grad_norm": 0.16113749146461487, "learning_rate": 2.8196276175598367e-05, "loss": 0.005, "step": 13230 }, { "grad_norm": 0.1010737493634224, "learning_rate": 2.8121907265688884e-05, "loss": 0.0041, "step": 13240 }, { "grad_norm": 0.2032041996717453, "learning_rate": 2.804759816964957e-05, "loss": 0.0093, "step": 13250 }, { "grad_norm": 0.24636270105838776, "learning_rate": 2.797334909063857e-05, "loss": 0.008, "step": 13260 }, { "grad_norm": 0.23668859899044037, "learning_rate": 2.7899160231650056e-05, "loss": 0.0081, "step": 13270 }, { "grad_norm": 0.18992413580417633, "learning_rate": 2.7825031795513585e-05, "loss": 0.0067, "step": 13280 }, { "grad_norm": 0.23678800463676453, "learning_rate": 2.775096398489341e-05, "loss": 0.0054, "step": 13290 }, { "grad_norm": 0.18498824536800385, "learning_rate": 2.7676957002288163e-05, "loss": 0.0066, "step": 13300 }, { "grad_norm": 0.16668982803821564, "learning_rate": 2.760301105003003e-05, "loss": 0.0097, "step": 13310 }, { "grad_norm": 0.2797868847846985, "learning_rate": 2.752912633028446e-05, "loss": 0.0065, "step": 13320 }, { "grad_norm": 0.1960904598236084, "learning_rate": 2.7455303045049474e-05, "loss": 0.0087, "step": 13330 }, { "grad_norm": 0.17691780626773834, "learning_rate": 2.7381541396155098e-05, "loss": 0.008, "step": 13340 }, { "grad_norm": 0.18473778665065765, "learning_rate": 2.730784158526286e-05, "loss": 0.0078, "step": 13350 }, { "grad_norm": 0.20810966193675995, "learning_rate": 2.723420381386521e-05, "loss": 0.0069, "step": 13360 }, { "grad_norm": 0.15738220512866974, "learning_rate": 2.7160628283285018e-05, "loss": 0.0073, "step": 13370 }, { "grad_norm": 0.2037407010793686, "learning_rate": 2.7087115194675007e-05, "loss": 0.0079, "step": 13380 }, { "grad_norm": 0.2184944599866867, "learning_rate": 2.701366474901712e-05, "loss": 0.0082, "step": 13390 }, { "grad_norm": 0.15587963163852692, "learning_rate": 2.6940277147122085e-05, "loss": 0.0059, "step": 13400 }, { "grad_norm": 0.16783714294433594, "learning_rate": 2.686695258962878e-05, "loss": 0.007, "step": 13410 }, { "grad_norm": 0.17781268060207367, "learning_rate": 2.679369127700375e-05, "loss": 0.007, "step": 13420 }, { "grad_norm": 0.18915528059005737, "learning_rate": 2.672049340954067e-05, "loss": 0.0087, "step": 13430 }, { "grad_norm": 0.12304963916540146, "learning_rate": 2.6647359187359676e-05, "loss": 0.0064, "step": 13440 }, { "grad_norm": 0.14121095836162567, "learning_rate": 2.6574288810406946e-05, "loss": 0.0057, "step": 13450 }, { "grad_norm": 0.16274137794971466, "learning_rate": 2.6501282478454083e-05, "loss": 0.0051, "step": 13460 }, { "grad_norm": 0.10483109205961227, "learning_rate": 2.6428340391097618e-05, "loss": 0.0062, "step": 13470 }, { "grad_norm": 0.2279292345046997, "learning_rate": 2.6355462747758485e-05, "loss": 0.0073, "step": 13480 }, { "grad_norm": 0.18314263224601746, "learning_rate": 2.6282649747681304e-05, "loss": 0.0051, "step": 13490 }, { "grad_norm": 0.1461828052997589, "learning_rate": 2.620990158993406e-05, "loss": 0.0058, "step": 13500 }, { "grad_norm": 0.20838026702404022, "learning_rate": 2.6137218473407477e-05, "loss": 0.0061, "step": 13510 }, { "grad_norm": 0.19161008298397064, "learning_rate": 2.606460059681436e-05, "loss": 0.0057, "step": 13520 }, { "grad_norm": 0.1648252159357071, "learning_rate": 2.599204815868928e-05, "loss": 0.0057, "step": 13530 }, { "grad_norm": 0.13666298985481262, "learning_rate": 2.5919561357387756e-05, "loss": 0.0054, "step": 13540 }, { "grad_norm": 0.13384267687797546, "learning_rate": 2.5847140391085972e-05, "loss": 0.0048, "step": 13550 }, { "grad_norm": 0.11047980189323425, "learning_rate": 2.5774785457780103e-05, "loss": 0.0077, "step": 13560 }, { "grad_norm": 0.13502903282642365, "learning_rate": 2.5702496755285753e-05, "loss": 0.0051, "step": 13570 }, { "grad_norm": 0.1964835673570633, "learning_rate": 2.5630274481237483e-05, "loss": 0.0066, "step": 13580 }, { "grad_norm": 0.1709243208169937, "learning_rate": 2.5558118833088197e-05, "loss": 0.006, "step": 13590 }, { "grad_norm": 0.19570232927799225, "learning_rate": 2.548603000810872e-05, "loss": 0.0072, "step": 13600 }, { "grad_norm": 0.1558714509010315, "learning_rate": 2.5414008203387152e-05, "loss": 0.0076, "step": 13610 }, { "grad_norm": 0.18019653856754303, "learning_rate": 2.534205361582834e-05, "loss": 0.0053, "step": 13620 }, { "grad_norm": 0.16142991185188293, "learning_rate": 2.527016644215338e-05, "loss": 0.0088, "step": 13630 }, { "grad_norm": 0.13523423671722412, "learning_rate": 2.519834687889905e-05, "loss": 0.0062, "step": 13640 }, { "grad_norm": 0.15555405616760254, "learning_rate": 2.5126595122417295e-05, "loss": 0.0051, "step": 13650 }, { "grad_norm": 0.1433010846376419, "learning_rate": 2.5054911368874713e-05, "loss": 0.0054, "step": 13660 }, { "grad_norm": 0.1764252781867981, "learning_rate": 2.4983295814251916e-05, "loss": 0.0052, "step": 13670 }, { "grad_norm": 0.2680763006210327, "learning_rate": 2.4911748654343105e-05, "loss": 0.0092, "step": 13680 }, { "grad_norm": 0.2182915061712265, "learning_rate": 2.4840270084755463e-05, "loss": 0.0066, "step": 13690 }, { "grad_norm": 0.21932873129844666, "learning_rate": 2.4768860300908685e-05, "loss": 0.0081, "step": 13700 }, { "grad_norm": 0.16487343609333038, "learning_rate": 2.469751949803443e-05, "loss": 0.0094, "step": 13710 }, { "grad_norm": 0.1933409422636032, "learning_rate": 2.4626247871175666e-05, "loss": 0.0084, "step": 13720 }, { "grad_norm": 0.22529985010623932, "learning_rate": 2.4555045615186346e-05, "loss": 0.006, "step": 13730 }, { "grad_norm": 0.14443622529506683, "learning_rate": 2.4483912924730677e-05, "loss": 0.0048, "step": 13740 }, { "grad_norm": 0.29311496019363403, "learning_rate": 2.4412849994282742e-05, "loss": 0.0077, "step": 13750 }, { "grad_norm": 0.1520247906446457, "learning_rate": 2.434185701812592e-05, "loss": 0.0091, "step": 13760 }, { "grad_norm": 0.2153899073600769, "learning_rate": 2.4270934190352218e-05, "loss": 0.0052, "step": 13770 }, { "grad_norm": 0.18413366377353668, "learning_rate": 2.4200081704861998e-05, "loss": 0.0046, "step": 13780 }, { "grad_norm": 0.15907230973243713, "learning_rate": 2.412929975536321e-05, "loss": 0.0067, "step": 13790 }, { "grad_norm": 0.1581483632326126, "learning_rate": 2.4058588535371017e-05, "loss": 0.0048, "step": 13800 }, { "grad_norm": 0.11827662587165833, "learning_rate": 2.3987948238207243e-05, "loss": 0.0044, "step": 13810 }, { "grad_norm": 0.147860586643219, "learning_rate": 2.3917379056999678e-05, "loss": 0.0054, "step": 13820 }, { "grad_norm": 0.13074715435504913, "learning_rate": 2.3846881184681824e-05, "loss": 0.0064, "step": 13830 }, { "grad_norm": 0.23691308498382568, "learning_rate": 2.377645481399214e-05, "loss": 0.0063, "step": 13840 }, { "grad_norm": 0.1883758157491684, "learning_rate": 2.3706100137473667e-05, "loss": 0.005, "step": 13850 }, { "grad_norm": 0.1975412368774414, "learning_rate": 2.3635817347473394e-05, "loss": 0.0094, "step": 13860 }, { "grad_norm": 0.1934109926223755, "learning_rate": 2.3565606636141757e-05, "loss": 0.0054, "step": 13870 }, { "grad_norm": 0.21237199008464813, "learning_rate": 2.3495468195432203e-05, "loss": 0.0067, "step": 13880 }, { "grad_norm": 0.1800747960805893, "learning_rate": 2.3425402217100507e-05, "loss": 0.0072, "step": 13890 }, { "grad_norm": 0.16725149750709534, "learning_rate": 2.3355408892704424e-05, "loss": 0.0049, "step": 13900 }, { "grad_norm": 0.1488545536994934, "learning_rate": 2.3285488413603003e-05, "loss": 0.0084, "step": 13910 }, { "grad_norm": 0.18190857768058777, "learning_rate": 2.321564097095615e-05, "loss": 0.0054, "step": 13920 }, { "grad_norm": 0.18374863266944885, "learning_rate": 2.3145866755724142e-05, "loss": 0.006, "step": 13930 }, { "grad_norm": 0.11474452167749405, "learning_rate": 2.307616595866699e-05, "loss": 0.0044, "step": 13940 }, { "grad_norm": 0.16919605433940887, "learning_rate": 2.3006538770344032e-05, "loss": 0.0062, "step": 13950 }, { "grad_norm": 0.14991624653339386, "learning_rate": 2.293698538111334e-05, "loss": 0.0081, "step": 13960 }, { "grad_norm": 0.20199425518512726, "learning_rate": 2.28675059811312e-05, "loss": 0.0068, "step": 13970 }, { "grad_norm": 0.15437351167201996, "learning_rate": 2.279810076035167e-05, "loss": 0.0049, "step": 13980 }, { "grad_norm": 0.16406899690628052, "learning_rate": 2.272876990852596e-05, "loss": 0.0073, "step": 13990 }, { "grad_norm": 0.13990004360675812, "learning_rate": 2.265951361520195e-05, "loss": 0.0058, "step": 14000 }, { "grad_norm": 0.1541512906551361, "learning_rate": 2.2590332069723748e-05, "loss": 0.0059, "step": 14010 }, { "grad_norm": 0.15612509846687317, "learning_rate": 2.2521225461231004e-05, "loss": 0.0118, "step": 14020 }, { "grad_norm": 0.10496876388788223, "learning_rate": 2.2452193978658597e-05, "loss": 0.0068, "step": 14030 }, { "grad_norm": 0.2270757555961609, "learning_rate": 2.238323781073594e-05, "loss": 0.0076, "step": 14040 }, { "grad_norm": 0.18619896471500397, "learning_rate": 2.2314357145986552e-05, "loss": 0.0056, "step": 14050 }, { "grad_norm": 0.1321837157011032, "learning_rate": 2.224555217272757e-05, "loss": 0.0044, "step": 14060 }, { "grad_norm": 0.15203258395195007, "learning_rate": 2.2176823079069127e-05, "loss": 0.0055, "step": 14070 }, { "grad_norm": 0.20918844640254974, "learning_rate": 2.210817005291398e-05, "loss": 0.006, "step": 14080 }, { "grad_norm": 0.22175315022468567, "learning_rate": 2.203959328195686e-05, "loss": 0.0061, "step": 14090 }, { "grad_norm": 0.09628665447235107, "learning_rate": 2.1971092953684026e-05, "loss": 0.0048, "step": 14100 }, { "grad_norm": 0.20158182084560394, "learning_rate": 2.1902669255372788e-05, "loss": 0.0045, "step": 14110 }, { "grad_norm": 0.1197918951511383, "learning_rate": 2.1834322374090897e-05, "loss": 0.0077, "step": 14120 }, { "grad_norm": 0.16859853267669678, "learning_rate": 2.1766052496696153e-05, "loss": 0.0061, "step": 14130 }, { "grad_norm": 0.1983485370874405, "learning_rate": 2.169785980983577e-05, "loss": 0.0046, "step": 14140 }, { "grad_norm": 0.18621300160884857, "learning_rate": 2.162974449994593e-05, "loss": 0.0063, "step": 14150 }, { "grad_norm": 0.1051415279507637, "learning_rate": 2.1561706753251337e-05, "loss": 0.0048, "step": 14160 }, { "grad_norm": 0.18732018768787384, "learning_rate": 2.1493746755764544e-05, "loss": 0.0068, "step": 14170 }, { "grad_norm": 0.13659246265888214, "learning_rate": 2.1425864693285635e-05, "loss": 0.0051, "step": 14180 }, { "grad_norm": 0.14370456337928772, "learning_rate": 2.1358060751401547e-05, "loss": 0.0044, "step": 14190 }, { "grad_norm": 0.14926820993423462, "learning_rate": 2.129033511548566e-05, "loss": 0.0052, "step": 14200 }, { "grad_norm": 0.15225178003311157, "learning_rate": 2.1222687970697315e-05, "loss": 0.0071, "step": 14210 }, { "grad_norm": 0.1812531054019928, "learning_rate": 2.1155119501981173e-05, "loss": 0.0063, "step": 14220 }, { "grad_norm": 0.21825918555259705, "learning_rate": 2.1087629894066895e-05, "loss": 0.0065, "step": 14230 }, { "grad_norm": 0.2558552026748657, "learning_rate": 2.1020219331468473e-05, "loss": 0.0077, "step": 14240 }, { "grad_norm": 0.16796202957630157, "learning_rate": 2.095288799848379e-05, "loss": 0.0045, "step": 14250 }, { "grad_norm": 0.15485888719558716, "learning_rate": 2.088563607919417e-05, "loss": 0.005, "step": 14260 }, { "grad_norm": 0.16367082297801971, "learning_rate": 2.0818463757463786e-05, "loss": 0.005, "step": 14270 }, { "grad_norm": 0.23867768049240112, "learning_rate": 2.0751371216939175e-05, "loss": 0.0092, "step": 14280 }, { "grad_norm": 0.2102659046649933, "learning_rate": 2.068435864104882e-05, "loss": 0.0043, "step": 14290 }, { "grad_norm": 0.18887297809123993, "learning_rate": 2.0617426213002506e-05, "loss": 0.0056, "step": 14300 }, { "grad_norm": 0.18655873835086823, "learning_rate": 2.055057411579097e-05, "loss": 0.0061, "step": 14310 }, { "grad_norm": 0.20691068470478058, "learning_rate": 2.0483802532185286e-05, "loss": 0.0061, "step": 14320 }, { "grad_norm": 0.15043525397777557, "learning_rate": 2.041711164473638e-05, "loss": 0.0055, "step": 14330 }, { "grad_norm": 0.14773303270339966, "learning_rate": 2.0350501635774637e-05, "loss": 0.0043, "step": 14340 }, { "grad_norm": 0.16621702909469604, "learning_rate": 2.0283972687409247e-05, "loss": 0.0057, "step": 14350 }, { "grad_norm": 0.14620442688465118, "learning_rate": 2.021752498152784e-05, "loss": 0.0062, "step": 14360 }, { "grad_norm": 0.17155922949314117, "learning_rate": 2.015115869979589e-05, "loss": 0.0054, "step": 14370 }, { "grad_norm": 0.16646042466163635, "learning_rate": 2.0084874023656265e-05, "loss": 0.0045, "step": 14380 }, { "grad_norm": 0.15475299954414368, "learning_rate": 2.001867113432877e-05, "loss": 0.0092, "step": 14390 }, { "grad_norm": 0.1650644838809967, "learning_rate": 1.995255021280954e-05, "loss": 0.006, "step": 14400 }, { "grad_norm": 0.21478888392448425, "learning_rate": 1.9886511439870688e-05, "loss": 0.0055, "step": 14410 }, { "grad_norm": 0.1912056803703308, "learning_rate": 1.9820554996059675e-05, "loss": 0.008, "step": 14420 }, { "grad_norm": 0.16896963119506836, "learning_rate": 1.9754681061698893e-05, "loss": 0.0064, "step": 14430 }, { "grad_norm": 0.17615389823913574, "learning_rate": 1.9688889816885185e-05, "loss": 0.0059, "step": 14440 }, { "grad_norm": 0.17325952649116516, "learning_rate": 1.962318144148928e-05, "loss": 0.0061, "step": 14450 }, { "grad_norm": 0.16517771780490875, "learning_rate": 1.955755611515539e-05, "loss": 0.007, "step": 14460 }, { "grad_norm": 0.11359621584415436, "learning_rate": 1.9492014017300642e-05, "loss": 0.0038, "step": 14470 }, { "grad_norm": 0.1052914410829544, "learning_rate": 1.942655532711461e-05, "loss": 0.0066, "step": 14480 }, { "grad_norm": 0.119536854326725, "learning_rate": 1.9361180223558882e-05, "loss": 0.0055, "step": 14490 }, { "grad_norm": 0.11046026647090912, "learning_rate": 1.929588888536647e-05, "loss": 0.0047, "step": 14500 }, { "grad_norm": 0.13877028226852417, "learning_rate": 1.9230681491041425e-05, "loss": 0.006, "step": 14510 }, { "grad_norm": 0.18733717501163483, "learning_rate": 1.9165558218858264e-05, "loss": 0.005, "step": 14520 }, { "grad_norm": 0.12102721631526947, "learning_rate": 1.9100519246861505e-05, "loss": 0.0052, "step": 14530 }, { "grad_norm": 0.14464715123176575, "learning_rate": 1.9035564752865248e-05, "loss": 0.0048, "step": 14540 }, { "grad_norm": 0.15074194967746735, "learning_rate": 1.897069491445258e-05, "loss": 0.0045, "step": 14550 }, { "grad_norm": 0.15628309547901154, "learning_rate": 1.890590990897515e-05, "loss": 0.0041, "step": 14560 }, { "grad_norm": 0.1759437769651413, "learning_rate": 1.884120991355272e-05, "loss": 0.008, "step": 14570 }, { "grad_norm": 0.14988595247268677, "learning_rate": 1.8776595105072576e-05, "loss": 0.0044, "step": 14580 }, { "grad_norm": 0.15233299136161804, "learning_rate": 1.8712065660189166e-05, "loss": 0.0058, "step": 14590 }, { "grad_norm": 0.118324413895607, "learning_rate": 1.8647621755323513e-05, "loss": 0.0049, "step": 14600 }, { "grad_norm": 0.11879850178956985, "learning_rate": 1.858326356666278e-05, "loss": 0.0049, "step": 14610 }, { "grad_norm": 0.14314360916614532, "learning_rate": 1.851899127015983e-05, "loss": 0.0048, "step": 14620 }, { "grad_norm": 0.12158121913671494, "learning_rate": 1.8454805041532626e-05, "loss": 0.0048, "step": 14630 }, { "grad_norm": 0.1919155865907669, "learning_rate": 1.8390705056263906e-05, "loss": 0.0064, "step": 14640 }, { "grad_norm": 0.2054741382598877, "learning_rate": 1.832669148960057e-05, "loss": 0.0072, "step": 14650 }, { "grad_norm": 0.16221170127391815, "learning_rate": 1.8262764516553233e-05, "loss": 0.0078, "step": 14660 }, { "grad_norm": 0.17063555121421814, "learning_rate": 1.8198924311895843e-05, "loss": 0.0066, "step": 14670 }, { "grad_norm": 0.14749523997306824, "learning_rate": 1.813517105016505e-05, "loss": 0.0044, "step": 14680 }, { "grad_norm": 0.15508879721164703, "learning_rate": 1.8071504905659888e-05, "loss": 0.0049, "step": 14690 }, { "grad_norm": 0.1870722770690918, "learning_rate": 1.800792605244109e-05, "loss": 0.0075, "step": 14700 }, { "grad_norm": 0.11199451237916946, "learning_rate": 1.7944434664330844e-05, "loss": 0.0047, "step": 14710 }, { "grad_norm": 0.15122726559638977, "learning_rate": 1.7881030914912212e-05, "loss": 0.0053, "step": 14720 }, { "grad_norm": 0.18994122743606567, "learning_rate": 1.7817714977528577e-05, "loss": 0.0063, "step": 14730 }, { "grad_norm": 0.14027075469493866, "learning_rate": 1.7754487025283332e-05, "loss": 0.0047, "step": 14740 }, { "grad_norm": 0.15433025360107422, "learning_rate": 1.7691347231039275e-05, "loss": 0.0045, "step": 14750 }, { "grad_norm": 0.12539328634738922, "learning_rate": 1.7628295767418164e-05, "loss": 0.0066, "step": 14760 }, { "grad_norm": 0.16263306140899658, "learning_rate": 1.7565332806800333e-05, "loss": 0.0057, "step": 14770 }, { "grad_norm": 0.20516878366470337, "learning_rate": 1.750245852132408e-05, "loss": 0.0069, "step": 14780 }, { "grad_norm": 0.155752494931221, "learning_rate": 1.7439673082885323e-05, "loss": 0.0042, "step": 14790 }, { "grad_norm": 0.21590086817741394, "learning_rate": 1.7376976663137047e-05, "loss": 0.0068, "step": 14800 }, { "grad_norm": 0.1473667174577713, "learning_rate": 1.7314369433488853e-05, "loss": 0.0086, "step": 14810 }, { "grad_norm": 0.18347211182117462, "learning_rate": 1.7251851565106548e-05, "loss": 0.0054, "step": 14820 }, { "grad_norm": 0.1619826704263687, "learning_rate": 1.7189423228911574e-05, "loss": 0.0052, "step": 14830 }, { "grad_norm": 0.10061666369438171, "learning_rate": 1.7127084595580606e-05, "loss": 0.0056, "step": 14840 }, { "grad_norm": 0.16300608217716217, "learning_rate": 1.706483583554513e-05, "loss": 0.009, "step": 14850 }, { "grad_norm": 0.16306783258914948, "learning_rate": 1.700267711899083e-05, "loss": 0.0121, "step": 14860 }, { "grad_norm": 0.15848280489444733, "learning_rate": 1.69406086158573e-05, "loss": 0.0065, "step": 14870 }, { "grad_norm": 0.1942242980003357, "learning_rate": 1.6878630495837455e-05, "loss": 0.0087, "step": 14880 }, { "grad_norm": 0.262200266122818, "learning_rate": 1.681674292837707e-05, "loss": 0.0063, "step": 14890 }, { "grad_norm": 0.21699029207229614, "learning_rate": 1.6754946082674444e-05, "loss": 0.0072, "step": 14900 }, { "grad_norm": 0.22985686361789703, "learning_rate": 1.6693240127679748e-05, "loss": 0.0044, "step": 14910 }, { "grad_norm": 0.15517984330654144, "learning_rate": 1.663162523209475e-05, "loss": 0.0104, "step": 14920 }, { "grad_norm": 0.1608007550239563, "learning_rate": 1.6570101564372193e-05, "loss": 0.0105, "step": 14930 }, { "grad_norm": 0.23195034265518188, "learning_rate": 1.650866929271543e-05, "loss": 0.0086, "step": 14940 }, { "grad_norm": 0.1510450690984726, "learning_rate": 1.644732858507797e-05, "loss": 0.0067, "step": 14950 }, { "grad_norm": 0.16938410699367523, "learning_rate": 1.6386079609162943e-05, "loss": 0.0054, "step": 14960 }, { "grad_norm": 0.17797565460205078, "learning_rate": 1.6324922532422742e-05, "loss": 0.0052, "step": 14970 }, { "grad_norm": 0.1584869772195816, "learning_rate": 1.6263857522058434e-05, "loss": 0.0079, "step": 14980 }, { "grad_norm": 0.17295604944229126, "learning_rate": 1.6202884745019443e-05, "loss": 0.0069, "step": 14990 }, { "grad_norm": 0.15914912521839142, "learning_rate": 1.614200436800304e-05, "loss": 0.0075, "step": 15000 }, { "grad_norm": 0.1304100751876831, "learning_rate": 1.6081216557453814e-05, "loss": 0.0056, "step": 15010 }, { "grad_norm": 0.1199740543961525, "learning_rate": 1.6020521479563367e-05, "loss": 0.0043, "step": 15020 }, { "grad_norm": 0.14026935398578644, "learning_rate": 1.5959919300269654e-05, "loss": 0.0041, "step": 15030 }, { "grad_norm": 0.12699109315872192, "learning_rate": 1.5899410185256764e-05, "loss": 0.005, "step": 15040 }, { "grad_norm": 0.13951744139194489, "learning_rate": 1.583899429995431e-05, "loss": 0.0044, "step": 15050 }, { "grad_norm": 0.1004301905632019, "learning_rate": 1.5778671809536993e-05, "loss": 0.0072, "step": 15060 }, { "grad_norm": 0.1335887461900711, "learning_rate": 1.5718442878924246e-05, "loss": 0.005, "step": 15070 }, { "grad_norm": 0.17112179100513458, "learning_rate": 1.5658307672779593e-05, "loss": 0.0067, "step": 15080 }, { "grad_norm": 0.16543389856815338, "learning_rate": 1.5598266355510427e-05, "loss": 0.0046, "step": 15090 }, { "grad_norm": 0.13843472301959991, "learning_rate": 1.553831909126744e-05, "loss": 0.0074, "step": 15100 }, { "grad_norm": 0.15110492706298828, "learning_rate": 1.5478466043944135e-05, "loss": 0.0054, "step": 15110 }, { "grad_norm": 0.22796016931533813, "learning_rate": 1.5418707377176468e-05, "loss": 0.0078, "step": 15120 }, { "grad_norm": 0.14223039150238037, "learning_rate": 1.535904325434233e-05, "loss": 0.0054, "step": 15130 }, { "grad_norm": 0.09044605493545532, "learning_rate": 1.529947383856118e-05, "loss": 0.005, "step": 15140 }, { "grad_norm": 0.18607757985591888, "learning_rate": 1.5239999292693524e-05, "loss": 0.0067, "step": 15150 }, { "grad_norm": 0.12722477316856384, "learning_rate": 1.5180619779340505e-05, "loss": 0.0077, "step": 15160 }, { "grad_norm": 0.134136363863945, "learning_rate": 1.5121335460843428e-05, "loss": 0.0045, "step": 15170 }, { "grad_norm": 0.09909097850322723, "learning_rate": 1.5062146499283347e-05, "loss": 0.0036, "step": 15180 }, { "grad_norm": 0.1600300371646881, "learning_rate": 1.5003053056480643e-05, "loss": 0.005, "step": 15190 }, { "grad_norm": 0.0979742482304573, "learning_rate": 1.4944055293994551e-05, "loss": 0.0045, "step": 15200 }, { "grad_norm": 0.1393214762210846, "learning_rate": 1.4885153373122656e-05, "loss": 0.0059, "step": 15210 }, { "grad_norm": 0.16196110844612122, "learning_rate": 1.482634745490059e-05, "loss": 0.0047, "step": 15220 }, { "grad_norm": 0.1884429156780243, "learning_rate": 1.4767637700101466e-05, "loss": 0.0072, "step": 15230 }, { "grad_norm": 0.20021089911460876, "learning_rate": 1.4709024269235528e-05, "loss": 0.0057, "step": 15240 }, { "grad_norm": 0.12900827825069427, "learning_rate": 1.4650507322549684e-05, "loss": 0.0047, "step": 15250 }, { "grad_norm": 0.14503392577171326, "learning_rate": 1.4592087020026972e-05, "loss": 0.0055, "step": 15260 }, { "grad_norm": 0.14478757977485657, "learning_rate": 1.4533763521386318e-05, "loss": 0.0079, "step": 15270 }, { "grad_norm": 0.1581815630197525, "learning_rate": 1.44755369860819e-05, "loss": 0.0098, "step": 15280 }, { "grad_norm": 0.16933442652225494, "learning_rate": 1.441740757330287e-05, "loss": 0.0071, "step": 15290 }, { "grad_norm": 0.1398744434118271, "learning_rate": 1.4359375441972844e-05, "loss": 0.0077, "step": 15300 }, { "grad_norm": 0.11288022249937057, "learning_rate": 1.4301440750749395e-05, "loss": 0.0038, "step": 15310 }, { "grad_norm": 0.12249251455068588, "learning_rate": 1.4243603658023808e-05, "loss": 0.0039, "step": 15320 }, { "grad_norm": 0.14226596057415009, "learning_rate": 1.4185864321920444e-05, "loss": 0.0075, "step": 15330 }, { "grad_norm": 0.16929765045642853, "learning_rate": 1.4128222900296485e-05, "loss": 0.0054, "step": 15340 }, { "grad_norm": 0.12506860494613647, "learning_rate": 1.407067955074135e-05, "loss": 0.0052, "step": 15350 }, { "grad_norm": 0.13736629486083984, "learning_rate": 1.4013234430576356e-05, "loss": 0.0074, "step": 15360 }, { "grad_norm": 0.1849488615989685, "learning_rate": 1.3955887696854286e-05, "loss": 0.0081, "step": 15370 }, { "grad_norm": 0.10855989158153534, "learning_rate": 1.38986395063589e-05, "loss": 0.0045, "step": 15380 }, { "grad_norm": 0.1332855224609375, "learning_rate": 1.3841490015604597e-05, "loss": 0.0051, "step": 15390 }, { "grad_norm": 0.13144734501838684, "learning_rate": 1.3784439380835879e-05, "loss": 0.0057, "step": 15400 }, { "grad_norm": 0.1505713164806366, "learning_rate": 1.3727487758026986e-05, "loss": 0.0046, "step": 15410 }, { "grad_norm": 0.18800579011440277, "learning_rate": 1.3670635302881525e-05, "loss": 0.0073, "step": 15420 }, { "grad_norm": 0.09939403086900711, "learning_rate": 1.3613882170831888e-05, "loss": 0.0035, "step": 15430 }, { "grad_norm": 0.14586606621742249, "learning_rate": 1.355722851703901e-05, "loss": 0.0041, "step": 15440 }, { "grad_norm": 0.15744535624980927, "learning_rate": 1.3500674496391814e-05, "loss": 0.0076, "step": 15450 }, { "grad_norm": 0.11027566343545914, "learning_rate": 1.3444220263506795e-05, "loss": 0.0044, "step": 15460 }, { "grad_norm": 0.16130055487155914, "learning_rate": 1.3387865972727714e-05, "loss": 0.005, "step": 15470 }, { "grad_norm": 0.11999835073947906, "learning_rate": 1.3331611778125036e-05, "loss": 0.0055, "step": 15480 }, { "grad_norm": 0.11079049110412598, "learning_rate": 1.3275457833495564e-05, "loss": 0.0063, "step": 15490 }, { "grad_norm": 0.11384254693984985, "learning_rate": 1.3219404292362065e-05, "loss": 0.005, "step": 15500 }, { "grad_norm": 0.13722357153892517, "learning_rate": 1.3163451307972751e-05, "loss": 0.0099, "step": 15510 }, { "grad_norm": 0.11740435659885406, "learning_rate": 1.3107599033300977e-05, "loss": 0.0037, "step": 15520 }, { "grad_norm": 0.09946759045124054, "learning_rate": 1.305184762104471e-05, "loss": 0.0047, "step": 15530 }, { "grad_norm": 0.09446703642606735, "learning_rate": 1.2996197223626178e-05, "loss": 0.0042, "step": 15540 }, { "grad_norm": 0.14761124551296234, "learning_rate": 1.2940647993191457e-05, "loss": 0.0045, "step": 15550 }, { "grad_norm": 0.15722501277923584, "learning_rate": 1.2885200081610005e-05, "loss": 0.0068, "step": 15560 }, { "grad_norm": 0.15671776235103607, "learning_rate": 1.2829853640474316e-05, "loss": 0.0043, "step": 15570 }, { "grad_norm": 0.08559279143810272, "learning_rate": 1.2774608821099438e-05, "loss": 0.0047, "step": 15580 }, { "grad_norm": 0.13336020708084106, "learning_rate": 1.2719465774522577e-05, "loss": 0.0066, "step": 15590 }, { "grad_norm": 0.17194277048110962, "learning_rate": 1.2664424651502755e-05, "loss": 0.0066, "step": 15600 }, { "grad_norm": 0.17069080471992493, "learning_rate": 1.260948560252026e-05, "loss": 0.0097, "step": 15610 }, { "grad_norm": 0.15968455374240875, "learning_rate": 1.2554648777776396e-05, "loss": 0.0039, "step": 15620 }, { "grad_norm": 0.13884195685386658, "learning_rate": 1.2499914327192919e-05, "loss": 0.004, "step": 15630 }, { "grad_norm": 0.1465432047843933, "learning_rate": 1.2445282400411722e-05, "loss": 0.0055, "step": 15640 }, { "grad_norm": 0.13197597861289978, "learning_rate": 1.2390753146794437e-05, "loss": 0.0052, "step": 15650 }, { "grad_norm": 0.15008197724819183, "learning_rate": 1.2336326715421925e-05, "loss": 0.0046, "step": 15660 }, { "grad_norm": 0.13557665050029755, "learning_rate": 1.2282003255094005e-05, "loss": 0.0045, "step": 15670 }, { "grad_norm": 0.1551276594400406, "learning_rate": 1.2227782914328928e-05, "loss": 0.0065, "step": 15680 }, { "grad_norm": 0.14345529675483704, "learning_rate": 1.2173665841363018e-05, "loss": 0.0033, "step": 15690 }, { "grad_norm": 0.12947356700897217, "learning_rate": 1.211965218415032e-05, "loss": 0.0042, "step": 15700 }, { "grad_norm": 0.08433026820421219, "learning_rate": 1.2065742090362082e-05, "loss": 0.0041, "step": 15710 }, { "grad_norm": 0.16045640408992767, "learning_rate": 1.2011935707386457e-05, "loss": 0.0058, "step": 15720 }, { "grad_norm": 0.10582105070352554, "learning_rate": 1.1958233182328044e-05, "loss": 0.0053, "step": 15730 }, { "grad_norm": 0.0928897112607956, "learning_rate": 1.1904634662007474e-05, "loss": 0.0041, "step": 15740 }, { "grad_norm": 0.13550110161304474, "learning_rate": 1.1851140292961088e-05, "loss": 0.0058, "step": 15750 }, { "grad_norm": 0.1278456449508667, "learning_rate": 1.1797750221440424e-05, "loss": 0.0074, "step": 15760 }, { "grad_norm": 0.194366917014122, "learning_rate": 1.1744464593411897e-05, "loss": 0.0055, "step": 15770 }, { "grad_norm": 0.1663769781589508, "learning_rate": 1.1691283554556399e-05, "loss": 0.0037, "step": 15780 }, { "grad_norm": 0.13579991459846497, "learning_rate": 1.1638207250268834e-05, "loss": 0.0087, "step": 15790 }, { "grad_norm": 0.12101404368877411, "learning_rate": 1.158523582565782e-05, "loss": 0.0059, "step": 15800 }, { "grad_norm": 0.1226835548877716, "learning_rate": 1.1532369425545192e-05, "loss": 0.0057, "step": 15810 }, { "grad_norm": 0.10354090481996536, "learning_rate": 1.1479608194465662e-05, "loss": 0.0097, "step": 15820 }, { "grad_norm": 0.10664427280426025, "learning_rate": 1.1426952276666442e-05, "loss": 0.0054, "step": 15830 }, { "grad_norm": 0.15973882377147675, "learning_rate": 1.1374401816106778e-05, "loss": 0.005, "step": 15840 }, { "grad_norm": 0.12123812735080719, "learning_rate": 1.1321956956457646e-05, "loss": 0.0052, "step": 15850 }, { "grad_norm": 0.15946902334690094, "learning_rate": 1.1269617841101277e-05, "loss": 0.0069, "step": 15860 }, { "grad_norm": 0.10081616789102554, "learning_rate": 1.1217384613130804e-05, "loss": 0.0091, "step": 15870 }, { "grad_norm": 0.11432350426912308, "learning_rate": 1.11652574153499e-05, "loss": 0.0065, "step": 15880 }, { "grad_norm": 0.08289428800344467, "learning_rate": 1.1113236390272303e-05, "loss": 0.0035, "step": 15890 }, { "grad_norm": 0.08323133736848831, "learning_rate": 1.106132168012155e-05, "loss": 0.0047, "step": 15900 }, { "grad_norm": 0.12318079173564911, "learning_rate": 1.1009513426830448e-05, "loss": 0.0044, "step": 15910 }, { "grad_norm": 0.16217155754566193, "learning_rate": 1.0957811772040777e-05, "loss": 0.0064, "step": 15920 }, { "grad_norm": 0.1420840173959732, "learning_rate": 1.0906216857102913e-05, "loss": 0.0059, "step": 15930 }, { "grad_norm": 0.16004914045333862, "learning_rate": 1.0854728823075355e-05, "loss": 0.0054, "step": 15940 }, { "grad_norm": 0.15781958401203156, "learning_rate": 1.0803347810724452e-05, "loss": 0.0096, "step": 15950 }, { "grad_norm": 0.12208251655101776, "learning_rate": 1.0752073960523911e-05, "loss": 0.0034, "step": 15960 }, { "grad_norm": 0.13120897114276886, "learning_rate": 1.070090741265447e-05, "loss": 0.0044, "step": 15970 }, { "grad_norm": 0.10382191836833954, "learning_rate": 1.0649848307003547e-05, "loss": 0.0039, "step": 15980 }, { "grad_norm": 0.09419625997543335, "learning_rate": 1.0598896783164757e-05, "loss": 0.0051, "step": 15990 }, { "grad_norm": 0.10848691314458847, "learning_rate": 1.0548052980437645e-05, "loss": 0.0042, "step": 16000 }, { "grad_norm": 0.13647383451461792, "learning_rate": 1.049731703782722e-05, "loss": 0.0061, "step": 16010 }, { "grad_norm": 0.1732291877269745, "learning_rate": 1.0446689094043587e-05, "loss": 0.0047, "step": 16020 }, { "grad_norm": 0.14692017436027527, "learning_rate": 1.039616928750165e-05, "loss": 0.0096, "step": 16030 }, { "grad_norm": 0.14247895777225494, "learning_rate": 1.0345757756320612e-05, "loss": 0.0047, "step": 16040 }, { "grad_norm": 0.10627839714288712, "learning_rate": 1.0295454638323666e-05, "loss": 0.0036, "step": 16050 }, { "grad_norm": 0.124620720744133, "learning_rate": 1.0245260071037632e-05, "loss": 0.0058, "step": 16060 }, { "grad_norm": 0.1555354744195938, "learning_rate": 1.0195174191692518e-05, "loss": 0.0046, "step": 16070 }, { "grad_norm": 0.16408278048038483, "learning_rate": 1.014519713722124e-05, "loss": 0.0085, "step": 16080 }, { "grad_norm": 0.09864310175180435, "learning_rate": 1.0095329044259132e-05, "loss": 0.0055, "step": 16090 }, { "grad_norm": 0.08889558166265488, "learning_rate": 1.004557004914365e-05, "loss": 0.004, "step": 16100 }, { "grad_norm": 0.10330936312675476, "learning_rate": 9.995920287914007e-06, "loss": 0.0037, "step": 16110 }, { "grad_norm": 0.11861933022737503, "learning_rate": 9.946379896310737e-06, "loss": 0.0048, "step": 16120 }, { "grad_norm": 0.12653601169586182, "learning_rate": 9.896949009775396e-06, "loss": 0.0075, "step": 16130 }, { "grad_norm": 0.09070475399494171, "learning_rate": 9.847627763450134e-06, "loss": 0.0036, "step": 16140 }, { "grad_norm": 0.12315920740365982, "learning_rate": 9.798416292177337e-06, "loss": 0.0055, "step": 16150 }, { "grad_norm": 0.111320361495018, "learning_rate": 9.74931473049932e-06, "loss": 0.0061, "step": 16160 }, { "grad_norm": 0.15200473368167877, "learning_rate": 9.700323212657847e-06, "loss": 0.0047, "step": 16170 }, { "grad_norm": 0.13636194169521332, "learning_rate": 9.65144187259388e-06, "loss": 0.0046, "step": 16180 }, { "grad_norm": 0.1087312325835228, "learning_rate": 9.602670843947132e-06, "loss": 0.007, "step": 16190 }, { "grad_norm": 0.13450048863887787, "learning_rate": 9.554010260055713e-06, "loss": 0.0049, "step": 16200 }, { "grad_norm": 0.09377999603748322, "learning_rate": 9.505460253955834e-06, "loss": 0.0127, "step": 16210 }, { "grad_norm": 0.1321907937526703, "learning_rate": 9.457020958381324e-06, "loss": 0.004, "step": 16220 }, { "grad_norm": 0.1684373915195465, "learning_rate": 9.408692505763395e-06, "loss": 0.0062, "step": 16230 }, { "grad_norm": 0.1340956687927246, "learning_rate": 9.360475028230181e-06, "loss": 0.0069, "step": 16240 }, { "grad_norm": 0.11114493757486343, "learning_rate": 9.312368657606412e-06, "loss": 0.0066, "step": 16250 }, { "grad_norm": 0.08800835907459259, "learning_rate": 9.264373525413096e-06, "loss": 0.0041, "step": 16260 }, { "grad_norm": 0.1255185306072235, "learning_rate": 9.216489762867058e-06, "loss": 0.0051, "step": 16270 }, { "grad_norm": 0.13304315507411957, "learning_rate": 9.168717500880708e-06, "loss": 0.0069, "step": 16280 }, { "grad_norm": 0.13777510821819305, "learning_rate": 9.121056870061574e-06, "loss": 0.0048, "step": 16290 }, { "grad_norm": 0.1482883095741272, "learning_rate": 9.073508000711983e-06, "loss": 0.0063, "step": 16300 }, { "grad_norm": 0.12836310267448425, "learning_rate": 9.026071022828758e-06, "loss": 0.0068, "step": 16310 }, { "grad_norm": 0.13698478043079376, "learning_rate": 8.978746066102771e-06, "loss": 0.0049, "step": 16320 }, { "grad_norm": 0.139384463429451, "learning_rate": 8.931533259918634e-06, "loss": 0.005, "step": 16330 }, { "grad_norm": 0.08790681511163712, "learning_rate": 8.884432733354382e-06, "loss": 0.0062, "step": 16340 }, { "grad_norm": 0.1132570132613182, "learning_rate": 8.837444615181029e-06, "loss": 0.0055, "step": 16350 }, { "grad_norm": 0.15096323192119598, "learning_rate": 8.790569033862323e-06, "loss": 0.0061, "step": 16360 }, { "grad_norm": 0.12137813121080399, "learning_rate": 8.7438061175543e-06, "loss": 0.0058, "step": 16370 }, { "grad_norm": 0.15718196332454681, "learning_rate": 8.697155994104978e-06, "loss": 0.0059, "step": 16380 }, { "grad_norm": 0.12548978626728058, "learning_rate": 8.650618791054033e-06, "loss": 0.0041, "step": 16390 }, { "grad_norm": 0.09397879242897034, "learning_rate": 8.604194635632373e-06, "loss": 0.0035, "step": 16400 }, { "grad_norm": 0.12385103106498718, "learning_rate": 8.557883654761906e-06, "loss": 0.0046, "step": 16410 }, { "grad_norm": 0.12034054100513458, "learning_rate": 8.511685975055061e-06, "loss": 0.0049, "step": 16420 }, { "grad_norm": 0.07118933647871017, "learning_rate": 8.46560172281452e-06, "loss": 0.0037, "step": 16430 }, { "grad_norm": 0.09793756902217865, "learning_rate": 8.419631024032893e-06, "loss": 0.0037, "step": 16440 }, { "grad_norm": 0.08896943926811218, "learning_rate": 8.373774004392293e-06, "loss": 0.0042, "step": 16450 }, { "grad_norm": 0.09761831164360046, "learning_rate": 8.32803078926409e-06, "loss": 0.0042, "step": 16460 }, { "grad_norm": 0.1202096939086914, "learning_rate": 8.282401503708454e-06, "loss": 0.0047, "step": 16470 }, { "grad_norm": 0.08817831426858902, "learning_rate": 8.23688627247412e-06, "loss": 0.0038, "step": 16480 }, { "grad_norm": 0.05380531772971153, "learning_rate": 8.191485219998007e-06, "loss": 0.0031, "step": 16490 }, { "grad_norm": 0.11282724887132645, "learning_rate": 8.146198470404843e-06, "loss": 0.0063, "step": 16500 }, { "grad_norm": 0.13265129923820496, "learning_rate": 8.101026147506897e-06, "loss": 0.0062, "step": 16510 }, { "grad_norm": 0.09475216269493103, "learning_rate": 8.05596837480353e-06, "loss": 0.0048, "step": 16520 }, { "grad_norm": 0.11130519211292267, "learning_rate": 8.011025275480998e-06, "loss": 0.0063, "step": 16530 }, { "grad_norm": 0.09376892447471619, "learning_rate": 7.966196972412027e-06, "loss": 0.0057, "step": 16540 }, { "grad_norm": 0.12341774255037308, "learning_rate": 7.92148358815547e-06, "loss": 0.005, "step": 16550 }, { "grad_norm": 0.1459421068429947, "learning_rate": 7.87688524495604e-06, "loss": 0.006, "step": 16560 }, { "grad_norm": 0.12143165618181229, "learning_rate": 7.83240206474386e-06, "loss": 0.0045, "step": 16570 }, { "grad_norm": 0.0978287011384964, "learning_rate": 7.788034169134272e-06, "loss": 0.0059, "step": 16580 }, { "grad_norm": 0.10168706625699997, "learning_rate": 7.743781679427414e-06, "loss": 0.0044, "step": 16590 }, { "grad_norm": 0.08256179839372635, "learning_rate": 7.699644716607895e-06, "loss": 0.0036, "step": 16600 }, { "grad_norm": 0.059778597205877304, "learning_rate": 7.655623401344486e-06, "loss": 0.0056, "step": 16610 }, { "grad_norm": 0.11977694928646088, "learning_rate": 7.611717853989775e-06, "loss": 0.0059, "step": 16620 }, { "grad_norm": 0.13426047563552856, "learning_rate": 7.567928194579854e-06, "loss": 0.0067, "step": 16630 }, { "grad_norm": 0.12195713818073273, "learning_rate": 7.524254542833997e-06, "loss": 0.0042, "step": 16640 }, { "grad_norm": 0.1642487794160843, "learning_rate": 7.480697018154286e-06, "loss": 0.0061, "step": 16650 }, { "grad_norm": 0.08334190398454666, "learning_rate": 7.437255739625332e-06, "loss": 0.0034, "step": 16660 }, { "grad_norm": 0.09278412163257599, "learning_rate": 7.393930826013923e-06, "loss": 0.0049, "step": 16670 }, { "grad_norm": 0.08380815386772156, "learning_rate": 7.350722395768722e-06, "loss": 0.004, "step": 16680 }, { "grad_norm": 0.08418401330709457, "learning_rate": 7.307630567019963e-06, "loss": 0.0041, "step": 16690 }, { "grad_norm": 0.10579981654882431, "learning_rate": 7.264655457579e-06, "loss": 0.0047, "step": 16700 }, { "grad_norm": 0.09246329963207245, "learning_rate": 7.221797184938184e-06, "loss": 0.0043, "step": 16710 }, { "grad_norm": 0.10854664444923401, "learning_rate": 7.179055866270373e-06, "loss": 0.0037, "step": 16720 }, { "grad_norm": 0.09156115353107452, "learning_rate": 7.136431618428707e-06, "loss": 0.0038, "step": 16730 }, { "grad_norm": 0.056634001433849335, "learning_rate": 7.09392455794628e-06, "loss": 0.0059, "step": 16740 }, { "grad_norm": 0.06214802712202072, "learning_rate": 7.051534801035725e-06, "loss": 0.003, "step": 16750 }, { "grad_norm": 0.11566188186407089, "learning_rate": 7.00926246358905e-06, "loss": 0.0078, "step": 16760 }, { "grad_norm": 0.10004517436027527, "learning_rate": 6.967107661177191e-06, "loss": 0.0043, "step": 16770 }, { "grad_norm": 0.14847685396671295, "learning_rate": 6.925070509049786e-06, "loss": 0.0049, "step": 16780 }, { "grad_norm": 0.08145119249820709, "learning_rate": 6.883151122134812e-06, "loss": 0.0035, "step": 16790 }, { "grad_norm": 0.07774826884269714, "learning_rate": 6.8413496150382394e-06, "loss": 0.0051, "step": 16800 }, { "grad_norm": 0.09139399975538254, "learning_rate": 6.7996661020438165e-06, "loss": 0.0048, "step": 16810 }, { "grad_norm": 0.11714893579483032, "learning_rate": 6.758100697112662e-06, "loss": 0.0049, "step": 16820 }, { "grad_norm": 0.10748549550771713, "learning_rate": 6.716653513883026e-06, "loss": 0.0034, "step": 16830 }, { "grad_norm": 0.13849052786827087, "learning_rate": 6.675324665669913e-06, "loss": 0.0074, "step": 16840 }, { "grad_norm": 0.10507303476333618, "learning_rate": 6.634114265464803e-06, "loss": 0.0037, "step": 16850 }, { "grad_norm": 0.11181323230266571, "learning_rate": 6.59302242593538e-06, "loss": 0.0046, "step": 16860 }, { "grad_norm": 0.0852261483669281, "learning_rate": 6.552049259425141e-06, "loss": 0.0035, "step": 16870 }, { "grad_norm": 0.12089522182941437, "learning_rate": 6.511194877953181e-06, "loss": 0.0071, "step": 16880 }, { "grad_norm": 0.10575058311223984, "learning_rate": 6.470459393213813e-06, "loss": 0.0047, "step": 16890 }, { "grad_norm": 0.1121765598654747, "learning_rate": 6.429842916576279e-06, "loss": 0.0064, "step": 16900 }, { "grad_norm": 0.11497557908296585, "learning_rate": 6.389345559084503e-06, "loss": 0.0038, "step": 16910 }, { "grad_norm": 0.10686445236206055, "learning_rate": 6.348967431456682e-06, "loss": 0.0078, "step": 16920 }, { "grad_norm": 0.11743766814470291, "learning_rate": 6.30870864408511e-06, "loss": 0.007, "step": 16930 }, { "grad_norm": 0.08018866926431656, "learning_rate": 6.268569307035754e-06, "loss": 0.0058, "step": 16940 }, { "grad_norm": 0.0866788849234581, "learning_rate": 6.228549530048022e-06, "loss": 0.0043, "step": 16950 }, { "grad_norm": 0.12864826619625092, "learning_rate": 6.1886494225344814e-06, "loss": 0.0053, "step": 16960 }, { "grad_norm": 0.08638875186443329, "learning_rate": 6.148869093580479e-06, "loss": 0.0041, "step": 16970 }, { "grad_norm": 0.11416413635015488, "learning_rate": 6.109208651943921e-06, "loss": 0.0041, "step": 16980 }, { "grad_norm": 0.08974229544401169, "learning_rate": 6.069668206054946e-06, "loss": 0.0059, "step": 16990 }, { "grad_norm": 0.08993878960609436, "learning_rate": 6.0302478640156145e-06, "loss": 0.0047, "step": 17000 }, { "grad_norm": 0.06425460427999496, "learning_rate": 5.990947733599644e-06, "loss": 0.0041, "step": 17010 }, { "grad_norm": 0.15895497798919678, "learning_rate": 5.951767922252105e-06, "loss": 0.0066, "step": 17020 }, { "grad_norm": 0.12571494281291962, "learning_rate": 5.912708537089068e-06, "loss": 0.004, "step": 17030 }, { "grad_norm": 0.13179023563861847, "learning_rate": 5.873769684897434e-06, "loss": 0.0088, "step": 17040 }, { "grad_norm": 0.12288176268339157, "learning_rate": 5.834951472134514e-06, "loss": 0.0062, "step": 17050 }, { "grad_norm": 0.12752611935138702, "learning_rate": 5.796254004927832e-06, "loss": 0.0082, "step": 17060 }, { "grad_norm": 0.10792287439107895, "learning_rate": 5.757677389074806e-06, "loss": 0.0079, "step": 17070 }, { "grad_norm": 0.12195702642202377, "learning_rate": 5.719221730042385e-06, "loss": 0.0041, "step": 17080 }, { "grad_norm": 0.10768016427755356, "learning_rate": 5.680887132966911e-06, "loss": 0.006, "step": 17090 }, { "grad_norm": 0.09891042858362198, "learning_rate": 5.642673702653683e-06, "loss": 0.004, "step": 17100 }, { "grad_norm": 0.0941566675901413, "learning_rate": 5.604581543576781e-06, "loss": 0.0056, "step": 17110 }, { "grad_norm": 0.10946481674909592, "learning_rate": 5.566610759878704e-06, "loss": 0.0054, "step": 17120 }, { "grad_norm": 0.07776937633752823, "learning_rate": 5.528761455370119e-06, "loss": 0.0032, "step": 17130 }, { "grad_norm": 0.07869040220975876, "learning_rate": 5.491033733529594e-06, "loss": 0.0036, "step": 17140 }, { "grad_norm": 0.08763402700424194, "learning_rate": 5.453427697503255e-06, "loss": 0.0044, "step": 17150 }, { "grad_norm": 0.12128516286611557, "learning_rate": 5.415943450104599e-06, "loss": 0.0061, "step": 17160 }, { "grad_norm": 0.1140231043100357, "learning_rate": 5.378581093814111e-06, "loss": 0.0048, "step": 17170 }, { "grad_norm": 0.07997678965330124, "learning_rate": 5.3413407307790375e-06, "loss": 0.005, "step": 17180 }, { "grad_norm": 0.13127312064170837, "learning_rate": 5.30422246281313e-06, "loss": 0.0057, "step": 17190 }, { "grad_norm": 0.058916617184877396, "learning_rate": 5.267226391396296e-06, "loss": 0.0029, "step": 17200 }, { "grad_norm": 0.1319904774427414, "learning_rate": 5.2303526176744e-06, "loss": 0.0044, "step": 17210 }, { "grad_norm": 0.11650247871875763, "learning_rate": 5.193601242458929e-06, "loss": 0.0051, "step": 17220 }, { "grad_norm": 0.07709518074989319, "learning_rate": 5.156972366226714e-06, "loss": 0.0055, "step": 17230 }, { "grad_norm": 0.08256576210260391, "learning_rate": 5.120466089119735e-06, "loss": 0.0043, "step": 17240 }, { "grad_norm": 0.06434403359889984, "learning_rate": 5.084082510944749e-06, "loss": 0.0027, "step": 17250 }, { "grad_norm": 0.06728126853704453, "learning_rate": 5.047821731173058e-06, "loss": 0.0038, "step": 17260 }, { "grad_norm": 0.090157650411129, "learning_rate": 5.011683848940274e-06, "loss": 0.0051, "step": 17270 }, { "grad_norm": 0.07788987457752228, "learning_rate": 4.975668963045954e-06, "loss": 0.0045, "step": 17280 }, { "grad_norm": 0.1306411623954773, "learning_rate": 4.9397771719534525e-06, "loss": 0.0064, "step": 17290 }, { "grad_norm": 0.10654694586992264, "learning_rate": 4.904008573789548e-06, "loss": 0.0054, "step": 17300 }, { "grad_norm": 0.09149842709302902, "learning_rate": 4.8683632663442005e-06, "loss": 0.0065, "step": 17310 }, { "grad_norm": 0.1496347337961197, "learning_rate": 4.832841347070343e-06, "loss": 0.0057, "step": 17320 }, { "grad_norm": 0.08486293256282806, "learning_rate": 4.797442913083539e-06, "loss": 0.004, "step": 17330 }, { "grad_norm": 0.06311454623937607, "learning_rate": 4.7621680611617596e-06, "loss": 0.0079, "step": 17340 }, { "grad_norm": 0.0772392749786377, "learning_rate": 4.727016887745095e-06, "loss": 0.0059, "step": 17350 }, { "grad_norm": 0.08348176628351212, "learning_rate": 4.691989488935511e-06, "loss": 0.0056, "step": 17360 }, { "grad_norm": 0.08476629853248596, "learning_rate": 4.657085960496588e-06, "loss": 0.0053, "step": 17370 }, { "grad_norm": 0.08960162848234177, "learning_rate": 4.6223063978532265e-06, "loss": 0.0054, "step": 17380 }, { "grad_norm": 0.06759439408779144, "learning_rate": 4.587650896091439e-06, "loss": 0.0028, "step": 17390 }, { "grad_norm": 0.08276814222335815, "learning_rate": 4.553119549958035e-06, "loss": 0.0057, "step": 17400 }, { "grad_norm": 0.10904081165790558, "learning_rate": 4.518712453860385e-06, "loss": 0.0059, "step": 17410 }, { "grad_norm": 0.06815115362405777, "learning_rate": 4.484429701866205e-06, "loss": 0.0032, "step": 17420 }, { "grad_norm": 0.058988507837057114, "learning_rate": 4.4502713877031975e-06, "loss": 0.003, "step": 17430 }, { "grad_norm": 0.07423502206802368, "learning_rate": 4.416237604758911e-06, "loss": 0.0035, "step": 17440 }, { "grad_norm": 0.08950015157461166, "learning_rate": 4.3823284460804025e-06, "loss": 0.0067, "step": 17450 }, { "grad_norm": 0.12428207695484161, "learning_rate": 4.348544004374011e-06, "loss": 0.0047, "step": 17460 }, { "grad_norm": 0.101078100502491, "learning_rate": 4.314884372005123e-06, "loss": 0.0044, "step": 17470 }, { "grad_norm": 0.09248964488506317, "learning_rate": 4.281349640997867e-06, "loss": 0.0046, "step": 17480 }, { "grad_norm": 0.05857797712087631, "learning_rate": 4.247939903034942e-06, "loss": 0.0056, "step": 17490 }, { "grad_norm": 0.08585874736309052, "learning_rate": 4.214655249457284e-06, "loss": 0.0084, "step": 17500 }, { "grad_norm": 0.09325850009918213, "learning_rate": 4.181495771263855e-06, "loss": 0.0038, "step": 17510 }, { "grad_norm": 0.08757645636796951, "learning_rate": 4.148461559111427e-06, "loss": 0.0047, "step": 17520 }, { "grad_norm": 0.0964057445526123, "learning_rate": 4.115552703314252e-06, "loss": 0.0045, "step": 17530 }, { "grad_norm": 0.06798765808343887, "learning_rate": 4.082769293843886e-06, "loss": 0.0033, "step": 17540 }, { "grad_norm": 0.05117792263627052, "learning_rate": 4.050111420328939e-06, "loss": 0.0038, "step": 17550 }, { "grad_norm": 0.1059410348534584, "learning_rate": 4.017579172054764e-06, "loss": 0.0057, "step": 17560 }, { "grad_norm": 0.08445511013269424, "learning_rate": 3.985172637963308e-06, "loss": 0.0041, "step": 17570 }, { "grad_norm": 0.06766512989997864, "learning_rate": 3.952891906652784e-06, "loss": 0.0032, "step": 17580 }, { "grad_norm": 0.10279787331819534, "learning_rate": 3.920737066377478e-06, "loss": 0.005, "step": 17590 }, { "grad_norm": 0.10965480655431747, "learning_rate": 3.888708205047509e-06, "loss": 0.0077, "step": 17600 }, { "grad_norm": 0.09529787302017212, "learning_rate": 3.856805410228542e-06, "loss": 0.005, "step": 17610 }, { "grad_norm": 0.07802815735340118, "learning_rate": 3.82502876914162e-06, "loss": 0.0044, "step": 17620 }, { "grad_norm": 0.063714899122715, "learning_rate": 3.7933783686628586e-06, "loss": 0.0064, "step": 17630 }, { "grad_norm": 0.0940946713089943, "learning_rate": 3.7618542953232306e-06, "loss": 0.0048, "step": 17640 }, { "grad_norm": 0.10446714609861374, "learning_rate": 3.7304566353083658e-06, "loss": 0.0095, "step": 17650 }, { "grad_norm": 0.07436424493789673, "learning_rate": 3.6991854744582555e-06, "loss": 0.0061, "step": 17660 }, { "grad_norm": 0.0746212750673294, "learning_rate": 3.6680408982670777e-06, "loss": 0.0039, "step": 17670 }, { "grad_norm": 0.1222284585237503, "learning_rate": 3.637022991882899e-06, "loss": 0.0084, "step": 17680 }, { "grad_norm": 0.09644754230976105, "learning_rate": 3.606131840107485e-06, "loss": 0.0046, "step": 17690 }, { "grad_norm": 0.15303337574005127, "learning_rate": 3.575367527396084e-06, "loss": 0.0084, "step": 17700 }, { "grad_norm": 0.07879474014043808, "learning_rate": 3.5447301378571386e-06, "loss": 0.0037, "step": 17710 }, { "grad_norm": 0.07187385112047195, "learning_rate": 3.514219755252113e-06, "loss": 0.0065, "step": 17720 }, { "grad_norm": 0.09277717024087906, "learning_rate": 3.4838364629952213e-06, "loss": 0.0074, "step": 17730 }, { "grad_norm": 0.09220196306705475, "learning_rate": 3.4535803441532123e-06, "loss": 0.0038, "step": 17740 }, { "grad_norm": 0.08987642079591751, "learning_rate": 3.4234514814451836e-06, "loss": 0.0043, "step": 17750 }, { "grad_norm": 0.08693262934684753, "learning_rate": 3.393449957242273e-06, "loss": 0.0049, "step": 17760 }, { "grad_norm": 0.12766656279563904, "learning_rate": 3.363575853567524e-06, "loss": 0.0059, "step": 17770 }, { "grad_norm": 0.1169864758849144, "learning_rate": 3.3338292520955826e-06, "loss": 0.0055, "step": 17780 }, { "grad_norm": 0.09048964083194733, "learning_rate": 3.304210234152516e-06, "loss": 0.0034, "step": 17790 }, { "grad_norm": 0.06841456145048141, "learning_rate": 3.2747188807155993e-06, "loss": 0.0031, "step": 17800 }, { "grad_norm": 0.051626693457365036, "learning_rate": 3.2453552724130643e-06, "loss": 0.0054, "step": 17810 }, { "grad_norm": 0.06671755760908127, "learning_rate": 3.216119489523889e-06, "loss": 0.0049, "step": 17820 }, { "grad_norm": 0.09001406282186508, "learning_rate": 3.1870116119775917e-06, "loss": 0.005, "step": 17830 }, { "grad_norm": 0.05401616171002388, "learning_rate": 3.158031719353999e-06, "loss": 0.0043, "step": 17840 }, { "grad_norm": 0.09218642860651016, "learning_rate": 3.1291798908830273e-06, "loss": 0.0066, "step": 17850 }, { "grad_norm": 0.05930738151073456, "learning_rate": 3.1004562054444853e-06, "loss": 0.0037, "step": 17860 }, { "grad_norm": 0.060345377773046494, "learning_rate": 3.071860741567806e-06, "loss": 0.0041, "step": 17870 }, { "grad_norm": 0.04504251852631569, "learning_rate": 3.04339357743193e-06, "loss": 0.0041, "step": 17880 }, { "grad_norm": 0.0834588035941124, "learning_rate": 3.0150547908649628e-06, "loss": 0.0049, "step": 17890 }, { "grad_norm": 0.06421680748462677, "learning_rate": 2.9868444593440957e-06, "loss": 0.0055, "step": 17900 }, { "grad_norm": 0.10178251564502716, "learning_rate": 2.9587626599952846e-06, "loss": 0.0068, "step": 17910 }, { "grad_norm": 0.06935402005910873, "learning_rate": 2.930809469593082e-06, "loss": 0.0067, "step": 17920 }, { "grad_norm": 0.09632204473018646, "learning_rate": 2.9029849645604733e-06, "loss": 0.0046, "step": 17930 }, { "grad_norm": 0.0971589982509613, "learning_rate": 2.8752892209685632e-06, "loss": 0.0053, "step": 17940 }, { "grad_norm": 0.06326834857463837, "learning_rate": 2.847722314536483e-06, "loss": 0.0066, "step": 17950 }, { "grad_norm": 0.08715000748634338, "learning_rate": 2.820284320631078e-06, "loss": 0.0057, "step": 17960 }, { "grad_norm": 0.11452065408229828, "learning_rate": 2.792975314266788e-06, "loss": 0.0046, "step": 17970 }, { "grad_norm": 0.10101866722106934, "learning_rate": 2.7657953701054007e-06, "loss": 0.0077, "step": 17980 }, { "grad_norm": 0.06383367627859116, "learning_rate": 2.7387445624558306e-06, "loss": 0.0084, "step": 17990 }, { "grad_norm": 0.0860927402973175, "learning_rate": 2.7118229652739747e-06, "loss": 0.0068, "step": 18000 }, { "grad_norm": 0.09062928706407547, "learning_rate": 2.6850306521624236e-06, "loss": 0.0077, "step": 18010 }, { "grad_norm": 0.0975007563829422, "learning_rate": 2.6583676963703507e-06, "loss": 0.0053, "step": 18020 }, { "grad_norm": 0.056053951382637024, "learning_rate": 2.631834170793268e-06, "loss": 0.0037, "step": 18030 }, { "grad_norm": 0.09389690309762955, "learning_rate": 2.6054301479728036e-06, "loss": 0.0039, "step": 18040 }, { "grad_norm": 0.08401983976364136, "learning_rate": 2.579155700096575e-06, "loss": 0.0057, "step": 18050 }, { "grad_norm": 0.06630973517894745, "learning_rate": 2.5530108989978873e-06, "loss": 0.007, "step": 18060 }, { "grad_norm": 0.08125243335962296, "learning_rate": 2.5269958161556416e-06, "loss": 0.0059, "step": 18070 }, { "grad_norm": 0.0948253944516182, "learning_rate": 2.5011105226940888e-06, "loss": 0.0065, "step": 18080 }, { "grad_norm": 0.06006864085793495, "learning_rate": 2.4753550893826248e-06, "loss": 0.0025, "step": 18090 }, { "grad_norm": 0.06576282531023026, "learning_rate": 2.4497295866356296e-06, "loss": 0.0082, "step": 18100 }, { "grad_norm": 0.0732223391532898, "learning_rate": 2.424234084512228e-06, "loss": 0.0062, "step": 18110 }, { "grad_norm": 0.0863489881157875, "learning_rate": 2.3988686527161687e-06, "loss": 0.0033, "step": 18120 }, { "grad_norm": 0.05276194587349892, "learning_rate": 2.373633360595573e-06, "loss": 0.0056, "step": 18130 }, { "grad_norm": 0.06433786451816559, "learning_rate": 2.3485282771427585e-06, "loss": 0.0048, "step": 18140 }, { "grad_norm": 0.04832305759191513, "learning_rate": 2.3235534709940665e-06, "loss": 0.0062, "step": 18150 }, { "grad_norm": 0.05605657771229744, "learning_rate": 2.2987090104296617e-06, "loss": 0.0047, "step": 18160 }, { "grad_norm": 0.06013290956616402, "learning_rate": 2.273994963373355e-06, "loss": 0.0055, "step": 18170 }, { "grad_norm": 0.054950859397649765, "learning_rate": 2.249411397392409e-06, "loss": 0.0049, "step": 18180 }, { "grad_norm": 0.04367905482649803, "learning_rate": 2.2249583796973506e-06, "loss": 0.0067, "step": 18190 }, { "grad_norm": 0.06013049557805061, "learning_rate": 2.200635977141796e-06, "loss": 0.004, "step": 18200 }, { "grad_norm": 0.05286560580134392, "learning_rate": 2.17644425622226e-06, "loss": 0.0072, "step": 18210 }, { "grad_norm": 0.04693790152668953, "learning_rate": 2.152383283077991e-06, "loss": 0.0033, "step": 18220 }, { "grad_norm": 0.08133646100759506, "learning_rate": 2.128453123490781e-06, "loss": 0.0049, "step": 18230 }, { "grad_norm": 0.11808016151189804, "learning_rate": 2.1046538428847462e-06, "loss": 0.0075, "step": 18240 }, { "grad_norm": 0.13297493755817413, "learning_rate": 2.0809855063262273e-06, "loss": 0.0039, "step": 18250 }, { "grad_norm": 0.06416473537683487, "learning_rate": 2.057448178523558e-06, "loss": 0.0042, "step": 18260 }, { "grad_norm": 0.08213549107313156, "learning_rate": 2.034041923826885e-06, "loss": 0.004, "step": 18270 }, { "grad_norm": 0.08681398630142212, "learning_rate": 2.0107668062280204e-06, "loss": 0.0062, "step": 18280 }, { "grad_norm": 0.04406105354428291, "learning_rate": 1.9876228893602357e-06, "loss": 0.0026, "step": 18290 }, { "grad_norm": 0.09231877326965332, "learning_rate": 1.9646102364981266e-06, "loss": 0.0048, "step": 18300 }, { "grad_norm": 0.09094957262277603, "learning_rate": 1.9417289105574053e-06, "loss": 0.0045, "step": 18310 }, { "grad_norm": 0.07368715852499008, "learning_rate": 1.9189789740947427e-06, "loss": 0.0048, "step": 18320 }, { "grad_norm": 0.05653895437717438, "learning_rate": 1.896360489307597e-06, "loss": 0.0071, "step": 18330 }, { "grad_norm": 0.061923108994960785, "learning_rate": 1.8738735180340362e-06, "loss": 0.0049, "step": 18340 }, { "grad_norm": 0.06822076439857483, "learning_rate": 1.8515181217525824e-06, "loss": 0.0044, "step": 18350 }, { "grad_norm": 0.06153898686170578, "learning_rate": 1.8292943615820457e-06, "loss": 0.0038, "step": 18360 }, { "grad_norm": 0.0566900372505188, "learning_rate": 1.8072022982813296e-06, "loss": 0.0047, "step": 18370 }, { "grad_norm": 0.10939788818359375, "learning_rate": 1.7852419922492925e-06, "loss": 0.0052, "step": 18380 }, { "grad_norm": 0.06535179167985916, "learning_rate": 1.763413503524569e-06, "loss": 0.0033, "step": 18390 }, { "grad_norm": 0.11079724133014679, "learning_rate": 1.7417168917854165e-06, "loss": 0.0075, "step": 18400 }, { "grad_norm": 0.08610212802886963, "learning_rate": 1.720152216349552e-06, "loss": 0.0078, "step": 18410 }, { "grad_norm": 0.05728074163198471, "learning_rate": 1.6987195361739595e-06, "loss": 0.003, "step": 18420 }, { "grad_norm": 0.08600207418203354, "learning_rate": 1.6774189098547832e-06, "loss": 0.0081, "step": 18430 }, { "grad_norm": 0.08694013953208923, "learning_rate": 1.6562503956271069e-06, "loss": 0.0043, "step": 18440 }, { "grad_norm": 0.06410788744688034, "learning_rate": 1.6352140513648417e-06, "loss": 0.0041, "step": 18450 }, { "grad_norm": 0.035524092614650726, "learning_rate": 1.6143099345805712e-06, "loss": 0.0046, "step": 18460 }, { "grad_norm": 0.06390909850597382, "learning_rate": 1.5935381024253293e-06, "loss": 0.0055, "step": 18470 }, { "grad_norm": 0.05754880607128143, "learning_rate": 1.572898611688517e-06, "loss": 0.0066, "step": 18480 }, { "grad_norm": 0.048745110630989075, "learning_rate": 1.5523915187977133e-06, "loss": 0.0034, "step": 18490 }, { "grad_norm": 0.051464542746543884, "learning_rate": 1.532016879818532e-06, "loss": 0.0052, "step": 18500 }, { "grad_norm": 0.04558296129107475, "learning_rate": 1.51177475045447e-06, "loss": 0.0039, "step": 18510 }, { "grad_norm": 0.050785310566425323, "learning_rate": 1.4916651860467035e-06, "loss": 0.0044, "step": 18520 }, { "grad_norm": 0.07850246131420135, "learning_rate": 1.471688241574043e-06, "loss": 0.0051, "step": 18530 }, { "grad_norm": 0.043779075145721436, "learning_rate": 1.451843971652672e-06, "loss": 0.0056, "step": 18540 }, { "grad_norm": 0.09892037510871887, "learning_rate": 1.432132430536076e-06, "loss": 0.005, "step": 18550 }, { "grad_norm": 0.0711660087108612, "learning_rate": 1.412553672114869e-06, "loss": 0.0049, "step": 18560 }, { "grad_norm": 0.040360577404499054, "learning_rate": 1.3931077499166056e-06, "loss": 0.0041, "step": 18570 }, { "grad_norm": 0.10091473162174225, "learning_rate": 1.3737947171057085e-06, "loss": 0.0053, "step": 18580 }, { "grad_norm": 0.06029286980628967, "learning_rate": 1.3546146264832582e-06, "loss": 0.0051, "step": 18590 }, { "grad_norm": 0.06340325623750687, "learning_rate": 1.3355675304869086e-06, "loss": 0.0054, "step": 18600 }, { "grad_norm": 0.059224601835012436, "learning_rate": 1.3166534811906827e-06, "loss": 0.0037, "step": 18610 }, { "grad_norm": 0.06791342049837112, "learning_rate": 1.2978725303048666e-06, "loss": 0.0042, "step": 18620 }, { "grad_norm": 0.0800284743309021, "learning_rate": 1.2792247291758762e-06, "loss": 0.0041, "step": 18630 }, { "grad_norm": 0.08240664750337601, "learning_rate": 1.2607101287860635e-06, "loss": 0.0038, "step": 18640 }, { "grad_norm": 0.047691259533166885, "learning_rate": 1.2423287797536654e-06, "loss": 0.0047, "step": 18650 }, { "grad_norm": 0.06712372601032257, "learning_rate": 1.2240807323325776e-06, "loss": 0.004, "step": 18660 }, { "grad_norm": 0.049060530960559845, "learning_rate": 1.205966036412254e-06, "loss": 0.0031, "step": 18670 }, { "grad_norm": 0.03923507779836655, "learning_rate": 1.1879847415175949e-06, "loss": 0.0057, "step": 18680 }, { "grad_norm": 0.11399254202842712, "learning_rate": 1.1701368968087712e-06, "loss": 0.0053, "step": 18690 }, { "grad_norm": 0.0680961012840271, "learning_rate": 1.1524225510811116e-06, "loss": 0.0046, "step": 18700 }, { "grad_norm": 0.04934504255652428, "learning_rate": 1.1348417527649535e-06, "loss": 0.0041, "step": 18710 }, { "grad_norm": 0.0424002967774868, "learning_rate": 1.1173945499255268e-06, "loss": 0.0048, "step": 18720 }, { "grad_norm": 0.037728987634181976, "learning_rate": 1.1000809902628307e-06, "loss": 0.003, "step": 18730 }, { "grad_norm": 0.03390166163444519, "learning_rate": 1.082901121111468e-06, "loss": 0.0041, "step": 18740 }, { "grad_norm": 0.044707078486680984, "learning_rate": 1.0658549894405456e-06, "loss": 0.0039, "step": 18750 }, { "grad_norm": 0.0984845906496048, "learning_rate": 1.0489426418535342e-06, "loss": 0.0056, "step": 18760 }, { "grad_norm": 0.03741137310862541, "learning_rate": 1.0321641245881474e-06, "loss": 0.0042, "step": 18770 }, { "grad_norm": 0.03353625163435936, "learning_rate": 1.015519483516214e-06, "loss": 0.0029, "step": 18780 }, { "grad_norm": 0.03772426396608353, "learning_rate": 9.990087641435443e-07, "loss": 0.0035, "step": 18790 }, { "grad_norm": 0.10662414133548737, "learning_rate": 9.826320116098132e-07, "loss": 0.0068, "step": 18800 }, { "grad_norm": 0.03115723840892315, "learning_rate": 9.663892706884447e-07, "loss": 0.0024, "step": 18810 }, { "grad_norm": 0.042881883680820465, "learning_rate": 9.502805857864616e-07, "loss": 0.004, "step": 18820 }, { "grad_norm": 0.0726175531744957, "learning_rate": 9.34306000944396e-07, "loss": 0.0039, "step": 18830 }, { "grad_norm": 0.10121230781078339, "learning_rate": 9.184655598361624e-07, "loss": 0.0064, "step": 18840 }, { "grad_norm": 0.07452309876680374, "learning_rate": 9.027593057689076e-07, "loss": 0.0077, "step": 18850 }, { "grad_norm": 0.047714851796627045, "learning_rate": 8.871872816829441e-07, "loss": 0.0043, "step": 18860 }, { "grad_norm": 0.05202336609363556, "learning_rate": 8.717495301515777e-07, "loss": 0.0063, "step": 18870 }, { "grad_norm": 0.026929423213005066, "learning_rate": 8.564460933810415e-07, "loss": 0.0032, "step": 18880 }, { "grad_norm": 0.1266394555568695, "learning_rate": 8.412770132103453e-07, "loss": 0.0078, "step": 18890 }, { "grad_norm": 0.04397597163915634, "learning_rate": 8.262423311111711e-07, "loss": 0.0046, "step": 18900 }, { "grad_norm": 0.04359564557671547, "learning_rate": 8.113420881877665e-07, "loss": 0.0032, "step": 18910 }, { "grad_norm": 0.05606939271092415, "learning_rate": 7.965763251768288e-07, "loss": 0.0047, "step": 18920 }, { "grad_norm": 0.032073911279439926, "learning_rate": 7.819450824473995e-07, "loss": 0.0044, "step": 18930 }, { "grad_norm": 0.052406635135412216, "learning_rate": 7.674484000007198e-07, "loss": 0.0043, "step": 18940 }, { "grad_norm": 0.06158546730875969, "learning_rate": 7.530863174701752e-07, "loss": 0.0052, "step": 18950 }, { "grad_norm": 0.06597442924976349, "learning_rate": 7.38858874121151e-07, "loss": 0.0061, "step": 18960 }, { "grad_norm": 0.04494604840874672, "learning_rate": 7.247661088509328e-07, "loss": 0.0054, "step": 18970 }, { "grad_norm": 0.03968639299273491, "learning_rate": 7.108080601886002e-07, "loss": 0.0034, "step": 18980 }, { "grad_norm": 0.04516436532139778, "learning_rate": 6.969847662949336e-07, "loss": 0.0046, "step": 18990 }, { "grad_norm": 0.03492464870214462, "learning_rate": 6.832962649622798e-07, "loss": 0.0041, "step": 19000 }, { "grad_norm": 0.03456168621778488, "learning_rate": 6.697425936144863e-07, "loss": 0.007, "step": 19010 }, { "grad_norm": 0.059465594589710236, "learning_rate": 6.563237893067731e-07, "loss": 0.0047, "step": 19020 }, { "grad_norm": 0.08866643160581589, "learning_rate": 6.430398887256328e-07, "loss": 0.0032, "step": 19030 }, { "grad_norm": 0.037074267864227295, "learning_rate": 6.298909281887478e-07, "loss": 0.0043, "step": 19040 }, { "grad_norm": 0.03368697687983513, "learning_rate": 6.168769436448673e-07, "loss": 0.0033, "step": 19050 }, { "grad_norm": 0.058988627046346664, "learning_rate": 6.03997970673742e-07, "loss": 0.003, "step": 19060 }, { "grad_norm": 0.04054597020149231, "learning_rate": 5.912540444859782e-07, "loss": 0.0054, "step": 19070 }, { "grad_norm": 0.04888832941651344, "learning_rate": 5.786451999229837e-07, "loss": 0.0043, "step": 19080 }, { "grad_norm": 0.09000938385725021, "learning_rate": 5.661714714568722e-07, "loss": 0.0075, "step": 19090 }, { "grad_norm": 0.08032537251710892, "learning_rate": 5.538328931903259e-07, "loss": 0.0065, "step": 19100 }, { "grad_norm": 0.03778580576181412, "learning_rate": 5.416294988565551e-07, "loss": 0.0038, "step": 19110 }, { "grad_norm": 0.0301471296697855, "learning_rate": 5.29561321819172e-07, "loss": 0.0064, "step": 19120 }, { "grad_norm": 0.04304756596684456, "learning_rate": 5.176283950721061e-07, "loss": 0.0037, "step": 19130 }, { "grad_norm": 0.08168292790651321, "learning_rate": 5.058307512395332e-07, "loss": 0.006, "step": 19140 }, { "grad_norm": 0.03314193710684776, "learning_rate": 4.941684225757526e-07, "loss": 0.0072, "step": 19150 }, { "grad_norm": 0.03319231793284416, "learning_rate": 4.826414409651314e-07, "loss": 0.0051, "step": 19160 }, { "grad_norm": 0.04385793209075928, "learning_rate": 4.712498379219943e-07, "loss": 0.0063, "step": 19170 }, { "grad_norm": 0.08488308638334274, "learning_rate": 4.599936445905506e-07, "loss": 0.0073, "step": 19180 }, { "grad_norm": 0.026116928085684776, "learning_rate": 4.4887289174480594e-07, "loss": 0.0029, "step": 19190 }, { "grad_norm": 0.038966987282037735, "learning_rate": 4.378876097884621e-07, "loss": 0.005, "step": 19200 }, { "grad_norm": 0.055600617080926895, "learning_rate": 4.2703782875487264e-07, "loss": 0.0062, "step": 19210 }, { "grad_norm": 0.029152419418096542, "learning_rate": 4.163235783069208e-07, "loss": 0.0047, "step": 19220 }, { "grad_norm": 0.02960650622844696, "learning_rate": 4.057448877369585e-07, "loss": 0.0037, "step": 19230 }, { "grad_norm": 0.05459560081362724, "learning_rate": 3.9530178596672295e-07, "loss": 0.0035, "step": 19240 }, { "grad_norm": 0.0414460264146328, "learning_rate": 3.849943015472479e-07, "loss": 0.0036, "step": 19250 }, { "grad_norm": 0.05241796001791954, "learning_rate": 3.748224626588137e-07, "loss": 0.0052, "step": 19260 }, { "grad_norm": 0.06404601037502289, "learning_rate": 3.647862971108307e-07, "loss": 0.0092, "step": 19270 }, { "grad_norm": 0.03816540539264679, "learning_rate": 3.5488583234179473e-07, "loss": 0.0034, "step": 19280 }, { "grad_norm": 0.042802464216947556, "learning_rate": 3.4512109541920413e-07, "loss": 0.0067, "step": 19290 }, { "grad_norm": 0.0605001300573349, "learning_rate": 3.354921130394706e-07, "loss": 0.0053, "step": 19300 }, { "grad_norm": 0.08837027847766876, "learning_rate": 3.259989115278639e-07, "loss": 0.0087, "step": 19310 }, { "grad_norm": 0.10174474865198135, "learning_rate": 3.1664151683843403e-07, "loss": 0.0043, "step": 19320 }, { "grad_norm": 0.03807970881462097, "learning_rate": 3.074199545539447e-07, "loss": 0.0027, "step": 19330 }, { "grad_norm": 0.02359367161989212, "learning_rate": 2.983342498857955e-07, "loss": 0.0043, "step": 19340 }, { "grad_norm": 0.034910764545202255, "learning_rate": 2.893844276739499e-07, "loss": 0.0035, "step": 19350 }, { "grad_norm": 0.08157561719417572, "learning_rate": 2.8057051238688514e-07, "loss": 0.0036, "step": 19360 }, { "grad_norm": 0.03558242693543434, "learning_rate": 2.71892528121509e-07, "loss": 0.0056, "step": 19370 }, { "grad_norm": 0.04014763608574867, "learning_rate": 2.633504986030988e-07, "loss": 0.006, "step": 19380 }, { "grad_norm": 0.049834150820970535, "learning_rate": 2.549444471852347e-07, "loss": 0.0035, "step": 19390 }, { "grad_norm": 0.04006095975637436, "learning_rate": 2.4667439684974423e-07, "loss": 0.0067, "step": 19400 }, { "grad_norm": 0.034812238067388535, "learning_rate": 2.3854037020662467e-07, "loss": 0.0041, "step": 19410 }, { "grad_norm": 0.04246693477034569, "learning_rate": 2.3054238949399288e-07, "loss": 0.0056, "step": 19420 }, { "grad_norm": 0.061797671020030975, "learning_rate": 2.2268047657802993e-07, "loss": 0.005, "step": 19430 }, { "grad_norm": 0.029228825122117996, "learning_rate": 2.149546529529034e-07, "loss": 0.005, "step": 19440 }, { "grad_norm": 0.030314629897475243, "learning_rate": 2.0736493974071736e-07, "loss": 0.0042, "step": 19450 }, { "grad_norm": 0.054420553147792816, "learning_rate": 1.9991135769145686e-07, "loss": 0.0044, "step": 19460 }, { "grad_norm": 0.026409875601530075, "learning_rate": 1.9259392718293245e-07, "loss": 0.004, "step": 19470 }, { "grad_norm": 0.026196150109171867, "learning_rate": 1.8541266822072467e-07, "loss": 0.0058, "step": 19480 }, { "grad_norm": 0.07092944532632828, "learning_rate": 1.7836760043811184e-07, "loss": 0.0037, "step": 19490 }, { "grad_norm": 0.10308902710676193, "learning_rate": 1.7145874309604792e-07, "loss": 0.008, "step": 19500 }, { "grad_norm": 0.06316528469324112, "learning_rate": 1.6468611508308474e-07, "loss": 0.0056, "step": 19510 }, { "grad_norm": 0.04925818368792534, "learning_rate": 1.5804973491532204e-07, "loss": 0.0041, "step": 19520 }, { "grad_norm": 0.06427190452814102, "learning_rate": 1.5154962073637424e-07, "loss": 0.0044, "step": 19530 }, { "grad_norm": 0.027820497751235962, "learning_rate": 1.4518579031730372e-07, "loss": 0.0035, "step": 19540 }, { "grad_norm": 0.08228770643472672, "learning_rate": 1.389582610565876e-07, "loss": 0.0038, "step": 19550 }, { "grad_norm": 0.035630207508802414, "learning_rate": 1.3286704998003995e-07, "loss": 0.0029, "step": 19560 }, { "grad_norm": 0.06858760118484497, "learning_rate": 1.2691217374080632e-07, "loss": 0.0059, "step": 19570 }, { "grad_norm": 0.030942440032958984, "learning_rate": 1.2109364861929705e-07, "loss": 0.0028, "step": 19580 }, { "grad_norm": 0.03261679783463478, "learning_rate": 1.1541149052312628e-07, "loss": 0.0048, "step": 19590 }, { "grad_norm": 0.02540130726993084, "learning_rate": 1.0986571498710074e-07, "loss": 0.0046, "step": 19600 }, { "grad_norm": 0.023928454145789146, "learning_rate": 1.0445633717316438e-07, "loss": 0.0136, "step": 19610 }, { "grad_norm": 0.03369027003645897, "learning_rate": 9.918337187034277e-08, "loss": 0.0058, "step": 19620 }, { "grad_norm": 0.03948524594306946, "learning_rate": 9.404683349472643e-08, "loss": 0.0055, "step": 19630 }, { "grad_norm": 0.03470523655414581, "learning_rate": 8.904673608940983e-08, "loss": 0.0056, "step": 19640 }, { "grad_norm": 0.024040725082159042, "learning_rate": 8.418309332447471e-08, "loss": 0.0092, "step": 19650 }, { "grad_norm": 0.023585395887494087, "learning_rate": 7.945591849692902e-08, "loss": 0.0042, "step": 19660 }, { "grad_norm": 0.029787050560116768, "learning_rate": 7.486522453069578e-08, "loss": 0.0058, "step": 19670 }, { "grad_norm": 0.016384869813919067, "learning_rate": 7.041102397655208e-08, "loss": 0.0067, "step": 19680 }, { "grad_norm": 0.054265912622213364, "learning_rate": 6.609332901210685e-08, "loss": 0.0058, "step": 19690 }, { "grad_norm": 0.04399024322628975, "learning_rate": 6.191215144178419e-08, "loss": 0.0055, "step": 19700 }, { "grad_norm": 0.04143797978758812, "learning_rate": 5.786750269675678e-08, "loss": 0.0045, "step": 19710 }, { "grad_norm": 0.03326734900474548, "learning_rate": 5.395939383494031e-08, "loss": 0.0058, "step": 19720 }, { "grad_norm": 0.015909343957901, "learning_rate": 5.018783554095463e-08, "loss": 0.003, "step": 19730 }, { "grad_norm": 0.08597715944051743, "learning_rate": 4.655283812610156e-08, "loss": 0.0037, "step": 19740 }, { "grad_norm": 0.031575653702020645, "learning_rate": 4.305441152831491e-08, "loss": 0.0038, "step": 19750 }, { "grad_norm": 0.0561877079308033, "learning_rate": 3.9692565312171584e-08, "loss": 0.004, "step": 19760 }, { "grad_norm": 0.08059482276439667, "learning_rate": 3.6467308668824975e-08, "loss": 0.0045, "step": 19770 }, { "grad_norm": 0.02128739282488823, "learning_rate": 3.3378650416004964e-08, "loss": 0.0034, "step": 19780 }, { "grad_norm": 0.030546288937330246, "learning_rate": 3.042659899797906e-08, "loss": 0.0049, "step": 19790 }, { "grad_norm": 0.023038653656840324, "learning_rate": 2.76111624855524e-08, "loss": 0.0035, "step": 19800 }, { "grad_norm": 0.03481736406683922, "learning_rate": 2.4932348576017784e-08, "loss": 0.0049, "step": 19810 }, { "grad_norm": 0.028808515518903732, "learning_rate": 2.239016459314458e-08, "loss": 0.0041, "step": 19820 }, { "grad_norm": 0.06456957757472992, "learning_rate": 1.9984617487173174e-08, "loss": 0.004, "step": 19830 }, { "grad_norm": 0.027450889348983765, "learning_rate": 1.7715713834776105e-08, "loss": 0.0039, "step": 19840 }, { "grad_norm": 0.07811158150434494, "learning_rate": 1.5583459839046964e-08, "loss": 0.0059, "step": 19850 }, { "grad_norm": 0.053056903183460236, "learning_rate": 1.3587861329489304e-08, "loss": 0.0042, "step": 19860 }, { "grad_norm": 0.02446102909743786, "learning_rate": 1.1728923761994415e-08, "loss": 0.0023, "step": 19870 }, { "grad_norm": 0.047315411269664764, "learning_rate": 1.0006652218819135e-08, "loss": 0.0042, "step": 19880 }, { "grad_norm": 0.035163022577762604, "learning_rate": 8.421051408596947e-09, "loss": 0.0077, "step": 19890 }, { "grad_norm": 0.020078569650650024, "learning_rate": 6.972125666299123e-09, "loss": 0.0069, "step": 19900 }, { "grad_norm": 0.045009635388851166, "learning_rate": 5.659878953229169e-09, "loss": 0.0045, "step": 19910 }, { "grad_norm": 0.036355242133140564, "learning_rate": 4.48431485701728e-09, "loss": 0.0043, "step": 19920 }, { "grad_norm": 0.044731684029102325, "learning_rate": 3.4454365916203322e-09, "loss": 0.0055, "step": 19930 }, { "grad_norm": 0.02352859079837799, "learning_rate": 2.5432469972830332e-09, "loss": 0.0029, "step": 19940 }, { "grad_norm": 0.04830089956521988, "learning_rate": 1.7777485405601203e-09, "loss": 0.0031, "step": 19950 }, { "grad_norm": 0.01771678775548935, "learning_rate": 1.1489433142941597e-09, "loss": 0.0034, "step": 19960 }, { "grad_norm": 0.021128326654434204, "learning_rate": 6.568330376210963e-10, "loss": 0.0041, "step": 19970 }, { "grad_norm": 0.04460067301988602, "learning_rate": 3.0141905594249787e-10, "loss": 0.0051, "step": 19980 }, { "grad_norm": 0.021370982751250267, "learning_rate": 8.270234094776008e-11, "loss": 0.0049, "step": 19990 }, { "grad_norm": 0.022540045902132988, "learning_rate": 6.834906085551041e-13, "loss": 0.0044, "step": 20000 } ], "logging_steps": 10, "max_steps": 20000, "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": 24, "trial_name": null, "trial_params": null }