{ "best_global_step": 4000, "best_metric": 2.031379968359076, "best_model_checkpoint": "./SALAMA_NEW4/checkpoint-4000", "epoch": 2.396644697423607, "eval_steps": 2000, "global_step": 4000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.005991611743559017, "grad_norm": 6.641392230987549, "learning_rate": 1.8e-07, "loss": 0.1132, "step": 10 }, { "epoch": 0.011983223487118035, "grad_norm": 6.161131858825684, "learning_rate": 3.8e-07, "loss": 0.149, "step": 20 }, { "epoch": 0.017974835230677052, "grad_norm": 5.680447578430176, "learning_rate": 5.800000000000001e-07, "loss": 0.0882, "step": 30 }, { "epoch": 0.02396644697423607, "grad_norm": 3.6827754974365234, "learning_rate": 7.8e-07, "loss": 0.0995, "step": 40 }, { "epoch": 0.029958058717795086, "grad_norm": 5.75164794921875, "learning_rate": 9.800000000000001e-07, "loss": 0.1368, "step": 50 }, { "epoch": 0.035949670461354104, "grad_norm": 5.719022274017334, "learning_rate": 1.1800000000000001e-06, "loss": 0.0938, "step": 60 }, { "epoch": 0.041941282204913125, "grad_norm": 4.832603931427002, "learning_rate": 1.3800000000000001e-06, "loss": 0.1163, "step": 70 }, { "epoch": 0.04793289394847214, "grad_norm": 5.619574546813965, "learning_rate": 1.5800000000000001e-06, "loss": 0.1008, "step": 80 }, { "epoch": 0.05392450569203116, "grad_norm": 4.749862194061279, "learning_rate": 1.7800000000000001e-06, "loss": 0.1126, "step": 90 }, { "epoch": 0.05991611743559017, "grad_norm": 6.061086177825928, "learning_rate": 1.98e-06, "loss": 0.1134, "step": 100 }, { "epoch": 0.0659077291791492, "grad_norm": 5.554771423339844, "learning_rate": 2.1800000000000003e-06, "loss": 0.1178, "step": 110 }, { "epoch": 0.07189934092270821, "grad_norm": 4.674679756164551, "learning_rate": 2.38e-06, "loss": 0.0758, "step": 120 }, { "epoch": 0.07789095266626722, "grad_norm": 4.852492332458496, "learning_rate": 2.5800000000000003e-06, "loss": 0.0836, "step": 130 }, { "epoch": 0.08388256440982625, "grad_norm": 5.85217809677124, "learning_rate": 2.7800000000000005e-06, "loss": 0.1097, "step": 140 }, { "epoch": 0.08987417615338526, "grad_norm": 6.0412421226501465, "learning_rate": 2.9800000000000003e-06, "loss": 0.0921, "step": 150 }, { "epoch": 0.09586578789694428, "grad_norm": 5.854034423828125, "learning_rate": 3.1800000000000005e-06, "loss": 0.1144, "step": 160 }, { "epoch": 0.10185739964050329, "grad_norm": 4.274904251098633, "learning_rate": 3.3800000000000007e-06, "loss": 0.1028, "step": 170 }, { "epoch": 0.10784901138406232, "grad_norm": 4.002969741821289, "learning_rate": 3.58e-06, "loss": 0.0759, "step": 180 }, { "epoch": 0.11384062312762133, "grad_norm": 4.920216083526611, "learning_rate": 3.7800000000000002e-06, "loss": 0.1074, "step": 190 }, { "epoch": 0.11983223487118035, "grad_norm": 5.039630889892578, "learning_rate": 3.980000000000001e-06, "loss": 0.1081, "step": 200 }, { "epoch": 0.12582384661473936, "grad_norm": 5.63913631439209, "learning_rate": 4.18e-06, "loss": 0.0974, "step": 210 }, { "epoch": 0.1318154583582984, "grad_norm": 4.254915714263916, "learning_rate": 4.38e-06, "loss": 0.1017, "step": 220 }, { "epoch": 0.1378070701018574, "grad_norm": 5.292032241821289, "learning_rate": 4.58e-06, "loss": 0.0982, "step": 230 }, { "epoch": 0.14379868184541642, "grad_norm": 3.856678009033203, "learning_rate": 4.78e-06, "loss": 0.1015, "step": 240 }, { "epoch": 0.14979029358897544, "grad_norm": 4.549379825592041, "learning_rate": 4.980000000000001e-06, "loss": 0.1245, "step": 250 }, { "epoch": 0.15578190533253444, "grad_norm": 4.482922077178955, "learning_rate": 5.18e-06, "loss": 0.104, "step": 260 }, { "epoch": 0.16177351707609347, "grad_norm": 3.305623769760132, "learning_rate": 5.380000000000001e-06, "loss": 0.0918, "step": 270 }, { "epoch": 0.1677651288196525, "grad_norm": 3.838818073272705, "learning_rate": 5.580000000000001e-06, "loss": 0.085, "step": 280 }, { "epoch": 0.1737567405632115, "grad_norm": 4.304590702056885, "learning_rate": 5.78e-06, "loss": 0.1025, "step": 290 }, { "epoch": 0.17974835230677053, "grad_norm": 5.107855796813965, "learning_rate": 5.98e-06, "loss": 0.1035, "step": 300 }, { "epoch": 0.18573996405032953, "grad_norm": 4.67608118057251, "learning_rate": 6.18e-06, "loss": 0.1292, "step": 310 }, { "epoch": 0.19173157579388855, "grad_norm": 3.5006017684936523, "learning_rate": 6.380000000000001e-06, "loss": 0.1232, "step": 320 }, { "epoch": 0.19772318753744758, "grad_norm": 4.693627834320068, "learning_rate": 6.5800000000000005e-06, "loss": 0.117, "step": 330 }, { "epoch": 0.20371479928100658, "grad_norm": 3.6866164207458496, "learning_rate": 6.780000000000001e-06, "loss": 0.1016, "step": 340 }, { "epoch": 0.2097064110245656, "grad_norm": 4.933455944061279, "learning_rate": 6.98e-06, "loss": 0.108, "step": 350 }, { "epoch": 0.21569802276812464, "grad_norm": 2.64819073677063, "learning_rate": 7.180000000000001e-06, "loss": 0.1101, "step": 360 }, { "epoch": 0.22168963451168364, "grad_norm": 5.819120407104492, "learning_rate": 7.3800000000000005e-06, "loss": 0.1305, "step": 370 }, { "epoch": 0.22768124625524266, "grad_norm": 4.504668235778809, "learning_rate": 7.58e-06, "loss": 0.0993, "step": 380 }, { "epoch": 0.23367285799880166, "grad_norm": 3.761913776397705, "learning_rate": 7.78e-06, "loss": 0.1033, "step": 390 }, { "epoch": 0.2396644697423607, "grad_norm": 3.8250932693481445, "learning_rate": 7.980000000000002e-06, "loss": 0.0917, "step": 400 }, { "epoch": 0.24565608148591972, "grad_norm": 4.284841537475586, "learning_rate": 8.18e-06, "loss": 0.1012, "step": 410 }, { "epoch": 0.2516476932294787, "grad_norm": 4.682767391204834, "learning_rate": 8.380000000000001e-06, "loss": 0.0861, "step": 420 }, { "epoch": 0.2576393049730378, "grad_norm": 3.7932474613189697, "learning_rate": 8.580000000000001e-06, "loss": 0.1151, "step": 430 }, { "epoch": 0.2636309167165968, "grad_norm": 3.588685989379883, "learning_rate": 8.78e-06, "loss": 0.1062, "step": 440 }, { "epoch": 0.2696225284601558, "grad_norm": 3.6409924030303955, "learning_rate": 8.98e-06, "loss": 0.0987, "step": 450 }, { "epoch": 0.2756141402037148, "grad_norm": 3.630150318145752, "learning_rate": 9.180000000000002e-06, "loss": 0.0952, "step": 460 }, { "epoch": 0.28160575194727383, "grad_norm": 3.858980417251587, "learning_rate": 9.38e-06, "loss": 0.0992, "step": 470 }, { "epoch": 0.28759736369083283, "grad_norm": 4.108921527862549, "learning_rate": 9.58e-06, "loss": 0.1153, "step": 480 }, { "epoch": 0.29358897543439183, "grad_norm": 3.802626848220825, "learning_rate": 9.780000000000001e-06, "loss": 0.1149, "step": 490 }, { "epoch": 0.2995805871779509, "grad_norm": 4.295905113220215, "learning_rate": 9.980000000000001e-06, "loss": 0.1118, "step": 500 }, { "epoch": 0.3055721989215099, "grad_norm": 4.09583854675293, "learning_rate": 9.985427461139897e-06, "loss": 0.0955, "step": 510 }, { "epoch": 0.3115638106650689, "grad_norm": 3.3187222480773926, "learning_rate": 9.969235751295337e-06, "loss": 0.1075, "step": 520 }, { "epoch": 0.31755542240862794, "grad_norm": 8.15339183807373, "learning_rate": 9.953044041450778e-06, "loss": 0.1113, "step": 530 }, { "epoch": 0.32354703415218694, "grad_norm": 5.273709297180176, "learning_rate": 9.936852331606218e-06, "loss": 0.1015, "step": 540 }, { "epoch": 0.32953864589574594, "grad_norm": 3.836277484893799, "learning_rate": 9.920660621761659e-06, "loss": 0.0927, "step": 550 }, { "epoch": 0.335530257639305, "grad_norm": 5.473774433135986, "learning_rate": 9.9044689119171e-06, "loss": 0.1045, "step": 560 }, { "epoch": 0.341521869382864, "grad_norm": 5.008555889129639, "learning_rate": 9.888277202072539e-06, "loss": 0.1193, "step": 570 }, { "epoch": 0.347513481126423, "grad_norm": 8.994523048400879, "learning_rate": 9.87208549222798e-06, "loss": 0.1154, "step": 580 }, { "epoch": 0.35350509286998205, "grad_norm": 4.773490905761719, "learning_rate": 9.85589378238342e-06, "loss": 0.1222, "step": 590 }, { "epoch": 0.35949670461354105, "grad_norm": 3.776129722595215, "learning_rate": 9.839702072538862e-06, "loss": 0.0999, "step": 600 }, { "epoch": 0.36548831635710005, "grad_norm": 3.550184726715088, "learning_rate": 9.823510362694301e-06, "loss": 0.1012, "step": 610 }, { "epoch": 0.37147992810065905, "grad_norm": 4.164742469787598, "learning_rate": 9.807318652849742e-06, "loss": 0.0978, "step": 620 }, { "epoch": 0.3774715398442181, "grad_norm": 3.2359657287597656, "learning_rate": 9.791126943005183e-06, "loss": 0.0781, "step": 630 }, { "epoch": 0.3834631515877771, "grad_norm": 3.4513018131256104, "learning_rate": 9.774935233160622e-06, "loss": 0.1098, "step": 640 }, { "epoch": 0.3894547633313361, "grad_norm": 3.489149808883667, "learning_rate": 9.758743523316063e-06, "loss": 0.1316, "step": 650 }, { "epoch": 0.39544637507489516, "grad_norm": 4.0016374588012695, "learning_rate": 9.742551813471504e-06, "loss": 0.092, "step": 660 }, { "epoch": 0.40143798681845416, "grad_norm": 7.272388935089111, "learning_rate": 9.726360103626944e-06, "loss": 0.1273, "step": 670 }, { "epoch": 0.40742959856201316, "grad_norm": 3.443683385848999, "learning_rate": 9.710168393782385e-06, "loss": 0.1053, "step": 680 }, { "epoch": 0.4134212103055722, "grad_norm": 4.35030460357666, "learning_rate": 9.693976683937824e-06, "loss": 0.114, "step": 690 }, { "epoch": 0.4194128220491312, "grad_norm": 3.6015312671661377, "learning_rate": 9.677784974093265e-06, "loss": 0.1018, "step": 700 }, { "epoch": 0.4254044337926902, "grad_norm": 4.1978840827941895, "learning_rate": 9.661593264248706e-06, "loss": 0.1151, "step": 710 }, { "epoch": 0.4313960455362493, "grad_norm": 4.4482879638671875, "learning_rate": 9.645401554404145e-06, "loss": 0.1114, "step": 720 }, { "epoch": 0.4373876572798083, "grad_norm": 4.560624599456787, "learning_rate": 9.629209844559586e-06, "loss": 0.1207, "step": 730 }, { "epoch": 0.4433792690233673, "grad_norm": 5.984576225280762, "learning_rate": 9.613018134715027e-06, "loss": 0.1262, "step": 740 }, { "epoch": 0.44937088076692633, "grad_norm": 4.926331043243408, "learning_rate": 9.596826424870466e-06, "loss": 0.1041, "step": 750 }, { "epoch": 0.45536249251048533, "grad_norm": 4.679685115814209, "learning_rate": 9.580634715025907e-06, "loss": 0.1196, "step": 760 }, { "epoch": 0.46135410425404433, "grad_norm": 4.890758991241455, "learning_rate": 9.564443005181347e-06, "loss": 0.1145, "step": 770 }, { "epoch": 0.46734571599760333, "grad_norm": 4.479861259460449, "learning_rate": 9.548251295336788e-06, "loss": 0.1223, "step": 780 }, { "epoch": 0.4733373277411624, "grad_norm": 4.720247745513916, "learning_rate": 9.532059585492229e-06, "loss": 0.1056, "step": 790 }, { "epoch": 0.4793289394847214, "grad_norm": 3.5392069816589355, "learning_rate": 9.51586787564767e-06, "loss": 0.0992, "step": 800 }, { "epoch": 0.4853205512282804, "grad_norm": 3.7479641437530518, "learning_rate": 9.49967616580311e-06, "loss": 0.1007, "step": 810 }, { "epoch": 0.49131216297183944, "grad_norm": 4.393255710601807, "learning_rate": 9.48348445595855e-06, "loss": 0.1394, "step": 820 }, { "epoch": 0.49730377471539844, "grad_norm": 3.2709712982177734, "learning_rate": 9.467292746113991e-06, "loss": 0.0992, "step": 830 }, { "epoch": 0.5032953864589574, "grad_norm": 4.891329288482666, "learning_rate": 9.451101036269432e-06, "loss": 0.1123, "step": 840 }, { "epoch": 0.5092869982025164, "grad_norm": 4.337294101715088, "learning_rate": 9.434909326424871e-06, "loss": 0.0924, "step": 850 }, { "epoch": 0.5152786099460755, "grad_norm": 3.9342563152313232, "learning_rate": 9.418717616580312e-06, "loss": 0.1099, "step": 860 }, { "epoch": 0.5212702216896345, "grad_norm": 3.220473527908325, "learning_rate": 9.402525906735751e-06, "loss": 0.0935, "step": 870 }, { "epoch": 0.5272618334331935, "grad_norm": 4.463641166687012, "learning_rate": 9.386334196891192e-06, "loss": 0.1101, "step": 880 }, { "epoch": 0.5332534451767525, "grad_norm": 4.686647415161133, "learning_rate": 9.370142487046633e-06, "loss": 0.1161, "step": 890 }, { "epoch": 0.5392450569203115, "grad_norm": 3.4189178943634033, "learning_rate": 9.353950777202073e-06, "loss": 0.1144, "step": 900 }, { "epoch": 0.5452366686638705, "grad_norm": 3.407076597213745, "learning_rate": 9.337759067357514e-06, "loss": 0.0926, "step": 910 }, { "epoch": 0.5512282804074295, "grad_norm": 5.300261974334717, "learning_rate": 9.321567357512955e-06, "loss": 0.1047, "step": 920 }, { "epoch": 0.5572198921509887, "grad_norm": 2.8152244091033936, "learning_rate": 9.305375647668394e-06, "loss": 0.1082, "step": 930 }, { "epoch": 0.5632115038945477, "grad_norm": 4.786292552947998, "learning_rate": 9.289183937823835e-06, "loss": 0.1206, "step": 940 }, { "epoch": 0.5692031156381067, "grad_norm": 3.642893075942993, "learning_rate": 9.272992227979276e-06, "loss": 0.0863, "step": 950 }, { "epoch": 0.5751947273816657, "grad_norm": 4.487881660461426, "learning_rate": 9.256800518134715e-06, "loss": 0.1171, "step": 960 }, { "epoch": 0.5811863391252247, "grad_norm": 4.288365364074707, "learning_rate": 9.240608808290156e-06, "loss": 0.1095, "step": 970 }, { "epoch": 0.5871779508687837, "grad_norm": 3.9558470249176025, "learning_rate": 9.224417098445595e-06, "loss": 0.1185, "step": 980 }, { "epoch": 0.5931695626123428, "grad_norm": 2.573941707611084, "learning_rate": 9.208225388601038e-06, "loss": 0.1157, "step": 990 }, { "epoch": 0.5991611743559018, "grad_norm": 3.7058804035186768, "learning_rate": 9.192033678756477e-06, "loss": 0.1173, "step": 1000 }, { "epoch": 0.6051527860994608, "grad_norm": 4.578528881072998, "learning_rate": 9.175841968911918e-06, "loss": 0.1049, "step": 1010 }, { "epoch": 0.6111443978430198, "grad_norm": 4.096612930297852, "learning_rate": 9.15965025906736e-06, "loss": 0.102, "step": 1020 }, { "epoch": 0.6171360095865788, "grad_norm": 2.6690211296081543, "learning_rate": 9.143458549222799e-06, "loss": 0.1219, "step": 1030 }, { "epoch": 0.6231276213301378, "grad_norm": 4.082022190093994, "learning_rate": 9.12726683937824e-06, "loss": 0.0958, "step": 1040 }, { "epoch": 0.6291192330736968, "grad_norm": 5.364941596984863, "learning_rate": 9.11107512953368e-06, "loss": 0.1152, "step": 1050 }, { "epoch": 0.6351108448172559, "grad_norm": 3.664750099182129, "learning_rate": 9.09488341968912e-06, "loss": 0.107, "step": 1060 }, { "epoch": 0.6411024565608149, "grad_norm": 4.337752342224121, "learning_rate": 9.07869170984456e-06, "loss": 0.1053, "step": 1070 }, { "epoch": 0.6470940683043739, "grad_norm": 4.239952564239502, "learning_rate": 9.0625e-06, "loss": 0.0965, "step": 1080 }, { "epoch": 0.6530856800479329, "grad_norm": 3.8299310207366943, "learning_rate": 9.046308290155441e-06, "loss": 0.0995, "step": 1090 }, { "epoch": 0.6590772917914919, "grad_norm": 3.6321651935577393, "learning_rate": 9.030116580310882e-06, "loss": 0.0991, "step": 1100 }, { "epoch": 0.6650689035350509, "grad_norm": 3.9516804218292236, "learning_rate": 9.013924870466321e-06, "loss": 0.1068, "step": 1110 }, { "epoch": 0.67106051527861, "grad_norm": 4.31940221786499, "learning_rate": 8.997733160621762e-06, "loss": 0.1065, "step": 1120 }, { "epoch": 0.677052127022169, "grad_norm": 4.485226154327393, "learning_rate": 8.981541450777203e-06, "loss": 0.1318, "step": 1130 }, { "epoch": 0.683043738765728, "grad_norm": 2.977614402770996, "learning_rate": 8.965349740932643e-06, "loss": 0.1111, "step": 1140 }, { "epoch": 0.689035350509287, "grad_norm": 4.644229888916016, "learning_rate": 8.949158031088084e-06, "loss": 0.1119, "step": 1150 }, { "epoch": 0.695026962252846, "grad_norm": 5.099035263061523, "learning_rate": 8.932966321243523e-06, "loss": 0.1033, "step": 1160 }, { "epoch": 0.701018573996405, "grad_norm": 3.9480855464935303, "learning_rate": 8.916774611398964e-06, "loss": 0.0895, "step": 1170 }, { "epoch": 0.7070101857399641, "grad_norm": 4.888814449310303, "learning_rate": 8.900582901554405e-06, "loss": 0.107, "step": 1180 }, { "epoch": 0.7130017974835231, "grad_norm": 2.8887240886688232, "learning_rate": 8.884391191709846e-06, "loss": 0.0974, "step": 1190 }, { "epoch": 0.7189934092270821, "grad_norm": 4.195169448852539, "learning_rate": 8.868199481865287e-06, "loss": 0.1186, "step": 1200 }, { "epoch": 0.7249850209706411, "grad_norm": 4.0084710121154785, "learning_rate": 8.852007772020726e-06, "loss": 0.1081, "step": 1210 }, { "epoch": 0.7309766327142001, "grad_norm": 3.978888511657715, "learning_rate": 8.835816062176167e-06, "loss": 0.1271, "step": 1220 }, { "epoch": 0.7369682444577591, "grad_norm": 2.9603512287139893, "learning_rate": 8.819624352331608e-06, "loss": 0.1017, "step": 1230 }, { "epoch": 0.7429598562013181, "grad_norm": 3.808750629425049, "learning_rate": 8.803432642487047e-06, "loss": 0.1119, "step": 1240 }, { "epoch": 0.7489514679448772, "grad_norm": 4.731956481933594, "learning_rate": 8.787240932642488e-06, "loss": 0.1015, "step": 1250 }, { "epoch": 0.7549430796884362, "grad_norm": 4.4244866371154785, "learning_rate": 8.771049222797927e-06, "loss": 0.1009, "step": 1260 }, { "epoch": 0.7609346914319952, "grad_norm": 5.3592305183410645, "learning_rate": 8.754857512953368e-06, "loss": 0.1175, "step": 1270 }, { "epoch": 0.7669263031755542, "grad_norm": 3.1789183616638184, "learning_rate": 8.73866580310881e-06, "loss": 0.1079, "step": 1280 }, { "epoch": 0.7729179149191132, "grad_norm": 4.719349384307861, "learning_rate": 8.722474093264249e-06, "loss": 0.1034, "step": 1290 }, { "epoch": 0.7789095266626722, "grad_norm": 3.4696717262268066, "learning_rate": 8.70628238341969e-06, "loss": 0.1042, "step": 1300 }, { "epoch": 0.7849011384062313, "grad_norm": 3.1757519245147705, "learning_rate": 8.69009067357513e-06, "loss": 0.0919, "step": 1310 }, { "epoch": 0.7908927501497903, "grad_norm": 5.520872592926025, "learning_rate": 8.67389896373057e-06, "loss": 0.1011, "step": 1320 }, { "epoch": 0.7968843618933493, "grad_norm": 4.574864387512207, "learning_rate": 8.657707253886011e-06, "loss": 0.1123, "step": 1330 }, { "epoch": 0.8028759736369083, "grad_norm": 3.809239625930786, "learning_rate": 8.641515544041452e-06, "loss": 0.1056, "step": 1340 }, { "epoch": 0.8088675853804673, "grad_norm": 4.405616283416748, "learning_rate": 8.625323834196891e-06, "loss": 0.1141, "step": 1350 }, { "epoch": 0.8148591971240263, "grad_norm": 4.816131114959717, "learning_rate": 8.609132124352332e-06, "loss": 0.1042, "step": 1360 }, { "epoch": 0.8208508088675854, "grad_norm": 4.3509602546691895, "learning_rate": 8.592940414507773e-06, "loss": 0.098, "step": 1370 }, { "epoch": 0.8268424206111444, "grad_norm": 3.7490243911743164, "learning_rate": 8.576748704663214e-06, "loss": 0.1049, "step": 1380 }, { "epoch": 0.8328340323547034, "grad_norm": 5.330031394958496, "learning_rate": 8.560556994818653e-06, "loss": 0.105, "step": 1390 }, { "epoch": 0.8388256440982624, "grad_norm": 4.027228355407715, "learning_rate": 8.544365284974094e-06, "loss": 0.1249, "step": 1400 }, { "epoch": 0.8448172558418214, "grad_norm": 4.0217790603637695, "learning_rate": 8.528173575129535e-06, "loss": 0.1259, "step": 1410 }, { "epoch": 0.8508088675853804, "grad_norm": 4.059916973114014, "learning_rate": 8.511981865284975e-06, "loss": 0.1173, "step": 1420 }, { "epoch": 0.8568004793289394, "grad_norm": 4.621293544769287, "learning_rate": 8.495790155440416e-06, "loss": 0.1037, "step": 1430 }, { "epoch": 0.8627920910724985, "grad_norm": 3.6666407585144043, "learning_rate": 8.479598445595855e-06, "loss": 0.1177, "step": 1440 }, { "epoch": 0.8687837028160575, "grad_norm": 4.930758953094482, "learning_rate": 8.463406735751296e-06, "loss": 0.1055, "step": 1450 }, { "epoch": 0.8747753145596165, "grad_norm": 4.322690010070801, "learning_rate": 8.447215025906737e-06, "loss": 0.1052, "step": 1460 }, { "epoch": 0.8807669263031755, "grad_norm": 3.768627405166626, "learning_rate": 8.431023316062176e-06, "loss": 0.1099, "step": 1470 }, { "epoch": 0.8867585380467345, "grad_norm": 3.582303047180176, "learning_rate": 8.414831606217617e-06, "loss": 0.1061, "step": 1480 }, { "epoch": 0.8927501497902935, "grad_norm": 3.874319076538086, "learning_rate": 8.398639896373058e-06, "loss": 0.126, "step": 1490 }, { "epoch": 0.8987417615338527, "grad_norm": 4.040960311889648, "learning_rate": 8.382448186528497e-06, "loss": 0.089, "step": 1500 }, { "epoch": 0.9047333732774117, "grad_norm": 4.185450077056885, "learning_rate": 8.366256476683938e-06, "loss": 0.0955, "step": 1510 }, { "epoch": 0.9107249850209707, "grad_norm": 3.388444185256958, "learning_rate": 8.35006476683938e-06, "loss": 0.1123, "step": 1520 }, { "epoch": 0.9167165967645297, "grad_norm": 4.470240592956543, "learning_rate": 8.333873056994819e-06, "loss": 0.1213, "step": 1530 }, { "epoch": 0.9227082085080887, "grad_norm": 4.528209686279297, "learning_rate": 8.31768134715026e-06, "loss": 0.1152, "step": 1540 }, { "epoch": 0.9286998202516477, "grad_norm": 4.039801597595215, "learning_rate": 8.301489637305699e-06, "loss": 0.0907, "step": 1550 }, { "epoch": 0.9346914319952067, "grad_norm": 2.667508840560913, "learning_rate": 8.28529792746114e-06, "loss": 0.1103, "step": 1560 }, { "epoch": 0.9406830437387658, "grad_norm": 4.166356086730957, "learning_rate": 8.269106217616581e-06, "loss": 0.1034, "step": 1570 }, { "epoch": 0.9466746554823248, "grad_norm": 4.047358989715576, "learning_rate": 8.252914507772022e-06, "loss": 0.1086, "step": 1580 }, { "epoch": 0.9526662672258838, "grad_norm": 3.4767513275146484, "learning_rate": 8.236722797927463e-06, "loss": 0.0982, "step": 1590 }, { "epoch": 0.9586578789694428, "grad_norm": 4.475752830505371, "learning_rate": 8.220531088082902e-06, "loss": 0.0879, "step": 1600 }, { "epoch": 0.9646494907130018, "grad_norm": 3.3900721073150635, "learning_rate": 8.204339378238343e-06, "loss": 0.0861, "step": 1610 }, { "epoch": 0.9706411024565608, "grad_norm": 3.077188014984131, "learning_rate": 8.188147668393784e-06, "loss": 0.1024, "step": 1620 }, { "epoch": 0.9766327142001199, "grad_norm": 4.166396141052246, "learning_rate": 8.171955958549223e-06, "loss": 0.0942, "step": 1630 }, { "epoch": 0.9826243259436789, "grad_norm": 2.931767463684082, "learning_rate": 8.155764248704664e-06, "loss": 0.0833, "step": 1640 }, { "epoch": 0.9886159376872379, "grad_norm": 3.9457390308380127, "learning_rate": 8.139572538860104e-06, "loss": 0.108, "step": 1650 }, { "epoch": 0.9946075494307969, "grad_norm": 4.256587505340576, "learning_rate": 8.123380829015545e-06, "loss": 0.0926, "step": 1660 }, { "epoch": 1.0005991611743559, "grad_norm": 1.9296483993530273, "learning_rate": 8.107189119170986e-06, "loss": 0.0814, "step": 1670 }, { "epoch": 1.0065907729179149, "grad_norm": 2.8497297763824463, "learning_rate": 8.090997409326425e-06, "loss": 0.0399, "step": 1680 }, { "epoch": 1.0125823846614739, "grad_norm": 1.6683580875396729, "learning_rate": 8.074805699481866e-06, "loss": 0.0465, "step": 1690 }, { "epoch": 1.0185739964050329, "grad_norm": 2.067464828491211, "learning_rate": 8.058613989637307e-06, "loss": 0.0388, "step": 1700 }, { "epoch": 1.0245656081485919, "grad_norm": 3.085991382598877, "learning_rate": 8.042422279792746e-06, "loss": 0.0389, "step": 1710 }, { "epoch": 1.030557219892151, "grad_norm": 2.972982883453369, "learning_rate": 8.026230569948187e-06, "loss": 0.0432, "step": 1720 }, { "epoch": 1.03654883163571, "grad_norm": 3.642639398574829, "learning_rate": 8.010038860103628e-06, "loss": 0.0474, "step": 1730 }, { "epoch": 1.042540443379269, "grad_norm": 3.922886848449707, "learning_rate": 7.993847150259067e-06, "loss": 0.0362, "step": 1740 }, { "epoch": 1.048532055122828, "grad_norm": 2.143171787261963, "learning_rate": 7.977655440414508e-06, "loss": 0.0453, "step": 1750 }, { "epoch": 1.054523666866387, "grad_norm": 2.2521893978118896, "learning_rate": 7.96146373056995e-06, "loss": 0.0347, "step": 1760 }, { "epoch": 1.060515278609946, "grad_norm": 2.5530147552490234, "learning_rate": 7.94527202072539e-06, "loss": 0.0321, "step": 1770 }, { "epoch": 1.066506890353505, "grad_norm": 1.3185819387435913, "learning_rate": 7.92908031088083e-06, "loss": 0.0352, "step": 1780 }, { "epoch": 1.072498502097064, "grad_norm": 4.018706321716309, "learning_rate": 7.91288860103627e-06, "loss": 0.0528, "step": 1790 }, { "epoch": 1.078490113840623, "grad_norm": 2.9235482215881348, "learning_rate": 7.896696891191711e-06, "loss": 0.0291, "step": 1800 }, { "epoch": 1.084481725584182, "grad_norm": 3.683203935623169, "learning_rate": 7.88050518134715e-06, "loss": 0.0398, "step": 1810 }, { "epoch": 1.090473337327741, "grad_norm": 7.694185256958008, "learning_rate": 7.864313471502592e-06, "loss": 0.0428, "step": 1820 }, { "epoch": 1.0964649490713, "grad_norm": 2.377105474472046, "learning_rate": 7.848121761658031e-06, "loss": 0.0474, "step": 1830 }, { "epoch": 1.102456560814859, "grad_norm": 2.613847494125366, "learning_rate": 7.831930051813472e-06, "loss": 0.0346, "step": 1840 }, { "epoch": 1.1084481725584183, "grad_norm": 2.805191993713379, "learning_rate": 7.815738341968913e-06, "loss": 0.043, "step": 1850 }, { "epoch": 1.1144397843019773, "grad_norm": 3.0600552558898926, "learning_rate": 7.799546632124352e-06, "loss": 0.0422, "step": 1860 }, { "epoch": 1.1204313960455363, "grad_norm": 1.6137948036193848, "learning_rate": 7.783354922279793e-06, "loss": 0.036, "step": 1870 }, { "epoch": 1.1264230077890953, "grad_norm": 1.492092251777649, "learning_rate": 7.767163212435234e-06, "loss": 0.0315, "step": 1880 }, { "epoch": 1.1324146195326543, "grad_norm": 2.410632610321045, "learning_rate": 7.750971502590674e-06, "loss": 0.0361, "step": 1890 }, { "epoch": 1.1384062312762133, "grad_norm": 2.9959256649017334, "learning_rate": 7.734779792746114e-06, "loss": 0.0248, "step": 1900 }, { "epoch": 1.1443978430197723, "grad_norm": 1.860141634941101, "learning_rate": 7.718588082901555e-06, "loss": 0.0337, "step": 1910 }, { "epoch": 1.1503894547633313, "grad_norm": 3.2396111488342285, "learning_rate": 7.702396373056995e-06, "loss": 0.0319, "step": 1920 }, { "epoch": 1.1563810665068903, "grad_norm": 3.930166721343994, "learning_rate": 7.686204663212436e-06, "loss": 0.0427, "step": 1930 }, { "epoch": 1.1623726782504493, "grad_norm": 2.2750892639160156, "learning_rate": 7.670012953367875e-06, "loss": 0.0417, "step": 1940 }, { "epoch": 1.1683642899940083, "grad_norm": 3.2963342666625977, "learning_rate": 7.653821243523318e-06, "loss": 0.0415, "step": 1950 }, { "epoch": 1.1743559017375673, "grad_norm": 2.719001531600952, "learning_rate": 7.637629533678757e-06, "loss": 0.0308, "step": 1960 }, { "epoch": 1.1803475134811263, "grad_norm": 3.191821813583374, "learning_rate": 7.621437823834198e-06, "loss": 0.0323, "step": 1970 }, { "epoch": 1.1863391252246855, "grad_norm": 2.4749386310577393, "learning_rate": 7.605246113989638e-06, "loss": 0.035, "step": 1980 }, { "epoch": 1.1923307369682445, "grad_norm": 2.652811050415039, "learning_rate": 7.589054404145079e-06, "loss": 0.0456, "step": 1990 }, { "epoch": 1.1983223487118035, "grad_norm": 2.6856987476348877, "learning_rate": 7.572862694300519e-06, "loss": 0.0313, "step": 2000 }, { "epoch": 1.1983223487118035, "eval_loss": 0.04815296828746796, "eval_runtime": 4874.046, "eval_samples_per_second": 2.739, "eval_steps_per_second": 0.342, "eval_wer": 3.9869195765987637, "step": 2000 }, { "epoch": 1.2043139604553625, "grad_norm": 2.2129974365234375, "learning_rate": 7.556670984455959e-06, "loss": 0.0343, "step": 2010 }, { "epoch": 1.2103055721989215, "grad_norm": 2.4016990661621094, "learning_rate": 7.5404792746113994e-06, "loss": 0.043, "step": 2020 }, { "epoch": 1.2162971839424805, "grad_norm": 2.5401735305786133, "learning_rate": 7.52428756476684e-06, "loss": 0.0452, "step": 2030 }, { "epoch": 1.2222887956860395, "grad_norm": 2.0397777557373047, "learning_rate": 7.5080958549222805e-06, "loss": 0.039, "step": 2040 }, { "epoch": 1.2282804074295985, "grad_norm": 3.3085644245147705, "learning_rate": 7.491904145077721e-06, "loss": 0.043, "step": 2050 }, { "epoch": 1.2342720191731575, "grad_norm": 1.1905828714370728, "learning_rate": 7.475712435233161e-06, "loss": 0.0406, "step": 2060 }, { "epoch": 1.2402636309167165, "grad_norm": 4.510779857635498, "learning_rate": 7.459520725388602e-06, "loss": 0.0494, "step": 2070 }, { "epoch": 1.2462552426602755, "grad_norm": 2.616729497909546, "learning_rate": 7.443329015544042e-06, "loss": 0.0436, "step": 2080 }, { "epoch": 1.2522468544038348, "grad_norm": 2.109017848968506, "learning_rate": 7.427137305699482e-06, "loss": 0.042, "step": 2090 }, { "epoch": 1.2582384661473935, "grad_norm": 2.5082507133483887, "learning_rate": 7.410945595854922e-06, "loss": 0.0307, "step": 2100 }, { "epoch": 1.2642300778909528, "grad_norm": 2.2173380851745605, "learning_rate": 7.394753886010363e-06, "loss": 0.052, "step": 2110 }, { "epoch": 1.2702216896345118, "grad_norm": 2.330674171447754, "learning_rate": 7.378562176165803e-06, "loss": 0.0357, "step": 2120 }, { "epoch": 1.2762133013780708, "grad_norm": 2.3817524909973145, "learning_rate": 7.362370466321243e-06, "loss": 0.0443, "step": 2130 }, { "epoch": 1.2822049131216298, "grad_norm": 1.9407250881195068, "learning_rate": 7.346178756476684e-06, "loss": 0.0469, "step": 2140 }, { "epoch": 1.2881965248651888, "grad_norm": 1.5077818632125854, "learning_rate": 7.329987046632125e-06, "loss": 0.0351, "step": 2150 }, { "epoch": 1.2941881366087478, "grad_norm": 1.8476344347000122, "learning_rate": 7.3137953367875655e-06, "loss": 0.0435, "step": 2160 }, { "epoch": 1.3001797483523068, "grad_norm": 1.7587623596191406, "learning_rate": 7.2976036269430065e-06, "loss": 0.0381, "step": 2170 }, { "epoch": 1.3061713600958658, "grad_norm": 3.003218173980713, "learning_rate": 7.281411917098447e-06, "loss": 0.0306, "step": 2180 }, { "epoch": 1.3121629718394248, "grad_norm": 2.0112128257751465, "learning_rate": 7.265220207253887e-06, "loss": 0.0401, "step": 2190 }, { "epoch": 1.3181545835829838, "grad_norm": 2.3190035820007324, "learning_rate": 7.249028497409327e-06, "loss": 0.0457, "step": 2200 }, { "epoch": 1.3241461953265428, "grad_norm": 2.1521100997924805, "learning_rate": 7.232836787564768e-06, "loss": 0.0403, "step": 2210 }, { "epoch": 1.330137807070102, "grad_norm": 2.4302561283111572, "learning_rate": 7.216645077720208e-06, "loss": 0.04, "step": 2220 }, { "epoch": 1.3361294188136608, "grad_norm": 3.34147047996521, "learning_rate": 7.200453367875648e-06, "loss": 0.0379, "step": 2230 }, { "epoch": 1.34212103055722, "grad_norm": 2.458252191543579, "learning_rate": 7.184261658031088e-06, "loss": 0.0464, "step": 2240 }, { "epoch": 1.348112642300779, "grad_norm": 3.149097204208374, "learning_rate": 7.168069948186529e-06, "loss": 0.0395, "step": 2250 }, { "epoch": 1.354104254044338, "grad_norm": 3.174875020980835, "learning_rate": 7.151878238341969e-06, "loss": 0.036, "step": 2260 }, { "epoch": 1.360095865787897, "grad_norm": 2.373333215713501, "learning_rate": 7.1356865284974095e-06, "loss": 0.042, "step": 2270 }, { "epoch": 1.366087477531456, "grad_norm": 2.6968131065368652, "learning_rate": 7.1194948186528505e-06, "loss": 0.0377, "step": 2280 }, { "epoch": 1.372079089275015, "grad_norm": 1.8244385719299316, "learning_rate": 7.103303108808291e-06, "loss": 0.0373, "step": 2290 }, { "epoch": 1.378070701018574, "grad_norm": 2.684811592102051, "learning_rate": 7.087111398963731e-06, "loss": 0.0514, "step": 2300 }, { "epoch": 1.384062312762133, "grad_norm": 2.1189537048339844, "learning_rate": 7.070919689119171e-06, "loss": 0.0327, "step": 2310 }, { "epoch": 1.390053924505692, "grad_norm": 2.272768497467041, "learning_rate": 7.054727979274612e-06, "loss": 0.0405, "step": 2320 }, { "epoch": 1.396045536249251, "grad_norm": 2.2334699630737305, "learning_rate": 7.038536269430052e-06, "loss": 0.0425, "step": 2330 }, { "epoch": 1.40203714799281, "grad_norm": 2.5346498489379883, "learning_rate": 7.022344559585493e-06, "loss": 0.0343, "step": 2340 }, { "epoch": 1.4080287597363692, "grad_norm": 2.477675676345825, "learning_rate": 7.006152849740934e-06, "loss": 0.0333, "step": 2350 }, { "epoch": 1.414020371479928, "grad_norm": 3.833782911300659, "learning_rate": 6.989961139896374e-06, "loss": 0.0482, "step": 2360 }, { "epoch": 1.4200119832234872, "grad_norm": 2.571751117706299, "learning_rate": 6.973769430051814e-06, "loss": 0.0421, "step": 2370 }, { "epoch": 1.4260035949670462, "grad_norm": 2.7097127437591553, "learning_rate": 6.957577720207255e-06, "loss": 0.0447, "step": 2380 }, { "epoch": 1.4319952067106052, "grad_norm": 1.97878098487854, "learning_rate": 6.941386010362695e-06, "loss": 0.0445, "step": 2390 }, { "epoch": 1.4379868184541642, "grad_norm": 2.1338934898376465, "learning_rate": 6.925194300518135e-06, "loss": 0.0484, "step": 2400 }, { "epoch": 1.4439784301977232, "grad_norm": 2.3054728507995605, "learning_rate": 6.9090025906735755e-06, "loss": 0.0327, "step": 2410 }, { "epoch": 1.4499700419412822, "grad_norm": 3.713162422180176, "learning_rate": 6.8928108808290165e-06, "loss": 0.0388, "step": 2420 }, { "epoch": 1.4559616536848412, "grad_norm": 2.2444119453430176, "learning_rate": 6.876619170984457e-06, "loss": 0.0364, "step": 2430 }, { "epoch": 1.4619532654284002, "grad_norm": 2.1193532943725586, "learning_rate": 6.860427461139897e-06, "loss": 0.0439, "step": 2440 }, { "epoch": 1.4679448771719592, "grad_norm": 2.278380870819092, "learning_rate": 6.844235751295337e-06, "loss": 0.0516, "step": 2450 }, { "epoch": 1.4739364889155182, "grad_norm": 2.936154365539551, "learning_rate": 6.828044041450778e-06, "loss": 0.0386, "step": 2460 }, { "epoch": 1.4799281006590772, "grad_norm": 3.983410358428955, "learning_rate": 6.811852331606218e-06, "loss": 0.0493, "step": 2470 }, { "epoch": 1.4859197124026364, "grad_norm": 3.1823031902313232, "learning_rate": 6.795660621761658e-06, "loss": 0.0339, "step": 2480 }, { "epoch": 1.4919113241461952, "grad_norm": 3.02040696144104, "learning_rate": 6.779468911917098e-06, "loss": 0.0377, "step": 2490 }, { "epoch": 1.4979029358897544, "grad_norm": 2.253005266189575, "learning_rate": 6.763277202072539e-06, "loss": 0.0385, "step": 2500 }, { "epoch": 1.5038945476333132, "grad_norm": 2.6622331142425537, "learning_rate": 6.747085492227979e-06, "loss": 0.0388, "step": 2510 }, { "epoch": 1.5098861593768724, "grad_norm": 2.394500732421875, "learning_rate": 6.7308937823834195e-06, "loss": 0.037, "step": 2520 }, { "epoch": 1.5158777711204314, "grad_norm": 1.885865330696106, "learning_rate": 6.714702072538861e-06, "loss": 0.0314, "step": 2530 }, { "epoch": 1.5218693828639904, "grad_norm": 2.989285469055176, "learning_rate": 6.6985103626943015e-06, "loss": 0.038, "step": 2540 }, { "epoch": 1.5278609946075494, "grad_norm": 3.051121950149536, "learning_rate": 6.682318652849742e-06, "loss": 0.0438, "step": 2550 }, { "epoch": 1.5338526063511084, "grad_norm": 3.2008793354034424, "learning_rate": 6.666126943005183e-06, "loss": 0.0391, "step": 2560 }, { "epoch": 1.5398442180946674, "grad_norm": 2.4947457313537598, "learning_rate": 6.649935233160623e-06, "loss": 0.0369, "step": 2570 }, { "epoch": 1.5458358298382264, "grad_norm": 2.38167405128479, "learning_rate": 6.633743523316063e-06, "loss": 0.0443, "step": 2580 }, { "epoch": 1.5518274415817856, "grad_norm": 2.149912118911743, "learning_rate": 6.617551813471503e-06, "loss": 0.0414, "step": 2590 }, { "epoch": 1.5578190533253444, "grad_norm": 1.8637117147445679, "learning_rate": 6.601360103626944e-06, "loss": 0.0321, "step": 2600 }, { "epoch": 1.5638106650689036, "grad_norm": 4.388784408569336, "learning_rate": 6.585168393782384e-06, "loss": 0.0443, "step": 2610 }, { "epoch": 1.5698022768124624, "grad_norm": 3.057321548461914, "learning_rate": 6.568976683937824e-06, "loss": 0.0358, "step": 2620 }, { "epoch": 1.5757938885560216, "grad_norm": 1.5130738019943237, "learning_rate": 6.552784974093264e-06, "loss": 0.0341, "step": 2630 }, { "epoch": 1.5817855002995804, "grad_norm": 2.506476879119873, "learning_rate": 6.536593264248705e-06, "loss": 0.0369, "step": 2640 }, { "epoch": 1.5877771120431396, "grad_norm": 1.917100429534912, "learning_rate": 6.5204015544041455e-06, "loss": 0.0436, "step": 2650 }, { "epoch": 1.5937687237866986, "grad_norm": 2.6610848903656006, "learning_rate": 6.504209844559586e-06, "loss": 0.0371, "step": 2660 }, { "epoch": 1.5997603355302576, "grad_norm": 2.3574564456939697, "learning_rate": 6.4880181347150266e-06, "loss": 0.0466, "step": 2670 }, { "epoch": 1.6057519472738166, "grad_norm": 3.3509860038757324, "learning_rate": 6.471826424870467e-06, "loss": 0.0351, "step": 2680 }, { "epoch": 1.6117435590173756, "grad_norm": 2.6591145992279053, "learning_rate": 6.455634715025907e-06, "loss": 0.0445, "step": 2690 }, { "epoch": 1.6177351707609346, "grad_norm": 3.161813735961914, "learning_rate": 6.439443005181347e-06, "loss": 0.0315, "step": 2700 }, { "epoch": 1.6237267825044936, "grad_norm": 2.351088047027588, "learning_rate": 6.423251295336788e-06, "loss": 0.0428, "step": 2710 }, { "epoch": 1.6297183942480529, "grad_norm": 2.1950671672821045, "learning_rate": 6.407059585492228e-06, "loss": 0.0322, "step": 2720 }, { "epoch": 1.6357100059916116, "grad_norm": 3.3831064701080322, "learning_rate": 6.390867875647669e-06, "loss": 0.0465, "step": 2730 }, { "epoch": 1.6417016177351709, "grad_norm": 3.514268636703491, "learning_rate": 6.37467616580311e-06, "loss": 0.0458, "step": 2740 }, { "epoch": 1.6476932294787296, "grad_norm": 1.8238720893859863, "learning_rate": 6.35848445595855e-06, "loss": 0.0398, "step": 2750 }, { "epoch": 1.6536848412222889, "grad_norm": 2.247701644897461, "learning_rate": 6.34229274611399e-06, "loss": 0.0617, "step": 2760 }, { "epoch": 1.6596764529658476, "grad_norm": 2.080018997192383, "learning_rate": 6.326101036269431e-06, "loss": 0.0434, "step": 2770 }, { "epoch": 1.6656680647094069, "grad_norm": 2.6713249683380127, "learning_rate": 6.309909326424871e-06, "loss": 0.0353, "step": 2780 }, { "epoch": 1.6716596764529659, "grad_norm": 2.1419765949249268, "learning_rate": 6.2937176165803115e-06, "loss": 0.0432, "step": 2790 }, { "epoch": 1.6776512881965249, "grad_norm": 2.145838737487793, "learning_rate": 6.277525906735752e-06, "loss": 0.0489, "step": 2800 }, { "epoch": 1.6836428999400839, "grad_norm": 3.2524642944335938, "learning_rate": 6.261334196891193e-06, "loss": 0.0401, "step": 2810 }, { "epoch": 1.6896345116836429, "grad_norm": 2.0695011615753174, "learning_rate": 6.245142487046633e-06, "loss": 0.0387, "step": 2820 }, { "epoch": 1.6956261234272019, "grad_norm": 2.829993963241577, "learning_rate": 6.228950777202073e-06, "loss": 0.0307, "step": 2830 }, { "epoch": 1.7016177351707609, "grad_norm": 2.6294639110565186, "learning_rate": 6.212759067357513e-06, "loss": 0.0478, "step": 2840 }, { "epoch": 1.70760934691432, "grad_norm": 3.4384214878082275, "learning_rate": 6.196567357512954e-06, "loss": 0.0433, "step": 2850 }, { "epoch": 1.7136009586578789, "grad_norm": 2.4208130836486816, "learning_rate": 6.180375647668394e-06, "loss": 0.0441, "step": 2860 }, { "epoch": 1.719592570401438, "grad_norm": 2.825674533843994, "learning_rate": 6.164183937823834e-06, "loss": 0.0516, "step": 2870 }, { "epoch": 1.7255841821449969, "grad_norm": 2.239126682281494, "learning_rate": 6.147992227979274e-06, "loss": 0.0318, "step": 2880 }, { "epoch": 1.731575793888556, "grad_norm": 2.4009127616882324, "learning_rate": 6.131800518134715e-06, "loss": 0.0358, "step": 2890 }, { "epoch": 1.737567405632115, "grad_norm": 1.753083348274231, "learning_rate": 6.1156088082901555e-06, "loss": 0.035, "step": 2900 }, { "epoch": 1.743559017375674, "grad_norm": 1.7808746099472046, "learning_rate": 6.099417098445596e-06, "loss": 0.038, "step": 2910 }, { "epoch": 1.749550629119233, "grad_norm": 2.457718849182129, "learning_rate": 6.0832253886010375e-06, "loss": 0.0339, "step": 2920 }, { "epoch": 1.755542240862792, "grad_norm": 2.7318003177642822, "learning_rate": 6.067033678756478e-06, "loss": 0.0386, "step": 2930 }, { "epoch": 1.761533852606351, "grad_norm": 2.1199822425842285, "learning_rate": 6.050841968911918e-06, "loss": 0.0357, "step": 2940 }, { "epoch": 1.76752546434991, "grad_norm": 3.083209753036499, "learning_rate": 6.034650259067359e-06, "loss": 0.0401, "step": 2950 }, { "epoch": 1.773517076093469, "grad_norm": 2.8726754188537598, "learning_rate": 6.018458549222799e-06, "loss": 0.047, "step": 2960 }, { "epoch": 1.779508687837028, "grad_norm": 2.853578805923462, "learning_rate": 6.002266839378239e-06, "loss": 0.0367, "step": 2970 }, { "epoch": 1.7855002995805873, "grad_norm": 1.9339247941970825, "learning_rate": 5.986075129533679e-06, "loss": 0.0324, "step": 2980 }, { "epoch": 1.791491911324146, "grad_norm": 2.1655240058898926, "learning_rate": 5.96988341968912e-06, "loss": 0.0345, "step": 2990 }, { "epoch": 1.7974835230677053, "grad_norm": 1.8844058513641357, "learning_rate": 5.95369170984456e-06, "loss": 0.0489, "step": 3000 }, { "epoch": 1.803475134811264, "grad_norm": 1.8927345275878906, "learning_rate": 5.9375e-06, "loss": 0.0392, "step": 3010 }, { "epoch": 1.8094667465548233, "grad_norm": 2.797908306121826, "learning_rate": 5.9213082901554405e-06, "loss": 0.0429, "step": 3020 }, { "epoch": 1.8154583582983823, "grad_norm": 2.138853073120117, "learning_rate": 5.9051165803108814e-06, "loss": 0.0366, "step": 3030 }, { "epoch": 1.8214499700419413, "grad_norm": 2.9621222019195557, "learning_rate": 5.8889248704663216e-06, "loss": 0.0364, "step": 3040 }, { "epoch": 1.8274415817855003, "grad_norm": 1.9806824922561646, "learning_rate": 5.872733160621762e-06, "loss": 0.0292, "step": 3050 }, { "epoch": 1.8334331935290593, "grad_norm": 1.8473615646362305, "learning_rate": 5.856541450777203e-06, "loss": 0.0315, "step": 3060 }, { "epoch": 1.8394248052726183, "grad_norm": 2.8493854999542236, "learning_rate": 5.840349740932643e-06, "loss": 0.0329, "step": 3070 }, { "epoch": 1.8454164170161773, "grad_norm": 4.222199440002441, "learning_rate": 5.824158031088083e-06, "loss": 0.0417, "step": 3080 }, { "epoch": 1.8514080287597365, "grad_norm": 2.80318546295166, "learning_rate": 5.807966321243523e-06, "loss": 0.0453, "step": 3090 }, { "epoch": 1.8573996405032953, "grad_norm": 3.1252284049987793, "learning_rate": 5.791774611398964e-06, "loss": 0.044, "step": 3100 }, { "epoch": 1.8633912522468545, "grad_norm": 2.628920555114746, "learning_rate": 5.775582901554405e-06, "loss": 0.0388, "step": 3110 }, { "epoch": 1.8693828639904133, "grad_norm": 2.189281940460205, "learning_rate": 5.759391191709845e-06, "loss": 0.0378, "step": 3120 }, { "epoch": 1.8753744757339725, "grad_norm": 2.4642791748046875, "learning_rate": 5.743199481865286e-06, "loss": 0.0418, "step": 3130 }, { "epoch": 1.8813660874775313, "grad_norm": 2.589334726333618, "learning_rate": 5.727007772020726e-06, "loss": 0.0367, "step": 3140 }, { "epoch": 1.8873576992210905, "grad_norm": 3.040709972381592, "learning_rate": 5.710816062176166e-06, "loss": 0.035, "step": 3150 }, { "epoch": 1.8933493109646495, "grad_norm": 2.8135108947753906, "learning_rate": 5.6946243523316065e-06, "loss": 0.0435, "step": 3160 }, { "epoch": 1.8993409227082085, "grad_norm": 1.8438613414764404, "learning_rate": 5.6784326424870475e-06, "loss": 0.0416, "step": 3170 }, { "epoch": 1.9053325344517675, "grad_norm": 2.7345640659332275, "learning_rate": 5.662240932642488e-06, "loss": 0.0367, "step": 3180 }, { "epoch": 1.9113241461953265, "grad_norm": 2.2093093395233154, "learning_rate": 5.646049222797928e-06, "loss": 0.0486, "step": 3190 }, { "epoch": 1.9173157579388855, "grad_norm": 2.6385855674743652, "learning_rate": 5.629857512953369e-06, "loss": 0.037, "step": 3200 }, { "epoch": 1.9233073696824445, "grad_norm": 4.1110124588012695, "learning_rate": 5.613665803108809e-06, "loss": 0.0418, "step": 3210 }, { "epoch": 1.9292989814260038, "grad_norm": 2.6954362392425537, "learning_rate": 5.597474093264249e-06, "loss": 0.0406, "step": 3220 }, { "epoch": 1.9352905931695625, "grad_norm": 3.4439690113067627, "learning_rate": 5.581282383419689e-06, "loss": 0.0373, "step": 3230 }, { "epoch": 1.9412822049131218, "grad_norm": 2.6216893196105957, "learning_rate": 5.56509067357513e-06, "loss": 0.0319, "step": 3240 }, { "epoch": 1.9472738166566805, "grad_norm": 2.924895763397217, "learning_rate": 5.54889896373057e-06, "loss": 0.0407, "step": 3250 }, { "epoch": 1.9532654284002398, "grad_norm": 2.488950729370117, "learning_rate": 5.53270725388601e-06, "loss": 0.0341, "step": 3260 }, { "epoch": 1.9592570401437985, "grad_norm": 2.4867169857025146, "learning_rate": 5.5165155440414505e-06, "loss": 0.0453, "step": 3270 }, { "epoch": 1.9652486518873578, "grad_norm": 5.1045427322387695, "learning_rate": 5.5003238341968915e-06, "loss": 0.0429, "step": 3280 }, { "epoch": 1.9712402636309168, "grad_norm": 2.7689225673675537, "learning_rate": 5.484132124352332e-06, "loss": 0.0309, "step": 3290 }, { "epoch": 1.9772318753744758, "grad_norm": 1.944758415222168, "learning_rate": 5.4679404145077734e-06, "loss": 0.0357, "step": 3300 }, { "epoch": 1.9832234871180348, "grad_norm": 2.512436628341675, "learning_rate": 5.4517487046632136e-06, "loss": 0.0436, "step": 3310 }, { "epoch": 1.9892150988615938, "grad_norm": 2.7214131355285645, "learning_rate": 5.435556994818654e-06, "loss": 0.0329, "step": 3320 }, { "epoch": 1.9952067106051528, "grad_norm": 2.3871612548828125, "learning_rate": 5.419365284974094e-06, "loss": 0.0368, "step": 3330 }, { "epoch": 2.0011983223487118, "grad_norm": 1.4264804124832153, "learning_rate": 5.403173575129535e-06, "loss": 0.0283, "step": 3340 }, { "epoch": 2.007189934092271, "grad_norm": 0.9049192070960999, "learning_rate": 5.386981865284975e-06, "loss": 0.0141, "step": 3350 }, { "epoch": 2.0131815458358298, "grad_norm": 1.757786512374878, "learning_rate": 5.370790155440415e-06, "loss": 0.0092, "step": 3360 }, { "epoch": 2.019173157579389, "grad_norm": 0.922551691532135, "learning_rate": 5.354598445595855e-06, "loss": 0.0108, "step": 3370 }, { "epoch": 2.0251647693229478, "grad_norm": 1.6763967275619507, "learning_rate": 5.338406735751296e-06, "loss": 0.0081, "step": 3380 }, { "epoch": 2.031156381066507, "grad_norm": 1.023136854171753, "learning_rate": 5.322215025906736e-06, "loss": 0.0106, "step": 3390 }, { "epoch": 2.0371479928100658, "grad_norm": 1.0758904218673706, "learning_rate": 5.3060233160621764e-06, "loss": 0.0127, "step": 3400 }, { "epoch": 2.043139604553625, "grad_norm": 0.85313481092453, "learning_rate": 5.2898316062176166e-06, "loss": 0.0112, "step": 3410 }, { "epoch": 2.0491312162971838, "grad_norm": 1.0233789682388306, "learning_rate": 5.2736398963730575e-06, "loss": 0.0115, "step": 3420 }, { "epoch": 2.055122828040743, "grad_norm": 1.594497561454773, "learning_rate": 5.257448186528498e-06, "loss": 0.0106, "step": 3430 }, { "epoch": 2.061114439784302, "grad_norm": 2.3627138137817383, "learning_rate": 5.241256476683938e-06, "loss": 0.0116, "step": 3440 }, { "epoch": 2.067106051527861, "grad_norm": 0.5149250030517578, "learning_rate": 5.225064766839378e-06, "loss": 0.0171, "step": 3450 }, { "epoch": 2.07309766327142, "grad_norm": 0.6787377595901489, "learning_rate": 5.208873056994819e-06, "loss": 0.0129, "step": 3460 }, { "epoch": 2.079089275014979, "grad_norm": 0.44210371375083923, "learning_rate": 5.192681347150259e-06, "loss": 0.0079, "step": 3470 }, { "epoch": 2.085080886758538, "grad_norm": 1.6856039762496948, "learning_rate": 5.176489637305699e-06, "loss": 0.0136, "step": 3480 }, { "epoch": 2.091072498502097, "grad_norm": 1.147712230682373, "learning_rate": 5.16029792746114e-06, "loss": 0.0084, "step": 3490 }, { "epoch": 2.097064110245656, "grad_norm": 0.7007279396057129, "learning_rate": 5.144106217616581e-06, "loss": 0.0143, "step": 3500 }, { "epoch": 2.103055721989215, "grad_norm": 1.0783849954605103, "learning_rate": 5.127914507772021e-06, "loss": 0.015, "step": 3510 }, { "epoch": 2.109047333732774, "grad_norm": 1.133612036705017, "learning_rate": 5.111722797927462e-06, "loss": 0.0117, "step": 3520 }, { "epoch": 2.115038945476333, "grad_norm": 0.5530627965927124, "learning_rate": 5.095531088082902e-06, "loss": 0.0098, "step": 3530 }, { "epoch": 2.121030557219892, "grad_norm": 1.3948954343795776, "learning_rate": 5.0793393782383425e-06, "loss": 0.009, "step": 3540 }, { "epoch": 2.127022168963451, "grad_norm": 1.2842837572097778, "learning_rate": 5.063147668393783e-06, "loss": 0.0117, "step": 3550 }, { "epoch": 2.13301378070701, "grad_norm": 0.8959969878196716, "learning_rate": 5.046955958549224e-06, "loss": 0.0063, "step": 3560 }, { "epoch": 2.139005392450569, "grad_norm": 0.8143625855445862, "learning_rate": 5.030764248704664e-06, "loss": 0.0087, "step": 3570 }, { "epoch": 2.144997004194128, "grad_norm": 1.388192057609558, "learning_rate": 5.014572538860104e-06, "loss": 0.0148, "step": 3580 }, { "epoch": 2.1509886159376874, "grad_norm": 1.5829776525497437, "learning_rate": 4.998380829015545e-06, "loss": 0.0129, "step": 3590 }, { "epoch": 2.156980227681246, "grad_norm": 1.3284603357315063, "learning_rate": 4.982189119170985e-06, "loss": 0.0161, "step": 3600 }, { "epoch": 2.1629718394248054, "grad_norm": 0.4669111371040344, "learning_rate": 4.965997409326425e-06, "loss": 0.0093, "step": 3610 }, { "epoch": 2.168963451168364, "grad_norm": 2.0282106399536133, "learning_rate": 4.949805699481865e-06, "loss": 0.0107, "step": 3620 }, { "epoch": 2.1749550629119234, "grad_norm": 1.9077093601226807, "learning_rate": 4.933613989637306e-06, "loss": 0.0099, "step": 3630 }, { "epoch": 2.180946674655482, "grad_norm": 0.4548993408679962, "learning_rate": 4.917422279792747e-06, "loss": 0.0086, "step": 3640 }, { "epoch": 2.1869382863990414, "grad_norm": 1.3280978202819824, "learning_rate": 4.901230569948187e-06, "loss": 0.0119, "step": 3650 }, { "epoch": 2.1929298981426, "grad_norm": 0.618032693862915, "learning_rate": 4.8850388601036275e-06, "loss": 0.01, "step": 3660 }, { "epoch": 2.1989215098861594, "grad_norm": 1.7678560018539429, "learning_rate": 4.868847150259068e-06, "loss": 0.0163, "step": 3670 }, { "epoch": 2.204913121629718, "grad_norm": 1.3157169818878174, "learning_rate": 4.8526554404145086e-06, "loss": 0.0117, "step": 3680 }, { "epoch": 2.2109047333732774, "grad_norm": 2.100522518157959, "learning_rate": 4.836463730569949e-06, "loss": 0.0082, "step": 3690 }, { "epoch": 2.2168963451168366, "grad_norm": 1.703456163406372, "learning_rate": 4.820272020725389e-06, "loss": 0.0082, "step": 3700 }, { "epoch": 2.2228879568603954, "grad_norm": 1.569329023361206, "learning_rate": 4.804080310880829e-06, "loss": 0.0119, "step": 3710 }, { "epoch": 2.2288795686039546, "grad_norm": 1.2286568880081177, "learning_rate": 4.78788860103627e-06, "loss": 0.0109, "step": 3720 }, { "epoch": 2.2348711803475134, "grad_norm": 1.1602758169174194, "learning_rate": 4.77169689119171e-06, "loss": 0.0088, "step": 3730 }, { "epoch": 2.2408627920910726, "grad_norm": 1.0114362239837646, "learning_rate": 4.755505181347151e-06, "loss": 0.0085, "step": 3740 }, { "epoch": 2.2468544038346314, "grad_norm": 1.3842345476150513, "learning_rate": 4.739313471502591e-06, "loss": 0.0179, "step": 3750 }, { "epoch": 2.2528460155781906, "grad_norm": 1.3407899141311646, "learning_rate": 4.723121761658031e-06, "loss": 0.0119, "step": 3760 }, { "epoch": 2.2588376273217494, "grad_norm": 1.4571481943130493, "learning_rate": 4.706930051813472e-06, "loss": 0.0093, "step": 3770 }, { "epoch": 2.2648292390653086, "grad_norm": 0.9713364243507385, "learning_rate": 4.690738341968912e-06, "loss": 0.0098, "step": 3780 }, { "epoch": 2.2708208508088674, "grad_norm": 1.1515699625015259, "learning_rate": 4.6745466321243525e-06, "loss": 0.0088, "step": 3790 }, { "epoch": 2.2768124625524266, "grad_norm": 1.1340457201004028, "learning_rate": 4.658354922279793e-06, "loss": 0.0121, "step": 3800 }, { "epoch": 2.282804074295986, "grad_norm": 1.0477025508880615, "learning_rate": 4.642163212435234e-06, "loss": 0.0107, "step": 3810 }, { "epoch": 2.2887956860395446, "grad_norm": 1.2503265142440796, "learning_rate": 4.625971502590674e-06, "loss": 0.0108, "step": 3820 }, { "epoch": 2.2947872977831034, "grad_norm": 0.959955096244812, "learning_rate": 4.609779792746114e-06, "loss": 0.0115, "step": 3830 }, { "epoch": 2.3007789095266626, "grad_norm": 6.084997177124023, "learning_rate": 4.593588082901555e-06, "loss": 0.0175, "step": 3840 }, { "epoch": 2.306770521270222, "grad_norm": 1.329868197441101, "learning_rate": 4.577396373056995e-06, "loss": 0.0104, "step": 3850 }, { "epoch": 2.3127621330137806, "grad_norm": 0.42435407638549805, "learning_rate": 4.561204663212436e-06, "loss": 0.0075, "step": 3860 }, { "epoch": 2.31875374475734, "grad_norm": 0.8225539922714233, "learning_rate": 4.545012953367876e-06, "loss": 0.0084, "step": 3870 }, { "epoch": 2.3247453565008986, "grad_norm": 1.0772991180419922, "learning_rate": 4.528821243523316e-06, "loss": 0.0098, "step": 3880 }, { "epoch": 2.330736968244458, "grad_norm": 1.085127592086792, "learning_rate": 4.512629533678756e-06, "loss": 0.0092, "step": 3890 }, { "epoch": 2.3367285799880166, "grad_norm": 1.8185174465179443, "learning_rate": 4.496437823834197e-06, "loss": 0.0095, "step": 3900 }, { "epoch": 2.342720191731576, "grad_norm": 1.0053589344024658, "learning_rate": 4.4802461139896375e-06, "loss": 0.0144, "step": 3910 }, { "epoch": 2.3487118034751346, "grad_norm": 2.0738162994384766, "learning_rate": 4.464054404145078e-06, "loss": 0.0166, "step": 3920 }, { "epoch": 2.354703415218694, "grad_norm": 1.0764700174331665, "learning_rate": 4.447862694300519e-06, "loss": 0.0096, "step": 3930 }, { "epoch": 2.3606950269622526, "grad_norm": 2.6137382984161377, "learning_rate": 4.431670984455959e-06, "loss": 0.0117, "step": 3940 }, { "epoch": 2.366686638705812, "grad_norm": 1.5871500968933105, "learning_rate": 4.4154792746114e-06, "loss": 0.0085, "step": 3950 }, { "epoch": 2.372678250449371, "grad_norm": 1.3385639190673828, "learning_rate": 4.39928756476684e-06, "loss": 0.0102, "step": 3960 }, { "epoch": 2.37866986219293, "grad_norm": 1.6860779523849487, "learning_rate": 4.38309585492228e-06, "loss": 0.0136, "step": 3970 }, { "epoch": 2.384661473936489, "grad_norm": 0.7180661559104919, "learning_rate": 4.366904145077721e-06, "loss": 0.0082, "step": 3980 }, { "epoch": 2.390653085680048, "grad_norm": 2.4058666229248047, "learning_rate": 4.350712435233161e-06, "loss": 0.0119, "step": 3990 }, { "epoch": 2.396644697423607, "grad_norm": 1.4926502704620361, "learning_rate": 4.334520725388601e-06, "loss": 0.0162, "step": 4000 }, { "epoch": 2.396644697423607, "eval_loss": 0.028091730549931526, "eval_runtime": 4824.2853, "eval_samples_per_second": 2.767, "eval_steps_per_second": 0.346, "eval_wer": 2.031379968359076, "step": 4000 } ], "logging_steps": 10, "max_steps": 6676, "num_input_tokens_seen": 0, "num_train_epochs": 4, "save_steps": 2000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 3.693431387455488e+19, "train_batch_size": 16, "trial_name": null, "trial_params": null }