diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,28093 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.999922968147329, + "global_step": 45434, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0, + "learning_rate": 2.197802197802198e-05, + "loss": 1.7193, + "step": 10 + }, + { + "epoch": 0.0, + "learning_rate": 4.395604395604396e-05, + "loss": 1.4377, + "step": 20 + }, + { + "epoch": 0.0, + "learning_rate": 6.593406593406594e-05, + "loss": 1.1605, + "step": 30 + }, + { + "epoch": 0.0, + "learning_rate": 8.791208791208792e-05, + "loss": 1.0569, + "step": 40 + }, + { + "epoch": 0.0, + "learning_rate": 0.00010989010989010989, + "loss": 0.9816, + "step": 50 + }, + { + "epoch": 0.0, + "learning_rate": 0.00013186813186813188, + "loss": 0.9185, + "step": 60 + }, + { + "epoch": 0.0, + "learning_rate": 0.00015384615384615385, + "loss": 0.9019, + "step": 70 + }, + { + "epoch": 0.0, + "learning_rate": 0.00017582417582417585, + "loss": 0.8693, + "step": 80 + }, + { + "epoch": 0.0, + "learning_rate": 0.00019780219780219779, + "loss": 0.8435, + "step": 90 + }, + { + "epoch": 0.0, + "learning_rate": 0.00021978021978021978, + "loss": 0.8014, + "step": 100 + }, + { + "epoch": 0.0, + "learning_rate": 0.00024175824175824177, + "loss": 0.7911, + "step": 110 + }, + { + "epoch": 0.01, + "learning_rate": 0.00026373626373626377, + "loss": 0.7755, + "step": 120 + }, + { + "epoch": 0.01, + "learning_rate": 0.0002857142857142857, + "loss": 0.7548, + "step": 130 + }, + { + "epoch": 0.01, + "learning_rate": 0.0003076923076923077, + "loss": 0.7425, + "step": 140 + }, + { + "epoch": 0.01, + "learning_rate": 0.00032967032967032967, + "loss": 0.7122, + "step": 150 + }, + { + "epoch": 0.01, + "learning_rate": 0.0003516483516483517, + "loss": 0.7016, + "step": 160 + }, + { + "epoch": 0.01, + "learning_rate": 0.00037362637362637366, + "loss": 0.6982, + "step": 170 + }, + { + "epoch": 0.01, + "learning_rate": 0.00039560439560439557, + "loss": 0.6823, + "step": 180 + }, + { + "epoch": 0.01, + "learning_rate": 0.0004175824175824176, + "loss": 0.667, + "step": 190 + }, + { + "epoch": 0.01, + "learning_rate": 0.00043956043956043956, + "loss": 0.6597, + "step": 200 + }, + { + "epoch": 0.01, + "learning_rate": 0.0004615384615384616, + "loss": 0.6479, + "step": 210 + }, + { + "epoch": 0.01, + "learning_rate": 0.00048351648351648355, + "loss": 0.632, + "step": 220 + }, + { + "epoch": 0.01, + "learning_rate": 0.0005054945054945055, + "loss": 0.6405, + "step": 230 + }, + { + "epoch": 0.01, + "learning_rate": 0.0005274725274725275, + "loss": 0.6277, + "step": 240 + }, + { + "epoch": 0.01, + "learning_rate": 0.0005494505494505496, + "loss": 0.6182, + "step": 250 + }, + { + "epoch": 0.01, + "learning_rate": 0.0005714285714285714, + "loss": 0.6058, + "step": 260 + }, + { + "epoch": 0.01, + "learning_rate": 0.0005934065934065934, + "loss": 0.6067, + "step": 270 + }, + { + "epoch": 0.01, + "learning_rate": 0.0006153846153846154, + "loss": 0.5889, + "step": 280 + }, + { + "epoch": 0.01, + "learning_rate": 0.0006373626373626373, + "loss": 0.5959, + "step": 290 + }, + { + "epoch": 0.01, + "learning_rate": 0.0006593406593406593, + "loss": 0.5787, + "step": 300 + }, + { + "epoch": 0.01, + "learning_rate": 0.0006813186813186814, + "loss": 0.5684, + "step": 310 + }, + { + "epoch": 0.01, + "learning_rate": 0.0007032967032967034, + "loss": 0.5666, + "step": 320 + }, + { + "epoch": 0.01, + "learning_rate": 0.0007252747252747253, + "loss": 0.5644, + "step": 330 + }, + { + "epoch": 0.01, + "learning_rate": 0.0007472527472527473, + "loss": 0.5524, + "step": 340 + }, + { + "epoch": 0.02, + "learning_rate": 0.0007692307692307693, + "loss": 0.544, + "step": 350 + }, + { + "epoch": 0.02, + "learning_rate": 0.0007912087912087911, + "loss": 0.5469, + "step": 360 + }, + { + "epoch": 0.02, + "learning_rate": 0.0008131868131868132, + "loss": 0.5363, + "step": 370 + }, + { + "epoch": 0.02, + "learning_rate": 0.0008351648351648352, + "loss": 0.5416, + "step": 380 + }, + { + "epoch": 0.02, + "learning_rate": 0.0008571428571428571, + "loss": 0.5389, + "step": 390 + }, + { + "epoch": 0.02, + "learning_rate": 0.0008791208791208791, + "loss": 0.5113, + "step": 400 + }, + { + "epoch": 0.02, + "learning_rate": 0.0009010989010989011, + "loss": 0.5216, + "step": 410 + }, + { + "epoch": 0.02, + "learning_rate": 0.0009230769230769232, + "loss": 0.5126, + "step": 420 + }, + { + "epoch": 0.02, + "learning_rate": 0.0009450549450549451, + "loss": 0.5039, + "step": 430 + }, + { + "epoch": 0.02, + "learning_rate": 0.0009670329670329671, + "loss": 0.503, + "step": 440 + }, + { + "epoch": 0.02, + "learning_rate": 0.000989010989010989, + "loss": 0.4883, + "step": 450 + }, + { + "epoch": 0.02, + "learning_rate": 0.0009998888370128282, + "loss": 0.4962, + "step": 460 + }, + { + "epoch": 0.02, + "learning_rate": 0.0009996665110384845, + "loss": 0.5009, + "step": 470 + }, + { + "epoch": 0.02, + "learning_rate": 0.000999444185064141, + "loss": 0.4939, + "step": 480 + }, + { + "epoch": 0.02, + "learning_rate": 0.0009992218590897974, + "loss": 0.4931, + "step": 490 + }, + { + "epoch": 0.02, + "learning_rate": 0.0009989995331154539, + "loss": 0.4675, + "step": 500 + }, + { + "epoch": 0.02, + "eval_cer": 0.5152791661179975, + "eval_loss": 0.33519384264945984, + "eval_runtime": 84.3103, + "eval_samples_per_second": 118.538, + "eval_steps_per_second": 7.413, + "step": 500 + }, + { + "epoch": 0.02, + "learning_rate": 0.0009987772071411102, + "loss": 0.4714, + "step": 510 + }, + { + "epoch": 0.02, + "learning_rate": 0.0009985548811667667, + "loss": 0.4869, + "step": 520 + }, + { + "epoch": 0.02, + "learning_rate": 0.000998332555192423, + "loss": 0.4702, + "step": 530 + }, + { + "epoch": 0.02, + "learning_rate": 0.0009981102292180796, + "loss": 0.453, + "step": 540 + }, + { + "epoch": 0.02, + "learning_rate": 0.000997887903243736, + "loss": 0.4609, + "step": 550 + }, + { + "epoch": 0.02, + "learning_rate": 0.0009976655772693925, + "loss": 0.4517, + "step": 560 + }, + { + "epoch": 0.03, + "learning_rate": 0.0009974432512950488, + "loss": 0.4407, + "step": 570 + }, + { + "epoch": 0.03, + "learning_rate": 0.000997220925320705, + "loss": 0.4433, + "step": 580 + }, + { + "epoch": 0.03, + "learning_rate": 0.0009969985993463616, + "loss": 0.4385, + "step": 590 + }, + { + "epoch": 0.03, + "learning_rate": 0.000996776273372018, + "loss": 0.4397, + "step": 600 + }, + { + "epoch": 0.03, + "learning_rate": 0.0009965539473976745, + "loss": 0.4297, + "step": 610 + }, + { + "epoch": 0.03, + "learning_rate": 0.0009963316214233308, + "loss": 0.4362, + "step": 620 + }, + { + "epoch": 0.03, + "learning_rate": 0.0009961092954489873, + "loss": 0.4164, + "step": 630 + }, + { + "epoch": 0.03, + "learning_rate": 0.0009958869694746437, + "loss": 0.4156, + "step": 640 + }, + { + "epoch": 0.03, + "learning_rate": 0.0009956646435003002, + "loss": 0.4145, + "step": 650 + }, + { + "epoch": 0.03, + "learning_rate": 0.0009954423175259565, + "loss": 0.411, + "step": 660 + }, + { + "epoch": 0.03, + "learning_rate": 0.000995219991551613, + "loss": 0.418, + "step": 670 + }, + { + "epoch": 0.03, + "learning_rate": 0.0009949976655772694, + "loss": 0.4242, + "step": 680 + }, + { + "epoch": 0.03, + "learning_rate": 0.000994775339602926, + "loss": 0.4123, + "step": 690 + }, + { + "epoch": 0.03, + "learning_rate": 0.0009945530136285822, + "loss": 0.42, + "step": 700 + }, + { + "epoch": 0.03, + "learning_rate": 0.0009943306876542385, + "loss": 0.4007, + "step": 710 + }, + { + "epoch": 0.03, + "learning_rate": 0.000994108361679895, + "loss": 0.4011, + "step": 720 + }, + { + "epoch": 0.03, + "learning_rate": 0.0009938860357055514, + "loss": 0.3976, + "step": 730 + }, + { + "epoch": 0.03, + "learning_rate": 0.000993663709731208, + "loss": 0.3933, + "step": 740 + }, + { + "epoch": 0.03, + "learning_rate": 0.0009934413837568642, + "loss": 0.4023, + "step": 750 + }, + { + "epoch": 0.03, + "learning_rate": 0.0009932190577825208, + "loss": 0.396, + "step": 760 + }, + { + "epoch": 0.03, + "learning_rate": 0.000992996731808177, + "loss": 0.4006, + "step": 770 + }, + { + "epoch": 0.03, + "learning_rate": 0.0009927744058338336, + "loss": 0.3934, + "step": 780 + }, + { + "epoch": 0.03, + "learning_rate": 0.00099255207985949, + "loss": 0.3931, + "step": 790 + }, + { + "epoch": 0.04, + "learning_rate": 0.0009923297538851465, + "loss": 0.3875, + "step": 800 + }, + { + "epoch": 0.04, + "learning_rate": 0.0009921074279108028, + "loss": 0.387, + "step": 810 + }, + { + "epoch": 0.04, + "learning_rate": 0.0009918851019364593, + "loss": 0.3832, + "step": 820 + }, + { + "epoch": 0.04, + "learning_rate": 0.0009916627759621157, + "loss": 0.3815, + "step": 830 + }, + { + "epoch": 0.04, + "learning_rate": 0.000991440449987772, + "loss": 0.3874, + "step": 840 + }, + { + "epoch": 0.04, + "learning_rate": 0.0009912181240134285, + "loss": 0.3864, + "step": 850 + }, + { + "epoch": 0.04, + "learning_rate": 0.0009909957980390848, + "loss": 0.3744, + "step": 860 + }, + { + "epoch": 0.04, + "learning_rate": 0.0009907734720647414, + "loss": 0.3641, + "step": 870 + }, + { + "epoch": 0.04, + "learning_rate": 0.0009905511460903977, + "loss": 0.366, + "step": 880 + }, + { + "epoch": 0.04, + "learning_rate": 0.0009903288201160542, + "loss": 0.3745, + "step": 890 + }, + { + "epoch": 0.04, + "learning_rate": 0.0009901064941417106, + "loss": 0.3639, + "step": 900 + }, + { + "epoch": 0.04, + "learning_rate": 0.000989884168167367, + "loss": 0.3585, + "step": 910 + }, + { + "epoch": 0.04, + "learning_rate": 0.0009896618421930234, + "loss": 0.3665, + "step": 920 + }, + { + "epoch": 0.04, + "learning_rate": 0.00098943951621868, + "loss": 0.3578, + "step": 930 + }, + { + "epoch": 0.04, + "learning_rate": 0.0009892171902443363, + "loss": 0.364, + "step": 940 + }, + { + "epoch": 0.04, + "learning_rate": 0.0009889948642699926, + "loss": 0.3632, + "step": 950 + }, + { + "epoch": 0.04, + "learning_rate": 0.0009887725382956491, + "loss": 0.3515, + "step": 960 + }, + { + "epoch": 0.04, + "learning_rate": 0.0009885502123213054, + "loss": 0.3569, + "step": 970 + }, + { + "epoch": 0.04, + "learning_rate": 0.000988327886346962, + "loss": 0.3553, + "step": 980 + }, + { + "epoch": 0.04, + "learning_rate": 0.0009881055603726183, + "loss": 0.3514, + "step": 990 + }, + { + "epoch": 0.04, + "learning_rate": 0.0009878832343982748, + "loss": 0.3607, + "step": 1000 + }, + { + "epoch": 0.04, + "eval_cer": 0.5093334302018216, + "eval_loss": 0.2502743601799011, + "eval_runtime": 83.7014, + "eval_samples_per_second": 119.401, + "eval_steps_per_second": 7.467, + "step": 1000 + }, + { + "epoch": 0.04, + "learning_rate": 0.0009876609084239311, + "loss": 0.3553, + "step": 1010 + }, + { + "epoch": 0.04, + "learning_rate": 0.0009874385824495877, + "loss": 0.3491, + "step": 1020 + }, + { + "epoch": 0.05, + "learning_rate": 0.000987216256475244, + "loss": 0.3467, + "step": 1030 + }, + { + "epoch": 0.05, + "learning_rate": 0.0009869939305009005, + "loss": 0.3518, + "step": 1040 + }, + { + "epoch": 0.05, + "learning_rate": 0.0009867716045265569, + "loss": 0.3432, + "step": 1050 + }, + { + "epoch": 0.05, + "learning_rate": 0.0009865492785522134, + "loss": 0.3437, + "step": 1060 + }, + { + "epoch": 0.05, + "learning_rate": 0.0009863269525778697, + "loss": 0.342, + "step": 1070 + }, + { + "epoch": 0.05, + "learning_rate": 0.000986104626603526, + "loss": 0.3442, + "step": 1080 + }, + { + "epoch": 0.05, + "learning_rate": 0.0009858823006291826, + "loss": 0.3447, + "step": 1090 + }, + { + "epoch": 0.05, + "learning_rate": 0.0009856599746548389, + "loss": 0.3369, + "step": 1100 + }, + { + "epoch": 0.05, + "learning_rate": 0.0009854376486804954, + "loss": 0.3376, + "step": 1110 + }, + { + "epoch": 0.05, + "learning_rate": 0.0009852153227061517, + "loss": 0.3383, + "step": 1120 + }, + { + "epoch": 0.05, + "learning_rate": 0.0009849929967318083, + "loss": 0.3374, + "step": 1130 + }, + { + "epoch": 0.05, + "learning_rate": 0.0009847706707574646, + "loss": 0.3333, + "step": 1140 + }, + { + "epoch": 0.05, + "learning_rate": 0.0009845483447831211, + "loss": 0.3298, + "step": 1150 + }, + { + "epoch": 0.05, + "learning_rate": 0.0009843260188087774, + "loss": 0.3352, + "step": 1160 + }, + { + "epoch": 0.05, + "learning_rate": 0.000984103692834434, + "loss": 0.3367, + "step": 1170 + }, + { + "epoch": 0.05, + "learning_rate": 0.0009838813668600903, + "loss": 0.3261, + "step": 1180 + }, + { + "epoch": 0.05, + "learning_rate": 0.0009836590408857466, + "loss": 0.3302, + "step": 1190 + }, + { + "epoch": 0.05, + "learning_rate": 0.0009834367149114032, + "loss": 0.3276, + "step": 1200 + }, + { + "epoch": 0.05, + "learning_rate": 0.0009832143889370595, + "loss": 0.3364, + "step": 1210 + }, + { + "epoch": 0.05, + "learning_rate": 0.000982992062962716, + "loss": 0.3282, + "step": 1220 + }, + { + "epoch": 0.05, + "learning_rate": 0.0009827697369883723, + "loss": 0.3361, + "step": 1230 + }, + { + "epoch": 0.05, + "learning_rate": 0.0009825474110140289, + "loss": 0.3277, + "step": 1240 + }, + { + "epoch": 0.06, + "learning_rate": 0.0009823250850396852, + "loss": 0.3127, + "step": 1250 + }, + { + "epoch": 0.06, + "learning_rate": 0.0009821027590653417, + "loss": 0.3218, + "step": 1260 + }, + { + "epoch": 0.06, + "learning_rate": 0.000981880433090998, + "loss": 0.3177, + "step": 1270 + }, + { + "epoch": 0.06, + "learning_rate": 0.0009816581071166546, + "loss": 0.3277, + "step": 1280 + }, + { + "epoch": 0.06, + "learning_rate": 0.000981435781142311, + "loss": 0.3073, + "step": 1290 + }, + { + "epoch": 0.06, + "learning_rate": 0.0009812134551679674, + "loss": 0.3276, + "step": 1300 + }, + { + "epoch": 0.06, + "learning_rate": 0.0009809911291936237, + "loss": 0.3185, + "step": 1310 + }, + { + "epoch": 0.06, + "learning_rate": 0.00098076880321928, + "loss": 0.3169, + "step": 1320 + }, + { + "epoch": 0.06, + "learning_rate": 0.0009805464772449366, + "loss": 0.3116, + "step": 1330 + }, + { + "epoch": 0.06, + "learning_rate": 0.000980324151270593, + "loss": 0.326, + "step": 1340 + }, + { + "epoch": 0.06, + "learning_rate": 0.0009801018252962495, + "loss": 0.3139, + "step": 1350 + }, + { + "epoch": 0.06, + "learning_rate": 0.0009798794993219058, + "loss": 0.3064, + "step": 1360 + }, + { + "epoch": 0.06, + "learning_rate": 0.0009796571733475623, + "loss": 0.3041, + "step": 1370 + }, + { + "epoch": 0.06, + "learning_rate": 0.0009794348473732186, + "loss": 0.3058, + "step": 1380 + }, + { + "epoch": 0.06, + "learning_rate": 0.0009792125213988752, + "loss": 0.3174, + "step": 1390 + }, + { + "epoch": 0.06, + "learning_rate": 0.0009789901954245315, + "loss": 0.3126, + "step": 1400 + }, + { + "epoch": 0.06, + "learning_rate": 0.000978767869450188, + "loss": 0.3137, + "step": 1410 + }, + { + "epoch": 0.06, + "learning_rate": 0.0009785455434758443, + "loss": 0.3143, + "step": 1420 + }, + { + "epoch": 0.06, + "learning_rate": 0.0009783232175015009, + "loss": 0.3136, + "step": 1430 + }, + { + "epoch": 0.06, + "learning_rate": 0.0009781008915271572, + "loss": 0.3194, + "step": 1440 + }, + { + "epoch": 0.06, + "learning_rate": 0.0009778785655528135, + "loss": 0.3029, + "step": 1450 + }, + { + "epoch": 0.06, + "learning_rate": 0.0009776562395784698, + "loss": 0.3088, + "step": 1460 + }, + { + "epoch": 0.06, + "learning_rate": 0.0009774339136041264, + "loss": 0.3008, + "step": 1470 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009772115876297827, + "loss": 0.3142, + "step": 1480 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009769892616554392, + "loss": 0.299, + "step": 1490 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009767669356810955, + "loss": 0.2985, + "step": 1500 + }, + { + "epoch": 0.07, + "eval_cer": 0.5073828707538682, + "eval_loss": 0.21521282196044922, + "eval_runtime": 82.87, + "eval_samples_per_second": 120.599, + "eval_steps_per_second": 7.542, + "step": 1500 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009765446097067521, + "loss": 0.2969, + "step": 1510 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009763222837324085, + "loss": 0.2939, + "step": 1520 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009760999577580649, + "loss": 0.3024, + "step": 1530 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009758776317837214, + "loss": 0.2968, + "step": 1540 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009756553058093778, + "loss": 0.2927, + "step": 1550 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009754329798350341, + "loss": 0.2935, + "step": 1560 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009752106538606905, + "loss": 0.3003, + "step": 1570 + }, + { + "epoch": 0.07, + "learning_rate": 0.000974988327886347, + "loss": 0.2987, + "step": 1580 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009747660019120034, + "loss": 0.299, + "step": 1590 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009745436759376598, + "loss": 0.2931, + "step": 1600 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009743213499633162, + "loss": 0.2941, + "step": 1610 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009740990239889727, + "loss": 0.2963, + "step": 1620 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009738766980146291, + "loss": 0.3031, + "step": 1630 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009736543720402855, + "loss": 0.288, + "step": 1640 + }, + { + "epoch": 0.07, + "learning_rate": 0.000973432046065942, + "loss": 0.2947, + "step": 1650 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009732097200915984, + "loss": 0.2985, + "step": 1660 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009729873941172548, + "loss": 0.2898, + "step": 1670 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009727650681429111, + "loss": 0.2951, + "step": 1680 + }, + { + "epoch": 0.07, + "learning_rate": 0.0009725427421685676, + "loss": 0.2842, + "step": 1690 + }, + { + "epoch": 0.07, + "learning_rate": 0.000972320416194224, + "loss": 0.2919, + "step": 1700 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009720980902198804, + "loss": 0.2834, + "step": 1710 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009718757642455368, + "loss": 0.2903, + "step": 1720 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009716534382711933, + "loss": 0.2878, + "step": 1730 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009714311122968497, + "loss": 0.2896, + "step": 1740 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009712087863225061, + "loss": 0.2907, + "step": 1750 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009709864603481625, + "loss": 0.2828, + "step": 1760 + }, + { + "epoch": 0.08, + "learning_rate": 0.000970764134373819, + "loss": 0.2849, + "step": 1770 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009705418083994754, + "loss": 0.271, + "step": 1780 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009703194824251318, + "loss": 0.2779, + "step": 1790 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009700971564507883, + "loss": 0.2721, + "step": 1800 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009698748304764446, + "loss": 0.2793, + "step": 1810 + }, + { + "epoch": 0.08, + "learning_rate": 0.000969652504502101, + "loss": 0.2822, + "step": 1820 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009694301785277574, + "loss": 0.2868, + "step": 1830 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009692078525534139, + "loss": 0.2776, + "step": 1840 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009689855265790703, + "loss": 0.2831, + "step": 1850 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009687632006047267, + "loss": 0.2673, + "step": 1860 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009685408746303831, + "loss": 0.2785, + "step": 1870 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009683185486560396, + "loss": 0.2764, + "step": 1880 + }, + { + "epoch": 0.08, + "learning_rate": 0.000968096222681696, + "loss": 0.2725, + "step": 1890 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009678738967073524, + "loss": 0.281, + "step": 1900 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009676515707330088, + "loss": 0.2715, + "step": 1910 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009674292447586653, + "loss": 0.2792, + "step": 1920 + }, + { + "epoch": 0.08, + "learning_rate": 0.0009672069187843216, + "loss": 0.278, + "step": 1930 + }, + { + "epoch": 0.09, + "learning_rate": 0.000966984592809978, + "loss": 0.2707, + "step": 1940 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009667622668356343, + "loss": 0.2697, + "step": 1950 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009665399408612908, + "loss": 0.2681, + "step": 1960 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009663176148869472, + "loss": 0.2605, + "step": 1970 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009660952889126036, + "loss": 0.2824, + "step": 1980 + }, + { + "epoch": 0.09, + "learning_rate": 0.00096587296293826, + "loss": 0.2676, + "step": 1990 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009656506369639165, + "loss": 0.261, + "step": 2000 + }, + { + "epoch": 0.09, + "eval_cer": 0.5057105919853867, + "eval_loss": 0.19225744903087616, + "eval_runtime": 83.8316, + "eval_samples_per_second": 119.215, + "eval_steps_per_second": 7.455, + "step": 2000 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009654283109895729, + "loss": 0.2581, + "step": 2010 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009652059850152293, + "loss": 0.2657, + "step": 2020 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009649836590408858, + "loss": 0.2714, + "step": 2030 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009647613330665422, + "loss": 0.2605, + "step": 2040 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009645390070921985, + "loss": 0.2773, + "step": 2050 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009643166811178549, + "loss": 0.2708, + "step": 2060 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009640943551435114, + "loss": 0.2646, + "step": 2070 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009638720291691678, + "loss": 0.2742, + "step": 2080 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009636497031948242, + "loss": 0.2676, + "step": 2090 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009634273772204806, + "loss": 0.2787, + "step": 2100 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009632050512461371, + "loss": 0.265, + "step": 2110 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009629827252717935, + "loss": 0.267, + "step": 2120 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009627603992974499, + "loss": 0.2573, + "step": 2130 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009625380733231064, + "loss": 0.2636, + "step": 2140 + }, + { + "epoch": 0.09, + "learning_rate": 0.0009623157473487628, + "loss": 0.2637, + "step": 2150 + }, + { + "epoch": 0.1, + "learning_rate": 0.0009620934213744192, + "loss": 0.2721, + "step": 2160 + }, + { + "epoch": 0.1, + "learning_rate": 0.0009618710954000755, + "loss": 0.2621, + "step": 2170 + }, + { + "epoch": 0.1, + "learning_rate": 0.000961648769425732, + "loss": 0.2614, + "step": 2180 + }, + { + "epoch": 0.1, + "learning_rate": 0.0009614264434513884, + "loss": 0.2675, + "step": 2190 + }, + { + "epoch": 0.1, + "learning_rate": 0.0009612041174770448, + "loss": 0.2623, + "step": 2200 + }, + { + "epoch": 0.1, + "learning_rate": 0.0009609817915027012, + "loss": 0.2617, + "step": 2210 + }, + { + "epoch": 0.1, + "learning_rate": 0.0009607594655283577, + "loss": 0.2596, + "step": 2220 + }, + { + "epoch": 0.1, + "learning_rate": 0.0009605371395540141, + "loss": 0.2653, + "step": 2230 + }, + { + "epoch": 0.1, + "learning_rate": 0.0009603148135796705, + "loss": 0.2551, + "step": 2240 + }, + { + "epoch": 0.1, + "learning_rate": 0.000960092487605327, + "loss": 0.2596, + "step": 2250 + }, + { + "epoch": 0.1, + "learning_rate": 0.0009598701616309834, + "loss": 0.259, + "step": 2260 + }, + { + "epoch": 0.1, + "learning_rate": 0.0009596478356566398, + "loss": 0.2609, + "step": 2270 + }, + { + "epoch": 0.1, + "learning_rate": 0.0009594255096822962, + "loss": 0.2532, + "step": 2280 + }, + { + "epoch": 0.1, + "learning_rate": 0.0009592031837079525, + "loss": 0.2623, + "step": 2290 + }, + { + "epoch": 0.1, + "learning_rate": 0.000958980857733609, + "loss": 0.2503, + "step": 2300 + }, + { + "epoch": 0.1, + "learning_rate": 0.0009587585317592654, + "loss": 0.2576, + "step": 2310 + }, + { + "epoch": 0.1, + "learning_rate": 0.0009585362057849218, + "loss": 0.2539, + "step": 2320 + }, + { + "epoch": 0.1, + "learning_rate": 0.0009583138798105783, + "loss": 0.2632, + "step": 2330 + }, + { + "epoch": 0.1, + "learning_rate": 0.0009580915538362347, + "loss": 0.2554, + "step": 2340 + }, + { + "epoch": 0.1, + "learning_rate": 0.0009578692278618911, + "loss": 0.2566, + "step": 2350 + }, + { + "epoch": 0.1, + "learning_rate": 0.0009576469018875475, + "loss": 0.2592, + "step": 2360 + }, + { + "epoch": 0.1, + "learning_rate": 0.000957424575913204, + "loss": 0.2575, + "step": 2370 + }, + { + "epoch": 0.1, + "learning_rate": 0.0009572022499388604, + "loss": 0.2549, + "step": 2380 + }, + { + "epoch": 0.11, + "learning_rate": 0.0009569799239645168, + "loss": 0.2519, + "step": 2390 + }, + { + "epoch": 0.11, + "learning_rate": 0.0009567575979901732, + "loss": 0.2554, + "step": 2400 + }, + { + "epoch": 0.11, + "learning_rate": 0.0009565352720158297, + "loss": 0.2544, + "step": 2410 + }, + { + "epoch": 0.11, + "learning_rate": 0.000956312946041486, + "loss": 0.2452, + "step": 2420 + }, + { + "epoch": 0.11, + "learning_rate": 0.0009560906200671424, + "loss": 0.258, + "step": 2430 + }, + { + "epoch": 0.11, + "learning_rate": 0.0009558682940927989, + "loss": 0.2457, + "step": 2440 + }, + { + "epoch": 0.11, + "learning_rate": 0.0009556459681184553, + "loss": 0.2532, + "step": 2450 + }, + { + "epoch": 0.11, + "learning_rate": 0.0009554236421441117, + "loss": 0.2432, + "step": 2460 + }, + { + "epoch": 0.11, + "learning_rate": 0.0009552013161697681, + "loss": 0.2497, + "step": 2470 + }, + { + "epoch": 0.11, + "learning_rate": 0.0009549789901954246, + "loss": 0.2515, + "step": 2480 + }, + { + "epoch": 0.11, + "learning_rate": 0.000954756664221081, + "loss": 0.2565, + "step": 2490 + }, + { + "epoch": 0.11, + "learning_rate": 0.0009545343382467374, + "loss": 0.2426, + "step": 2500 + }, + { + "epoch": 0.11, + "eval_cer": 0.5044463284788815, + "eval_loss": 0.1803441345691681, + "eval_runtime": 82.4693, + "eval_samples_per_second": 121.185, + "eval_steps_per_second": 7.579, + "step": 2500 + }, + { + "epoch": 0.11, + "learning_rate": 0.0009543120122723938, + "loss": 0.2449, + "step": 2510 + }, + { + "epoch": 0.11, + "learning_rate": 0.0009540896862980503, + "loss": 0.244, + "step": 2520 + }, + { + "epoch": 0.11, + "learning_rate": 0.0009538673603237067, + "loss": 0.2542, + "step": 2530 + }, + { + "epoch": 0.11, + "learning_rate": 0.000953645034349363, + "loss": 0.2491, + "step": 2540 + }, + { + "epoch": 0.11, + "learning_rate": 0.0009534227083750194, + "loss": 0.2402, + "step": 2550 + }, + { + "epoch": 0.11, + "learning_rate": 0.0009532003824006759, + "loss": 0.2444, + "step": 2560 + }, + { + "epoch": 0.11, + "learning_rate": 0.0009529780564263323, + "loss": 0.2534, + "step": 2570 + }, + { + "epoch": 0.11, + "learning_rate": 0.0009527557304519887, + "loss": 0.2506, + "step": 2580 + }, + { + "epoch": 0.11, + "learning_rate": 0.0009525334044776452, + "loss": 0.2472, + "step": 2590 + }, + { + "epoch": 0.11, + "learning_rate": 0.0009523110785033016, + "loss": 0.2436, + "step": 2600 + }, + { + "epoch": 0.11, + "learning_rate": 0.000952088752528958, + "loss": 0.2463, + "step": 2610 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009518664265546144, + "loss": 0.2499, + "step": 2620 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009516441005802709, + "loss": 0.2458, + "step": 2630 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009514217746059273, + "loss": 0.2395, + "step": 2640 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009511994486315837, + "loss": 0.2437, + "step": 2650 + }, + { + "epoch": 0.12, + "learning_rate": 0.00095097712265724, + "loss": 0.2509, + "step": 2660 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009507547966828965, + "loss": 0.2394, + "step": 2670 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009505324707085529, + "loss": 0.2412, + "step": 2680 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009503101447342093, + "loss": 0.2404, + "step": 2690 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009500878187598657, + "loss": 0.2405, + "step": 2700 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009498654927855222, + "loss": 0.2416, + "step": 2710 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009496431668111786, + "loss": 0.2455, + "step": 2720 + }, + { + "epoch": 0.12, + "learning_rate": 0.000949420840836835, + "loss": 0.2416, + "step": 2730 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009491985148624915, + "loss": 0.2323, + "step": 2740 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009489761888881479, + "loss": 0.2305, + "step": 2750 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009487538629138043, + "loss": 0.2395, + "step": 2760 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009485315369394607, + "loss": 0.237, + "step": 2770 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009483092109651171, + "loss": 0.2431, + "step": 2780 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009480868849907735, + "loss": 0.2343, + "step": 2790 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009478645590164299, + "loss": 0.2384, + "step": 2800 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009476422330420863, + "loss": 0.2336, + "step": 2810 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009474199070677428, + "loss": 0.2334, + "step": 2820 + }, + { + "epoch": 0.12, + "learning_rate": 0.0009471975810933992, + "loss": 0.2411, + "step": 2830 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009469752551190556, + "loss": 0.2405, + "step": 2840 + }, + { + "epoch": 0.13, + "learning_rate": 0.000946752929144712, + "loss": 0.238, + "step": 2850 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009465306031703685, + "loss": 0.2393, + "step": 2860 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009463082771960249, + "loss": 0.2344, + "step": 2870 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009460859512216813, + "loss": 0.2299, + "step": 2880 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009458636252473378, + "loss": 0.2323, + "step": 2890 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009456412992729942, + "loss": 0.2347, + "step": 2900 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009454189732986505, + "loss": 0.2315, + "step": 2910 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009451966473243069, + "loss": 0.2327, + "step": 2920 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009449743213499634, + "loss": 0.2374, + "step": 2930 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009447519953756198, + "loss": 0.2388, + "step": 2940 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009445296694012762, + "loss": 0.2298, + "step": 2950 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009443073434269325, + "loss": 0.2267, + "step": 2960 + }, + { + "epoch": 0.13, + "learning_rate": 0.000944085017452589, + "loss": 0.2372, + "step": 2970 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009438626914782454, + "loss": 0.2357, + "step": 2980 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009436403655039018, + "loss": 0.243, + "step": 2990 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009434180395295582, + "loss": 0.236, + "step": 3000 + }, + { + "epoch": 0.13, + "eval_cer": 0.5032359048241156, + "eval_loss": 0.16923315823078156, + "eval_runtime": 83.4141, + "eval_samples_per_second": 119.812, + "eval_steps_per_second": 7.493, + "step": 3000 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009431957135552147, + "loss": 0.2309, + "step": 3010 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009429733875808711, + "loss": 0.2285, + "step": 3020 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009427510616065274, + "loss": 0.2362, + "step": 3030 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009425287356321838, + "loss": 0.2299, + "step": 3040 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009423064096578403, + "loss": 0.234, + "step": 3050 + }, + { + "epoch": 0.13, + "learning_rate": 0.0009420840836834967, + "loss": 0.2337, + "step": 3060 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009418617577091531, + "loss": 0.2282, + "step": 3070 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009416394317348096, + "loss": 0.2293, + "step": 3080 + }, + { + "epoch": 0.14, + "learning_rate": 0.000941417105760466, + "loss": 0.2309, + "step": 3090 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009411947797861224, + "loss": 0.2255, + "step": 3100 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009409724538117788, + "loss": 0.2365, + "step": 3110 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009407501278374353, + "loss": 0.2394, + "step": 3120 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009405278018630917, + "loss": 0.2354, + "step": 3130 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009403054758887481, + "loss": 0.2326, + "step": 3140 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009400831499144044, + "loss": 0.2278, + "step": 3150 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009398608239400609, + "loss": 0.2228, + "step": 3160 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009396384979657173, + "loss": 0.2417, + "step": 3170 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009394161719913737, + "loss": 0.2262, + "step": 3180 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009391938460170301, + "loss": 0.2254, + "step": 3190 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009389715200426866, + "loss": 0.229, + "step": 3200 + }, + { + "epoch": 0.14, + "learning_rate": 0.000938749194068343, + "loss": 0.2253, + "step": 3210 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009385268680939994, + "loss": 0.2205, + "step": 3220 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009383045421196559, + "loss": 0.2288, + "step": 3230 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009380822161453123, + "loss": 0.2298, + "step": 3240 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009378598901709687, + "loss": 0.2237, + "step": 3250 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009376375641966251, + "loss": 0.2265, + "step": 3260 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009374152382222815, + "loss": 0.2272, + "step": 3270 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009371929122479379, + "loss": 0.2291, + "step": 3280 + }, + { + "epoch": 0.14, + "learning_rate": 0.0009369705862735943, + "loss": 0.2317, + "step": 3290 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009367482602992507, + "loss": 0.228, + "step": 3300 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009365259343249072, + "loss": 0.2246, + "step": 3310 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009363036083505636, + "loss": 0.223, + "step": 3320 + }, + { + "epoch": 0.15, + "learning_rate": 0.00093608128237622, + "loss": 0.227, + "step": 3330 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009358589564018764, + "loss": 0.2376, + "step": 3340 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009356366304275329, + "loss": 0.2173, + "step": 3350 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009354143044531893, + "loss": 0.2268, + "step": 3360 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009351919784788457, + "loss": 0.2177, + "step": 3370 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009349696525045022, + "loss": 0.2204, + "step": 3380 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009347473265301586, + "loss": 0.2329, + "step": 3390 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009345250005558149, + "loss": 0.2131, + "step": 3400 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009343026745814713, + "loss": 0.227, + "step": 3410 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009340803486071278, + "loss": 0.2232, + "step": 3420 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009338580226327842, + "loss": 0.2181, + "step": 3430 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009336356966584406, + "loss": 0.2221, + "step": 3440 + }, + { + "epoch": 0.15, + "learning_rate": 0.000933413370684097, + "loss": 0.2286, + "step": 3450 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009331910447097535, + "loss": 0.2177, + "step": 3460 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009329687187354099, + "loss": 0.2227, + "step": 3470 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009327463927610663, + "loss": 0.2175, + "step": 3480 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009325240667867227, + "loss": 0.2175, + "step": 3490 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009323017408123792, + "loss": 0.224, + "step": 3500 + }, + { + "epoch": 0.15, + "eval_cer": 0.5037554918270456, + "eval_loss": 0.1586163341999054, + "eval_runtime": 83.9694, + "eval_samples_per_second": 119.02, + "eval_steps_per_second": 7.443, + "step": 3500 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009320794148380356, + "loss": 0.225, + "step": 3510 + }, + { + "epoch": 0.15, + "learning_rate": 0.0009318570888636919, + "loss": 0.2166, + "step": 3520 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009316347628893484, + "loss": 0.223, + "step": 3530 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009314124369150048, + "loss": 0.2252, + "step": 3540 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009311901109406612, + "loss": 0.2169, + "step": 3550 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009309677849663176, + "loss": 0.2172, + "step": 3560 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009307454589919741, + "loss": 0.2154, + "step": 3570 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009305231330176305, + "loss": 0.2211, + "step": 3580 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009303008070432869, + "loss": 0.2168, + "step": 3590 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009300784810689433, + "loss": 0.2156, + "step": 3600 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009298561550945998, + "loss": 0.2172, + "step": 3610 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009296338291202562, + "loss": 0.2162, + "step": 3620 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009294115031459126, + "loss": 0.2112, + "step": 3630 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009291891771715689, + "loss": 0.2143, + "step": 3640 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009289668511972254, + "loss": 0.2156, + "step": 3650 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009287445252228818, + "loss": 0.2213, + "step": 3660 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009285221992485382, + "loss": 0.2213, + "step": 3670 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009282998732741947, + "loss": 0.2188, + "step": 3680 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009280775472998511, + "loss": 0.215, + "step": 3690 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009278552213255075, + "loss": 0.2207, + "step": 3700 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009276328953511639, + "loss": 0.2146, + "step": 3710 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009274105693768204, + "loss": 0.2239, + "step": 3720 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009271882434024768, + "loss": 0.2155, + "step": 3730 + }, + { + "epoch": 0.16, + "learning_rate": 0.0009269659174281332, + "loss": 0.2199, + "step": 3740 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009267435914537896, + "loss": 0.2152, + "step": 3750 + }, + { + "epoch": 0.17, + "learning_rate": 0.000926521265479446, + "loss": 0.2128, + "step": 3760 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009262989395051024, + "loss": 0.2163, + "step": 3770 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009260766135307588, + "loss": 0.2142, + "step": 3780 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009258542875564152, + "loss": 0.2124, + "step": 3790 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009256319615820717, + "loss": 0.2098, + "step": 3800 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009254096356077281, + "loss": 0.2186, + "step": 3810 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009251873096333845, + "loss": 0.2146, + "step": 3820 + }, + { + "epoch": 0.17, + "learning_rate": 0.000924964983659041, + "loss": 0.2126, + "step": 3830 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009247426576846974, + "loss": 0.2175, + "step": 3840 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009245203317103538, + "loss": 0.2148, + "step": 3850 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009242980057360102, + "loss": 0.2166, + "step": 3860 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009240756797616667, + "loss": 0.2188, + "step": 3870 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009238533537873231, + "loss": 0.2164, + "step": 3880 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009236310278129794, + "loss": 0.2133, + "step": 3890 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009234087018386358, + "loss": 0.2182, + "step": 3900 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009231863758642923, + "loss": 0.2094, + "step": 3910 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009229640498899487, + "loss": 0.2145, + "step": 3920 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009227417239156051, + "loss": 0.2105, + "step": 3930 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009225193979412615, + "loss": 0.2094, + "step": 3940 + }, + { + "epoch": 0.17, + "learning_rate": 0.000922297071966918, + "loss": 0.2082, + "step": 3950 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009220747459925743, + "loss": 0.2122, + "step": 3960 + }, + { + "epoch": 0.17, + "learning_rate": 0.0009218524200182307, + "loss": 0.2163, + "step": 3970 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009216300940438871, + "loss": 0.2138, + "step": 3980 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009214077680695436, + "loss": 0.2082, + "step": 3990 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009211854420952, + "loss": 0.2126, + "step": 4000 + }, + { + "epoch": 0.18, + "eval_cer": 0.5031178463540366, + "eval_loss": 0.1523096263408661, + "eval_runtime": 82.3386, + "eval_samples_per_second": 121.377, + "eval_steps_per_second": 7.591, + "step": 4000 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009209631161208563, + "loss": 0.2096, + "step": 4010 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009207407901465128, + "loss": 0.2035, + "step": 4020 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009205184641721692, + "loss": 0.2135, + "step": 4030 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009202961381978256, + "loss": 0.2217, + "step": 4040 + }, + { + "epoch": 0.18, + "learning_rate": 0.000920073812223482, + "loss": 0.2019, + "step": 4050 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009198514862491385, + "loss": 0.2059, + "step": 4060 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009196291602747949, + "loss": 0.2101, + "step": 4070 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009194068343004513, + "loss": 0.2052, + "step": 4080 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009191845083261077, + "loss": 0.2063, + "step": 4090 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009189621823517642, + "loss": 0.2069, + "step": 4100 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009187398563774206, + "loss": 0.1967, + "step": 4110 + }, + { + "epoch": 0.18, + "learning_rate": 0.000918517530403077, + "loss": 0.2108, + "step": 4120 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009182952044287333, + "loss": 0.2026, + "step": 4130 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009180728784543898, + "loss": 0.2066, + "step": 4140 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009178505524800462, + "loss": 0.2112, + "step": 4150 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009176282265057026, + "loss": 0.211, + "step": 4160 + }, + { + "epoch": 0.18, + "learning_rate": 0.000917405900531359, + "loss": 0.2024, + "step": 4170 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009171835745570155, + "loss": 0.2031, + "step": 4180 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009169612485826719, + "loss": 0.2056, + "step": 4190 + }, + { + "epoch": 0.18, + "learning_rate": 0.0009167389226083283, + "loss": 0.1978, + "step": 4200 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009165165966339848, + "loss": 0.2052, + "step": 4210 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009162942706596412, + "loss": 0.2085, + "step": 4220 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009160719446852976, + "loss": 0.213, + "step": 4230 + }, + { + "epoch": 0.19, + "learning_rate": 0.000915849618710954, + "loss": 0.207, + "step": 4240 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009156272927366104, + "loss": 0.2052, + "step": 4250 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009154049667622668, + "loss": 0.2105, + "step": 4260 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009151826407879232, + "loss": 0.2051, + "step": 4270 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009149603148135796, + "loss": 0.2101, + "step": 4280 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009147379888392361, + "loss": 0.2071, + "step": 4290 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009145156628648925, + "loss": 0.204, + "step": 4300 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009142933368905489, + "loss": 0.2066, + "step": 4310 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009140710109162054, + "loss": 0.2082, + "step": 4320 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009138486849418618, + "loss": 0.2036, + "step": 4330 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009136263589675182, + "loss": 0.2093, + "step": 4340 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009134040329931746, + "loss": 0.2076, + "step": 4350 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009131817070188311, + "loss": 0.207, + "step": 4360 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009129593810444874, + "loss": 0.2013, + "step": 4370 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009127370550701438, + "loss": 0.2053, + "step": 4380 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009125147290958002, + "loss": 0.2009, + "step": 4390 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009122924031214567, + "loss": 0.2029, + "step": 4400 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009120700771471131, + "loss": 0.1995, + "step": 4410 + }, + { + "epoch": 0.19, + "learning_rate": 0.0009118477511727695, + "loss": 0.2044, + "step": 4420 + }, + { + "epoch": 0.2, + "learning_rate": 0.000911625425198426, + "loss": 0.2046, + "step": 4430 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009114030992240824, + "loss": 0.2092, + "step": 4440 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009111807732497388, + "loss": 0.2065, + "step": 4450 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009109584472753952, + "loss": 0.2036, + "step": 4460 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009107361213010517, + "loss": 0.1995, + "step": 4470 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009105137953267081, + "loss": 0.2007, + "step": 4480 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009102914693523645, + "loss": 0.1991, + "step": 4490 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009100691433780208, + "loss": 0.1977, + "step": 4500 + }, + { + "epoch": 0.2, + "eval_cer": 0.5025755557991682, + "eval_loss": 0.1470678597688675, + "eval_runtime": 84.0526, + "eval_samples_per_second": 118.902, + "eval_steps_per_second": 7.436, + "step": 4500 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009098468174036773, + "loss": 0.2026, + "step": 4510 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009096244914293337, + "loss": 0.2052, + "step": 4520 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009094021654549901, + "loss": 0.1984, + "step": 4530 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009091798394806465, + "loss": 0.208, + "step": 4540 + }, + { + "epoch": 0.2, + "learning_rate": 0.000908957513506303, + "loss": 0.2079, + "step": 4550 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009087351875319594, + "loss": 0.2039, + "step": 4560 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009085128615576158, + "loss": 0.204, + "step": 4570 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009082905355832723, + "loss": 0.1962, + "step": 4580 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009080682096089287, + "loss": 0.2059, + "step": 4590 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009078458836345851, + "loss": 0.1967, + "step": 4600 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009076235576602415, + "loss": 0.2056, + "step": 4610 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009074012316858979, + "loss": 0.2011, + "step": 4620 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009071789057115543, + "loss": 0.2022, + "step": 4630 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009069565797372107, + "loss": 0.2017, + "step": 4640 + }, + { + "epoch": 0.2, + "learning_rate": 0.0009067342537628671, + "loss": 0.1918, + "step": 4650 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009065119277885236, + "loss": 0.1958, + "step": 4660 + }, + { + "epoch": 0.21, + "learning_rate": 0.00090628960181418, + "loss": 0.1968, + "step": 4670 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009060672758398364, + "loss": 0.2035, + "step": 4680 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009058449498654928, + "loss": 0.2091, + "step": 4690 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009056226238911493, + "loss": 0.2004, + "step": 4700 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009054002979168057, + "loss": 0.196, + "step": 4710 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009051779719424621, + "loss": 0.2014, + "step": 4720 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009049556459681186, + "loss": 0.1993, + "step": 4730 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009047333199937749, + "loss": 0.197, + "step": 4740 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009045109940194313, + "loss": 0.2068, + "step": 4750 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009042886680450877, + "loss": 0.1959, + "step": 4760 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009040663420707442, + "loss": 0.2008, + "step": 4770 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009038440160964006, + "loss": 0.1965, + "step": 4780 + }, + { + "epoch": 0.21, + "learning_rate": 0.000903621690122057, + "loss": 0.1947, + "step": 4790 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009033993641477134, + "loss": 0.2076, + "step": 4800 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009031770381733699, + "loss": 0.1915, + "step": 4810 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009029547121990263, + "loss": 0.1932, + "step": 4820 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009027323862246827, + "loss": 0.1924, + "step": 4830 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009025100602503391, + "loss": 0.1936, + "step": 4840 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009022877342759956, + "loss": 0.2015, + "step": 4850 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009020654083016519, + "loss": 0.1998, + "step": 4860 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009018430823273083, + "loss": 0.196, + "step": 4870 + }, + { + "epoch": 0.21, + "learning_rate": 0.0009016207563529647, + "loss": 0.1955, + "step": 4880 + }, + { + "epoch": 0.22, + "learning_rate": 0.0009013984303786212, + "loss": 0.1997, + "step": 4890 + }, + { + "epoch": 0.22, + "learning_rate": 0.0009011761044042776, + "loss": 0.2027, + "step": 4900 + }, + { + "epoch": 0.22, + "learning_rate": 0.000900953778429934, + "loss": 0.1994, + "step": 4910 + }, + { + "epoch": 0.22, + "learning_rate": 0.0009007314524555905, + "loss": 0.1956, + "step": 4920 + }, + { + "epoch": 0.22, + "learning_rate": 0.0009005091264812469, + "loss": 0.206, + "step": 4930 + }, + { + "epoch": 0.22, + "learning_rate": 0.0009002868005069033, + "loss": 0.1996, + "step": 4940 + }, + { + "epoch": 0.22, + "learning_rate": 0.0009000644745325597, + "loss": 0.1999, + "step": 4950 + }, + { + "epoch": 0.22, + "learning_rate": 0.0008998421485582162, + "loss": 0.1921, + "step": 4960 + }, + { + "epoch": 0.22, + "learning_rate": 0.0008996198225838725, + "loss": 0.2006, + "step": 4970 + }, + { + "epoch": 0.22, + "learning_rate": 0.0008993974966095289, + "loss": 0.196, + "step": 4980 + }, + { + "epoch": 0.22, + "learning_rate": 0.0008991751706351852, + "loss": 0.1939, + "step": 4990 + }, + { + "epoch": 0.22, + "learning_rate": 0.0008989528446608417, + "loss": 0.1907, + "step": 5000 + }, + { + "epoch": 0.22, + "eval_cer": 0.5015973570470852, + "eval_loss": 0.14370913803577423, + "eval_runtime": 84.1359, + "eval_samples_per_second": 118.784, + "eval_steps_per_second": 7.428, + "step": 5000 + }, + { + "epoch": 0.22, + "learning_rate": 0.0008987305186864981, + "loss": 0.1987, + "step": 5010 + }, + { + "epoch": 0.22, + "learning_rate": 0.0008985081927121545, + "loss": 0.1932, + "step": 5020 + }, + { + "epoch": 0.22, + "learning_rate": 0.0008982858667378109, + "loss": 0.1977, + "step": 5030 + }, + { + "epoch": 0.22, + "learning_rate": 0.0008980635407634674, + "loss": 0.1895, + "step": 5040 + }, + { + "epoch": 0.22, + "learning_rate": 0.0008978412147891238, + "loss": 0.1931, + "step": 5050 + }, + { + "epoch": 0.22, + "learning_rate": 0.0008976188888147802, + "loss": 0.1958, + "step": 5060 + }, + { + "epoch": 0.22, + "learning_rate": 0.0008973965628404367, + "loss": 0.1939, + "step": 5070 + }, + { + "epoch": 0.22, + "learning_rate": 0.0008971742368660931, + "loss": 0.1962, + "step": 5080 + }, + { + "epoch": 0.22, + "learning_rate": 0.0008969519108917495, + "loss": 0.1899, + "step": 5090 + }, + { + "epoch": 0.22, + "learning_rate": 0.0008967295849174059, + "loss": 0.2, + "step": 5100 + }, + { + "epoch": 0.22, + "learning_rate": 0.0008965072589430623, + "loss": 0.1944, + "step": 5110 + }, + { + "epoch": 0.23, + "learning_rate": 0.0008962849329687187, + "loss": 0.1982, + "step": 5120 + }, + { + "epoch": 0.23, + "learning_rate": 0.0008960626069943751, + "loss": 0.2024, + "step": 5130 + }, + { + "epoch": 0.23, + "learning_rate": 0.0008958402810200315, + "loss": 0.1927, + "step": 5140 + }, + { + "epoch": 0.23, + "learning_rate": 0.000895617955045688, + "loss": 0.2034, + "step": 5150 + }, + { + "epoch": 0.23, + "learning_rate": 0.0008953956290713444, + "loss": 0.1964, + "step": 5160 + }, + { + "epoch": 0.23, + "learning_rate": 0.0008951733030970008, + "loss": 0.1947, + "step": 5170 + }, + { + "epoch": 0.23, + "learning_rate": 0.0008949509771226572, + "loss": 0.1963, + "step": 5180 + }, + { + "epoch": 0.23, + "learning_rate": 0.0008947286511483137, + "loss": 0.1885, + "step": 5190 + }, + { + "epoch": 0.23, + "learning_rate": 0.0008945063251739701, + "loss": 0.1947, + "step": 5200 + }, + { + "epoch": 0.23, + "learning_rate": 0.0008942839991996265, + "loss": 0.1931, + "step": 5210 + }, + { + "epoch": 0.23, + "learning_rate": 0.000894061673225283, + "loss": 0.1978, + "step": 5220 + }, + { + "epoch": 0.23, + "learning_rate": 0.0008938393472509393, + "loss": 0.1938, + "step": 5230 + }, + { + "epoch": 0.23, + "learning_rate": 0.0008936170212765957, + "loss": 0.1916, + "step": 5240 + }, + { + "epoch": 0.23, + "learning_rate": 0.0008933946953022521, + "loss": 0.19, + "step": 5250 + }, + { + "epoch": 0.23, + "learning_rate": 0.0008931723693279086, + "loss": 0.1939, + "step": 5260 + }, + { + "epoch": 0.23, + "learning_rate": 0.000892950043353565, + "loss": 0.2006, + "step": 5270 + }, + { + "epoch": 0.23, + "learning_rate": 0.0008927277173792214, + "loss": 0.1928, + "step": 5280 + }, + { + "epoch": 0.23, + "learning_rate": 0.0008925053914048778, + "loss": 0.19, + "step": 5290 + }, + { + "epoch": 0.23, + "learning_rate": 0.0008922830654305343, + "loss": 0.1884, + "step": 5300 + }, + { + "epoch": 0.23, + "learning_rate": 0.0008920607394561907, + "loss": 0.1959, + "step": 5310 + }, + { + "epoch": 0.23, + "learning_rate": 0.0008918384134818471, + "loss": 0.1928, + "step": 5320 + }, + { + "epoch": 0.23, + "learning_rate": 0.0008916160875075035, + "loss": 0.1916, + "step": 5330 + }, + { + "epoch": 0.24, + "learning_rate": 0.00089139376153316, + "loss": 0.1923, + "step": 5340 + }, + { + "epoch": 0.24, + "learning_rate": 0.0008911714355588163, + "loss": 0.1941, + "step": 5350 + }, + { + "epoch": 0.24, + "learning_rate": 0.0008909491095844727, + "loss": 0.1845, + "step": 5360 + }, + { + "epoch": 0.24, + "learning_rate": 0.0008907267836101291, + "loss": 0.1894, + "step": 5370 + }, + { + "epoch": 0.24, + "learning_rate": 0.0008905044576357856, + "loss": 0.194, + "step": 5380 + }, + { + "epoch": 0.24, + "learning_rate": 0.000890282131661442, + "loss": 0.1895, + "step": 5390 + }, + { + "epoch": 0.24, + "learning_rate": 0.0008900598056870984, + "loss": 0.1897, + "step": 5400 + }, + { + "epoch": 0.24, + "learning_rate": 0.0008898374797127549, + "loss": 0.192, + "step": 5410 + }, + { + "epoch": 0.24, + "learning_rate": 0.0008896151537384113, + "loss": 0.1918, + "step": 5420 + }, + { + "epoch": 0.24, + "learning_rate": 0.0008893928277640677, + "loss": 0.1961, + "step": 5430 + }, + { + "epoch": 0.24, + "learning_rate": 0.0008891705017897241, + "loss": 0.1928, + "step": 5440 + }, + { + "epoch": 0.24, + "learning_rate": 0.0008889481758153806, + "loss": 0.189, + "step": 5450 + }, + { + "epoch": 0.24, + "learning_rate": 0.000888725849841037, + "loss": 0.1896, + "step": 5460 + }, + { + "epoch": 0.24, + "learning_rate": 0.0008885035238666934, + "loss": 0.1897, + "step": 5470 + }, + { + "epoch": 0.24, + "learning_rate": 0.0008882811978923497, + "loss": 0.1963, + "step": 5480 + }, + { + "epoch": 0.24, + "learning_rate": 0.0008880588719180062, + "loss": 0.1904, + "step": 5490 + }, + { + "epoch": 0.24, + "learning_rate": 0.0008878365459436626, + "loss": 0.1923, + "step": 5500 + }, + { + "epoch": 0.24, + "eval_cer": 0.5013469693028517, + "eval_loss": 0.13830307126045227, + "eval_runtime": 83.9757, + "eval_samples_per_second": 119.011, + "eval_steps_per_second": 7.443, + "step": 5500 + }, + { + "epoch": 0.24, + "learning_rate": 0.000887614219969319, + "loss": 0.1884, + "step": 5510 + }, + { + "epoch": 0.24, + "learning_rate": 0.0008873918939949754, + "loss": 0.1878, + "step": 5520 + }, + { + "epoch": 0.24, + "learning_rate": 0.0008871695680206319, + "loss": 0.1988, + "step": 5530 + }, + { + "epoch": 0.24, + "learning_rate": 0.0008869472420462883, + "loss": 0.1942, + "step": 5540 + }, + { + "epoch": 0.24, + "learning_rate": 0.0008867249160719447, + "loss": 0.1905, + "step": 5550 + }, + { + "epoch": 0.24, + "learning_rate": 0.0008865025900976012, + "loss": 0.1846, + "step": 5560 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008862802641232576, + "loss": 0.1853, + "step": 5570 + }, + { + "epoch": 0.25, + "learning_rate": 0.000886057938148914, + "loss": 0.1821, + "step": 5580 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008858356121745704, + "loss": 0.1926, + "step": 5590 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008856132862002268, + "loss": 0.1839, + "step": 5600 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008853909602258832, + "loss": 0.1886, + "step": 5610 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008851686342515396, + "loss": 0.1892, + "step": 5620 + }, + { + "epoch": 0.25, + "learning_rate": 0.000884946308277196, + "loss": 0.19, + "step": 5630 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008847239823028525, + "loss": 0.1868, + "step": 5640 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008845016563285089, + "loss": 0.1867, + "step": 5650 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008842793303541653, + "loss": 0.1899, + "step": 5660 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008840570043798218, + "loss": 0.1928, + "step": 5670 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008838346784054782, + "loss": 0.187, + "step": 5680 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008836123524311346, + "loss": 0.1822, + "step": 5690 + }, + { + "epoch": 0.25, + "learning_rate": 0.000883390026456791, + "loss": 0.1883, + "step": 5700 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008831677004824475, + "loss": 0.1892, + "step": 5710 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008829453745081038, + "loss": 0.1825, + "step": 5720 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008827230485337602, + "loss": 0.1873, + "step": 5730 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008825007225594166, + "loss": 0.1756, + "step": 5740 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008822783965850731, + "loss": 0.1854, + "step": 5750 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008820560706107295, + "loss": 0.1802, + "step": 5760 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008818337446363859, + "loss": 0.1847, + "step": 5770 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008816114186620423, + "loss": 0.1811, + "step": 5780 + }, + { + "epoch": 0.25, + "learning_rate": 0.0008813890926876988, + "loss": 0.1878, + "step": 5790 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008811667667133552, + "loss": 0.1854, + "step": 5800 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008809444407390116, + "loss": 0.1858, + "step": 5810 + }, + { + "epoch": 0.26, + "learning_rate": 0.000880722114764668, + "loss": 0.1835, + "step": 5820 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008804997887903245, + "loss": 0.1858, + "step": 5830 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008802774628159808, + "loss": 0.1856, + "step": 5840 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008800551368416372, + "loss": 0.1894, + "step": 5850 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008798328108672937, + "loss": 0.1852, + "step": 5860 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008796104848929501, + "loss": 0.1897, + "step": 5870 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008793881589186065, + "loss": 0.1906, + "step": 5880 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008791658329442629, + "loss": 0.1835, + "step": 5890 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008789435069699194, + "loss": 0.1836, + "step": 5900 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008787211809955758, + "loss": 0.1715, + "step": 5910 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008784988550212322, + "loss": 0.1877, + "step": 5920 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008782765290468886, + "loss": 0.1828, + "step": 5930 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008780542030725451, + "loss": 0.1847, + "step": 5940 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008778318770982015, + "loss": 0.1848, + "step": 5950 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008776095511238577, + "loss": 0.1818, + "step": 5960 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008773872251495141, + "loss": 0.1809, + "step": 5970 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008771648991751706, + "loss": 0.1839, + "step": 5980 + }, + { + "epoch": 0.26, + "learning_rate": 0.000876942573200827, + "loss": 0.187, + "step": 5990 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008767202472264834, + "loss": 0.1805, + "step": 6000 + }, + { + "epoch": 0.26, + "eval_cer": 0.5021279714895281, + "eval_loss": 0.13357099890708923, + "eval_runtime": 83.859, + "eval_samples_per_second": 119.176, + "eval_steps_per_second": 7.453, + "step": 6000 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008764979212521398, + "loss": 0.1915, + "step": 6010 + }, + { + "epoch": 0.26, + "learning_rate": 0.0008762755952777963, + "loss": 0.1855, + "step": 6020 + }, + { + "epoch": 0.27, + "learning_rate": 0.0008760532693034527, + "loss": 0.1883, + "step": 6030 + }, + { + "epoch": 0.27, + "learning_rate": 0.0008758309433291091, + "loss": 0.1829, + "step": 6040 + }, + { + "epoch": 0.27, + "learning_rate": 0.0008756086173547656, + "loss": 0.1815, + "step": 6050 + }, + { + "epoch": 0.27, + "learning_rate": 0.000875386291380422, + "loss": 0.1811, + "step": 6060 + }, + { + "epoch": 0.27, + "learning_rate": 0.0008751639654060784, + "loss": 0.1865, + "step": 6070 + }, + { + "epoch": 0.27, + "learning_rate": 0.0008749416394317348, + "loss": 0.1845, + "step": 6080 + }, + { + "epoch": 0.27, + "learning_rate": 0.0008747193134573912, + "loss": 0.1811, + "step": 6090 + }, + { + "epoch": 0.27, + "learning_rate": 0.0008744969874830476, + "loss": 0.1754, + "step": 6100 + }, + { + "epoch": 0.27, + "learning_rate": 0.000874274661508704, + "loss": 0.1832, + "step": 6110 + }, + { + "epoch": 0.27, + "learning_rate": 0.0008740523355343604, + "loss": 0.1869, + "step": 6120 + }, + { + "epoch": 0.27, + "learning_rate": 0.0008738300095600169, + "loss": 0.1869, + "step": 6130 + }, + { + "epoch": 0.27, + "learning_rate": 0.0008736076835856733, + "loss": 0.1818, + "step": 6140 + }, + { + "epoch": 0.27, + "learning_rate": 0.0008733853576113297, + "loss": 0.1761, + "step": 6150 + }, + { + "epoch": 0.27, + "learning_rate": 0.0008731630316369862, + "loss": 0.1878, + "step": 6160 + }, + { + "epoch": 0.27, + "learning_rate": 0.0008729407056626426, + "loss": 0.1777, + "step": 6170 + }, + { + "epoch": 0.27, + "learning_rate": 0.000872718379688299, + "loss": 0.1776, + "step": 6180 + }, + { + "epoch": 0.27, + "learning_rate": 0.0008724960537139554, + "loss": 0.1807, + "step": 6190 + }, + { + "epoch": 0.27, + "learning_rate": 0.0008722737277396119, + "loss": 0.1776, + "step": 6200 + }, + { + "epoch": 0.27, + "learning_rate": 0.0008720514017652682, + "loss": 0.1805, + "step": 6210 + }, + { + "epoch": 0.27, + "learning_rate": 0.0008718290757909246, + "loss": 0.182, + "step": 6220 + }, + { + "epoch": 0.27, + "learning_rate": 0.000871606749816581, + "loss": 0.1817, + "step": 6230 + }, + { + "epoch": 0.27, + "learning_rate": 0.0008713844238422375, + "loss": 0.1798, + "step": 6240 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008711620978678939, + "loss": 0.1818, + "step": 6250 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008709397718935503, + "loss": 0.183, + "step": 6260 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008707174459192067, + "loss": 0.1842, + "step": 6270 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008704951199448632, + "loss": 0.1788, + "step": 6280 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008702727939705196, + "loss": 0.1849, + "step": 6290 + }, + { + "epoch": 0.28, + "learning_rate": 0.000870050467996176, + "loss": 0.1809, + "step": 6300 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008698281420218325, + "loss": 0.1834, + "step": 6310 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008696058160474889, + "loss": 0.1824, + "step": 6320 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008693834900731452, + "loss": 0.1871, + "step": 6330 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008691611640988016, + "loss": 0.1782, + "step": 6340 + }, + { + "epoch": 0.28, + "learning_rate": 0.000868938838124458, + "loss": 0.1778, + "step": 6350 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008687165121501145, + "loss": 0.1766, + "step": 6360 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008684941861757709, + "loss": 0.1784, + "step": 6370 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008682718602014273, + "loss": 0.1683, + "step": 6380 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008680495342270838, + "loss": 0.1846, + "step": 6390 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008678272082527402, + "loss": 0.1841, + "step": 6400 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008676048822783966, + "loss": 0.1799, + "step": 6410 + }, + { + "epoch": 0.28, + "learning_rate": 0.000867382556304053, + "loss": 0.1847, + "step": 6420 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008671602303297095, + "loss": 0.182, + "step": 6430 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008669379043553659, + "loss": 0.1813, + "step": 6440 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008667155783810222, + "loss": 0.1812, + "step": 6450 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008664932524066786, + "loss": 0.179, + "step": 6460 + }, + { + "epoch": 0.28, + "learning_rate": 0.0008662709264323351, + "loss": 0.1841, + "step": 6470 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008660486004579915, + "loss": 0.171, + "step": 6480 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008658262744836479, + "loss": 0.1754, + "step": 6490 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008656039485093044, + "loss": 0.1719, + "step": 6500 + }, + { + "epoch": 0.29, + "eval_cer": 0.5016615756654249, + "eval_loss": 0.1321278065443039, + "eval_runtime": 83.8834, + "eval_samples_per_second": 119.142, + "eval_steps_per_second": 7.451, + "step": 6500 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008653816225349608, + "loss": 0.1792, + "step": 6510 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008651592965606172, + "loss": 0.1786, + "step": 6520 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008649369705862736, + "loss": 0.1762, + "step": 6530 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008647146446119301, + "loss": 0.1718, + "step": 6540 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008644923186375865, + "loss": 0.1716, + "step": 6550 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008642699926632429, + "loss": 0.1822, + "step": 6560 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008640476666888993, + "loss": 0.1751, + "step": 6570 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008638253407145557, + "loss": 0.1772, + "step": 6580 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008636030147402121, + "loss": 0.1735, + "step": 6590 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008633806887658685, + "loss": 0.1864, + "step": 6600 + }, + { + "epoch": 0.29, + "learning_rate": 0.000863158362791525, + "loss": 0.1793, + "step": 6610 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008629360368171814, + "loss": 0.1737, + "step": 6620 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008627137108428378, + "loss": 0.1747, + "step": 6630 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008624913848684942, + "loss": 0.1734, + "step": 6640 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008622690588941507, + "loss": 0.1731, + "step": 6650 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008620467329198071, + "loss": 0.1725, + "step": 6660 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008618244069454635, + "loss": 0.1758, + "step": 6670 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008616020809711199, + "loss": 0.1745, + "step": 6680 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008613797549967764, + "loss": 0.1708, + "step": 6690 + }, + { + "epoch": 0.29, + "learning_rate": 0.0008611574290224327, + "loss": 0.1809, + "step": 6700 + }, + { + "epoch": 0.3, + "learning_rate": 0.0008609351030480891, + "loss": 0.1738, + "step": 6710 + }, + { + "epoch": 0.3, + "learning_rate": 0.0008607127770737455, + "loss": 0.179, + "step": 6720 + }, + { + "epoch": 0.3, + "learning_rate": 0.000860490451099402, + "loss": 0.185, + "step": 6730 + }, + { + "epoch": 0.3, + "learning_rate": 0.0008602681251250584, + "loss": 0.1873, + "step": 6740 + }, + { + "epoch": 0.3, + "learning_rate": 0.0008600457991507148, + "loss": 0.185, + "step": 6750 + }, + { + "epoch": 0.3, + "learning_rate": 0.0008598234731763713, + "loss": 0.1754, + "step": 6760 + }, + { + "epoch": 0.3, + "learning_rate": 0.0008596011472020277, + "loss": 0.1751, + "step": 6770 + }, + { + "epoch": 0.3, + "learning_rate": 0.0008593788212276841, + "loss": 0.1758, + "step": 6780 + }, + { + "epoch": 0.3, + "learning_rate": 0.0008591564952533405, + "loss": 0.1829, + "step": 6790 + }, + { + "epoch": 0.3, + "learning_rate": 0.000858934169278997, + "loss": 0.1749, + "step": 6800 + }, + { + "epoch": 0.3, + "learning_rate": 0.0008587118433046534, + "loss": 0.1742, + "step": 6810 + }, + { + "epoch": 0.3, + "learning_rate": 0.0008584895173303097, + "loss": 0.1785, + "step": 6820 + }, + { + "epoch": 0.3, + "learning_rate": 0.0008582671913559661, + "loss": 0.1805, + "step": 6830 + }, + { + "epoch": 0.3, + "learning_rate": 0.0008580448653816226, + "loss": 0.1788, + "step": 6840 + }, + { + "epoch": 0.3, + "learning_rate": 0.000857822539407279, + "loss": 0.1719, + "step": 6850 + }, + { + "epoch": 0.3, + "learning_rate": 0.0008576002134329354, + "loss": 0.1745, + "step": 6860 + }, + { + "epoch": 0.3, + "learning_rate": 0.0008573778874585918, + "loss": 0.17, + "step": 6870 + }, + { + "epoch": 0.3, + "learning_rate": 0.0008571555614842483, + "loss": 0.176, + "step": 6880 + }, + { + "epoch": 0.3, + "learning_rate": 0.0008569332355099047, + "loss": 0.1734, + "step": 6890 + }, + { + "epoch": 0.3, + "learning_rate": 0.0008567109095355611, + "loss": 0.1692, + "step": 6900 + }, + { + "epoch": 0.3, + "learning_rate": 0.0008564885835612176, + "loss": 0.1717, + "step": 6910 + }, + { + "epoch": 0.3, + "learning_rate": 0.000856266257586874, + "loss": 0.1843, + "step": 6920 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008560439316125304, + "loss": 0.1652, + "step": 6930 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008558216056381867, + "loss": 0.18, + "step": 6940 + }, + { + "epoch": 0.31, + "learning_rate": 0.000855599279663843, + "loss": 0.1732, + "step": 6950 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008553769536894995, + "loss": 0.1747, + "step": 6960 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008551546277151559, + "loss": 0.1701, + "step": 6970 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008549323017408123, + "loss": 0.1793, + "step": 6980 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008547099757664688, + "loss": 0.1751, + "step": 6990 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008544876497921252, + "loss": 0.1782, + "step": 7000 + }, + { + "epoch": 0.31, + "eval_cer": 0.502301815830084, + "eval_loss": 0.12710104882717133, + "eval_runtime": 84.1137, + "eval_samples_per_second": 118.815, + "eval_steps_per_second": 7.43, + "step": 7000 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008542653238177816, + "loss": 0.1728, + "step": 7010 + }, + { + "epoch": 0.31, + "learning_rate": 0.000854042997843438, + "loss": 0.1687, + "step": 7020 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008538206718690945, + "loss": 0.1741, + "step": 7030 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008535983458947509, + "loss": 0.1732, + "step": 7040 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008533760199204073, + "loss": 0.1771, + "step": 7050 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008531536939460637, + "loss": 0.1725, + "step": 7060 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008529313679717201, + "loss": 0.168, + "step": 7070 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008527090419973765, + "loss": 0.1705, + "step": 7080 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008524867160230329, + "loss": 0.1675, + "step": 7090 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008522643900486893, + "loss": 0.1692, + "step": 7100 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008520420640743458, + "loss": 0.173, + "step": 7110 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008518197381000022, + "loss": 0.1745, + "step": 7120 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008515974121256586, + "loss": 0.169, + "step": 7130 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008513750861513151, + "loss": 0.1742, + "step": 7140 + }, + { + "epoch": 0.31, + "learning_rate": 0.0008511527601769715, + "loss": 0.1749, + "step": 7150 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008509304342026279, + "loss": 0.1719, + "step": 7160 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008507081082282843, + "loss": 0.1757, + "step": 7170 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008504857822539408, + "loss": 0.1784, + "step": 7180 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008502634562795971, + "loss": 0.1729, + "step": 7190 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008500411303052535, + "loss": 0.1699, + "step": 7200 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008498188043309099, + "loss": 0.169, + "step": 7210 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008495964783565664, + "loss": 0.1685, + "step": 7220 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008493741523822228, + "loss": 0.1754, + "step": 7230 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008491518264078792, + "loss": 0.1718, + "step": 7240 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008489295004335357, + "loss": 0.1721, + "step": 7250 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008487071744591921, + "loss": 0.1771, + "step": 7260 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008484848484848485, + "loss": 0.1635, + "step": 7270 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008482625225105049, + "loss": 0.1747, + "step": 7280 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008480401965361614, + "loss": 0.163, + "step": 7290 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008478178705618178, + "loss": 0.1729, + "step": 7300 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008475955445874741, + "loss": 0.1687, + "step": 7310 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008473732186131305, + "loss": 0.1747, + "step": 7320 + }, + { + "epoch": 0.32, + "learning_rate": 0.000847150892638787, + "loss": 0.1733, + "step": 7330 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008469285666644434, + "loss": 0.1676, + "step": 7340 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008467062406900998, + "loss": 0.1661, + "step": 7350 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008464839147157562, + "loss": 0.1713, + "step": 7360 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008462615887414127, + "loss": 0.1772, + "step": 7370 + }, + { + "epoch": 0.32, + "learning_rate": 0.0008460392627670691, + "loss": 0.1733, + "step": 7380 + }, + { + "epoch": 0.33, + "learning_rate": 0.0008458169367927255, + "loss": 0.1694, + "step": 7390 + }, + { + "epoch": 0.33, + "learning_rate": 0.000845594610818382, + "loss": 0.1728, + "step": 7400 + }, + { + "epoch": 0.33, + "learning_rate": 0.0008453722848440384, + "loss": 0.1766, + "step": 7410 + }, + { + "epoch": 0.33, + "learning_rate": 0.0008451499588696948, + "loss": 0.1719, + "step": 7420 + }, + { + "epoch": 0.33, + "learning_rate": 0.0008449276328953511, + "loss": 0.1683, + "step": 7430 + }, + { + "epoch": 0.33, + "learning_rate": 0.0008447053069210076, + "loss": 0.1683, + "step": 7440 + }, + { + "epoch": 0.33, + "learning_rate": 0.000844482980946664, + "loss": 0.1684, + "step": 7450 + }, + { + "epoch": 0.33, + "learning_rate": 0.0008442606549723204, + "loss": 0.1716, + "step": 7460 + }, + { + "epoch": 0.33, + "learning_rate": 0.0008440383289979768, + "loss": 0.1674, + "step": 7470 + }, + { + "epoch": 0.33, + "learning_rate": 0.0008438160030236333, + "loss": 0.1732, + "step": 7480 + }, + { + "epoch": 0.33, + "learning_rate": 0.0008435936770492897, + "loss": 0.1738, + "step": 7490 + }, + { + "epoch": 0.33, + "learning_rate": 0.0008433713510749461, + "loss": 0.1762, + "step": 7500 + }, + { + "epoch": 0.33, + "eval_cer": 0.5015791942055347, + "eval_loss": 0.12540031969547272, + "eval_runtime": 83.978, + "eval_samples_per_second": 119.007, + "eval_steps_per_second": 7.442, + "step": 7500 + }, + { + "epoch": 0.33, + "learning_rate": 0.0008431490251006025, + "loss": 0.1652, + "step": 7510 + }, + { + "epoch": 0.33, + "learning_rate": 0.000842926699126259, + "loss": 0.1635, + "step": 7520 + }, + { + "epoch": 0.33, + "learning_rate": 0.0008427043731519154, + "loss": 0.1705, + "step": 7530 + }, + { + "epoch": 0.33, + "learning_rate": 0.0008424820471775718, + "loss": 0.1682, + "step": 7540 + }, + { + "epoch": 0.33, + "learning_rate": 0.0008422597212032281, + "loss": 0.1659, + "step": 7550 + }, + { + "epoch": 0.33, + "learning_rate": 0.0008420373952288846, + "loss": 0.1716, + "step": 7560 + }, + { + "epoch": 0.33, + "learning_rate": 0.000841815069254541, + "loss": 0.1707, + "step": 7570 + }, + { + "epoch": 0.33, + "learning_rate": 0.0008415927432801974, + "loss": 0.1644, + "step": 7580 + }, + { + "epoch": 0.33, + "learning_rate": 0.0008413704173058539, + "loss": 0.1687, + "step": 7590 + }, + { + "epoch": 0.33, + "learning_rate": 0.0008411480913315103, + "loss": 0.1649, + "step": 7600 + }, + { + "epoch": 0.33, + "learning_rate": 0.0008409257653571667, + "loss": 0.1652, + "step": 7610 + }, + { + "epoch": 0.34, + "learning_rate": 0.0008407034393828231, + "loss": 0.1647, + "step": 7620 + }, + { + "epoch": 0.34, + "learning_rate": 0.0008404811134084796, + "loss": 0.1679, + "step": 7630 + }, + { + "epoch": 0.34, + "learning_rate": 0.000840258787434136, + "loss": 0.1745, + "step": 7640 + }, + { + "epoch": 0.34, + "learning_rate": 0.0008400364614597924, + "loss": 0.1632, + "step": 7650 + }, + { + "epoch": 0.34, + "learning_rate": 0.0008398141354854488, + "loss": 0.1743, + "step": 7660 + }, + { + "epoch": 0.34, + "learning_rate": 0.0008395918095111053, + "loss": 0.1694, + "step": 7670 + }, + { + "epoch": 0.34, + "learning_rate": 0.0008393694835367616, + "loss": 0.1683, + "step": 7680 + }, + { + "epoch": 0.34, + "learning_rate": 0.000839147157562418, + "loss": 0.1686, + "step": 7690 + }, + { + "epoch": 0.34, + "learning_rate": 0.0008389248315880744, + "loss": 0.1677, + "step": 7700 + }, + { + "epoch": 0.34, + "learning_rate": 0.0008387025056137309, + "loss": 0.1658, + "step": 7710 + }, + { + "epoch": 0.34, + "learning_rate": 0.0008384801796393873, + "loss": 0.1638, + "step": 7720 + }, + { + "epoch": 0.34, + "learning_rate": 0.0008382578536650437, + "loss": 0.1659, + "step": 7730 + }, + { + "epoch": 0.34, + "learning_rate": 0.0008380355276907002, + "loss": 0.1668, + "step": 7740 + }, + { + "epoch": 0.34, + "learning_rate": 0.0008378132017163566, + "loss": 0.1681, + "step": 7750 + }, + { + "epoch": 0.34, + "learning_rate": 0.000837590875742013, + "loss": 0.1715, + "step": 7760 + }, + { + "epoch": 0.34, + "learning_rate": 0.0008373685497676694, + "loss": 0.1721, + "step": 7770 + }, + { + "epoch": 0.34, + "learning_rate": 0.0008371462237933259, + "loss": 0.1676, + "step": 7780 + }, + { + "epoch": 0.34, + "learning_rate": 0.0008369238978189823, + "loss": 0.1658, + "step": 7790 + }, + { + "epoch": 0.34, + "learning_rate": 0.0008367015718446386, + "loss": 0.1677, + "step": 7800 + }, + { + "epoch": 0.34, + "learning_rate": 0.000836479245870295, + "loss": 0.167, + "step": 7810 + }, + { + "epoch": 0.34, + "learning_rate": 0.0008362569198959515, + "loss": 0.1672, + "step": 7820 + }, + { + "epoch": 0.34, + "learning_rate": 0.0008360345939216079, + "loss": 0.1663, + "step": 7830 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008358122679472643, + "loss": 0.1645, + "step": 7840 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008355899419729208, + "loss": 0.1693, + "step": 7850 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008353676159985772, + "loss": 0.1693, + "step": 7860 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008351452900242336, + "loss": 0.1617, + "step": 7870 + }, + { + "epoch": 0.35, + "learning_rate": 0.00083492296404989, + "loss": 0.1602, + "step": 7880 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008347006380755465, + "loss": 0.1709, + "step": 7890 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008344783121012029, + "loss": 0.1633, + "step": 7900 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008342559861268593, + "loss": 0.1681, + "step": 7910 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008340336601525156, + "loss": 0.1688, + "step": 7920 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008338113341781721, + "loss": 0.1697, + "step": 7930 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008335890082038285, + "loss": 0.1662, + "step": 7940 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008333666822294848, + "loss": 0.1711, + "step": 7950 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008331443562551412, + "loss": 0.1611, + "step": 7960 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008329220302807977, + "loss": 0.1648, + "step": 7970 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008326997043064541, + "loss": 0.171, + "step": 7980 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008324773783321105, + "loss": 0.1638, + "step": 7990 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008322550523577669, + "loss": 0.1621, + "step": 8000 + }, + { + "epoch": 0.35, + "eval_cer": 0.5012444789826733, + "eval_loss": 0.12311427295207977, + "eval_runtime": 83.9199, + "eval_samples_per_second": 119.09, + "eval_steps_per_second": 7.448, + "step": 8000 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008320327263834234, + "loss": 0.1662, + "step": 8010 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008318104004090798, + "loss": 0.1672, + "step": 8020 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008315880744347362, + "loss": 0.1755, + "step": 8030 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008313657484603925, + "loss": 0.1663, + "step": 8040 + }, + { + "epoch": 0.35, + "learning_rate": 0.000831143422486049, + "loss": 0.1693, + "step": 8050 + }, + { + "epoch": 0.35, + "learning_rate": 0.0008309210965117054, + "loss": 0.1655, + "step": 8060 + }, + { + "epoch": 0.36, + "learning_rate": 0.0008306987705373618, + "loss": 0.1717, + "step": 8070 + }, + { + "epoch": 0.36, + "learning_rate": 0.0008304764445630183, + "loss": 0.1633, + "step": 8080 + }, + { + "epoch": 0.36, + "learning_rate": 0.0008302541185886747, + "loss": 0.1686, + "step": 8090 + }, + { + "epoch": 0.36, + "learning_rate": 0.0008300317926143311, + "loss": 0.1656, + "step": 8100 + }, + { + "epoch": 0.36, + "learning_rate": 0.0008298094666399875, + "loss": 0.1642, + "step": 8110 + }, + { + "epoch": 0.36, + "learning_rate": 0.000829587140665644, + "loss": 0.1628, + "step": 8120 + }, + { + "epoch": 0.36, + "learning_rate": 0.0008293648146913004, + "loss": 0.1661, + "step": 8130 + }, + { + "epoch": 0.36, + "learning_rate": 0.0008291424887169568, + "loss": 0.1647, + "step": 8140 + }, + { + "epoch": 0.36, + "learning_rate": 0.0008289201627426132, + "loss": 0.1654, + "step": 8150 + }, + { + "epoch": 0.36, + "learning_rate": 0.0008286978367682697, + "loss": 0.1635, + "step": 8160 + }, + { + "epoch": 0.36, + "learning_rate": 0.000828475510793926, + "loss": 0.1665, + "step": 8170 + }, + { + "epoch": 0.36, + "learning_rate": 0.0008282531848195824, + "loss": 0.1635, + "step": 8180 + }, + { + "epoch": 0.36, + "learning_rate": 0.0008280308588452388, + "loss": 0.1603, + "step": 8190 + }, + { + "epoch": 0.36, + "learning_rate": 0.0008278085328708953, + "loss": 0.16, + "step": 8200 + }, + { + "epoch": 0.36, + "learning_rate": 0.0008275862068965517, + "loss": 0.1622, + "step": 8210 + }, + { + "epoch": 0.36, + "learning_rate": 0.0008273638809222081, + "loss": 0.1673, + "step": 8220 + }, + { + "epoch": 0.36, + "learning_rate": 0.0008271415549478646, + "loss": 0.1695, + "step": 8230 + }, + { + "epoch": 0.36, + "learning_rate": 0.000826919228973521, + "loss": 0.1619, + "step": 8240 + }, + { + "epoch": 0.36, + "learning_rate": 0.0008266969029991774, + "loss": 0.1658, + "step": 8250 + }, + { + "epoch": 0.36, + "learning_rate": 0.0008264745770248338, + "loss": 0.1663, + "step": 8260 + }, + { + "epoch": 0.36, + "learning_rate": 0.0008262522510504903, + "loss": 0.1636, + "step": 8270 + }, + { + "epoch": 0.36, + "learning_rate": 0.0008260299250761467, + "loss": 0.1669, + "step": 8280 + }, + { + "epoch": 0.36, + "learning_rate": 0.000825807599101803, + "loss": 0.1654, + "step": 8290 + }, + { + "epoch": 0.37, + "learning_rate": 0.0008255852731274594, + "loss": 0.1698, + "step": 8300 + }, + { + "epoch": 0.37, + "learning_rate": 0.0008253629471531159, + "loss": 0.1624, + "step": 8310 + }, + { + "epoch": 0.37, + "learning_rate": 0.0008251406211787723, + "loss": 0.1644, + "step": 8320 + }, + { + "epoch": 0.37, + "learning_rate": 0.0008249182952044287, + "loss": 0.1595, + "step": 8330 + }, + { + "epoch": 0.37, + "learning_rate": 0.0008246959692300852, + "loss": 0.1656, + "step": 8340 + }, + { + "epoch": 0.37, + "learning_rate": 0.0008244736432557416, + "loss": 0.1579, + "step": 8350 + }, + { + "epoch": 0.37, + "learning_rate": 0.000824251317281398, + "loss": 0.1666, + "step": 8360 + }, + { + "epoch": 0.37, + "learning_rate": 0.0008240289913070544, + "loss": 0.1633, + "step": 8370 + }, + { + "epoch": 0.37, + "learning_rate": 0.0008238066653327109, + "loss": 0.1635, + "step": 8380 + }, + { + "epoch": 0.37, + "learning_rate": 0.0008235843393583673, + "loss": 0.1666, + "step": 8390 + }, + { + "epoch": 0.37, + "learning_rate": 0.0008233620133840237, + "loss": 0.159, + "step": 8400 + }, + { + "epoch": 0.37, + "learning_rate": 0.00082313968740968, + "loss": 0.1658, + "step": 8410 + }, + { + "epoch": 0.37, + "learning_rate": 0.0008229173614353365, + "loss": 0.1603, + "step": 8420 + }, + { + "epoch": 0.37, + "learning_rate": 0.0008226950354609929, + "loss": 0.1683, + "step": 8430 + }, + { + "epoch": 0.37, + "learning_rate": 0.0008224727094866493, + "loss": 0.1652, + "step": 8440 + }, + { + "epoch": 0.37, + "learning_rate": 0.0008222503835123057, + "loss": 0.1626, + "step": 8450 + }, + { + "epoch": 0.37, + "learning_rate": 0.0008220280575379622, + "loss": 0.1628, + "step": 8460 + }, + { + "epoch": 0.37, + "learning_rate": 0.0008218057315636186, + "loss": 0.1633, + "step": 8470 + }, + { + "epoch": 0.37, + "learning_rate": 0.000821583405589275, + "loss": 0.1605, + "step": 8480 + }, + { + "epoch": 0.37, + "learning_rate": 0.0008213610796149315, + "loss": 0.1628, + "step": 8490 + }, + { + "epoch": 0.37, + "learning_rate": 0.0008211387536405879, + "loss": 0.162, + "step": 8500 + }, + { + "epoch": 0.37, + "eval_cer": 0.5010771213712426, + "eval_loss": 0.12126629054546356, + "eval_runtime": 83.851, + "eval_samples_per_second": 119.188, + "eval_steps_per_second": 7.454, + "step": 8500 + }, + { + "epoch": 0.37, + "learning_rate": 0.0008209164276662443, + "loss": 0.163, + "step": 8510 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008206941016919007, + "loss": 0.1609, + "step": 8520 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008204717757175571, + "loss": 0.1622, + "step": 8530 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008202494497432135, + "loss": 0.1659, + "step": 8540 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008200271237688699, + "loss": 0.1651, + "step": 8550 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008198047977945263, + "loss": 0.1628, + "step": 8560 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008195824718201828, + "loss": 0.1558, + "step": 8570 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008193601458458392, + "loss": 0.1613, + "step": 8580 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008191378198714956, + "loss": 0.1626, + "step": 8590 + }, + { + "epoch": 0.38, + "learning_rate": 0.000818915493897152, + "loss": 0.1555, + "step": 8600 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008186931679228085, + "loss": 0.1623, + "step": 8610 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008184708419484649, + "loss": 0.1655, + "step": 8620 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008182485159741213, + "loss": 0.1633, + "step": 8630 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008180261899997778, + "loss": 0.1639, + "step": 8640 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008178038640254342, + "loss": 0.1644, + "step": 8650 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008175815380510905, + "loss": 0.1632, + "step": 8660 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008173592120767469, + "loss": 0.1612, + "step": 8670 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008171368861024034, + "loss": 0.1638, + "step": 8680 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008169145601280598, + "loss": 0.1658, + "step": 8690 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008166922341537162, + "loss": 0.1577, + "step": 8700 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008164699081793726, + "loss": 0.1654, + "step": 8710 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008162475822050291, + "loss": 0.1635, + "step": 8720 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008160252562306855, + "loss": 0.158, + "step": 8730 + }, + { + "epoch": 0.38, + "learning_rate": 0.0008158029302563419, + "loss": 0.1622, + "step": 8740 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008155806042819983, + "loss": 0.1639, + "step": 8750 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008153582783076548, + "loss": 0.1584, + "step": 8760 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008151359523333112, + "loss": 0.1594, + "step": 8770 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008149136263589675, + "loss": 0.1578, + "step": 8780 + }, + { + "epoch": 0.39, + "learning_rate": 0.000814691300384624, + "loss": 0.1616, + "step": 8790 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008144689744102804, + "loss": 0.1577, + "step": 8800 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008142466484359368, + "loss": 0.1537, + "step": 8810 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008140243224615932, + "loss": 0.1593, + "step": 8820 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008138019964872497, + "loss": 0.1573, + "step": 8830 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008135796705129061, + "loss": 0.1615, + "step": 8840 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008133573445385625, + "loss": 0.162, + "step": 8850 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008131350185642189, + "loss": 0.158, + "step": 8860 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008129126925898754, + "loss": 0.1617, + "step": 8870 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008126903666155318, + "loss": 0.1567, + "step": 8880 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008124680406411882, + "loss": 0.1605, + "step": 8890 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008122457146668445, + "loss": 0.1596, + "step": 8900 + }, + { + "epoch": 0.39, + "learning_rate": 0.000812023388692501, + "loss": 0.1629, + "step": 8910 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008118010627181574, + "loss": 0.1625, + "step": 8920 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008115787367438138, + "loss": 0.1597, + "step": 8930 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008113564107694703, + "loss": 0.1626, + "step": 8940 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008111340847951266, + "loss": 0.1647, + "step": 8950 + }, + { + "epoch": 0.39, + "learning_rate": 0.000810911758820783, + "loss": 0.1591, + "step": 8960 + }, + { + "epoch": 0.39, + "learning_rate": 0.0008106894328464394, + "loss": 0.1622, + "step": 8970 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008104671068720959, + "loss": 0.1563, + "step": 8980 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008102447808977523, + "loss": 0.1625, + "step": 8990 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008100224549234087, + "loss": 0.162, + "step": 9000 + }, + { + "epoch": 0.4, + "eval_cer": 0.5010038213321276, + "eval_loss": 0.11899059265851974, + "eval_runtime": 84.3326, + "eval_samples_per_second": 118.507, + "eval_steps_per_second": 7.411, + "step": 9000 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008098001289490651, + "loss": 0.1613, + "step": 9010 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008095778029747215, + "loss": 0.1597, + "step": 9020 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008093554770003779, + "loss": 0.16, + "step": 9030 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008091331510260343, + "loss": 0.1567, + "step": 9040 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008089108250516907, + "loss": 0.1603, + "step": 9050 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008086884990773472, + "loss": 0.1597, + "step": 9060 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008084661731030036, + "loss": 0.1562, + "step": 9070 + }, + { + "epoch": 0.4, + "learning_rate": 0.00080824384712866, + "loss": 0.1545, + "step": 9080 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008080215211543164, + "loss": 0.1638, + "step": 9090 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008077991951799729, + "loss": 0.1631, + "step": 9100 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008075768692056293, + "loss": 0.1585, + "step": 9110 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008073545432312857, + "loss": 0.1578, + "step": 9120 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008071322172569422, + "loss": 0.1544, + "step": 9130 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008069098912825986, + "loss": 0.1589, + "step": 9140 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008066875653082549, + "loss": 0.1561, + "step": 9150 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008064652393339113, + "loss": 0.157, + "step": 9160 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008062429133595678, + "loss": 0.1632, + "step": 9170 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008060205873852242, + "loss": 0.1623, + "step": 9180 + }, + { + "epoch": 0.4, + "learning_rate": 0.0008057982614108806, + "loss": 0.1584, + "step": 9190 + }, + { + "epoch": 0.4, + "learning_rate": 0.000805575935436537, + "loss": 0.1559, + "step": 9200 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008053536094621935, + "loss": 0.1607, + "step": 9210 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008051312834878499, + "loss": 0.1559, + "step": 9220 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008049089575135063, + "loss": 0.159, + "step": 9230 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008046866315391627, + "loss": 0.161, + "step": 9240 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008044643055648192, + "loss": 0.1546, + "step": 9250 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008042419795904756, + "loss": 0.1533, + "step": 9260 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008040196536161319, + "loss": 0.1554, + "step": 9270 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008037973276417883, + "loss": 0.1561, + "step": 9280 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008035750016674448, + "loss": 0.1598, + "step": 9290 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008033526756931012, + "loss": 0.1563, + "step": 9300 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008031303497187576, + "loss": 0.1551, + "step": 9310 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008029080237444141, + "loss": 0.1574, + "step": 9320 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008026856977700705, + "loss": 0.1546, + "step": 9330 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008024633717957269, + "loss": 0.1558, + "step": 9340 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008022410458213833, + "loss": 0.1582, + "step": 9350 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008020187198470398, + "loss": 0.1586, + "step": 9360 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008017963938726962, + "loss": 0.1574, + "step": 9370 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008015740678983526, + "loss": 0.1585, + "step": 9380 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008013517419240089, + "loss": 0.1571, + "step": 9390 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008011294159496654, + "loss": 0.1598, + "step": 9400 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008009070899753218, + "loss": 0.1601, + "step": 9410 + }, + { + "epoch": 0.41, + "learning_rate": 0.0008006847640009782, + "loss": 0.161, + "step": 9420 + }, + { + "epoch": 0.42, + "learning_rate": 0.0008004624380266347, + "loss": 0.1577, + "step": 9430 + }, + { + "epoch": 0.42, + "learning_rate": 0.0008002401120522911, + "loss": 0.1576, + "step": 9440 + }, + { + "epoch": 0.42, + "learning_rate": 0.0008000177860779475, + "loss": 0.1591, + "step": 9450 + }, + { + "epoch": 0.42, + "learning_rate": 0.0007997954601036039, + "loss": 0.1596, + "step": 9460 + }, + { + "epoch": 0.42, + "learning_rate": 0.0007995731341292604, + "loss": 0.1554, + "step": 9470 + }, + { + "epoch": 0.42, + "learning_rate": 0.0007993508081549168, + "loss": 0.1545, + "step": 9480 + }, + { + "epoch": 0.42, + "learning_rate": 0.0007991284821805732, + "loss": 0.1567, + "step": 9490 + }, + { + "epoch": 0.42, + "learning_rate": 0.0007989061562062296, + "loss": 0.1522, + "step": 9500 + }, + { + "epoch": 0.42, + "eval_cer": 0.5011225284751192, + "eval_loss": 0.1170419380068779, + "eval_runtime": 83.9997, + "eval_samples_per_second": 118.977, + "eval_steps_per_second": 7.441, + "step": 9500 + }, + { + "epoch": 0.42, + "learning_rate": 0.000798683830231886, + "loss": 0.1618, + "step": 9510 + }, + { + "epoch": 0.42, + "learning_rate": 0.0007984615042575424, + "loss": 0.162, + "step": 9520 + }, + { + "epoch": 0.42, + "learning_rate": 0.0007982391782831988, + "loss": 0.1558, + "step": 9530 + }, + { + "epoch": 0.42, + "learning_rate": 0.0007980168523088552, + "loss": 0.1549, + "step": 9540 + }, + { + "epoch": 0.42, + "learning_rate": 0.0007977945263345117, + "loss": 0.1582, + "step": 9550 + }, + { + "epoch": 0.42, + "learning_rate": 0.0007975722003601681, + "loss": 0.156, + "step": 9560 + }, + { + "epoch": 0.42, + "learning_rate": 0.0007973498743858245, + "loss": 0.1504, + "step": 9570 + }, + { + "epoch": 0.42, + "learning_rate": 0.000797127548411481, + "loss": 0.1547, + "step": 9580 + }, + { + "epoch": 0.42, + "learning_rate": 0.0007969052224371374, + "loss": 0.1578, + "step": 9590 + }, + { + "epoch": 0.42, + "learning_rate": 0.0007966828964627938, + "loss": 0.1527, + "step": 9600 + }, + { + "epoch": 0.42, + "learning_rate": 0.0007964605704884502, + "loss": 0.1516, + "step": 9610 + }, + { + "epoch": 0.42, + "learning_rate": 0.0007962382445141067, + "loss": 0.1544, + "step": 9620 + }, + { + "epoch": 0.42, + "learning_rate": 0.000796015918539763, + "loss": 0.1527, + "step": 9630 + }, + { + "epoch": 0.42, + "learning_rate": 0.0007957935925654194, + "loss": 0.1632, + "step": 9640 + }, + { + "epoch": 0.42, + "learning_rate": 0.0007955712665910758, + "loss": 0.1537, + "step": 9650 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007953489406167323, + "loss": 0.1632, + "step": 9660 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007951266146423887, + "loss": 0.1541, + "step": 9670 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007949042886680451, + "loss": 0.1513, + "step": 9680 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007946819626937015, + "loss": 0.1602, + "step": 9690 + }, + { + "epoch": 0.43, + "learning_rate": 0.000794459636719358, + "loss": 0.1603, + "step": 9700 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007942373107450144, + "loss": 0.1562, + "step": 9710 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007940149847706708, + "loss": 0.157, + "step": 9720 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007937926587963273, + "loss": 0.154, + "step": 9730 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007935703328219837, + "loss": 0.1581, + "step": 9740 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007933480068476401, + "loss": 0.1557, + "step": 9750 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007931256808732964, + "loss": 0.1543, + "step": 9760 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007929033548989529, + "loss": 0.1542, + "step": 9770 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007926810289246093, + "loss": 0.1502, + "step": 9780 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007924587029502657, + "loss": 0.158, + "step": 9790 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007922363769759221, + "loss": 0.1527, + "step": 9800 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007920140510015786, + "loss": 0.1511, + "step": 9810 + }, + { + "epoch": 0.43, + "learning_rate": 0.000791791725027235, + "loss": 0.1575, + "step": 9820 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007915693990528914, + "loss": 0.1569, + "step": 9830 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007913470730785478, + "loss": 0.1519, + "step": 9840 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007911247471042043, + "loss": 0.1506, + "step": 9850 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007909024211298607, + "loss": 0.1579, + "step": 9860 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007906800951555171, + "loss": 0.1569, + "step": 9870 + }, + { + "epoch": 0.43, + "learning_rate": 0.0007904577691811735, + "loss": 0.1596, + "step": 9880 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007902354432068299, + "loss": 0.1522, + "step": 9890 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007900131172324863, + "loss": 0.1541, + "step": 9900 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007897907912581427, + "loss": 0.1525, + "step": 9910 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007895684652837992, + "loss": 0.1498, + "step": 9920 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007893461393094556, + "loss": 0.1511, + "step": 9930 + }, + { + "epoch": 0.44, + "learning_rate": 0.000789123813335112, + "loss": 0.1545, + "step": 9940 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007889014873607684, + "loss": 0.155, + "step": 9950 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007886791613864248, + "loss": 0.1634, + "step": 9960 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007884568354120812, + "loss": 0.1624, + "step": 9970 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007882345094377376, + "loss": 0.1571, + "step": 9980 + }, + { + "epoch": 0.44, + "learning_rate": 0.000788012183463394, + "loss": 0.156, + "step": 9990 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007877898574890504, + "loss": 0.1493, + "step": 10000 + }, + { + "epoch": 0.44, + "eval_cer": 0.5012976701615001, + "eval_loss": 0.1143530085682869, + "eval_runtime": 84.0673, + "eval_samples_per_second": 118.881, + "eval_steps_per_second": 7.435, + "step": 10000 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007875675315147068, + "loss": 0.1541, + "step": 10010 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007873452055403632, + "loss": 0.1483, + "step": 10020 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007871228795660196, + "loss": 0.1542, + "step": 10030 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007869005535916761, + "loss": 0.1527, + "step": 10040 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007866782276173325, + "loss": 0.1503, + "step": 10050 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007864559016429889, + "loss": 0.1567, + "step": 10060 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007862335756686454, + "loss": 0.1534, + "step": 10070 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007860112496943018, + "loss": 0.1513, + "step": 10080 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007857889237199582, + "loss": 0.1563, + "step": 10090 + }, + { + "epoch": 0.44, + "learning_rate": 0.0007855665977456146, + "loss": 0.156, + "step": 10100 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007853442717712711, + "loss": 0.159, + "step": 10110 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007851219457969274, + "loss": 0.1577, + "step": 10120 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007848996198225838, + "loss": 0.1576, + "step": 10130 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007846772938482402, + "loss": 0.1528, + "step": 10140 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007844549678738967, + "loss": 0.1548, + "step": 10150 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007842326418995531, + "loss": 0.1534, + "step": 10160 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007840103159252095, + "loss": 0.1455, + "step": 10170 + }, + { + "epoch": 0.45, + "learning_rate": 0.000783787989950866, + "loss": 0.1531, + "step": 10180 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007835656639765224, + "loss": 0.1605, + "step": 10190 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007833433380021788, + "loss": 0.1491, + "step": 10200 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007831210120278352, + "loss": 0.157, + "step": 10210 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007828986860534917, + "loss": 0.1463, + "step": 10220 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007826763600791481, + "loss": 0.1495, + "step": 10230 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007824540341048045, + "loss": 0.1546, + "step": 10240 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007822317081304608, + "loss": 0.1539, + "step": 10250 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007820093821561173, + "loss": 0.1485, + "step": 10260 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007817870561817737, + "loss": 0.1529, + "step": 10270 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007815647302074301, + "loss": 0.1533, + "step": 10280 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007813424042330865, + "loss": 0.1502, + "step": 10290 + }, + { + "epoch": 0.45, + "learning_rate": 0.000781120078258743, + "loss": 0.1447, + "step": 10300 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007808977522843994, + "loss": 0.1502, + "step": 10310 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007806754263100558, + "loss": 0.1563, + "step": 10320 + }, + { + "epoch": 0.45, + "learning_rate": 0.0007804531003357122, + "loss": 0.1616, + "step": 10330 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007802307743613687, + "loss": 0.1599, + "step": 10340 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007800084483870251, + "loss": 0.1459, + "step": 10350 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007797861224126815, + "loss": 0.1496, + "step": 10360 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007795637964383379, + "loss": 0.1518, + "step": 10370 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007793414704639943, + "loss": 0.1483, + "step": 10380 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007791191444896507, + "loss": 0.1555, + "step": 10390 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007788968185153071, + "loss": 0.1461, + "step": 10400 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007786744925409636, + "loss": 0.1517, + "step": 10410 + }, + { + "epoch": 0.46, + "learning_rate": 0.00077845216656662, + "loss": 0.154, + "step": 10420 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007782298405922764, + "loss": 0.1571, + "step": 10430 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007780075146179328, + "loss": 0.1562, + "step": 10440 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007777851886435893, + "loss": 0.149, + "step": 10450 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007775628626692457, + "loss": 0.1482, + "step": 10460 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007773405366949021, + "loss": 0.1525, + "step": 10470 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007771182107205586, + "loss": 0.1477, + "step": 10480 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007768958847462149, + "loss": 0.1552, + "step": 10490 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007766735587718713, + "loss": 0.1563, + "step": 10500 + }, + { + "epoch": 0.46, + "eval_cer": 0.500866302674673, + "eval_loss": 0.11313384771347046, + "eval_runtime": 83.8469, + "eval_samples_per_second": 119.193, + "eval_steps_per_second": 7.454, + "step": 10500 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007764512327975277, + "loss": 0.1547, + "step": 10510 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007762289068231842, + "loss": 0.1478, + "step": 10520 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007760065808488406, + "loss": 0.1533, + "step": 10530 + }, + { + "epoch": 0.46, + "learning_rate": 0.000775784254874497, + "loss": 0.1513, + "step": 10540 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007755619289001534, + "loss": 0.1446, + "step": 10550 + }, + { + "epoch": 0.46, + "learning_rate": 0.0007753396029258099, + "loss": 0.1482, + "step": 10560 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007751172769514663, + "loss": 0.1565, + "step": 10570 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007748949509771227, + "loss": 0.1501, + "step": 10580 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007746726250027791, + "loss": 0.151, + "step": 10590 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007744502990284356, + "loss": 0.1451, + "step": 10600 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007742279730540919, + "loss": 0.1493, + "step": 10610 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007740056470797483, + "loss": 0.1483, + "step": 10620 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007737833211054047, + "loss": 0.1546, + "step": 10630 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007735609951310612, + "loss": 0.1482, + "step": 10640 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007733386691567176, + "loss": 0.1428, + "step": 10650 + }, + { + "epoch": 0.47, + "learning_rate": 0.000773116343182374, + "loss": 0.1522, + "step": 10660 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007728940172080305, + "loss": 0.1501, + "step": 10670 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007726716912336869, + "loss": 0.1567, + "step": 10680 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007724493652593433, + "loss": 0.1486, + "step": 10690 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007722270392849997, + "loss": 0.1582, + "step": 10700 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007720047133106562, + "loss": 0.1543, + "step": 10710 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007717823873363126, + "loss": 0.1498, + "step": 10720 + }, + { + "epoch": 0.47, + "learning_rate": 0.000771560061361969, + "loss": 0.1463, + "step": 10730 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007713377353876253, + "loss": 0.1456, + "step": 10740 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007711154094132818, + "loss": 0.1461, + "step": 10750 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007708930834389382, + "loss": 0.1508, + "step": 10760 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007706707574645946, + "loss": 0.146, + "step": 10770 + }, + { + "epoch": 0.47, + "learning_rate": 0.000770448431490251, + "loss": 0.1492, + "step": 10780 + }, + { + "epoch": 0.47, + "learning_rate": 0.0007702261055159075, + "loss": 0.1491, + "step": 10790 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007700037795415639, + "loss": 0.1454, + "step": 10800 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007697814535672203, + "loss": 0.1508, + "step": 10810 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007695591275928768, + "loss": 0.1495, + "step": 10820 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007693368016185332, + "loss": 0.1467, + "step": 10830 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007691144756441896, + "loss": 0.1492, + "step": 10840 + }, + { + "epoch": 0.48, + "learning_rate": 0.000768892149669846, + "loss": 0.1521, + "step": 10850 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007686698236955024, + "loss": 0.1503, + "step": 10860 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007684474977211588, + "loss": 0.1487, + "step": 10870 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007682251717468152, + "loss": 0.1522, + "step": 10880 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007680028457724716, + "loss": 0.153, + "step": 10890 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007677805197981281, + "loss": 0.1568, + "step": 10900 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007675581938237845, + "loss": 0.1486, + "step": 10910 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007673358678494409, + "loss": 0.1555, + "step": 10920 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007671135418750973, + "loss": 0.1476, + "step": 10930 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007668912159007538, + "loss": 0.1469, + "step": 10940 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007666688899264102, + "loss": 0.1483, + "step": 10950 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007664465639520665, + "loss": 0.1482, + "step": 10960 + }, + { + "epoch": 0.48, + "learning_rate": 0.000766224237977723, + "loss": 0.1584, + "step": 10970 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007660019120033793, + "loss": 0.1446, + "step": 10980 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007657795860290357, + "loss": 0.1534, + "step": 10990 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007655572600546921, + "loss": 0.1455, + "step": 11000 + }, + { + "epoch": 0.48, + "eval_cer": 0.5009006823390367, + "eval_loss": 0.11143666505813599, + "eval_runtime": 83.7751, + "eval_samples_per_second": 119.296, + "eval_steps_per_second": 7.46, + "step": 11000 + }, + { + "epoch": 0.48, + "learning_rate": 0.0007653349340803486, + "loss": 0.154, + "step": 11010 + }, + { + "epoch": 0.49, + "learning_rate": 0.000765112608106005, + "loss": 0.1484, + "step": 11020 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007648902821316614, + "loss": 0.149, + "step": 11030 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007646679561573178, + "loss": 0.1519, + "step": 11040 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007644456301829743, + "loss": 0.154, + "step": 11050 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007642233042086307, + "loss": 0.1548, + "step": 11060 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007640009782342871, + "loss": 0.1444, + "step": 11070 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007637786522599435, + "loss": 0.1471, + "step": 11080 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007635563262856, + "loss": 0.1483, + "step": 11090 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007633340003112563, + "loss": 0.1523, + "step": 11100 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007631116743369127, + "loss": 0.1486, + "step": 11110 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007628893483625691, + "loss": 0.1472, + "step": 11120 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007626670223882256, + "loss": 0.1498, + "step": 11130 + }, + { + "epoch": 0.49, + "learning_rate": 0.000762444696413882, + "loss": 0.1491, + "step": 11140 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007622223704395384, + "loss": 0.1549, + "step": 11150 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007620000444651949, + "loss": 0.1557, + "step": 11160 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007617777184908513, + "loss": 0.1399, + "step": 11170 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007615553925165077, + "loss": 0.1476, + "step": 11180 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007613330665421641, + "loss": 0.1475, + "step": 11190 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007611107405678206, + "loss": 0.1473, + "step": 11200 + }, + { + "epoch": 0.49, + "learning_rate": 0.000760888414593477, + "loss": 0.1548, + "step": 11210 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007606660886191333, + "loss": 0.1492, + "step": 11220 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007604437626447897, + "loss": 0.1514, + "step": 11230 + }, + { + "epoch": 0.49, + "learning_rate": 0.0007602214366704462, + "loss": 0.1453, + "step": 11240 + }, + { + "epoch": 0.5, + "learning_rate": 0.0007599991106961026, + "loss": 0.1502, + "step": 11250 + }, + { + "epoch": 0.5, + "learning_rate": 0.000759776784721759, + "loss": 0.1481, + "step": 11260 + }, + { + "epoch": 0.5, + "learning_rate": 0.0007595544587474154, + "loss": 0.1415, + "step": 11270 + }, + { + "epoch": 0.5, + "learning_rate": 0.0007593321327730719, + "loss": 0.1384, + "step": 11280 + }, + { + "epoch": 0.5, + "learning_rate": 0.0007591098067987283, + "loss": 0.147, + "step": 11290 + }, + { + "epoch": 0.5, + "learning_rate": 0.0007588874808243847, + "loss": 0.1468, + "step": 11300 + }, + { + "epoch": 0.5, + "learning_rate": 0.0007586651548500412, + "loss": 0.1507, + "step": 11310 + }, + { + "epoch": 0.5, + "learning_rate": 0.0007584428288756976, + "loss": 0.1488, + "step": 11320 + }, + { + "epoch": 0.5, + "learning_rate": 0.000758220502901354, + "loss": 0.1523, + "step": 11330 + }, + { + "epoch": 0.5, + "learning_rate": 0.0007579981769270104, + "loss": 0.1447, + "step": 11340 + }, + { + "epoch": 0.5, + "learning_rate": 0.0007577758509526668, + "loss": 0.1454, + "step": 11350 + }, + { + "epoch": 0.5, + "learning_rate": 0.0007575535249783232, + "loss": 0.1504, + "step": 11360 + }, + { + "epoch": 0.5, + "learning_rate": 0.0007573311990039796, + "loss": 0.145, + "step": 11370 + }, + { + "epoch": 0.5, + "learning_rate": 0.000757108873029636, + "loss": 0.145, + "step": 11380 + }, + { + "epoch": 0.5, + "learning_rate": 0.0007568865470552925, + "loss": 0.1483, + "step": 11390 + }, + { + "epoch": 0.5, + "learning_rate": 0.0007566642210809489, + "loss": 0.1433, + "step": 11400 + }, + { + "epoch": 0.5, + "learning_rate": 0.0007564418951066053, + "loss": 0.1458, + "step": 11410 + }, + { + "epoch": 0.5, + "learning_rate": 0.0007562195691322617, + "loss": 0.154, + "step": 11420 + }, + { + "epoch": 0.5, + "learning_rate": 0.0007559972431579182, + "loss": 0.1468, + "step": 11430 + }, + { + "epoch": 0.5, + "learning_rate": 0.0007557749171835746, + "loss": 0.1491, + "step": 11440 + }, + { + "epoch": 0.5, + "learning_rate": 0.000755552591209231, + "loss": 0.1426, + "step": 11450 + }, + { + "epoch": 0.5, + "learning_rate": 0.0007553302652348875, + "loss": 0.1474, + "step": 11460 + }, + { + "epoch": 0.5, + "learning_rate": 0.0007551079392605438, + "loss": 0.147, + "step": 11470 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007548856132862002, + "loss": 0.1473, + "step": 11480 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007546632873118566, + "loss": 0.1483, + "step": 11490 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007544409613375131, + "loss": 0.1497, + "step": 11500 + }, + { + "epoch": 0.51, + "eval_cer": 0.500616563603352, + "eval_loss": 0.10954929888248444, + "eval_runtime": 83.806, + "eval_samples_per_second": 119.252, + "eval_steps_per_second": 7.458, + "step": 11500 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007542186353631695, + "loss": 0.1504, + "step": 11510 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007539963093888259, + "loss": 0.1453, + "step": 11520 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007537739834144823, + "loss": 0.1453, + "step": 11530 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007535516574401388, + "loss": 0.1463, + "step": 11540 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007533293314657952, + "loss": 0.1449, + "step": 11550 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007531070054914516, + "loss": 0.148, + "step": 11560 + }, + { + "epoch": 0.51, + "learning_rate": 0.000752884679517108, + "loss": 0.1464, + "step": 11570 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007526623535427645, + "loss": 0.1448, + "step": 11580 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007524400275684208, + "loss": 0.1514, + "step": 11590 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007522177015940772, + "loss": 0.1402, + "step": 11600 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007519953756197337, + "loss": 0.1516, + "step": 11610 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007517730496453901, + "loss": 0.1494, + "step": 11620 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007515507236710465, + "loss": 0.1431, + "step": 11630 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007513283976967029, + "loss": 0.1457, + "step": 11640 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007511060717223594, + "loss": 0.1407, + "step": 11650 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007508837457480158, + "loss": 0.1438, + "step": 11660 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007506614197736722, + "loss": 0.1414, + "step": 11670 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007504390937993286, + "loss": 0.1499, + "step": 11680 + }, + { + "epoch": 0.51, + "learning_rate": 0.0007502167678249851, + "loss": 0.1414, + "step": 11690 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007499944418506415, + "loss": 0.1441, + "step": 11700 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007497721158762978, + "loss": 0.1482, + "step": 11710 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007495497899019542, + "loss": 0.1458, + "step": 11720 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007493274639276107, + "loss": 0.1435, + "step": 11730 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007491051379532671, + "loss": 0.1419, + "step": 11740 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007488828119789235, + "loss": 0.1462, + "step": 11750 + }, + { + "epoch": 0.52, + "learning_rate": 0.00074866048600458, + "loss": 0.1482, + "step": 11760 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007484381600302364, + "loss": 0.1491, + "step": 11770 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007482158340558928, + "loss": 0.1478, + "step": 11780 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007479935080815492, + "loss": 0.148, + "step": 11790 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007477711821072057, + "loss": 0.1455, + "step": 11800 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007475488561328621, + "loss": 0.148, + "step": 11810 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007473265301585185, + "loss": 0.1497, + "step": 11820 + }, + { + "epoch": 0.52, + "learning_rate": 0.000747104204184175, + "loss": 0.1485, + "step": 11830 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007468818782098313, + "loss": 0.1481, + "step": 11840 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007466595522354877, + "loss": 0.1451, + "step": 11850 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007464372262611441, + "loss": 0.1399, + "step": 11860 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007462149002868005, + "loss": 0.1465, + "step": 11870 + }, + { + "epoch": 0.52, + "learning_rate": 0.000745992574312457, + "loss": 0.1431, + "step": 11880 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007457702483381134, + "loss": 0.1448, + "step": 11890 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007455479223637698, + "loss": 0.1452, + "step": 11900 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007453255963894263, + "loss": 0.1378, + "step": 11910 + }, + { + "epoch": 0.52, + "learning_rate": 0.0007451032704150827, + "loss": 0.1416, + "step": 11920 + }, + { + "epoch": 0.53, + "learning_rate": 0.0007448809444407391, + "loss": 0.1364, + "step": 11930 + }, + { + "epoch": 0.53, + "learning_rate": 0.0007446586184663955, + "loss": 0.1516, + "step": 11940 + }, + { + "epoch": 0.53, + "learning_rate": 0.000744436292492052, + "loss": 0.1488, + "step": 11950 + }, + { + "epoch": 0.53, + "learning_rate": 0.0007442139665177082, + "loss": 0.1432, + "step": 11960 + }, + { + "epoch": 0.53, + "learning_rate": 0.0007439916405433646, + "loss": 0.1454, + "step": 11970 + }, + { + "epoch": 0.53, + "learning_rate": 0.000743769314569021, + "loss": 0.1367, + "step": 11980 + }, + { + "epoch": 0.53, + "learning_rate": 0.0007435469885946775, + "loss": 0.1488, + "step": 11990 + }, + { + "epoch": 0.53, + "learning_rate": 0.0007433246626203339, + "loss": 0.1411, + "step": 12000 + }, + { + "epoch": 0.53, + "eval_cer": 0.5008163548604089, + "eval_loss": 0.10829149186611176, + "eval_runtime": 83.6996, + "eval_samples_per_second": 119.403, + "eval_steps_per_second": 7.467, + "step": 12000 + }, + { + "epoch": 0.53, + "learning_rate": 0.0007431023366459903, + "loss": 0.1514, + "step": 12010 + }, + { + "epoch": 0.53, + "learning_rate": 0.0007428800106716467, + "loss": 0.1433, + "step": 12020 + }, + { + "epoch": 0.53, + "learning_rate": 0.0007426576846973032, + "loss": 0.1457, + "step": 12030 + }, + { + "epoch": 0.53, + "learning_rate": 0.0007424353587229596, + "loss": 0.1412, + "step": 12040 + }, + { + "epoch": 0.53, + "learning_rate": 0.000742213032748616, + "loss": 0.1467, + "step": 12050 + }, + { + "epoch": 0.53, + "learning_rate": 0.0007419907067742725, + "loss": 0.149, + "step": 12060 + }, + { + "epoch": 0.53, + "learning_rate": 0.0007417683807999289, + "loss": 0.1407, + "step": 12070 + }, + { + "epoch": 0.53, + "learning_rate": 0.0007415460548255852, + "loss": 0.1459, + "step": 12080 + }, + { + "epoch": 0.53, + "learning_rate": 0.0007413237288512416, + "loss": 0.1475, + "step": 12090 + }, + { + "epoch": 0.53, + "learning_rate": 0.000741101402876898, + "loss": 0.1445, + "step": 12100 + }, + { + "epoch": 0.53, + "learning_rate": 0.0007408790769025545, + "loss": 0.1487, + "step": 12110 + }, + { + "epoch": 0.53, + "learning_rate": 0.0007406567509282109, + "loss": 0.148, + "step": 12120 + }, + { + "epoch": 0.53, + "learning_rate": 0.0007404344249538673, + "loss": 0.1444, + "step": 12130 + }, + { + "epoch": 0.53, + "learning_rate": 0.0007402120989795238, + "loss": 0.1466, + "step": 12140 + }, + { + "epoch": 0.53, + "learning_rate": 0.0007399897730051802, + "loss": 0.1499, + "step": 12150 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007397674470308366, + "loss": 0.1417, + "step": 12160 + }, + { + "epoch": 0.54, + "learning_rate": 0.000739545121056493, + "loss": 0.1427, + "step": 12170 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007393227950821495, + "loss": 0.1415, + "step": 12180 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007391004691078059, + "loss": 0.1462, + "step": 12190 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007388781431334622, + "loss": 0.1419, + "step": 12200 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007386558171591186, + "loss": 0.1425, + "step": 12210 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007384334911847751, + "loss": 0.1489, + "step": 12220 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007382111652104315, + "loss": 0.1447, + "step": 12230 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007379888392360879, + "loss": 0.1445, + "step": 12240 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007377665132617444, + "loss": 0.1484, + "step": 12250 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007375441872874008, + "loss": 0.1456, + "step": 12260 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007373218613130572, + "loss": 0.1465, + "step": 12270 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007370995353387136, + "loss": 0.1351, + "step": 12280 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007368772093643701, + "loss": 0.1467, + "step": 12290 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007366548833900265, + "loss": 0.1409, + "step": 12300 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007364325574156829, + "loss": 0.1474, + "step": 12310 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007362102314413393, + "loss": 0.1405, + "step": 12320 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007359879054669957, + "loss": 0.1455, + "step": 12330 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007357655794926521, + "loss": 0.1427, + "step": 12340 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007355432535183085, + "loss": 0.1462, + "step": 12350 + }, + { + "epoch": 0.54, + "learning_rate": 0.000735320927543965, + "loss": 0.1416, + "step": 12360 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007350986015696214, + "loss": 0.143, + "step": 12370 + }, + { + "epoch": 0.54, + "learning_rate": 0.0007348762755952778, + "loss": 0.1425, + "step": 12380 + }, + { + "epoch": 0.55, + "learning_rate": 0.0007346539496209342, + "loss": 0.144, + "step": 12390 + }, + { + "epoch": 0.55, + "learning_rate": 0.0007344316236465907, + "loss": 0.1442, + "step": 12400 + }, + { + "epoch": 0.55, + "learning_rate": 0.0007342092976722471, + "loss": 0.1448, + "step": 12410 + }, + { + "epoch": 0.55, + "learning_rate": 0.0007339869716979035, + "loss": 0.15, + "step": 12420 + }, + { + "epoch": 0.55, + "learning_rate": 0.0007337646457235599, + "loss": 0.1492, + "step": 12430 + }, + { + "epoch": 0.55, + "learning_rate": 0.0007335423197492164, + "loss": 0.1438, + "step": 12440 + }, + { + "epoch": 0.55, + "learning_rate": 0.0007333199937748727, + "loss": 0.145, + "step": 12450 + }, + { + "epoch": 0.55, + "learning_rate": 0.0007330976678005291, + "loss": 0.1376, + "step": 12460 + }, + { + "epoch": 0.55, + "learning_rate": 0.0007328753418261855, + "loss": 0.1382, + "step": 12470 + }, + { + "epoch": 0.55, + "learning_rate": 0.000732653015851842, + "loss": 0.144, + "step": 12480 + }, + { + "epoch": 0.55, + "learning_rate": 0.0007324306898774984, + "loss": 0.1426, + "step": 12490 + }, + { + "epoch": 0.55, + "learning_rate": 0.0007322083639031548, + "loss": 0.1395, + "step": 12500 + }, + { + "epoch": 0.55, + "eval_cer": 0.5005212086852113, + "eval_loss": 0.10718655586242676, + "eval_runtime": 83.9796, + "eval_samples_per_second": 119.005, + "eval_steps_per_second": 7.442, + "step": 12500 + }, + { + "epoch": 0.55, + "learning_rate": 0.0007319860379288113, + "loss": 0.1422, + "step": 12510 + }, + { + "epoch": 0.55, + "learning_rate": 0.0007317637119544677, + "loss": 0.139, + "step": 12520 + }, + { + "epoch": 0.55, + "learning_rate": 0.0007315413859801241, + "loss": 0.1411, + "step": 12530 + }, + { + "epoch": 0.55, + "learning_rate": 0.0007313190600057805, + "loss": 0.1418, + "step": 12540 + }, + { + "epoch": 0.55, + "learning_rate": 0.000731096734031437, + "loss": 0.1432, + "step": 12550 + }, + { + "epoch": 0.55, + "learning_rate": 0.0007308744080570934, + "loss": 0.1393, + "step": 12560 + }, + { + "epoch": 0.55, + "learning_rate": 0.0007306520820827497, + "loss": 0.1455, + "step": 12570 + }, + { + "epoch": 0.55, + "learning_rate": 0.0007304297561084061, + "loss": 0.1475, + "step": 12580 + }, + { + "epoch": 0.55, + "learning_rate": 0.0007302074301340626, + "loss": 0.1395, + "step": 12590 + }, + { + "epoch": 0.55, + "learning_rate": 0.000729985104159719, + "loss": 0.1382, + "step": 12600 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007297627781853754, + "loss": 0.148, + "step": 12610 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007295404522110318, + "loss": 0.1368, + "step": 12620 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007293181262366883, + "loss": 0.1389, + "step": 12630 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007290958002623447, + "loss": 0.1446, + "step": 12640 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007288734742880011, + "loss": 0.1459, + "step": 12650 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007286511483136576, + "loss": 0.1437, + "step": 12660 + }, + { + "epoch": 0.56, + "learning_rate": 0.000728428822339314, + "loss": 0.1419, + "step": 12670 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007282064963649704, + "loss": 0.1411, + "step": 12680 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007279841703906267, + "loss": 0.1397, + "step": 12690 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007277618444162832, + "loss": 0.1445, + "step": 12700 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007275395184419396, + "loss": 0.1411, + "step": 12710 + }, + { + "epoch": 0.56, + "learning_rate": 0.000727317192467596, + "loss": 0.142, + "step": 12720 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007270948664932524, + "loss": 0.1433, + "step": 12730 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007268725405189089, + "loss": 0.1447, + "step": 12740 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007266502145445653, + "loss": 0.1435, + "step": 12750 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007264278885702217, + "loss": 0.1445, + "step": 12760 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007262055625958781, + "loss": 0.1428, + "step": 12770 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007259832366215346, + "loss": 0.146, + "step": 12780 + }, + { + "epoch": 0.56, + "learning_rate": 0.000725760910647191, + "loss": 0.1451, + "step": 12790 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007255385846728474, + "loss": 0.137, + "step": 12800 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007253162586985037, + "loss": 0.1395, + "step": 12810 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007250939327241602, + "loss": 0.1398, + "step": 12820 + }, + { + "epoch": 0.56, + "learning_rate": 0.0007248716067498166, + "loss": 0.1427, + "step": 12830 + }, + { + "epoch": 0.57, + "learning_rate": 0.000724649280775473, + "loss": 0.1419, + "step": 12840 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007244269548011295, + "loss": 0.1402, + "step": 12850 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007242046288267859, + "loss": 0.1378, + "step": 12860 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007239823028524423, + "loss": 0.1425, + "step": 12870 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007237599768780987, + "loss": 0.1408, + "step": 12880 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007235376509037552, + "loss": 0.1398, + "step": 12890 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007233153249294116, + "loss": 0.1425, + "step": 12900 + }, + { + "epoch": 0.57, + "learning_rate": 0.000723092998955068, + "loss": 0.1447, + "step": 12910 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007228706729807244, + "loss": 0.1392, + "step": 12920 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007226483470063809, + "loss": 0.144, + "step": 12930 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007224260210320371, + "loss": 0.146, + "step": 12940 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007222036950576935, + "loss": 0.1411, + "step": 12950 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007219813690833499, + "loss": 0.1421, + "step": 12960 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007217590431090064, + "loss": 0.1444, + "step": 12970 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007215367171346628, + "loss": 0.1349, + "step": 12980 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007213143911603192, + "loss": 0.1424, + "step": 12990 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007210920651859756, + "loss": 0.1364, + "step": 13000 + }, + { + "epoch": 0.57, + "eval_cer": 0.5003687705507687, + "eval_loss": 0.1058240681886673, + "eval_runtime": 83.9573, + "eval_samples_per_second": 119.037, + "eval_steps_per_second": 7.444, + "step": 13000 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007208697392116321, + "loss": 0.1384, + "step": 13010 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007206474132372885, + "loss": 0.141, + "step": 13020 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007204250872629449, + "loss": 0.1419, + "step": 13030 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007202027612886014, + "loss": 0.1389, + "step": 13040 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007199804353142578, + "loss": 0.1373, + "step": 13050 + }, + { + "epoch": 0.57, + "learning_rate": 0.0007197581093399141, + "loss": 0.1347, + "step": 13060 + }, + { + "epoch": 0.58, + "learning_rate": 0.0007195357833655705, + "loss": 0.1395, + "step": 13070 + }, + { + "epoch": 0.58, + "learning_rate": 0.000719313457391227, + "loss": 0.1443, + "step": 13080 + }, + { + "epoch": 0.58, + "learning_rate": 0.0007190911314168834, + "loss": 0.1434, + "step": 13090 + }, + { + "epoch": 0.58, + "learning_rate": 0.0007188688054425398, + "loss": 0.1336, + "step": 13100 + }, + { + "epoch": 0.58, + "learning_rate": 0.0007186464794681962, + "loss": 0.1409, + "step": 13110 + }, + { + "epoch": 0.58, + "learning_rate": 0.0007184241534938527, + "loss": 0.1433, + "step": 13120 + }, + { + "epoch": 0.58, + "learning_rate": 0.0007182018275195091, + "loss": 0.1452, + "step": 13130 + }, + { + "epoch": 0.58, + "learning_rate": 0.0007179795015451655, + "loss": 0.1437, + "step": 13140 + }, + { + "epoch": 0.58, + "learning_rate": 0.000717757175570822, + "loss": 0.1429, + "step": 13150 + }, + { + "epoch": 0.58, + "learning_rate": 0.0007175348495964784, + "loss": 0.1442, + "step": 13160 + }, + { + "epoch": 0.58, + "learning_rate": 0.0007173125236221348, + "loss": 0.1441, + "step": 13170 + }, + { + "epoch": 0.58, + "learning_rate": 0.0007170901976477911, + "loss": 0.1379, + "step": 13180 + }, + { + "epoch": 0.58, + "learning_rate": 0.0007168678716734476, + "loss": 0.1397, + "step": 13190 + }, + { + "epoch": 0.58, + "learning_rate": 0.000716645545699104, + "loss": 0.1405, + "step": 13200 + }, + { + "epoch": 0.58, + "learning_rate": 0.0007164232197247604, + "loss": 0.1343, + "step": 13210 + }, + { + "epoch": 0.58, + "learning_rate": 0.0007162008937504168, + "loss": 0.1434, + "step": 13220 + }, + { + "epoch": 0.58, + "learning_rate": 0.0007159785677760733, + "loss": 0.1403, + "step": 13230 + }, + { + "epoch": 0.58, + "learning_rate": 0.0007157562418017297, + "loss": 0.1435, + "step": 13240 + }, + { + "epoch": 0.58, + "learning_rate": 0.0007155339158273861, + "loss": 0.1384, + "step": 13250 + }, + { + "epoch": 0.58, + "learning_rate": 0.0007153115898530425, + "loss": 0.1348, + "step": 13260 + }, + { + "epoch": 0.58, + "learning_rate": 0.000715089263878699, + "loss": 0.1414, + "step": 13270 + }, + { + "epoch": 0.58, + "learning_rate": 0.0007148669379043554, + "loss": 0.1398, + "step": 13280 + }, + { + "epoch": 0.59, + "learning_rate": 0.0007146446119300118, + "loss": 0.14, + "step": 13290 + }, + { + "epoch": 0.59, + "learning_rate": 0.0007144222859556681, + "loss": 0.14, + "step": 13300 + }, + { + "epoch": 0.59, + "learning_rate": 0.0007141999599813246, + "loss": 0.1405, + "step": 13310 + }, + { + "epoch": 0.59, + "learning_rate": 0.000713977634006981, + "loss": 0.1378, + "step": 13320 + }, + { + "epoch": 0.59, + "learning_rate": 0.0007137553080326374, + "loss": 0.1448, + "step": 13330 + }, + { + "epoch": 0.59, + "learning_rate": 0.0007135329820582939, + "loss": 0.1362, + "step": 13340 + }, + { + "epoch": 0.59, + "learning_rate": 0.0007133106560839503, + "loss": 0.1408, + "step": 13350 + }, + { + "epoch": 0.59, + "learning_rate": 0.0007130883301096067, + "loss": 0.1416, + "step": 13360 + }, + { + "epoch": 0.59, + "learning_rate": 0.0007128660041352631, + "loss": 0.1422, + "step": 13370 + }, + { + "epoch": 0.59, + "learning_rate": 0.0007126436781609196, + "loss": 0.1404, + "step": 13380 + }, + { + "epoch": 0.59, + "learning_rate": 0.000712421352186576, + "loss": 0.1406, + "step": 13390 + }, + { + "epoch": 0.59, + "learning_rate": 0.0007121990262122324, + "loss": 0.1416, + "step": 13400 + }, + { + "epoch": 0.59, + "learning_rate": 0.0007119767002378888, + "loss": 0.1369, + "step": 13410 + }, + { + "epoch": 0.59, + "learning_rate": 0.0007117543742635453, + "loss": 0.1437, + "step": 13420 + }, + { + "epoch": 0.59, + "learning_rate": 0.0007115320482892016, + "loss": 0.141, + "step": 13430 + }, + { + "epoch": 0.59, + "learning_rate": 0.000711309722314858, + "loss": 0.1387, + "step": 13440 + }, + { + "epoch": 0.59, + "learning_rate": 0.0007110873963405144, + "loss": 0.1348, + "step": 13450 + }, + { + "epoch": 0.59, + "learning_rate": 0.0007108650703661709, + "loss": 0.1365, + "step": 13460 + }, + { + "epoch": 0.59, + "learning_rate": 0.0007106427443918273, + "loss": 0.1403, + "step": 13470 + }, + { + "epoch": 0.59, + "learning_rate": 0.0007104204184174837, + "loss": 0.1386, + "step": 13480 + }, + { + "epoch": 0.59, + "learning_rate": 0.0007101980924431402, + "loss": 0.1384, + "step": 13490 + }, + { + "epoch": 0.59, + "learning_rate": 0.0007099757664687966, + "loss": 0.1379, + "step": 13500 + }, + { + "epoch": 0.59, + "eval_cer": 0.5005620750787002, + "eval_loss": 0.10582197457551956, + "eval_runtime": 84.0218, + "eval_samples_per_second": 118.945, + "eval_steps_per_second": 7.439, + "step": 13500 + }, + { + "epoch": 0.59, + "learning_rate": 0.000709753440494453, + "loss": 0.1447, + "step": 13510 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007095311145201094, + "loss": 0.1393, + "step": 13520 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007093087885457659, + "loss": 0.1436, + "step": 13530 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007090864625714223, + "loss": 0.1397, + "step": 13540 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007088641365970786, + "loss": 0.1391, + "step": 13550 + }, + { + "epoch": 0.6, + "learning_rate": 0.000708641810622735, + "loss": 0.1411, + "step": 13560 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007084194846483915, + "loss": 0.1431, + "step": 13570 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007081971586740479, + "loss": 0.1359, + "step": 13580 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007079748326997043, + "loss": 0.1441, + "step": 13590 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007077525067253608, + "loss": 0.1358, + "step": 13600 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007075301807510172, + "loss": 0.1403, + "step": 13610 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007073078547766736, + "loss": 0.1444, + "step": 13620 + }, + { + "epoch": 0.6, + "learning_rate": 0.00070708552880233, + "loss": 0.1385, + "step": 13630 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007068632028279865, + "loss": 0.1422, + "step": 13640 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007066408768536429, + "loss": 0.1416, + "step": 13650 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007064185508792993, + "loss": 0.1394, + "step": 13660 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007061962249049556, + "loss": 0.1367, + "step": 13670 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007059738989306121, + "loss": 0.1366, + "step": 13680 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007057515729562685, + "loss": 0.1422, + "step": 13690 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007055292469819249, + "loss": 0.1375, + "step": 13700 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007053069210075813, + "loss": 0.1436, + "step": 13710 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007050845950332378, + "loss": 0.136, + "step": 13720 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007048622690588942, + "loss": 0.1391, + "step": 13730 + }, + { + "epoch": 0.6, + "learning_rate": 0.0007046399430845506, + "loss": 0.1421, + "step": 13740 + }, + { + "epoch": 0.61, + "learning_rate": 0.000704417617110207, + "loss": 0.1387, + "step": 13750 + }, + { + "epoch": 0.61, + "learning_rate": 0.0007041952911358635, + "loss": 0.1343, + "step": 13760 + }, + { + "epoch": 0.61, + "learning_rate": 0.0007039729651615199, + "loss": 0.1402, + "step": 13770 + }, + { + "epoch": 0.61, + "learning_rate": 0.0007037506391871763, + "loss": 0.1445, + "step": 13780 + }, + { + "epoch": 0.61, + "learning_rate": 0.0007035283132128327, + "loss": 0.1399, + "step": 13790 + }, + { + "epoch": 0.61, + "learning_rate": 0.0007033059872384891, + "loss": 0.135, + "step": 13800 + }, + { + "epoch": 0.61, + "learning_rate": 0.0007030836612641455, + "loss": 0.1411, + "step": 13810 + }, + { + "epoch": 0.61, + "learning_rate": 0.0007028613352898019, + "loss": 0.1377, + "step": 13820 + }, + { + "epoch": 0.61, + "learning_rate": 0.0007026390093154584, + "loss": 0.1378, + "step": 13830 + }, + { + "epoch": 0.61, + "learning_rate": 0.0007024166833411148, + "loss": 0.1359, + "step": 13840 + }, + { + "epoch": 0.61, + "learning_rate": 0.0007021943573667712, + "loss": 0.1361, + "step": 13850 + }, + { + "epoch": 0.61, + "learning_rate": 0.0007019720313924276, + "loss": 0.1361, + "step": 13860 + }, + { + "epoch": 0.61, + "learning_rate": 0.0007017497054180841, + "loss": 0.1332, + "step": 13870 + }, + { + "epoch": 0.61, + "learning_rate": 0.0007015273794437405, + "loss": 0.1413, + "step": 13880 + }, + { + "epoch": 0.61, + "learning_rate": 0.0007013050534693969, + "loss": 0.1415, + "step": 13890 + }, + { + "epoch": 0.61, + "learning_rate": 0.0007010827274950534, + "loss": 0.1434, + "step": 13900 + }, + { + "epoch": 0.61, + "learning_rate": 0.0007008604015207098, + "loss": 0.1363, + "step": 13910 + }, + { + "epoch": 0.61, + "learning_rate": 0.0007006380755463661, + "loss": 0.1363, + "step": 13920 + }, + { + "epoch": 0.61, + "learning_rate": 0.0007004157495720225, + "loss": 0.1401, + "step": 13930 + }, + { + "epoch": 0.61, + "learning_rate": 0.000700193423597679, + "loss": 0.138, + "step": 13940 + }, + { + "epoch": 0.61, + "learning_rate": 0.0006999710976233353, + "loss": 0.1386, + "step": 13950 + }, + { + "epoch": 0.61, + "learning_rate": 0.0006997487716489917, + "loss": 0.1366, + "step": 13960 + }, + { + "epoch": 0.61, + "learning_rate": 0.0006995264456746481, + "loss": 0.1365, + "step": 13970 + }, + { + "epoch": 0.62, + "learning_rate": 0.0006993041197003046, + "loss": 0.1324, + "step": 13980 + }, + { + "epoch": 0.62, + "learning_rate": 0.000699081793725961, + "loss": 0.1388, + "step": 13990 + }, + { + "epoch": 0.62, + "learning_rate": 0.0006988594677516174, + "loss": 0.1346, + "step": 14000 + }, + { + "epoch": 0.62, + "eval_cer": 0.5001183828065352, + "eval_loss": 0.1031935065984726, + "eval_runtime": 84.222, + "eval_samples_per_second": 118.663, + "eval_steps_per_second": 7.421, + "step": 14000 + }, + { + "epoch": 0.62, + "learning_rate": 0.0006986371417772738, + "loss": 0.1375, + "step": 14010 + }, + { + "epoch": 0.62, + "learning_rate": 0.0006984148158029303, + "loss": 0.1338, + "step": 14020 + }, + { + "epoch": 0.62, + "learning_rate": 0.0006981924898285867, + "loss": 0.1358, + "step": 14030 + }, + { + "epoch": 0.62, + "learning_rate": 0.000697970163854243, + "loss": 0.1358, + "step": 14040 + }, + { + "epoch": 0.62, + "learning_rate": 0.0006977478378798994, + "loss": 0.1368, + "step": 14050 + }, + { + "epoch": 0.62, + "learning_rate": 0.0006975255119055559, + "loss": 0.1342, + "step": 14060 + }, + { + "epoch": 0.62, + "learning_rate": 0.0006973031859312123, + "loss": 0.1359, + "step": 14070 + }, + { + "epoch": 0.62, + "learning_rate": 0.0006970808599568687, + "loss": 0.1408, + "step": 14080 + }, + { + "epoch": 0.62, + "learning_rate": 0.0006968585339825252, + "loss": 0.1355, + "step": 14090 + }, + { + "epoch": 0.62, + "learning_rate": 0.0006966362080081816, + "loss": 0.1334, + "step": 14100 + }, + { + "epoch": 0.62, + "learning_rate": 0.000696413882033838, + "loss": 0.1329, + "step": 14110 + }, + { + "epoch": 0.62, + "learning_rate": 0.0006961915560594944, + "loss": 0.133, + "step": 14120 + }, + { + "epoch": 0.62, + "learning_rate": 0.0006959692300851509, + "loss": 0.135, + "step": 14130 + }, + { + "epoch": 0.62, + "learning_rate": 0.0006957469041108073, + "loss": 0.1382, + "step": 14140 + }, + { + "epoch": 0.62, + "learning_rate": 0.0006955245781364637, + "loss": 0.1366, + "step": 14150 + }, + { + "epoch": 0.62, + "learning_rate": 0.00069530225216212, + "loss": 0.132, + "step": 14160 + }, + { + "epoch": 0.62, + "learning_rate": 0.0006950799261877765, + "loss": 0.1375, + "step": 14170 + }, + { + "epoch": 0.62, + "learning_rate": 0.0006948576002134329, + "loss": 0.1381, + "step": 14180 + }, + { + "epoch": 0.62, + "learning_rate": 0.0006946352742390893, + "loss": 0.1376, + "step": 14190 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006944129482647457, + "loss": 0.1355, + "step": 14200 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006941906222904022, + "loss": 0.1402, + "step": 14210 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006939682963160586, + "loss": 0.1387, + "step": 14220 + }, + { + "epoch": 0.63, + "learning_rate": 0.000693745970341715, + "loss": 0.1361, + "step": 14230 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006935236443673715, + "loss": 0.1421, + "step": 14240 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006933013183930279, + "loss": 0.1375, + "step": 14250 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006930789924186843, + "loss": 0.1331, + "step": 14260 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006928566664443407, + "loss": 0.1329, + "step": 14270 + }, + { + "epoch": 0.63, + "learning_rate": 0.000692634340469997, + "loss": 0.1409, + "step": 14280 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006924120144956535, + "loss": 0.1342, + "step": 14290 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006921896885213099, + "loss": 0.1319, + "step": 14300 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006919673625469663, + "loss": 0.133, + "step": 14310 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006917450365726228, + "loss": 0.1403, + "step": 14320 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006915227105982792, + "loss": 0.1369, + "step": 14330 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006913003846239356, + "loss": 0.14, + "step": 14340 + }, + { + "epoch": 0.63, + "learning_rate": 0.000691078058649592, + "loss": 0.1313, + "step": 14350 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006908557326752485, + "loss": 0.1336, + "step": 14360 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006906334067009049, + "loss": 0.1319, + "step": 14370 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006904110807265613, + "loss": 0.1369, + "step": 14380 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006901887547522178, + "loss": 0.1367, + "step": 14390 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006899664287778742, + "loss": 0.1351, + "step": 14400 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006897441028035305, + "loss": 0.1332, + "step": 14410 + }, + { + "epoch": 0.63, + "learning_rate": 0.0006895217768291869, + "loss": 0.1347, + "step": 14420 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006892994508548434, + "loss": 0.1322, + "step": 14430 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006890771248804998, + "loss": 0.1367, + "step": 14440 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006888547989061562, + "loss": 0.1355, + "step": 14450 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006886324729318126, + "loss": 0.1394, + "step": 14460 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006884101469574691, + "loss": 0.1313, + "step": 14470 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006881878209831255, + "loss": 0.1355, + "step": 14480 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006879654950087819, + "loss": 0.1351, + "step": 14490 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006877431690344383, + "loss": 0.1325, + "step": 14500 + }, + { + "epoch": 0.64, + "eval_cer": 0.500228657201664, + "eval_loss": 0.10184630006551743, + "eval_runtime": 83.8541, + "eval_samples_per_second": 119.183, + "eval_steps_per_second": 7.453, + "step": 14500 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006875208430600948, + "loss": 0.137, + "step": 14510 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006872985170857512, + "loss": 0.1337, + "step": 14520 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006870761911114075, + "loss": 0.1348, + "step": 14530 + }, + { + "epoch": 0.64, + "learning_rate": 0.000686853865137064, + "loss": 0.1343, + "step": 14540 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006866315391627204, + "loss": 0.1317, + "step": 14550 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006864092131883768, + "loss": 0.1329, + "step": 14560 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006861868872140332, + "loss": 0.1396, + "step": 14570 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006859645612396897, + "loss": 0.1409, + "step": 14580 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006857422352653461, + "loss": 0.1346, + "step": 14590 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006855199092910025, + "loss": 0.1403, + "step": 14600 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006852975833166589, + "loss": 0.1262, + "step": 14610 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006850752573423154, + "loss": 0.139, + "step": 14620 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006848529313679718, + "loss": 0.1326, + "step": 14630 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006846306053936282, + "loss": 0.1462, + "step": 14640 + }, + { + "epoch": 0.64, + "learning_rate": 0.0006844082794192845, + "loss": 0.1305, + "step": 14650 + }, + { + "epoch": 0.65, + "learning_rate": 0.000684185953444941, + "loss": 0.1357, + "step": 14660 + }, + { + "epoch": 0.65, + "learning_rate": 0.0006839636274705974, + "loss": 0.1367, + "step": 14670 + }, + { + "epoch": 0.65, + "learning_rate": 0.0006837413014962538, + "loss": 0.1415, + "step": 14680 + }, + { + "epoch": 0.65, + "learning_rate": 0.0006835189755219103, + "loss": 0.1285, + "step": 14690 + }, + { + "epoch": 0.65, + "learning_rate": 0.0006832966495475667, + "loss": 0.1267, + "step": 14700 + }, + { + "epoch": 0.65, + "learning_rate": 0.0006830743235732231, + "loss": 0.1339, + "step": 14710 + }, + { + "epoch": 0.65, + "learning_rate": 0.0006828519975988795, + "loss": 0.1281, + "step": 14720 + }, + { + "epoch": 0.65, + "learning_rate": 0.000682629671624536, + "loss": 0.1346, + "step": 14730 + }, + { + "epoch": 0.65, + "learning_rate": 0.0006824073456501924, + "loss": 0.1334, + "step": 14740 + }, + { + "epoch": 0.65, + "learning_rate": 0.0006821850196758488, + "loss": 0.1353, + "step": 14750 + }, + { + "epoch": 0.65, + "learning_rate": 0.0006819626937015052, + "loss": 0.1391, + "step": 14760 + }, + { + "epoch": 0.65, + "learning_rate": 0.0006817403677271616, + "loss": 0.138, + "step": 14770 + }, + { + "epoch": 0.65, + "learning_rate": 0.000681518041752818, + "loss": 0.1382, + "step": 14780 + }, + { + "epoch": 0.65, + "learning_rate": 0.0006812957157784744, + "loss": 0.1335, + "step": 14790 + }, + { + "epoch": 0.65, + "learning_rate": 0.0006810733898041308, + "loss": 0.1383, + "step": 14800 + }, + { + "epoch": 0.65, + "learning_rate": 0.0006808510638297873, + "loss": 0.1403, + "step": 14810 + }, + { + "epoch": 0.65, + "learning_rate": 0.0006806287378554437, + "loss": 0.1331, + "step": 14820 + }, + { + "epoch": 0.65, + "learning_rate": 0.0006804064118811001, + "loss": 0.1362, + "step": 14830 + }, + { + "epoch": 0.65, + "learning_rate": 0.0006801840859067566, + "loss": 0.1368, + "step": 14840 + }, + { + "epoch": 0.65, + "learning_rate": 0.000679961759932413, + "loss": 0.1361, + "step": 14850 + }, + { + "epoch": 0.65, + "learning_rate": 0.0006797394339580694, + "loss": 0.1304, + "step": 14860 + }, + { + "epoch": 0.65, + "learning_rate": 0.0006795171079837258, + "loss": 0.1297, + "step": 14870 + }, + { + "epoch": 0.65, + "learning_rate": 0.0006792947820093823, + "loss": 0.1347, + "step": 14880 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006790724560350386, + "loss": 0.134, + "step": 14890 + }, + { + "epoch": 0.66, + "learning_rate": 0.000678850130060695, + "loss": 0.1363, + "step": 14900 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006786278040863514, + "loss": 0.1421, + "step": 14910 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006784054781120079, + "loss": 0.1383, + "step": 14920 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006781831521376643, + "loss": 0.1347, + "step": 14930 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006779608261633207, + "loss": 0.1347, + "step": 14940 + }, + { + "epoch": 0.66, + "learning_rate": 0.000677738500188977, + "loss": 0.1363, + "step": 14950 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006775161742146335, + "loss": 0.1272, + "step": 14960 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006772938482402899, + "loss": 0.1363, + "step": 14970 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006770715222659463, + "loss": 0.1419, + "step": 14980 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006768491962916027, + "loss": 0.1311, + "step": 14990 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006766268703172592, + "loss": 0.1339, + "step": 15000 + }, + { + "epoch": 0.66, + "eval_cer": 0.5001222748440104, + "eval_loss": 0.1005987599492073, + "eval_runtime": 83.7417, + "eval_samples_per_second": 119.343, + "eval_steps_per_second": 7.463, + "step": 15000 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006764045443429156, + "loss": 0.1273, + "step": 15010 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006761822183685719, + "loss": 0.1379, + "step": 15020 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006759598923942283, + "loss": 0.1339, + "step": 15030 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006757375664198848, + "loss": 0.1345, + "step": 15040 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006755152404455412, + "loss": 0.1341, + "step": 15050 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006752929144711976, + "loss": 0.1309, + "step": 15060 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006750705884968541, + "loss": 0.134, + "step": 15070 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006748482625225105, + "loss": 0.1336, + "step": 15080 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006746259365481669, + "loss": 0.1386, + "step": 15090 + }, + { + "epoch": 0.66, + "learning_rate": 0.0006744036105738233, + "loss": 0.1344, + "step": 15100 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006741812845994798, + "loss": 0.1286, + "step": 15110 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006739589586251362, + "loss": 0.1299, + "step": 15120 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006737366326507926, + "loss": 0.1353, + "step": 15130 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006735143066764489, + "loss": 0.1346, + "step": 15140 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006732919807021054, + "loss": 0.1335, + "step": 15150 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006730696547277618, + "loss": 0.1357, + "step": 15160 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006728473287534182, + "loss": 0.1349, + "step": 15170 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006726250027790747, + "loss": 0.1366, + "step": 15180 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006724026768047311, + "loss": 0.138, + "step": 15190 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006721803508303875, + "loss": 0.1336, + "step": 15200 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006719580248560439, + "loss": 0.1337, + "step": 15210 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006717356988817004, + "loss": 0.135, + "step": 15220 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006715133729073568, + "loss": 0.1352, + "step": 15230 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006712910469330132, + "loss": 0.1354, + "step": 15240 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006710687209586696, + "loss": 0.137, + "step": 15250 + }, + { + "epoch": 0.67, + "learning_rate": 0.000670846394984326, + "loss": 0.1307, + "step": 15260 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006706240690099824, + "loss": 0.136, + "step": 15270 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006704017430356388, + "loss": 0.1414, + "step": 15280 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006701794170612952, + "loss": 0.133, + "step": 15290 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006699570910869517, + "loss": 0.1353, + "step": 15300 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006697347651126081, + "loss": 0.1317, + "step": 15310 + }, + { + "epoch": 0.67, + "learning_rate": 0.0006695124391382645, + "loss": 0.132, + "step": 15320 + }, + { + "epoch": 0.67, + "learning_rate": 0.000669290113163921, + "loss": 0.1308, + "step": 15330 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006690677871895774, + "loss": 0.1284, + "step": 15340 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006688454612152338, + "loss": 0.1391, + "step": 15350 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006686231352408902, + "loss": 0.1279, + "step": 15360 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006684008092665467, + "loss": 0.1315, + "step": 15370 + }, + { + "epoch": 0.68, + "learning_rate": 0.000668178483292203, + "loss": 0.1335, + "step": 15380 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006679561573178594, + "loss": 0.1382, + "step": 15390 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006677338313435158, + "loss": 0.132, + "step": 15400 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006675115053691723, + "loss": 0.1303, + "step": 15410 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006672891793948287, + "loss": 0.1348, + "step": 15420 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006670668534204851, + "loss": 0.1407, + "step": 15430 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006668445274461415, + "loss": 0.135, + "step": 15440 + }, + { + "epoch": 0.68, + "learning_rate": 0.000666622201471798, + "loss": 0.132, + "step": 15450 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006663998754974544, + "loss": 0.132, + "step": 15460 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006661775495231108, + "loss": 0.1286, + "step": 15470 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006659552235487673, + "loss": 0.1324, + "step": 15480 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006657328975744237, + "loss": 0.1292, + "step": 15490 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006655105716000801, + "loss": 0.1351, + "step": 15500 + }, + { + "epoch": 0.68, + "eval_cer": 0.5004128803088203, + "eval_loss": 0.10005035996437073, + "eval_runtime": 83.6691, + "eval_samples_per_second": 119.447, + "eval_steps_per_second": 7.47, + "step": 15500 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006652882456257364, + "loss": 0.129, + "step": 15510 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006650659196513929, + "loss": 0.1344, + "step": 15520 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006648435936770493, + "loss": 0.1316, + "step": 15530 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006646212677027057, + "loss": 0.1314, + "step": 15540 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006643989417283621, + "loss": 0.1306, + "step": 15550 + }, + { + "epoch": 0.68, + "learning_rate": 0.0006641766157540186, + "loss": 0.1287, + "step": 15560 + }, + { + "epoch": 0.69, + "learning_rate": 0.000663954289779675, + "loss": 0.1305, + "step": 15570 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006637319638053314, + "loss": 0.1319, + "step": 15580 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006635096378309878, + "loss": 0.1285, + "step": 15590 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006632873118566443, + "loss": 0.1344, + "step": 15600 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006630649858823007, + "loss": 0.1373, + "step": 15610 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006628426599079571, + "loss": 0.1338, + "step": 15620 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006626203339336134, + "loss": 0.1302, + "step": 15630 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006623980079592699, + "loss": 0.134, + "step": 15640 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006621756819849263, + "loss": 0.1329, + "step": 15650 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006619533560105827, + "loss": 0.1259, + "step": 15660 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006617310300362392, + "loss": 0.1293, + "step": 15670 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006615087040618956, + "loss": 0.1288, + "step": 15680 + }, + { + "epoch": 0.69, + "learning_rate": 0.000661286378087552, + "loss": 0.1269, + "step": 15690 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006610640521132084, + "loss": 0.1286, + "step": 15700 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006608417261388649, + "loss": 0.1295, + "step": 15710 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006606194001645213, + "loss": 0.1344, + "step": 15720 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006603970741901777, + "loss": 0.1283, + "step": 15730 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006601747482158342, + "loss": 0.1333, + "step": 15740 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006599524222414905, + "loss": 0.1316, + "step": 15750 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006597300962671469, + "loss": 0.1299, + "step": 15760 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006595077702928033, + "loss": 0.1309, + "step": 15770 + }, + { + "epoch": 0.69, + "learning_rate": 0.0006592854443184598, + "loss": 0.1276, + "step": 15780 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006590631183441162, + "loss": 0.1316, + "step": 15790 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006588407923697726, + "loss": 0.1316, + "step": 15800 + }, + { + "epoch": 0.7, + "learning_rate": 0.000658618466395429, + "loss": 0.1331, + "step": 15810 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006583961404210855, + "loss": 0.1276, + "step": 15820 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006581738144467419, + "loss": 0.1358, + "step": 15830 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006579514884723983, + "loss": 0.1283, + "step": 15840 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006577291624980547, + "loss": 0.1324, + "step": 15850 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006575068365237112, + "loss": 0.1323, + "step": 15860 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006572845105493675, + "loss": 0.1406, + "step": 15870 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006570621845750239, + "loss": 0.1313, + "step": 15880 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006568398586006803, + "loss": 0.13, + "step": 15890 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006566175326263368, + "loss": 0.1319, + "step": 15900 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006563952066519932, + "loss": 0.1268, + "step": 15910 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006561728806776496, + "loss": 0.1277, + "step": 15920 + }, + { + "epoch": 0.7, + "learning_rate": 0.000655950554703306, + "loss": 0.1318, + "step": 15930 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006557282287289625, + "loss": 0.1316, + "step": 15940 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006555059027546189, + "loss": 0.1333, + "step": 15950 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006552835767802752, + "loss": 0.1382, + "step": 15960 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006550612508059317, + "loss": 0.1281, + "step": 15970 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006548389248315881, + "loss": 0.1362, + "step": 15980 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006546165988572445, + "loss": 0.1313, + "step": 15990 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006543942728829008, + "loss": 0.1305, + "step": 16000 + }, + { + "epoch": 0.7, + "eval_cer": 0.5004258537670706, + "eval_loss": 0.09878270328044891, + "eval_runtime": 83.8163, + "eval_samples_per_second": 119.237, + "eval_steps_per_second": 7.457, + "step": 16000 + }, + { + "epoch": 0.7, + "learning_rate": 0.0006541719469085573, + "loss": 0.1356, + "step": 16010 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006539496209342137, + "loss": 0.1321, + "step": 16020 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006537272949598701, + "loss": 0.1322, + "step": 16030 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006535049689855265, + "loss": 0.1295, + "step": 16040 + }, + { + "epoch": 0.71, + "learning_rate": 0.000653282643011183, + "loss": 0.1315, + "step": 16050 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006530603170368394, + "loss": 0.1327, + "step": 16060 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006528379910624958, + "loss": 0.1295, + "step": 16070 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006526156650881522, + "loss": 0.1348, + "step": 16080 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006523933391138087, + "loss": 0.1378, + "step": 16090 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006521710131394651, + "loss": 0.13, + "step": 16100 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006519486871651215, + "loss": 0.1288, + "step": 16110 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006517263611907778, + "loss": 0.1322, + "step": 16120 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006515040352164343, + "loss": 0.1265, + "step": 16130 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006512817092420907, + "loss": 0.13, + "step": 16140 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006510593832677471, + "loss": 0.1302, + "step": 16150 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006508370572934036, + "loss": 0.1272, + "step": 16160 + }, + { + "epoch": 0.71, + "learning_rate": 0.00065061473131906, + "loss": 0.1325, + "step": 16170 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006503924053447164, + "loss": 0.1281, + "step": 16180 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006501700793703728, + "loss": 0.1326, + "step": 16190 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006499477533960293, + "loss": 0.1272, + "step": 16200 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006497254274216857, + "loss": 0.1346, + "step": 16210 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006495031014473421, + "loss": 0.1336, + "step": 16220 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006492807754729986, + "loss": 0.1273, + "step": 16230 + }, + { + "epoch": 0.71, + "learning_rate": 0.0006490584494986549, + "loss": 0.1345, + "step": 16240 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006488361235243113, + "loss": 0.1285, + "step": 16250 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006486137975499677, + "loss": 0.1342, + "step": 16260 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006483914715756242, + "loss": 0.1341, + "step": 16270 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006481691456012806, + "loss": 0.1386, + "step": 16280 + }, + { + "epoch": 0.72, + "learning_rate": 0.000647946819626937, + "loss": 0.1315, + "step": 16290 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006477244936525934, + "loss": 0.1276, + "step": 16300 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006475021676782499, + "loss": 0.1216, + "step": 16310 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006472798417039063, + "loss": 0.1354, + "step": 16320 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006470575157295627, + "loss": 0.1301, + "step": 16330 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006468351897552191, + "loss": 0.1323, + "step": 16340 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006466128637808756, + "loss": 0.1302, + "step": 16350 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006463905378065319, + "loss": 0.1307, + "step": 16360 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006461682118321883, + "loss": 0.1287, + "step": 16370 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006459458858578447, + "loss": 0.1309, + "step": 16380 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006457235598835012, + "loss": 0.1328, + "step": 16390 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006455012339091576, + "loss": 0.1363, + "step": 16400 + }, + { + "epoch": 0.72, + "learning_rate": 0.000645278907934814, + "loss": 0.1268, + "step": 16410 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006450565819604705, + "loss": 0.1256, + "step": 16420 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006448342559861269, + "loss": 0.1368, + "step": 16430 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006446119300117833, + "loss": 0.1312, + "step": 16440 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006443896040374397, + "loss": 0.1276, + "step": 16450 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006441672780630962, + "loss": 0.1344, + "step": 16460 + }, + { + "epoch": 0.72, + "learning_rate": 0.0006439449520887526, + "loss": 0.13, + "step": 16470 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006437226261144089, + "loss": 0.123, + "step": 16480 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006435003001400653, + "loss": 0.1355, + "step": 16490 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006432779741657218, + "loss": 0.1275, + "step": 16500 + }, + { + "epoch": 0.73, + "eval_cer": 0.5000048650468439, + "eval_loss": 0.09856224805116653, + "eval_runtime": 83.4898, + "eval_samples_per_second": 119.703, + "eval_steps_per_second": 7.486, + "step": 16500 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006430556481913782, + "loss": 0.1298, + "step": 16510 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006428333222170346, + "loss": 0.1321, + "step": 16520 + }, + { + "epoch": 0.73, + "learning_rate": 0.000642610996242691, + "loss": 0.1285, + "step": 16530 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006423886702683475, + "loss": 0.1293, + "step": 16540 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006421663442940039, + "loss": 0.1271, + "step": 16550 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006419440183196603, + "loss": 0.1307, + "step": 16560 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006417216923453168, + "loss": 0.126, + "step": 16570 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006414993663709732, + "loss": 0.1315, + "step": 16580 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006412770403966296, + "loss": 0.1305, + "step": 16590 + }, + { + "epoch": 0.73, + "learning_rate": 0.000641054714422286, + "loss": 0.1289, + "step": 16600 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006408323884479424, + "loss": 0.129, + "step": 16610 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006406100624735988, + "loss": 0.1307, + "step": 16620 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006403877364992552, + "loss": 0.1244, + "step": 16630 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006401654105249116, + "loss": 0.1317, + "step": 16640 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006399430845505681, + "loss": 0.1263, + "step": 16650 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006397207585762245, + "loss": 0.1301, + "step": 16660 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006394984326018809, + "loss": 0.1321, + "step": 16670 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006392761066275373, + "loss": 0.1314, + "step": 16680 + }, + { + "epoch": 0.73, + "learning_rate": 0.0006390537806531938, + "loss": 0.1323, + "step": 16690 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006388314546788502, + "loss": 0.1346, + "step": 16700 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006386091287045066, + "loss": 0.1281, + "step": 16710 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006383868027301631, + "loss": 0.1318, + "step": 16720 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006381644767558194, + "loss": 0.1292, + "step": 16730 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006379421507814758, + "loss": 0.1307, + "step": 16740 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006377198248071322, + "loss": 0.1288, + "step": 16750 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006374974988327887, + "loss": 0.1291, + "step": 16760 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006372751728584451, + "loss": 0.1272, + "step": 16770 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006370528468841015, + "loss": 0.1296, + "step": 16780 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006368305209097579, + "loss": 0.1352, + "step": 16790 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006366081949354144, + "loss": 0.1323, + "step": 16800 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006363858689610708, + "loss": 0.1279, + "step": 16810 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006361635429867272, + "loss": 0.1331, + "step": 16820 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006359412170123837, + "loss": 0.1205, + "step": 16830 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006357188910380401, + "loss": 0.1286, + "step": 16840 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006354965650636964, + "loss": 0.128, + "step": 16850 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006352742390893528, + "loss": 0.1308, + "step": 16860 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006350519131150093, + "loss": 0.1263, + "step": 16870 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006348295871406657, + "loss": 0.1259, + "step": 16880 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006346072611663221, + "loss": 0.1327, + "step": 16890 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006343849351919785, + "loss": 0.1288, + "step": 16900 + }, + { + "epoch": 0.74, + "learning_rate": 0.000634162609217635, + "loss": 0.1279, + "step": 16910 + }, + { + "epoch": 0.74, + "learning_rate": 0.0006339402832432914, + "loss": 0.1267, + "step": 16920 + }, + { + "epoch": 0.75, + "learning_rate": 0.0006337179572689478, + "loss": 0.124, + "step": 16930 + }, + { + "epoch": 0.75, + "learning_rate": 0.0006334956312946042, + "loss": 0.1248, + "step": 16940 + }, + { + "epoch": 0.75, + "learning_rate": 0.0006332733053202607, + "loss": 0.1283, + "step": 16950 + }, + { + "epoch": 0.75, + "learning_rate": 0.000633050979345917, + "loss": 0.1245, + "step": 16960 + }, + { + "epoch": 0.75, + "learning_rate": 0.0006328286533715733, + "loss": 0.1273, + "step": 16970 + }, + { + "epoch": 0.75, + "learning_rate": 0.0006326063273972297, + "loss": 0.126, + "step": 16980 + }, + { + "epoch": 0.75, + "learning_rate": 0.0006323840014228862, + "loss": 0.1225, + "step": 16990 + }, + { + "epoch": 0.75, + "learning_rate": 0.0006321616754485426, + "loss": 0.1279, + "step": 17000 + }, + { + "epoch": 0.75, + "eval_cer": 0.49993480837229154, + "eval_loss": 0.09698430448770523, + "eval_runtime": 83.2871, + "eval_samples_per_second": 119.995, + "eval_steps_per_second": 7.504, + "step": 17000 + }, + { + "epoch": 0.75, + "learning_rate": 0.000631939349474199, + "loss": 0.1307, + "step": 17010 + }, + { + "epoch": 0.75, + "learning_rate": 0.0006317170234998554, + "loss": 0.1261, + "step": 17020 + }, + { + "epoch": 0.75, + "learning_rate": 0.0006314946975255119, + "loss": 0.1298, + "step": 17030 + }, + { + "epoch": 0.75, + "learning_rate": 0.0006312723715511683, + "loss": 0.1357, + "step": 17040 + }, + { + "epoch": 0.75, + "learning_rate": 0.0006310500455768247, + "loss": 0.1259, + "step": 17050 + }, + { + "epoch": 0.75, + "learning_rate": 0.0006308277196024812, + "loss": 0.1287, + "step": 17060 + }, + { + "epoch": 0.75, + "learning_rate": 0.0006306053936281376, + "loss": 0.1291, + "step": 17070 + }, + { + "epoch": 0.75, + "learning_rate": 0.000630383067653794, + "loss": 0.1294, + "step": 17080 + }, + { + "epoch": 0.75, + "learning_rate": 0.0006301607416794504, + "loss": 0.1271, + "step": 17090 + }, + { + "epoch": 0.75, + "learning_rate": 0.0006299384157051068, + "loss": 0.1259, + "step": 17100 + }, + { + "epoch": 0.75, + "learning_rate": 0.0006297160897307632, + "loss": 0.1291, + "step": 17110 + }, + { + "epoch": 0.75, + "learning_rate": 0.0006294937637564196, + "loss": 0.1273, + "step": 17120 + }, + { + "epoch": 0.75, + "learning_rate": 0.000629271437782076, + "loss": 0.1264, + "step": 17130 + }, + { + "epoch": 0.75, + "learning_rate": 0.0006290491118077325, + "loss": 0.1328, + "step": 17140 + }, + { + "epoch": 0.75, + "learning_rate": 0.0006288267858333889, + "loss": 0.1267, + "step": 17150 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006286044598590453, + "loss": 0.1321, + "step": 17160 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006283821338847017, + "loss": 0.1227, + "step": 17170 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006281598079103582, + "loss": 0.1241, + "step": 17180 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006279374819360146, + "loss": 0.1287, + "step": 17190 + }, + { + "epoch": 0.76, + "learning_rate": 0.000627715155961671, + "loss": 0.1281, + "step": 17200 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006274928299873275, + "loss": 0.1348, + "step": 17210 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006272705040129838, + "loss": 0.1218, + "step": 17220 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006270481780386402, + "loss": 0.1302, + "step": 17230 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006268258520642966, + "loss": 0.1304, + "step": 17240 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006266035260899531, + "loss": 0.1258, + "step": 17250 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006263812001156095, + "loss": 0.1256, + "step": 17260 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006261588741412659, + "loss": 0.1316, + "step": 17270 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006259365481669223, + "loss": 0.1293, + "step": 17280 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006257142221925788, + "loss": 0.1216, + "step": 17290 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006254918962182352, + "loss": 0.1272, + "step": 17300 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006252695702438916, + "loss": 0.1262, + "step": 17310 + }, + { + "epoch": 0.76, + "learning_rate": 0.000625047244269548, + "loss": 0.1327, + "step": 17320 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006248249182952045, + "loss": 0.1288, + "step": 17330 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006246025923208608, + "loss": 0.1234, + "step": 17340 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006243802663465172, + "loss": 0.1322, + "step": 17350 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006241579403721737, + "loss": 0.1306, + "step": 17360 + }, + { + "epoch": 0.76, + "learning_rate": 0.0006239356143978301, + "loss": 0.1233, + "step": 17370 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006237132884234865, + "loss": 0.1232, + "step": 17380 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006234909624491429, + "loss": 0.129, + "step": 17390 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006232686364747994, + "loss": 0.1289, + "step": 17400 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006230463105004558, + "loss": 0.1254, + "step": 17410 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006228239845261122, + "loss": 0.1279, + "step": 17420 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006226016585517686, + "loss": 0.1241, + "step": 17430 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006223793325774251, + "loss": 0.1257, + "step": 17440 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006221570066030815, + "loss": 0.1315, + "step": 17450 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006219346806287378, + "loss": 0.1279, + "step": 17460 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006217123546543942, + "loss": 0.1304, + "step": 17470 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006214900286800507, + "loss": 0.1352, + "step": 17480 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006212677027057071, + "loss": 0.1193, + "step": 17490 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006210453767313635, + "loss": 0.1315, + "step": 17500 + }, + { + "epoch": 0.77, + "eval_cer": 0.4999257269515162, + "eval_loss": 0.09700983017683029, + "eval_runtime": 83.7959, + "eval_samples_per_second": 119.266, + "eval_steps_per_second": 7.459, + "step": 17500 + }, + { + "epoch": 0.77, + "learning_rate": 0.00062082305075702, + "loss": 0.1199, + "step": 17510 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006206007247826764, + "loss": 0.1249, + "step": 17520 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006203783988083328, + "loss": 0.1232, + "step": 17530 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006201560728339892, + "loss": 0.1317, + "step": 17540 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006199337468596457, + "loss": 0.1288, + "step": 17550 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006197114208853021, + "loss": 0.1318, + "step": 17560 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006194890949109585, + "loss": 0.1292, + "step": 17570 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006192667689366149, + "loss": 0.1218, + "step": 17580 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006190444429622713, + "loss": 0.1261, + "step": 17590 + }, + { + "epoch": 0.77, + "learning_rate": 0.0006188221169879277, + "loss": 0.1271, + "step": 17600 + }, + { + "epoch": 0.78, + "learning_rate": 0.0006185997910135841, + "loss": 0.1226, + "step": 17610 + }, + { + "epoch": 0.78, + "learning_rate": 0.0006183774650392405, + "loss": 0.1315, + "step": 17620 + }, + { + "epoch": 0.78, + "learning_rate": 0.000618155139064897, + "loss": 0.1305, + "step": 17630 + }, + { + "epoch": 0.78, + "learning_rate": 0.0006179328130905534, + "loss": 0.1328, + "step": 17640 + }, + { + "epoch": 0.78, + "learning_rate": 0.0006177104871162098, + "loss": 0.1299, + "step": 17650 + }, + { + "epoch": 0.78, + "learning_rate": 0.0006174881611418663, + "loss": 0.1231, + "step": 17660 + }, + { + "epoch": 0.78, + "learning_rate": 0.0006172658351675227, + "loss": 0.1265, + "step": 17670 + }, + { + "epoch": 0.78, + "learning_rate": 0.0006170435091931791, + "loss": 0.1224, + "step": 17680 + }, + { + "epoch": 0.78, + "learning_rate": 0.0006168211832188355, + "loss": 0.1251, + "step": 17690 + }, + { + "epoch": 0.78, + "learning_rate": 0.000616598857244492, + "loss": 0.1242, + "step": 17700 + }, + { + "epoch": 0.78, + "learning_rate": 0.0006163765312701483, + "loss": 0.1328, + "step": 17710 + }, + { + "epoch": 0.78, + "learning_rate": 0.0006161542052958047, + "loss": 0.1257, + "step": 17720 + }, + { + "epoch": 0.78, + "learning_rate": 0.0006159318793214611, + "loss": 0.1224, + "step": 17730 + }, + { + "epoch": 0.78, + "learning_rate": 0.0006157095533471176, + "loss": 0.126, + "step": 17740 + }, + { + "epoch": 0.78, + "learning_rate": 0.000615487227372774, + "loss": 0.1272, + "step": 17750 + }, + { + "epoch": 0.78, + "learning_rate": 0.0006152649013984304, + "loss": 0.1259, + "step": 17760 + }, + { + "epoch": 0.78, + "learning_rate": 0.0006150425754240868, + "loss": 0.1263, + "step": 17770 + }, + { + "epoch": 0.78, + "learning_rate": 0.0006148202494497433, + "loss": 0.1307, + "step": 17780 + }, + { + "epoch": 0.78, + "learning_rate": 0.0006145979234753997, + "loss": 0.1285, + "step": 17790 + }, + { + "epoch": 0.78, + "learning_rate": 0.0006143755975010561, + "loss": 0.1251, + "step": 17800 + }, + { + "epoch": 0.78, + "learning_rate": 0.0006141532715267126, + "loss": 0.1283, + "step": 17810 + }, + { + "epoch": 0.78, + "learning_rate": 0.000613930945552369, + "loss": 0.1244, + "step": 17820 + }, + { + "epoch": 0.78, + "learning_rate": 0.0006137086195780253, + "loss": 0.1248, + "step": 17830 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006134862936036817, + "loss": 0.1258, + "step": 17840 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006132639676293382, + "loss": 0.1259, + "step": 17850 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006130416416549946, + "loss": 0.1231, + "step": 17860 + }, + { + "epoch": 0.79, + "learning_rate": 0.000612819315680651, + "loss": 0.122, + "step": 17870 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006125969897063074, + "loss": 0.1243, + "step": 17880 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006123746637319639, + "loss": 0.1269, + "step": 17890 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006121523377576203, + "loss": 0.1238, + "step": 17900 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006119300117832767, + "loss": 0.1251, + "step": 17910 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006117076858089332, + "loss": 0.1269, + "step": 17920 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006114853598345896, + "loss": 0.1296, + "step": 17930 + }, + { + "epoch": 0.79, + "learning_rate": 0.000611263033860246, + "loss": 0.1229, + "step": 17940 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006110407078859022, + "loss": 0.1255, + "step": 17950 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006108183819115586, + "loss": 0.1317, + "step": 17960 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006105960559372151, + "loss": 0.1194, + "step": 17970 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006103737299628715, + "loss": 0.1281, + "step": 17980 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006101514039885279, + "loss": 0.1273, + "step": 17990 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006099290780141844, + "loss": 0.1215, + "step": 18000 + }, + { + "epoch": 0.79, + "eval_cer": 0.4999737287470429, + "eval_loss": 0.09640804678201675, + "eval_runtime": 83.7615, + "eval_samples_per_second": 119.315, + "eval_steps_per_second": 7.462, + "step": 18000 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006097067520398408, + "loss": 0.1288, + "step": 18010 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006094844260654972, + "loss": 0.1279, + "step": 18020 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006092621000911536, + "loss": 0.1289, + "step": 18030 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006090397741168101, + "loss": 0.1248, + "step": 18040 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006088174481424665, + "loss": 0.1235, + "step": 18050 + }, + { + "epoch": 0.79, + "learning_rate": 0.0006085951221681229, + "loss": 0.1225, + "step": 18060 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006083727961937792, + "loss": 0.1289, + "step": 18070 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006081504702194357, + "loss": 0.1277, + "step": 18080 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006079281442450921, + "loss": 0.1318, + "step": 18090 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006077058182707485, + "loss": 0.1235, + "step": 18100 + }, + { + "epoch": 0.8, + "learning_rate": 0.000607483492296405, + "loss": 0.1258, + "step": 18110 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006072611663220614, + "loss": 0.1269, + "step": 18120 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006070388403477178, + "loss": 0.1197, + "step": 18130 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006068165143733742, + "loss": 0.1215, + "step": 18140 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006065941883990307, + "loss": 0.123, + "step": 18150 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006063718624246871, + "loss": 0.1245, + "step": 18160 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006061495364503435, + "loss": 0.124, + "step": 18170 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006059272104759999, + "loss": 0.1231, + "step": 18180 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006057048845016564, + "loss": 0.125, + "step": 18190 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006054825585273127, + "loss": 0.1218, + "step": 18200 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006052602325529691, + "loss": 0.1282, + "step": 18210 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006050379065786255, + "loss": 0.124, + "step": 18220 + }, + { + "epoch": 0.8, + "learning_rate": 0.000604815580604282, + "loss": 0.1256, + "step": 18230 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006045932546299384, + "loss": 0.1228, + "step": 18240 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006043709286555948, + "loss": 0.1271, + "step": 18250 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006041486026812512, + "loss": 0.1266, + "step": 18260 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006039262767069077, + "loss": 0.12, + "step": 18270 + }, + { + "epoch": 0.8, + "learning_rate": 0.0006037039507325641, + "loss": 0.1207, + "step": 18280 + }, + { + "epoch": 0.81, + "learning_rate": 0.0006034816247582205, + "loss": 0.1215, + "step": 18290 + }, + { + "epoch": 0.81, + "learning_rate": 0.000603259298783877, + "loss": 0.124, + "step": 18300 + }, + { + "epoch": 0.81, + "learning_rate": 0.0006030369728095334, + "loss": 0.1292, + "step": 18310 + }, + { + "epoch": 0.81, + "learning_rate": 0.0006028146468351897, + "loss": 0.1247, + "step": 18320 + }, + { + "epoch": 0.81, + "learning_rate": 0.0006025923208608461, + "loss": 0.1248, + "step": 18330 + }, + { + "epoch": 0.81, + "learning_rate": 0.0006023699948865026, + "loss": 0.1265, + "step": 18340 + }, + { + "epoch": 0.81, + "learning_rate": 0.000602147668912159, + "loss": 0.1257, + "step": 18350 + }, + { + "epoch": 0.81, + "learning_rate": 0.0006019253429378154, + "loss": 0.1229, + "step": 18360 + }, + { + "epoch": 0.81, + "learning_rate": 0.0006017030169634718, + "loss": 0.1237, + "step": 18370 + }, + { + "epoch": 0.81, + "learning_rate": 0.0006014806909891283, + "loss": 0.127, + "step": 18380 + }, + { + "epoch": 0.81, + "learning_rate": 0.0006012583650147847, + "loss": 0.1236, + "step": 18390 + }, + { + "epoch": 0.81, + "learning_rate": 0.0006010360390404411, + "loss": 0.1221, + "step": 18400 + }, + { + "epoch": 0.81, + "learning_rate": 0.0006008137130660976, + "loss": 0.1233, + "step": 18410 + }, + { + "epoch": 0.81, + "learning_rate": 0.000600591387091754, + "loss": 0.1214, + "step": 18420 + }, + { + "epoch": 0.81, + "learning_rate": 0.0006003690611174104, + "loss": 0.1262, + "step": 18430 + }, + { + "epoch": 0.81, + "learning_rate": 0.0006001467351430667, + "loss": 0.1276, + "step": 18440 + }, + { + "epoch": 0.81, + "learning_rate": 0.0005999244091687232, + "loss": 0.128, + "step": 18450 + }, + { + "epoch": 0.81, + "learning_rate": 0.0005997020831943796, + "loss": 0.1256, + "step": 18460 + }, + { + "epoch": 0.81, + "learning_rate": 0.000599479757220036, + "loss": 0.1295, + "step": 18470 + }, + { + "epoch": 0.81, + "learning_rate": 0.0005992574312456924, + "loss": 0.1269, + "step": 18480 + }, + { + "epoch": 0.81, + "learning_rate": 0.0005990351052713489, + "loss": 0.1256, + "step": 18490 + }, + { + "epoch": 0.81, + "learning_rate": 0.0005988127792970053, + "loss": 0.1218, + "step": 18500 + }, + { + "epoch": 0.81, + "eval_cer": 0.4998647516977392, + "eval_loss": 0.09529462456703186, + "eval_runtime": 83.8384, + "eval_samples_per_second": 119.206, + "eval_steps_per_second": 7.455, + "step": 18500 + }, + { + "epoch": 0.81, + "learning_rate": 0.0005985904533226617, + "loss": 0.1261, + "step": 18510 + }, + { + "epoch": 0.82, + "learning_rate": 0.0005983681273483181, + "loss": 0.1235, + "step": 18520 + }, + { + "epoch": 0.82, + "learning_rate": 0.0005981458013739746, + "loss": 0.1257, + "step": 18530 + }, + { + "epoch": 0.82, + "learning_rate": 0.000597923475399631, + "loss": 0.123, + "step": 18540 + }, + { + "epoch": 0.82, + "learning_rate": 0.0005977011494252874, + "loss": 0.1239, + "step": 18550 + }, + { + "epoch": 0.82, + "learning_rate": 0.0005974788234509437, + "loss": 0.1184, + "step": 18560 + }, + { + "epoch": 0.82, + "learning_rate": 0.0005972564974766002, + "loss": 0.1247, + "step": 18570 + }, + { + "epoch": 0.82, + "learning_rate": 0.0005970341715022566, + "loss": 0.1222, + "step": 18580 + }, + { + "epoch": 0.82, + "learning_rate": 0.000596811845527913, + "loss": 0.1256, + "step": 18590 + }, + { + "epoch": 0.82, + "learning_rate": 0.0005965895195535695, + "loss": 0.1275, + "step": 18600 + }, + { + "epoch": 0.82, + "learning_rate": 0.0005963671935792259, + "loss": 0.1209, + "step": 18610 + }, + { + "epoch": 0.82, + "learning_rate": 0.0005961448676048823, + "loss": 0.1229, + "step": 18620 + }, + { + "epoch": 0.82, + "learning_rate": 0.0005959225416305387, + "loss": 0.1199, + "step": 18630 + }, + { + "epoch": 0.82, + "learning_rate": 0.0005957002156561952, + "loss": 0.1213, + "step": 18640 + }, + { + "epoch": 0.82, + "learning_rate": 0.0005954778896818516, + "loss": 0.1199, + "step": 18650 + }, + { + "epoch": 0.82, + "learning_rate": 0.000595255563707508, + "loss": 0.125, + "step": 18660 + }, + { + "epoch": 0.82, + "learning_rate": 0.0005950332377331644, + "loss": 0.1238, + "step": 18670 + }, + { + "epoch": 0.82, + "learning_rate": 0.0005948109117588209, + "loss": 0.1249, + "step": 18680 + }, + { + "epoch": 0.82, + "learning_rate": 0.0005945885857844772, + "loss": 0.1259, + "step": 18690 + }, + { + "epoch": 0.82, + "learning_rate": 0.0005943662598101336, + "loss": 0.1254, + "step": 18700 + }, + { + "epoch": 0.82, + "learning_rate": 0.00059414393383579, + "loss": 0.1264, + "step": 18710 + }, + { + "epoch": 0.82, + "learning_rate": 0.0005939216078614465, + "loss": 0.1239, + "step": 18720 + }, + { + "epoch": 0.82, + "learning_rate": 0.0005936992818871029, + "loss": 0.1263, + "step": 18730 + }, + { + "epoch": 0.82, + "learning_rate": 0.0005934769559127593, + "loss": 0.123, + "step": 18740 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005932546299384158, + "loss": 0.1204, + "step": 18750 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005930323039640722, + "loss": 0.1194, + "step": 18760 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005928099779897286, + "loss": 0.1211, + "step": 18770 + }, + { + "epoch": 0.83, + "learning_rate": 0.000592587652015385, + "loss": 0.1184, + "step": 18780 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005923653260410415, + "loss": 0.122, + "step": 18790 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005921430000666979, + "loss": 0.1257, + "step": 18800 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005919206740923542, + "loss": 0.121, + "step": 18810 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005916983481180106, + "loss": 0.1251, + "step": 18820 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005914760221436671, + "loss": 0.1259, + "step": 18830 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005912536961693235, + "loss": 0.1253, + "step": 18840 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005910313701949799, + "loss": 0.1275, + "step": 18850 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005908090442206363, + "loss": 0.123, + "step": 18860 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005905867182462928, + "loss": 0.1212, + "step": 18870 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005903643922719492, + "loss": 0.1239, + "step": 18880 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005901420662976056, + "loss": 0.1218, + "step": 18890 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005899197403232621, + "loss": 0.1257, + "step": 18900 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005896974143489185, + "loss": 0.1297, + "step": 18910 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005894750883745749, + "loss": 0.1207, + "step": 18920 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005892527624002312, + "loss": 0.1244, + "step": 18930 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005890304364258876, + "loss": 0.1199, + "step": 18940 + }, + { + "epoch": 0.83, + "learning_rate": 0.000588808110451544, + "loss": 0.1248, + "step": 18950 + }, + { + "epoch": 0.83, + "learning_rate": 0.0005885857844772004, + "loss": 0.1229, + "step": 18960 + }, + { + "epoch": 0.84, + "learning_rate": 0.0005883634585028568, + "loss": 0.1265, + "step": 18970 + }, + { + "epoch": 0.84, + "learning_rate": 0.0005881411325285133, + "loss": 0.1215, + "step": 18980 + }, + { + "epoch": 0.84, + "learning_rate": 0.0005879188065541697, + "loss": 0.1195, + "step": 18990 + }, + { + "epoch": 0.84, + "learning_rate": 0.0005876964805798261, + "loss": 0.124, + "step": 19000 + }, + { + "epoch": 0.84, + "eval_cer": 0.49991210482035325, + "eval_loss": 0.09477666765451431, + "eval_runtime": 83.7389, + "eval_samples_per_second": 119.347, + "eval_steps_per_second": 7.464, + "step": 19000 + }, + { + "epoch": 0.84, + "learning_rate": 0.0005874741546054825, + "loss": 0.1195, + "step": 19010 + }, + { + "epoch": 0.84, + "learning_rate": 0.000587251828631139, + "loss": 0.1265, + "step": 19020 + }, + { + "epoch": 0.84, + "learning_rate": 0.0005870295026567954, + "loss": 0.1209, + "step": 19030 + }, + { + "epoch": 0.84, + "learning_rate": 0.0005868071766824518, + "loss": 0.1266, + "step": 19040 + }, + { + "epoch": 0.84, + "learning_rate": 0.0005865848507081081, + "loss": 0.1163, + "step": 19050 + }, + { + "epoch": 0.84, + "learning_rate": 0.0005863625247337646, + "loss": 0.1236, + "step": 19060 + }, + { + "epoch": 0.84, + "learning_rate": 0.000586140198759421, + "loss": 0.1206, + "step": 19070 + }, + { + "epoch": 0.84, + "learning_rate": 0.0005859178727850774, + "loss": 0.1258, + "step": 19080 + }, + { + "epoch": 0.84, + "learning_rate": 0.0005856955468107339, + "loss": 0.1235, + "step": 19090 + }, + { + "epoch": 0.84, + "learning_rate": 0.0005854732208363903, + "loss": 0.1289, + "step": 19100 + }, + { + "epoch": 0.84, + "learning_rate": 0.0005852508948620467, + "loss": 0.1192, + "step": 19110 + }, + { + "epoch": 0.84, + "learning_rate": 0.0005850285688877031, + "loss": 0.125, + "step": 19120 + }, + { + "epoch": 0.84, + "learning_rate": 0.0005848062429133596, + "loss": 0.1243, + "step": 19130 + }, + { + "epoch": 0.84, + "learning_rate": 0.000584583916939016, + "loss": 0.1179, + "step": 19140 + }, + { + "epoch": 0.84, + "learning_rate": 0.0005843615909646724, + "loss": 0.13, + "step": 19150 + }, + { + "epoch": 0.84, + "learning_rate": 0.0005841392649903288, + "loss": 0.1221, + "step": 19160 + }, + { + "epoch": 0.84, + "learning_rate": 0.0005839169390159853, + "loss": 0.1219, + "step": 19170 + }, + { + "epoch": 0.84, + "learning_rate": 0.0005836946130416416, + "loss": 0.1214, + "step": 19180 + }, + { + "epoch": 0.84, + "learning_rate": 0.000583472287067298, + "loss": 0.1228, + "step": 19190 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005832499610929544, + "loss": 0.1231, + "step": 19200 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005830276351186109, + "loss": 0.1258, + "step": 19210 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005828053091442673, + "loss": 0.1228, + "step": 19220 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005825829831699237, + "loss": 0.1205, + "step": 19230 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005823606571955802, + "loss": 0.1224, + "step": 19240 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005821383312212366, + "loss": 0.1241, + "step": 19250 + }, + { + "epoch": 0.85, + "learning_rate": 0.000581916005246893, + "loss": 0.1248, + "step": 19260 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005816936792725494, + "loss": 0.12, + "step": 19270 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005814713532982059, + "loss": 0.1206, + "step": 19280 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005812490273238623, + "loss": 0.129, + "step": 19290 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005810267013495186, + "loss": 0.1235, + "step": 19300 + }, + { + "epoch": 0.85, + "learning_rate": 0.000580804375375175, + "loss": 0.122, + "step": 19310 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005805820494008315, + "loss": 0.1186, + "step": 19320 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005803597234264879, + "loss": 0.1225, + "step": 19330 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005801373974521443, + "loss": 0.1237, + "step": 19340 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005799150714778007, + "loss": 0.122, + "step": 19350 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005796927455034572, + "loss": 0.1232, + "step": 19360 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005794704195291136, + "loss": 0.1276, + "step": 19370 + }, + { + "epoch": 0.85, + "learning_rate": 0.00057924809355477, + "loss": 0.1206, + "step": 19380 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005790257675804265, + "loss": 0.1286, + "step": 19390 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005788034416060829, + "loss": 0.1222, + "step": 19400 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005785811156317393, + "loss": 0.1215, + "step": 19410 + }, + { + "epoch": 0.85, + "learning_rate": 0.0005783587896573956, + "loss": 0.125, + "step": 19420 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005781364636830521, + "loss": 0.1235, + "step": 19430 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005779141377087085, + "loss": 0.1197, + "step": 19440 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005776918117343649, + "loss": 0.1224, + "step": 19450 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005774694857600213, + "loss": 0.1258, + "step": 19460 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005772471597856778, + "loss": 0.1232, + "step": 19470 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005770248338113342, + "loss": 0.1198, + "step": 19480 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005768025078369906, + "loss": 0.1229, + "step": 19490 + }, + { + "epoch": 0.86, + "learning_rate": 0.000576580181862647, + "loss": 0.121, + "step": 19500 + }, + { + "epoch": 0.86, + "eval_cer": 0.5001611952187617, + "eval_loss": 0.09304533898830414, + "eval_runtime": 83.7671, + "eval_samples_per_second": 119.307, + "eval_steps_per_second": 7.461, + "step": 19500 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005763578558883035, + "loss": 0.1222, + "step": 19510 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005761355299139599, + "loss": 0.121, + "step": 19520 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005759132039396163, + "loss": 0.1236, + "step": 19530 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005756908779652727, + "loss": 0.1212, + "step": 19540 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005754685519909291, + "loss": 0.1191, + "step": 19550 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005752462260165855, + "loss": 0.1219, + "step": 19560 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005750239000422419, + "loss": 0.1235, + "step": 19570 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005748015740678984, + "loss": 0.1263, + "step": 19580 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005745792480935548, + "loss": 0.1181, + "step": 19590 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005743569221192112, + "loss": 0.125, + "step": 19600 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005741345961448676, + "loss": 0.1185, + "step": 19610 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005739122701705241, + "loss": 0.1249, + "step": 19620 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005736899441961805, + "loss": 0.1221, + "step": 19630 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005734676182218369, + "loss": 0.124, + "step": 19640 + }, + { + "epoch": 0.86, + "learning_rate": 0.0005732452922474934, + "loss": 0.1195, + "step": 19650 + }, + { + "epoch": 0.87, + "learning_rate": 0.0005730229662731498, + "loss": 0.1191, + "step": 19660 + }, + { + "epoch": 0.87, + "learning_rate": 0.0005728006402988061, + "loss": 0.1171, + "step": 19670 + }, + { + "epoch": 0.87, + "learning_rate": 0.0005725783143244625, + "loss": 0.1213, + "step": 19680 + }, + { + "epoch": 0.87, + "learning_rate": 0.000572355988350119, + "loss": 0.115, + "step": 19690 + }, + { + "epoch": 0.87, + "learning_rate": 0.0005721336623757754, + "loss": 0.1209, + "step": 19700 + }, + { + "epoch": 0.87, + "learning_rate": 0.0005719113364014318, + "loss": 0.1236, + "step": 19710 + }, + { + "epoch": 0.87, + "learning_rate": 0.0005716890104270882, + "loss": 0.1238, + "step": 19720 + }, + { + "epoch": 0.87, + "learning_rate": 0.0005714666844527447, + "loss": 0.1258, + "step": 19730 + }, + { + "epoch": 0.87, + "learning_rate": 0.0005712443584784011, + "loss": 0.1211, + "step": 19740 + }, + { + "epoch": 0.87, + "learning_rate": 0.0005710220325040575, + "loss": 0.1227, + "step": 19750 + }, + { + "epoch": 0.87, + "learning_rate": 0.000570799706529714, + "loss": 0.1226, + "step": 19760 + }, + { + "epoch": 0.87, + "learning_rate": 0.0005705773805553704, + "loss": 0.1234, + "step": 19770 + }, + { + "epoch": 0.87, + "learning_rate": 0.0005703550545810268, + "loss": 0.1204, + "step": 19780 + }, + { + "epoch": 0.87, + "learning_rate": 0.0005701327286066831, + "loss": 0.1178, + "step": 19790 + }, + { + "epoch": 0.87, + "learning_rate": 0.0005699104026323395, + "loss": 0.1236, + "step": 19800 + }, + { + "epoch": 0.87, + "learning_rate": 0.000569688076657996, + "loss": 0.1222, + "step": 19810 + }, + { + "epoch": 0.87, + "learning_rate": 0.0005694657506836524, + "loss": 0.1233, + "step": 19820 + }, + { + "epoch": 0.87, + "learning_rate": 0.0005692434247093088, + "loss": 0.1232, + "step": 19830 + }, + { + "epoch": 0.87, + "learning_rate": 0.0005690210987349653, + "loss": 0.1253, + "step": 19840 + }, + { + "epoch": 0.87, + "learning_rate": 0.0005687987727606217, + "loss": 0.1284, + "step": 19850 + }, + { + "epoch": 0.87, + "learning_rate": 0.0005685764467862781, + "loss": 0.1201, + "step": 19860 + }, + { + "epoch": 0.87, + "learning_rate": 0.0005683541208119345, + "loss": 0.119, + "step": 19870 + }, + { + "epoch": 0.88, + "learning_rate": 0.000568131794837591, + "loss": 0.1244, + "step": 19880 + }, + { + "epoch": 0.88, + "learning_rate": 0.0005679094688632474, + "loss": 0.1234, + "step": 19890 + }, + { + "epoch": 0.88, + "learning_rate": 0.0005676871428889038, + "loss": 0.1163, + "step": 19900 + }, + { + "epoch": 0.88, + "learning_rate": 0.0005674648169145601, + "loss": 0.123, + "step": 19910 + }, + { + "epoch": 0.88, + "learning_rate": 0.0005672424909402166, + "loss": 0.1199, + "step": 19920 + }, + { + "epoch": 0.88, + "learning_rate": 0.000567020164965873, + "loss": 0.1237, + "step": 19930 + }, + { + "epoch": 0.88, + "learning_rate": 0.0005667978389915294, + "loss": 0.1228, + "step": 19940 + }, + { + "epoch": 0.88, + "learning_rate": 0.0005665755130171857, + "loss": 0.1196, + "step": 19950 + }, + { + "epoch": 0.88, + "learning_rate": 0.0005663531870428422, + "loss": 0.1162, + "step": 19960 + }, + { + "epoch": 0.88, + "learning_rate": 0.0005661308610684986, + "loss": 0.1185, + "step": 19970 + }, + { + "epoch": 0.88, + "learning_rate": 0.000565908535094155, + "loss": 0.1225, + "step": 19980 + }, + { + "epoch": 0.88, + "learning_rate": 0.0005656862091198115, + "loss": 0.1185, + "step": 19990 + }, + { + "epoch": 0.88, + "learning_rate": 0.0005654638831454679, + "loss": 0.1242, + "step": 20000 + }, + { + "epoch": 0.88, + "eval_cer": 0.4995546860455537, + "eval_loss": 0.09247354418039322, + "eval_runtime": 83.7624, + "eval_samples_per_second": 119.314, + "eval_steps_per_second": 7.462, + "step": 20000 + }, + { + "epoch": 0.88, + "learning_rate": 0.0005652415571711243, + "loss": 0.1184, + "step": 20010 + }, + { + "epoch": 0.88, + "learning_rate": 0.0005650192311967807, + "loss": 0.1233, + "step": 20020 + }, + { + "epoch": 0.88, + "learning_rate": 0.000564796905222437, + "loss": 0.1187, + "step": 20030 + }, + { + "epoch": 0.88, + "learning_rate": 0.0005645745792480935, + "loss": 0.1254, + "step": 20040 + }, + { + "epoch": 0.88, + "learning_rate": 0.0005643522532737499, + "loss": 0.1165, + "step": 20050 + }, + { + "epoch": 0.88, + "learning_rate": 0.0005641299272994063, + "loss": 0.1198, + "step": 20060 + }, + { + "epoch": 0.88, + "learning_rate": 0.0005639076013250628, + "loss": 0.1206, + "step": 20070 + }, + { + "epoch": 0.88, + "learning_rate": 0.0005636852753507192, + "loss": 0.1228, + "step": 20080 + }, + { + "epoch": 0.88, + "learning_rate": 0.0005634629493763756, + "loss": 0.1233, + "step": 20090 + }, + { + "epoch": 0.88, + "learning_rate": 0.000563240623402032, + "loss": 0.122, + "step": 20100 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005630182974276885, + "loss": 0.1158, + "step": 20110 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005627959714533449, + "loss": 0.1239, + "step": 20120 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005625736454790013, + "loss": 0.122, + "step": 20130 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005623513195046578, + "loss": 0.1229, + "step": 20140 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005621289935303141, + "loss": 0.12, + "step": 20150 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005619066675559705, + "loss": 0.1197, + "step": 20160 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005616843415816269, + "loss": 0.1167, + "step": 20170 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005614620156072834, + "loss": 0.12, + "step": 20180 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005612396896329398, + "loss": 0.123, + "step": 20190 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005610173636585962, + "loss": 0.1287, + "step": 20200 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005607950376842526, + "loss": 0.1164, + "step": 20210 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005605727117099091, + "loss": 0.1214, + "step": 20220 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005603503857355655, + "loss": 0.1234, + "step": 20230 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005601280597612219, + "loss": 0.128, + "step": 20240 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005599057337868783, + "loss": 0.1219, + "step": 20250 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005596834078125348, + "loss": 0.1194, + "step": 20260 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005594610818381912, + "loss": 0.1259, + "step": 20270 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005592387558638475, + "loss": 0.1224, + "step": 20280 + }, + { + "epoch": 0.89, + "learning_rate": 0.000559016429889504, + "loss": 0.1242, + "step": 20290 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005587941039151604, + "loss": 0.1193, + "step": 20300 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005585717779408168, + "loss": 0.123, + "step": 20310 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005583494519664732, + "loss": 0.1153, + "step": 20320 + }, + { + "epoch": 0.89, + "learning_rate": 0.0005581271259921297, + "loss": 0.124, + "step": 20330 + }, + { + "epoch": 0.9, + "learning_rate": 0.0005579048000177861, + "loss": 0.1182, + "step": 20340 + }, + { + "epoch": 0.9, + "learning_rate": 0.0005576824740434425, + "loss": 0.1127, + "step": 20350 + }, + { + "epoch": 0.9, + "learning_rate": 0.0005574601480690989, + "loss": 0.1198, + "step": 20360 + }, + { + "epoch": 0.9, + "learning_rate": 0.0005572378220947554, + "loss": 0.1229, + "step": 20370 + }, + { + "epoch": 0.9, + "learning_rate": 0.0005570154961204118, + "loss": 0.1214, + "step": 20380 + }, + { + "epoch": 0.9, + "learning_rate": 0.0005567931701460682, + "loss": 0.1186, + "step": 20390 + }, + { + "epoch": 0.9, + "learning_rate": 0.0005565708441717245, + "loss": 0.1222, + "step": 20400 + }, + { + "epoch": 0.9, + "learning_rate": 0.000556348518197381, + "loss": 0.1185, + "step": 20410 + }, + { + "epoch": 0.9, + "learning_rate": 0.0005561261922230374, + "loss": 0.125, + "step": 20420 + }, + { + "epoch": 0.9, + "learning_rate": 0.0005559038662486938, + "loss": 0.1188, + "step": 20430 + }, + { + "epoch": 0.9, + "learning_rate": 0.0005556815402743502, + "loss": 0.1214, + "step": 20440 + }, + { + "epoch": 0.9, + "learning_rate": 0.0005554592143000067, + "loss": 0.1205, + "step": 20450 + }, + { + "epoch": 0.9, + "learning_rate": 0.0005552368883256631, + "loss": 0.1183, + "step": 20460 + }, + { + "epoch": 0.9, + "learning_rate": 0.0005550145623513195, + "loss": 0.124, + "step": 20470 + }, + { + "epoch": 0.9, + "learning_rate": 0.000554792236376976, + "loss": 0.1219, + "step": 20480 + }, + { + "epoch": 0.9, + "learning_rate": 0.0005545699104026324, + "loss": 0.1218, + "step": 20490 + }, + { + "epoch": 0.9, + "learning_rate": 0.0005543475844282888, + "loss": 0.1228, + "step": 20500 + }, + { + "epoch": 0.9, + "eval_cer": 0.49985826496861396, + "eval_loss": 0.0920482724905014, + "eval_runtime": 83.8037, + "eval_samples_per_second": 119.255, + "eval_steps_per_second": 7.458, + "step": 20500 + }, + { + "epoch": 0.9, + "learning_rate": 0.0005541252584539452, + "loss": 0.1201, + "step": 20510 + }, + { + "epoch": 0.9, + "learning_rate": 0.0005539029324796016, + "loss": 0.1256, + "step": 20520 + }, + { + "epoch": 0.9, + "learning_rate": 0.000553680606505258, + "loss": 0.1191, + "step": 20530 + }, + { + "epoch": 0.9, + "learning_rate": 0.0005534582805309144, + "loss": 0.1188, + "step": 20540 + }, + { + "epoch": 0.9, + "learning_rate": 0.0005532359545565708, + "loss": 0.1141, + "step": 20550 + }, + { + "epoch": 0.91, + "learning_rate": 0.0005530136285822273, + "loss": 0.1224, + "step": 20560 + }, + { + "epoch": 0.91, + "learning_rate": 0.0005527913026078837, + "loss": 0.1166, + "step": 20570 + }, + { + "epoch": 0.91, + "learning_rate": 0.0005525689766335401, + "loss": 0.1224, + "step": 20580 + }, + { + "epoch": 0.91, + "learning_rate": 0.0005523466506591966, + "loss": 0.1213, + "step": 20590 + }, + { + "epoch": 0.91, + "learning_rate": 0.000552124324684853, + "loss": 0.1162, + "step": 20600 + }, + { + "epoch": 0.91, + "learning_rate": 0.0005519019987105094, + "loss": 0.123, + "step": 20610 + }, + { + "epoch": 0.91, + "learning_rate": 0.0005516796727361658, + "loss": 0.1217, + "step": 20620 + }, + { + "epoch": 0.91, + "learning_rate": 0.0005514573467618223, + "loss": 0.1158, + "step": 20630 + }, + { + "epoch": 0.91, + "learning_rate": 0.0005512350207874786, + "loss": 0.1197, + "step": 20640 + }, + { + "epoch": 0.91, + "learning_rate": 0.000551012694813135, + "loss": 0.1211, + "step": 20650 + }, + { + "epoch": 0.91, + "learning_rate": 0.0005507903688387914, + "loss": 0.1202, + "step": 20660 + }, + { + "epoch": 0.91, + "learning_rate": 0.0005505680428644479, + "loss": 0.1189, + "step": 20670 + }, + { + "epoch": 0.91, + "learning_rate": 0.0005503457168901043, + "loss": 0.122, + "step": 20680 + }, + { + "epoch": 0.91, + "learning_rate": 0.0005501233909157607, + "loss": 0.1232, + "step": 20690 + }, + { + "epoch": 0.91, + "learning_rate": 0.0005499010649414171, + "loss": 0.1194, + "step": 20700 + }, + { + "epoch": 0.91, + "learning_rate": 0.0005496787389670736, + "loss": 0.1194, + "step": 20710 + }, + { + "epoch": 0.91, + "learning_rate": 0.00054945641299273, + "loss": 0.1176, + "step": 20720 + }, + { + "epoch": 0.91, + "learning_rate": 0.0005492340870183864, + "loss": 0.1192, + "step": 20730 + }, + { + "epoch": 0.91, + "learning_rate": 0.0005490117610440429, + "loss": 0.1229, + "step": 20740 + }, + { + "epoch": 0.91, + "learning_rate": 0.0005487894350696993, + "loss": 0.121, + "step": 20750 + }, + { + "epoch": 0.91, + "learning_rate": 0.0005485671090953557, + "loss": 0.1221, + "step": 20760 + }, + { + "epoch": 0.91, + "learning_rate": 0.000548344783121012, + "loss": 0.1187, + "step": 20770 + }, + { + "epoch": 0.91, + "learning_rate": 0.0005481224571466685, + "loss": 0.1172, + "step": 20780 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005479001311723249, + "loss": 0.1245, + "step": 20790 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005476778051979813, + "loss": 0.1203, + "step": 20800 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005474554792236377, + "loss": 0.1172, + "step": 20810 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005472331532492942, + "loss": 0.1184, + "step": 20820 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005470108272749506, + "loss": 0.1181, + "step": 20830 + }, + { + "epoch": 0.92, + "learning_rate": 0.000546788501300607, + "loss": 0.1208, + "step": 20840 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005465661753262634, + "loss": 0.1204, + "step": 20850 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005463438493519199, + "loss": 0.1173, + "step": 20860 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005461215233775763, + "loss": 0.1237, + "step": 20870 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005458991974032327, + "loss": 0.1249, + "step": 20880 + }, + { + "epoch": 0.92, + "learning_rate": 0.000545676871428889, + "loss": 0.1232, + "step": 20890 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005454545454545455, + "loss": 0.1176, + "step": 20900 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005452322194802019, + "loss": 0.1184, + "step": 20910 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005450098935058583, + "loss": 0.1222, + "step": 20920 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005447875675315148, + "loss": 0.1267, + "step": 20930 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005445652415571712, + "loss": 0.1151, + "step": 20940 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005443429155828275, + "loss": 0.1134, + "step": 20950 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005441205896084839, + "loss": 0.1183, + "step": 20960 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005438982636341404, + "loss": 0.1186, + "step": 20970 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005436759376597968, + "loss": 0.1214, + "step": 20980 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005434536116854532, + "loss": 0.114, + "step": 20990 + }, + { + "epoch": 0.92, + "learning_rate": 0.0005432312857111096, + "loss": 0.1214, + "step": 21000 + }, + { + "epoch": 0.92, + "eval_cer": 0.4996909073571833, + "eval_loss": 0.09141510725021362, + "eval_runtime": 83.8345, + "eval_samples_per_second": 119.211, + "eval_steps_per_second": 7.455, + "step": 21000 + }, + { + "epoch": 0.92, + "learning_rate": 0.000543008959736766, + "loss": 0.1254, + "step": 21010 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005427866337624224, + "loss": 0.1214, + "step": 21020 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005425643077880788, + "loss": 0.1178, + "step": 21030 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005423419818137352, + "loss": 0.1163, + "step": 21040 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005421196558393917, + "loss": 0.1198, + "step": 21050 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005418973298650481, + "loss": 0.1228, + "step": 21060 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005416750038907045, + "loss": 0.125, + "step": 21070 + }, + { + "epoch": 0.93, + "learning_rate": 0.000541452677916361, + "loss": 0.1175, + "step": 21080 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005412303519420174, + "loss": 0.12, + "step": 21090 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005410080259676738, + "loss": 0.1165, + "step": 21100 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005407856999933302, + "loss": 0.1175, + "step": 21110 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005405633740189867, + "loss": 0.1231, + "step": 21120 + }, + { + "epoch": 0.93, + "learning_rate": 0.000540341048044643, + "loss": 0.12, + "step": 21130 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005401187220702994, + "loss": 0.1133, + "step": 21140 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005398963960959558, + "loss": 0.1219, + "step": 21150 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005396740701216123, + "loss": 0.1237, + "step": 21160 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005394517441472687, + "loss": 0.1183, + "step": 21170 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005392294181729251, + "loss": 0.1174, + "step": 21180 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005390070921985815, + "loss": 0.1167, + "step": 21190 + }, + { + "epoch": 0.93, + "learning_rate": 0.000538784766224238, + "loss": 0.1228, + "step": 21200 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005385624402498944, + "loss": 0.1183, + "step": 21210 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005383401142755508, + "loss": 0.1221, + "step": 21220 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005381177883012073, + "loss": 0.1223, + "step": 21230 + }, + { + "epoch": 0.93, + "learning_rate": 0.0005378954623268637, + "loss": 0.1208, + "step": 21240 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005376731363525201, + "loss": 0.1175, + "step": 21250 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005374508103781764, + "loss": 0.1223, + "step": 21260 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005372284844038329, + "loss": 0.1233, + "step": 21270 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005370061584294893, + "loss": 0.1187, + "step": 21280 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005367838324551457, + "loss": 0.1178, + "step": 21290 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005365615064808021, + "loss": 0.1189, + "step": 21300 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005363391805064586, + "loss": 0.1202, + "step": 21310 + }, + { + "epoch": 0.94, + "learning_rate": 0.000536116854532115, + "loss": 0.119, + "step": 21320 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005358945285577714, + "loss": 0.1201, + "step": 21330 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005356722025834278, + "loss": 0.1132, + "step": 21340 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005354498766090843, + "loss": 0.1224, + "step": 21350 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005352275506347407, + "loss": 0.1177, + "step": 21360 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005350052246603971, + "loss": 0.1158, + "step": 21370 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005347828986860534, + "loss": 0.1143, + "step": 21380 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005345605727117099, + "loss": 0.1195, + "step": 21390 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005343382467373663, + "loss": 0.1202, + "step": 21400 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005341159207630227, + "loss": 0.1245, + "step": 21410 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005338935947886792, + "loss": 0.1183, + "step": 21420 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005336712688143356, + "loss": 0.1208, + "step": 21430 + }, + { + "epoch": 0.94, + "learning_rate": 0.000533448942839992, + "loss": 0.1166, + "step": 21440 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005332266168656484, + "loss": 0.1161, + "step": 21450 + }, + { + "epoch": 0.94, + "learning_rate": 0.0005330042908913049, + "loss": 0.1186, + "step": 21460 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005327819649169613, + "loss": 0.1111, + "step": 21470 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005325596389426177, + "loss": 0.1166, + "step": 21480 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005323373129682741, + "loss": 0.1204, + "step": 21490 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005321149869939305, + "loss": 0.1189, + "step": 21500 + }, + { + "epoch": 0.95, + "eval_cer": 0.4994599798003255, + "eval_loss": 0.0908852219581604, + "eval_runtime": 83.7343, + "eval_samples_per_second": 119.354, + "eval_steps_per_second": 7.464, + "step": 21500 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005318926610195869, + "loss": 0.1151, + "step": 21510 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005316703350452433, + "loss": 0.1161, + "step": 21520 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005314480090708998, + "loss": 0.1192, + "step": 21530 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005312256830965562, + "loss": 0.1186, + "step": 21540 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005310033571222126, + "loss": 0.1177, + "step": 21550 + }, + { + "epoch": 0.95, + "learning_rate": 0.000530781031147869, + "loss": 0.1222, + "step": 21560 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005305587051735255, + "loss": 0.12, + "step": 21570 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005303363791991819, + "loss": 0.1143, + "step": 21580 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005301140532248383, + "loss": 0.1167, + "step": 21590 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005298917272504947, + "loss": 0.1194, + "step": 21600 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005296694012761512, + "loss": 0.1165, + "step": 21610 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005294470753018075, + "loss": 0.1192, + "step": 21620 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005292247493274639, + "loss": 0.1144, + "step": 21630 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005290024233531203, + "loss": 0.116, + "step": 21640 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005287800973787768, + "loss": 0.1133, + "step": 21650 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005285577714044332, + "loss": 0.1229, + "step": 21660 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005283354454300896, + "loss": 0.1168, + "step": 21670 + }, + { + "epoch": 0.95, + "learning_rate": 0.000528113119455746, + "loss": 0.1158, + "step": 21680 + }, + { + "epoch": 0.95, + "learning_rate": 0.0005278907934814025, + "loss": 0.1171, + "step": 21690 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005276684675070589, + "loss": 0.1179, + "step": 21700 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005274461415327153, + "loss": 0.1175, + "step": 21710 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005272238155583718, + "loss": 0.1189, + "step": 21720 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005270014895840282, + "loss": 0.115, + "step": 21730 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005267791636096845, + "loss": 0.1217, + "step": 21740 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005265568376353409, + "loss": 0.1141, + "step": 21750 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005263345116609974, + "loss": 0.1161, + "step": 21760 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005261121856866538, + "loss": 0.1191, + "step": 21770 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005258898597123102, + "loss": 0.1205, + "step": 21780 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005256675337379666, + "loss": 0.1195, + "step": 21790 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005254452077636231, + "loss": 0.1147, + "step": 21800 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005252228817892795, + "loss": 0.1162, + "step": 21810 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005250005558149359, + "loss": 0.1194, + "step": 21820 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005247782298405924, + "loss": 0.1187, + "step": 21830 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005245559038662488, + "loss": 0.1229, + "step": 21840 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005243335778919052, + "loss": 0.1145, + "step": 21850 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005241112519175616, + "loss": 0.1143, + "step": 21860 + }, + { + "epoch": 0.96, + "learning_rate": 0.000523888925943218, + "loss": 0.1168, + "step": 21870 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005236665999688744, + "loss": 0.1195, + "step": 21880 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005234442739945308, + "loss": 0.1175, + "step": 21890 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005232219480201872, + "loss": 0.115, + "step": 21900 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005229996220458437, + "loss": 0.1154, + "step": 21910 + }, + { + "epoch": 0.96, + "learning_rate": 0.0005227772960715001, + "loss": 0.1178, + "step": 21920 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005225549700971565, + "loss": 0.1234, + "step": 21930 + }, + { + "epoch": 0.97, + "learning_rate": 0.000522332644122813, + "loss": 0.117, + "step": 21940 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005221103181484694, + "loss": 0.1151, + "step": 21950 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005218879921741257, + "loss": 0.114, + "step": 21960 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005216656661997821, + "loss": 0.1185, + "step": 21970 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005214433402254385, + "loss": 0.1097, + "step": 21980 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005212210142510949, + "loss": 0.1142, + "step": 21990 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005209986882767513, + "loss": 0.1198, + "step": 22000 + }, + { + "epoch": 0.97, + "eval_cer": 0.5000911385442093, + "eval_loss": 0.08931281417608261, + "eval_runtime": 83.2991, + "eval_samples_per_second": 119.977, + "eval_steps_per_second": 7.503, + "step": 22000 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005207763623024077, + "loss": 0.1121, + "step": 22010 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005205540363280641, + "loss": 0.1104, + "step": 22020 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005203317103537206, + "loss": 0.1199, + "step": 22030 + }, + { + "epoch": 0.97, + "learning_rate": 0.000520109384379377, + "loss": 0.1152, + "step": 22040 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005198870584050334, + "loss": 0.1159, + "step": 22050 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005196647324306899, + "loss": 0.1197, + "step": 22060 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005194424064563463, + "loss": 0.1173, + "step": 22070 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005192200804820027, + "loss": 0.1175, + "step": 22080 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005189977545076591, + "loss": 0.1163, + "step": 22090 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005187754285333156, + "loss": 0.1181, + "step": 22100 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005185531025589719, + "loss": 0.1092, + "step": 22110 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005183307765846283, + "loss": 0.1154, + "step": 22120 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005181084506102847, + "loss": 0.1163, + "step": 22130 + }, + { + "epoch": 0.97, + "learning_rate": 0.0005178861246359412, + "loss": 0.1234, + "step": 22140 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005176637986615976, + "loss": 0.1159, + "step": 22150 + }, + { + "epoch": 0.98, + "learning_rate": 0.000517441472687254, + "loss": 0.1168, + "step": 22160 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005172191467129105, + "loss": 0.1184, + "step": 22170 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005169968207385669, + "loss": 0.116, + "step": 22180 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005167744947642233, + "loss": 0.116, + "step": 22190 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005165521687898797, + "loss": 0.1133, + "step": 22200 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005163298428155362, + "loss": 0.1131, + "step": 22210 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005161075168411926, + "loss": 0.1151, + "step": 22220 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005158851908668489, + "loss": 0.1129, + "step": 22230 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005156628648925053, + "loss": 0.1209, + "step": 22240 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005154405389181618, + "loss": 0.1174, + "step": 22250 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005152182129438182, + "loss": 0.1134, + "step": 22260 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005149958869694746, + "loss": 0.117, + "step": 22270 + }, + { + "epoch": 0.98, + "learning_rate": 0.000514773560995131, + "loss": 0.1164, + "step": 22280 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005145512350207875, + "loss": 0.1156, + "step": 22290 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005143289090464439, + "loss": 0.1207, + "step": 22300 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005141065830721003, + "loss": 0.1179, + "step": 22310 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005138842570977568, + "loss": 0.1164, + "step": 22320 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005136619311234132, + "loss": 0.1112, + "step": 22330 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005134396051490696, + "loss": 0.1151, + "step": 22340 + }, + { + "epoch": 0.98, + "learning_rate": 0.000513217279174726, + "loss": 0.1197, + "step": 22350 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005129949532003824, + "loss": 0.1174, + "step": 22360 + }, + { + "epoch": 0.98, + "learning_rate": 0.0005127726272260388, + "loss": 0.113, + "step": 22370 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005125503012516952, + "loss": 0.1189, + "step": 22380 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005123279752773516, + "loss": 0.1152, + "step": 22390 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005121056493030081, + "loss": 0.1187, + "step": 22400 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005118833233286645, + "loss": 0.1181, + "step": 22410 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005116609973543209, + "loss": 0.1165, + "step": 22420 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005114386713799773, + "loss": 0.1155, + "step": 22430 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005112163454056338, + "loss": 0.1209, + "step": 22440 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005109940194312902, + "loss": 0.1227, + "step": 22450 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005107716934569466, + "loss": 0.1161, + "step": 22460 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005105493674826031, + "loss": 0.116, + "step": 22470 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005103270415082594, + "loss": 0.1151, + "step": 22480 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005101047155339158, + "loss": 0.1208, + "step": 22490 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005098823895595722, + "loss": 0.116, + "step": 22500 + }, + { + "epoch": 0.99, + "eval_cer": 0.49957155154127925, + "eval_loss": 0.08863367140293121, + "eval_runtime": 83.2073, + "eval_samples_per_second": 120.11, + "eval_steps_per_second": 7.511, + "step": 22500 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005096600635852287, + "loss": 0.1173, + "step": 22510 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005094377376108851, + "loss": 0.118, + "step": 22520 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005092154116365415, + "loss": 0.116, + "step": 22530 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005089930856621979, + "loss": 0.1138, + "step": 22540 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005087707596878544, + "loss": 0.1132, + "step": 22550 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005085484337135108, + "loss": 0.1163, + "step": 22560 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005083261077391672, + "loss": 0.1127, + "step": 22570 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005081037817648236, + "loss": 0.1184, + "step": 22580 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005078814557904801, + "loss": 0.119, + "step": 22590 + }, + { + "epoch": 0.99, + "learning_rate": 0.0005076591298161364, + "loss": 0.1131, + "step": 22600 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005074368038417928, + "loss": 0.1161, + "step": 22610 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005072144778674493, + "loss": 0.1218, + "step": 22620 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005069921518931057, + "loss": 0.1176, + "step": 22630 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005067698259187621, + "loss": 0.1209, + "step": 22640 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005065474999444185, + "loss": 0.1207, + "step": 22650 + }, + { + "epoch": 1.0, + "learning_rate": 0.000506325173970075, + "loss": 0.117, + "step": 22660 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005061028479957314, + "loss": 0.1168, + "step": 22670 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005058805220213878, + "loss": 0.1144, + "step": 22680 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005056581960470442, + "loss": 0.1178, + "step": 22690 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005054358700727007, + "loss": 0.1125, + "step": 22700 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005052135440983571, + "loss": 0.1185, + "step": 22710 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005049912181240134, + "loss": 0.1163, + "step": 22720 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005047688921496698, + "loss": 0.1076, + "step": 22730 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005045465661753263, + "loss": 0.1012, + "step": 22740 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005043242402009827, + "loss": 0.1055, + "step": 22750 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005041019142266391, + "loss": 0.1067, + "step": 22760 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005038795882522956, + "loss": 0.1045, + "step": 22770 + }, + { + "epoch": 1.0, + "learning_rate": 0.000503657262277952, + "loss": 0.1062, + "step": 22780 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005034349363036084, + "loss": 0.1032, + "step": 22790 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005032126103292648, + "loss": 0.0991, + "step": 22800 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005029902843549213, + "loss": 0.1042, + "step": 22810 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005027679583805777, + "loss": 0.1049, + "step": 22820 + }, + { + "epoch": 1.0, + "learning_rate": 0.0005025456324062341, + "loss": 0.1079, + "step": 22830 + }, + { + "epoch": 1.01, + "learning_rate": 0.0005023233064318905, + "loss": 0.1068, + "step": 22840 + }, + { + "epoch": 1.01, + "learning_rate": 0.0005021009804575469, + "loss": 0.1035, + "step": 22850 + }, + { + "epoch": 1.01, + "learning_rate": 0.0005018786544832033, + "loss": 0.1077, + "step": 22860 + }, + { + "epoch": 1.01, + "learning_rate": 0.0005016563285088597, + "loss": 0.1076, + "step": 22870 + }, + { + "epoch": 1.01, + "learning_rate": 0.0005014340025345161, + "loss": 0.0995, + "step": 22880 + }, + { + "epoch": 1.01, + "learning_rate": 0.0005012116765601726, + "loss": 0.1062, + "step": 22890 + }, + { + "epoch": 1.01, + "learning_rate": 0.000500989350585829, + "loss": 0.1069, + "step": 22900 + }, + { + "epoch": 1.01, + "learning_rate": 0.0005007670246114854, + "loss": 0.1049, + "step": 22910 + }, + { + "epoch": 1.01, + "learning_rate": 0.0005005446986371419, + "loss": 0.1001, + "step": 22920 + }, + { + "epoch": 1.01, + "learning_rate": 0.0005003223726627983, + "loss": 0.1013, + "step": 22930 + }, + { + "epoch": 1.01, + "learning_rate": 0.0005001000466884547, + "loss": 0.104, + "step": 22940 + }, + { + "epoch": 1.01, + "learning_rate": 0.000499877720714111, + "loss": 0.107, + "step": 22950 + }, + { + "epoch": 1.01, + "learning_rate": 0.0004996553947397675, + "loss": 0.1035, + "step": 22960 + }, + { + "epoch": 1.01, + "learning_rate": 0.0004994330687654239, + "loss": 0.1042, + "step": 22970 + }, + { + "epoch": 1.01, + "learning_rate": 0.0004992107427910803, + "loss": 0.1056, + "step": 22980 + }, + { + "epoch": 1.01, + "learning_rate": 0.0004989884168167367, + "loss": 0.1063, + "step": 22990 + }, + { + "epoch": 1.01, + "learning_rate": 0.0004987660908423931, + "loss": 0.1062, + "step": 23000 + }, + { + "epoch": 1.01, + "eval_cer": 0.4995345771852655, + "eval_loss": 0.08944559842348099, + "eval_runtime": 83.207, + "eval_samples_per_second": 120.11, + "eval_steps_per_second": 7.511, + "step": 23000 + }, + { + "epoch": 1.01, + "learning_rate": 0.0004985437648680495, + "loss": 0.1012, + "step": 23010 + }, + { + "epoch": 1.01, + "learning_rate": 0.0004983214388937059, + "loss": 0.097, + "step": 23020 + }, + { + "epoch": 1.01, + "learning_rate": 0.0004980991129193623, + "loss": 0.1049, + "step": 23030 + }, + { + "epoch": 1.01, + "learning_rate": 0.0004978767869450188, + "loss": 0.1047, + "step": 23040 + }, + { + "epoch": 1.01, + "learning_rate": 0.0004976544609706752, + "loss": 0.1077, + "step": 23050 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004974321349963316, + "loss": 0.1057, + "step": 23060 + }, + { + "epoch": 1.02, + "learning_rate": 0.000497209809021988, + "loss": 0.1019, + "step": 23070 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004969874830476445, + "loss": 0.1064, + "step": 23080 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004967651570733009, + "loss": 0.1021, + "step": 23090 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004965428310989573, + "loss": 0.1072, + "step": 23100 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004963205051246138, + "loss": 0.1056, + "step": 23110 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004960981791502701, + "loss": 0.1055, + "step": 23120 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004958758531759265, + "loss": 0.106, + "step": 23130 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004956535272015829, + "loss": 0.1045, + "step": 23140 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004954312012272394, + "loss": 0.106, + "step": 23150 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004952088752528958, + "loss": 0.1061, + "step": 23160 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004949865492785522, + "loss": 0.1011, + "step": 23170 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004947642233042086, + "loss": 0.1054, + "step": 23180 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004945418973298651, + "loss": 0.1002, + "step": 23190 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004943195713555215, + "loss": 0.1045, + "step": 23200 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004940972453811779, + "loss": 0.1005, + "step": 23210 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004938749194068344, + "loss": 0.1066, + "step": 23220 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004936525934324908, + "loss": 0.1083, + "step": 23230 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004934302674581471, + "loss": 0.1044, + "step": 23240 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004932079414838035, + "loss": 0.1065, + "step": 23250 + }, + { + "epoch": 1.02, + "learning_rate": 0.00049298561550946, + "loss": 0.1006, + "step": 23260 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004927632895351164, + "loss": 0.1046, + "step": 23270 + }, + { + "epoch": 1.02, + "learning_rate": 0.0004925409635607728, + "loss": 0.1025, + "step": 23280 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004923186375864292, + "loss": 0.1023, + "step": 23290 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004920963116120857, + "loss": 0.1037, + "step": 23300 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004918739856377421, + "loss": 0.108, + "step": 23310 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004916516596633985, + "loss": 0.1062, + "step": 23320 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004914293336890549, + "loss": 0.0998, + "step": 23330 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004912070077147114, + "loss": 0.105, + "step": 23340 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004909846817403678, + "loss": 0.1048, + "step": 23350 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004907623557660242, + "loss": 0.1032, + "step": 23360 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004905400297916805, + "loss": 0.1036, + "step": 23370 + }, + { + "epoch": 1.03, + "learning_rate": 0.000490317703817337, + "loss": 0.1065, + "step": 23380 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004900953778429934, + "loss": 0.1019, + "step": 23390 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004898730518686498, + "loss": 0.1038, + "step": 23400 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004896507258943063, + "loss": 0.106, + "step": 23410 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004894283999199627, + "loss": 0.1018, + "step": 23420 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004892060739456191, + "loss": 0.1042, + "step": 23430 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004889837479712755, + "loss": 0.1047, + "step": 23440 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004887614219969319, + "loss": 0.1044, + "step": 23450 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004885390960225883, + "loss": 0.1068, + "step": 23460 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004883167700482447, + "loss": 0.1035, + "step": 23470 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004880944440739012, + "loss": 0.103, + "step": 23480 + }, + { + "epoch": 1.03, + "learning_rate": 0.00048787211809955757, + "loss": 0.1051, + "step": 23490 + }, + { + "epoch": 1.03, + "learning_rate": 0.000487649792125214, + "loss": 0.1059, + "step": 23500 + }, + { + "epoch": 1.03, + "eval_cer": 0.4994450603233375, + "eval_loss": 0.08834969252347946, + "eval_runtime": 83.1621, + "eval_samples_per_second": 120.175, + "eval_steps_per_second": 7.515, + "step": 23500 + }, + { + "epoch": 1.03, + "learning_rate": 0.0004874274661508704, + "loss": 0.103, + "step": 23510 + }, + { + "epoch": 1.04, + "learning_rate": 0.00048720514017652685, + "loss": 0.1015, + "step": 23520 + }, + { + "epoch": 1.04, + "learning_rate": 0.0004869828142021833, + "loss": 0.1089, + "step": 23530 + }, + { + "epoch": 1.04, + "learning_rate": 0.0004867604882278397, + "loss": 0.1025, + "step": 23540 + }, + { + "epoch": 1.04, + "learning_rate": 0.0004865381622534961, + "loss": 0.101, + "step": 23550 + }, + { + "epoch": 1.04, + "learning_rate": 0.0004863158362791525, + "loss": 0.1048, + "step": 23560 + }, + { + "epoch": 1.04, + "learning_rate": 0.00048609351030480893, + "loss": 0.1055, + "step": 23570 + }, + { + "epoch": 1.04, + "learning_rate": 0.00048587118433046536, + "loss": 0.1061, + "step": 23580 + }, + { + "epoch": 1.04, + "learning_rate": 0.0004856488583561218, + "loss": 0.1047, + "step": 23590 + }, + { + "epoch": 1.04, + "learning_rate": 0.0004854265323817782, + "loss": 0.1027, + "step": 23600 + }, + { + "epoch": 1.04, + "learning_rate": 0.0004852042064074346, + "loss": 0.1032, + "step": 23610 + }, + { + "epoch": 1.04, + "learning_rate": 0.000484981880433091, + "loss": 0.1059, + "step": 23620 + }, + { + "epoch": 1.04, + "learning_rate": 0.00048475955445874744, + "loss": 0.0985, + "step": 23630 + }, + { + "epoch": 1.04, + "learning_rate": 0.00048453722848440387, + "loss": 0.1028, + "step": 23640 + }, + { + "epoch": 1.04, + "learning_rate": 0.0004843149025100603, + "loss": 0.1021, + "step": 23650 + }, + { + "epoch": 1.04, + "learning_rate": 0.0004840925765357167, + "loss": 0.106, + "step": 23660 + }, + { + "epoch": 1.04, + "learning_rate": 0.0004838702505613731, + "loss": 0.1091, + "step": 23670 + }, + { + "epoch": 1.04, + "learning_rate": 0.0004836479245870295, + "loss": 0.0998, + "step": 23680 + }, + { + "epoch": 1.04, + "learning_rate": 0.00048342559861268595, + "loss": 0.1033, + "step": 23690 + }, + { + "epoch": 1.04, + "learning_rate": 0.0004832032726383423, + "loss": 0.1025, + "step": 23700 + }, + { + "epoch": 1.04, + "learning_rate": 0.00048298094666399875, + "loss": 0.1006, + "step": 23710 + }, + { + "epoch": 1.04, + "learning_rate": 0.0004827586206896552, + "loss": 0.1022, + "step": 23720 + }, + { + "epoch": 1.04, + "learning_rate": 0.00048253629471531155, + "loss": 0.1004, + "step": 23730 + }, + { + "epoch": 1.04, + "learning_rate": 0.000482313968740968, + "loss": 0.1099, + "step": 23740 + }, + { + "epoch": 1.05, + "learning_rate": 0.0004820916427666244, + "loss": 0.1024, + "step": 23750 + }, + { + "epoch": 1.05, + "learning_rate": 0.00048186931679228083, + "loss": 0.1042, + "step": 23760 + }, + { + "epoch": 1.05, + "learning_rate": 0.00048164699081793726, + "loss": 0.1059, + "step": 23770 + }, + { + "epoch": 1.05, + "learning_rate": 0.0004814246648435937, + "loss": 0.1028, + "step": 23780 + }, + { + "epoch": 1.05, + "learning_rate": 0.00048120233886925006, + "loss": 0.099, + "step": 23790 + }, + { + "epoch": 1.05, + "learning_rate": 0.0004809800128949065, + "loss": 0.1013, + "step": 23800 + }, + { + "epoch": 1.05, + "learning_rate": 0.0004807576869205629, + "loss": 0.1058, + "step": 23810 + }, + { + "epoch": 1.05, + "learning_rate": 0.00048053536094621934, + "loss": 0.1027, + "step": 23820 + }, + { + "epoch": 1.05, + "learning_rate": 0.00048031303497187577, + "loss": 0.1039, + "step": 23830 + }, + { + "epoch": 1.05, + "learning_rate": 0.0004800907089975322, + "loss": 0.1064, + "step": 23840 + }, + { + "epoch": 1.05, + "learning_rate": 0.00047986838302318857, + "loss": 0.1002, + "step": 23850 + }, + { + "epoch": 1.05, + "learning_rate": 0.000479646057048845, + "loss": 0.1059, + "step": 23860 + }, + { + "epoch": 1.05, + "learning_rate": 0.0004794237310745014, + "loss": 0.101, + "step": 23870 + }, + { + "epoch": 1.05, + "learning_rate": 0.00047920140510015785, + "loss": 0.1032, + "step": 23880 + }, + { + "epoch": 1.05, + "learning_rate": 0.0004789790791258143, + "loss": 0.1072, + "step": 23890 + }, + { + "epoch": 1.05, + "learning_rate": 0.0004787567531514707, + "loss": 0.1061, + "step": 23900 + }, + { + "epoch": 1.05, + "learning_rate": 0.00047853442717712714, + "loss": 0.0992, + "step": 23910 + }, + { + "epoch": 1.05, + "learning_rate": 0.0004783121012027835, + "loss": 0.1037, + "step": 23920 + }, + { + "epoch": 1.05, + "learning_rate": 0.00047808977522843994, + "loss": 0.1002, + "step": 23930 + }, + { + "epoch": 1.05, + "learning_rate": 0.00047786744925409636, + "loss": 0.1116, + "step": 23940 + }, + { + "epoch": 1.05, + "learning_rate": 0.0004776451232797528, + "loss": 0.0995, + "step": 23950 + }, + { + "epoch": 1.05, + "learning_rate": 0.0004774227973054092, + "loss": 0.1069, + "step": 23960 + }, + { + "epoch": 1.06, + "learning_rate": 0.00047720047133106565, + "loss": 0.0997, + "step": 23970 + }, + { + "epoch": 1.06, + "learning_rate": 0.000476978145356722, + "loss": 0.1029, + "step": 23980 + }, + { + "epoch": 1.06, + "learning_rate": 0.00047675581938237845, + "loss": 0.1018, + "step": 23990 + }, + { + "epoch": 1.06, + "learning_rate": 0.0004765334934080349, + "loss": 0.108, + "step": 24000 + }, + { + "epoch": 1.06, + "eval_cer": 0.4994106806589738, + "eval_loss": 0.08722981065511703, + "eval_runtime": 83.2531, + "eval_samples_per_second": 120.044, + "eval_steps_per_second": 7.507, + "step": 24000 + }, + { + "epoch": 1.06, + "learning_rate": 0.0004763111674336913, + "loss": 0.1033, + "step": 24010 + }, + { + "epoch": 1.06, + "learning_rate": 0.00047608884145934773, + "loss": 0.1035, + "step": 24020 + }, + { + "epoch": 1.06, + "learning_rate": 0.00047586651548500416, + "loss": 0.1026, + "step": 24030 + }, + { + "epoch": 1.06, + "learning_rate": 0.00047564418951066053, + "loss": 0.1022, + "step": 24040 + }, + { + "epoch": 1.06, + "learning_rate": 0.00047542186353631696, + "loss": 0.1056, + "step": 24050 + }, + { + "epoch": 1.06, + "learning_rate": 0.0004751995375619734, + "loss": 0.1009, + "step": 24060 + }, + { + "epoch": 1.06, + "learning_rate": 0.0004749772115876298, + "loss": 0.1019, + "step": 24070 + }, + { + "epoch": 1.06, + "learning_rate": 0.00047475488561328624, + "loss": 0.1048, + "step": 24080 + }, + { + "epoch": 1.06, + "learning_rate": 0.00047453255963894267, + "loss": 0.1049, + "step": 24090 + }, + { + "epoch": 1.06, + "learning_rate": 0.00047431023366459904, + "loss": 0.1, + "step": 24100 + }, + { + "epoch": 1.06, + "learning_rate": 0.00047408790769025547, + "loss": 0.1046, + "step": 24110 + }, + { + "epoch": 1.06, + "learning_rate": 0.0004738655817159119, + "loss": 0.1017, + "step": 24120 + }, + { + "epoch": 1.06, + "learning_rate": 0.0004736432557415683, + "loss": 0.1009, + "step": 24130 + }, + { + "epoch": 1.06, + "learning_rate": 0.00047342092976722475, + "loss": 0.1068, + "step": 24140 + }, + { + "epoch": 1.06, + "learning_rate": 0.0004731986037928812, + "loss": 0.1025, + "step": 24150 + }, + { + "epoch": 1.06, + "learning_rate": 0.00047297627781853755, + "loss": 0.0982, + "step": 24160 + }, + { + "epoch": 1.06, + "learning_rate": 0.000472753951844194, + "loss": 0.0996, + "step": 24170 + }, + { + "epoch": 1.06, + "learning_rate": 0.0004725316258698504, + "loss": 0.103, + "step": 24180 + }, + { + "epoch": 1.06, + "learning_rate": 0.00047230929989550683, + "loss": 0.1104, + "step": 24190 + }, + { + "epoch": 1.07, + "learning_rate": 0.0004720869739211632, + "loss": 0.1065, + "step": 24200 + }, + { + "epoch": 1.07, + "learning_rate": 0.00047186464794681963, + "loss": 0.1012, + "step": 24210 + }, + { + "epoch": 1.07, + "learning_rate": 0.000471642321972476, + "loss": 0.1007, + "step": 24220 + }, + { + "epoch": 1.07, + "learning_rate": 0.00047141999599813243, + "loss": 0.1076, + "step": 24230 + }, + { + "epoch": 1.07, + "learning_rate": 0.00047119767002378886, + "loss": 0.1007, + "step": 24240 + }, + { + "epoch": 1.07, + "learning_rate": 0.0004709753440494453, + "loss": 0.1051, + "step": 24250 + }, + { + "epoch": 1.07, + "learning_rate": 0.0004707530180751017, + "loss": 0.103, + "step": 24260 + }, + { + "epoch": 1.07, + "learning_rate": 0.00047053069210075814, + "loss": 0.104, + "step": 24270 + }, + { + "epoch": 1.07, + "learning_rate": 0.0004703083661264145, + "loss": 0.0994, + "step": 24280 + }, + { + "epoch": 1.07, + "learning_rate": 0.00047008604015207094, + "loss": 0.1046, + "step": 24290 + }, + { + "epoch": 1.07, + "learning_rate": 0.00046986371417772737, + "loss": 0.1045, + "step": 24300 + }, + { + "epoch": 1.07, + "learning_rate": 0.0004696413882033838, + "loss": 0.1009, + "step": 24310 + }, + { + "epoch": 1.07, + "learning_rate": 0.0004694190622290402, + "loss": 0.1074, + "step": 24320 + }, + { + "epoch": 1.07, + "learning_rate": 0.00046919673625469665, + "loss": 0.1076, + "step": 24330 + }, + { + "epoch": 1.07, + "learning_rate": 0.000468974410280353, + "loss": 0.1044, + "step": 24340 + }, + { + "epoch": 1.07, + "learning_rate": 0.00046875208430600945, + "loss": 0.1053, + "step": 24350 + }, + { + "epoch": 1.07, + "learning_rate": 0.0004685297583316659, + "loss": 0.1059, + "step": 24360 + }, + { + "epoch": 1.07, + "learning_rate": 0.0004683074323573223, + "loss": 0.1035, + "step": 24370 + }, + { + "epoch": 1.07, + "learning_rate": 0.00046808510638297874, + "loss": 0.103, + "step": 24380 + }, + { + "epoch": 1.07, + "learning_rate": 0.00046786278040863516, + "loss": 0.1009, + "step": 24390 + }, + { + "epoch": 1.07, + "learning_rate": 0.0004676404544342916, + "loss": 0.1012, + "step": 24400 + }, + { + "epoch": 1.07, + "learning_rate": 0.00046741812845994796, + "loss": 0.1042, + "step": 24410 + }, + { + "epoch": 1.07, + "learning_rate": 0.0004671958024856044, + "loss": 0.1077, + "step": 24420 + }, + { + "epoch": 1.08, + "learning_rate": 0.0004669734765112608, + "loss": 0.1055, + "step": 24430 + }, + { + "epoch": 1.08, + "learning_rate": 0.00046675115053691725, + "loss": 0.1027, + "step": 24440 + }, + { + "epoch": 1.08, + "learning_rate": 0.0004665288245625737, + "loss": 0.1095, + "step": 24450 + }, + { + "epoch": 1.08, + "learning_rate": 0.0004663064985882301, + "loss": 0.1052, + "step": 24460 + }, + { + "epoch": 1.08, + "learning_rate": 0.0004660841726138865, + "loss": 0.0993, + "step": 24470 + }, + { + "epoch": 1.08, + "learning_rate": 0.0004658618466395429, + "loss": 0.1038, + "step": 24480 + }, + { + "epoch": 1.08, + "learning_rate": 0.00046563952066519933, + "loss": 0.1061, + "step": 24490 + }, + { + "epoch": 1.08, + "learning_rate": 0.00046541719469085576, + "loss": 0.1039, + "step": 24500 + }, + { + "epoch": 1.08, + "eval_cer": 0.4997142595820341, + "eval_loss": 0.0873190239071846, + "eval_runtime": 83.3843, + "eval_samples_per_second": 119.855, + "eval_steps_per_second": 7.495, + "step": 24500 + }, + { + "epoch": 1.08, + "learning_rate": 0.0004651948687165122, + "loss": 0.104, + "step": 24510 + }, + { + "epoch": 1.08, + "learning_rate": 0.0004649725427421686, + "loss": 0.1012, + "step": 24520 + }, + { + "epoch": 1.08, + "learning_rate": 0.000464750216767825, + "loss": 0.1046, + "step": 24530 + }, + { + "epoch": 1.08, + "learning_rate": 0.0004645278907934814, + "loss": 0.1014, + "step": 24540 + }, + { + "epoch": 1.08, + "learning_rate": 0.00046430556481913784, + "loss": 0.1013, + "step": 24550 + }, + { + "epoch": 1.08, + "learning_rate": 0.00046408323884479427, + "loss": 0.1022, + "step": 24560 + }, + { + "epoch": 1.08, + "learning_rate": 0.0004638609128704507, + "loss": 0.1039, + "step": 24570 + }, + { + "epoch": 1.08, + "learning_rate": 0.0004636385868961071, + "loss": 0.1006, + "step": 24580 + }, + { + "epoch": 1.08, + "learning_rate": 0.0004634162609217635, + "loss": 0.1057, + "step": 24590 + }, + { + "epoch": 1.08, + "learning_rate": 0.0004631939349474199, + "loss": 0.1068, + "step": 24600 + }, + { + "epoch": 1.08, + "learning_rate": 0.00046297160897307635, + "loss": 0.0986, + "step": 24610 + }, + { + "epoch": 1.08, + "learning_rate": 0.0004627492829987328, + "loss": 0.104, + "step": 24620 + }, + { + "epoch": 1.08, + "learning_rate": 0.0004625269570243892, + "loss": 0.1021, + "step": 24630 + }, + { + "epoch": 1.08, + "learning_rate": 0.00046230463105004563, + "loss": 0.1033, + "step": 24640 + }, + { + "epoch": 1.09, + "learning_rate": 0.000462082305075702, + "loss": 0.1041, + "step": 24650 + }, + { + "epoch": 1.09, + "learning_rate": 0.00046185997910135843, + "loss": 0.1007, + "step": 24660 + }, + { + "epoch": 1.09, + "learning_rate": 0.00046163765312701486, + "loss": 0.1072, + "step": 24670 + }, + { + "epoch": 1.09, + "learning_rate": 0.0004614153271526713, + "loss": 0.1046, + "step": 24680 + }, + { + "epoch": 1.09, + "learning_rate": 0.0004611930011783277, + "loss": 0.1009, + "step": 24690 + }, + { + "epoch": 1.09, + "learning_rate": 0.0004609706752039841, + "loss": 0.1045, + "step": 24700 + }, + { + "epoch": 1.09, + "learning_rate": 0.00046074834922964046, + "loss": 0.1029, + "step": 24710 + }, + { + "epoch": 1.09, + "learning_rate": 0.0004605260232552969, + "loss": 0.0993, + "step": 24720 + }, + { + "epoch": 1.09, + "learning_rate": 0.0004603036972809533, + "loss": 0.1063, + "step": 24730 + }, + { + "epoch": 1.09, + "learning_rate": 0.00046008137130660974, + "loss": 0.1044, + "step": 24740 + }, + { + "epoch": 1.09, + "learning_rate": 0.00045985904533226617, + "loss": 0.102, + "step": 24750 + }, + { + "epoch": 1.09, + "learning_rate": 0.0004596367193579226, + "loss": 0.1001, + "step": 24760 + }, + { + "epoch": 1.09, + "learning_rate": 0.00045941439338357897, + "loss": 0.1086, + "step": 24770 + }, + { + "epoch": 1.09, + "learning_rate": 0.0004591920674092354, + "loss": 0.1044, + "step": 24780 + }, + { + "epoch": 1.09, + "learning_rate": 0.0004589697414348918, + "loss": 0.11, + "step": 24790 + }, + { + "epoch": 1.09, + "learning_rate": 0.00045874741546054825, + "loss": 0.1063, + "step": 24800 + }, + { + "epoch": 1.09, + "learning_rate": 0.0004585250894862047, + "loss": 0.1022, + "step": 24810 + }, + { + "epoch": 1.09, + "learning_rate": 0.0004583027635118611, + "loss": 0.101, + "step": 24820 + }, + { + "epoch": 1.09, + "learning_rate": 0.0004580804375375175, + "loss": 0.1044, + "step": 24830 + }, + { + "epoch": 1.09, + "learning_rate": 0.0004578581115631739, + "loss": 0.0995, + "step": 24840 + }, + { + "epoch": 1.09, + "learning_rate": 0.00045763578558883033, + "loss": 0.1038, + "step": 24850 + }, + { + "epoch": 1.09, + "learning_rate": 0.00045741345961448676, + "loss": 0.1036, + "step": 24860 + }, + { + "epoch": 1.09, + "learning_rate": 0.0004571911336401432, + "loss": 0.1084, + "step": 24870 + }, + { + "epoch": 1.1, + "learning_rate": 0.0004569688076657996, + "loss": 0.0991, + "step": 24880 + }, + { + "epoch": 1.1, + "learning_rate": 0.000456746481691456, + "loss": 0.1043, + "step": 24890 + }, + { + "epoch": 1.1, + "learning_rate": 0.0004565241557171124, + "loss": 0.1033, + "step": 24900 + }, + { + "epoch": 1.1, + "learning_rate": 0.00045630182974276884, + "loss": 0.101, + "step": 24910 + }, + { + "epoch": 1.1, + "learning_rate": 0.00045607950376842527, + "loss": 0.1027, + "step": 24920 + }, + { + "epoch": 1.1, + "learning_rate": 0.0004558571777940817, + "loss": 0.1062, + "step": 24930 + }, + { + "epoch": 1.1, + "learning_rate": 0.0004556348518197381, + "loss": 0.0998, + "step": 24940 + }, + { + "epoch": 1.1, + "learning_rate": 0.00045541252584539455, + "loss": 0.1026, + "step": 24950 + }, + { + "epoch": 1.1, + "learning_rate": 0.00045519019987105093, + "loss": 0.103, + "step": 24960 + }, + { + "epoch": 1.1, + "learning_rate": 0.00045496787389670735, + "loss": 0.1026, + "step": 24970 + }, + { + "epoch": 1.1, + "learning_rate": 0.0004547455479223638, + "loss": 0.1056, + "step": 24980 + }, + { + "epoch": 1.1, + "learning_rate": 0.0004545232219480202, + "loss": 0.0987, + "step": 24990 + }, + { + "epoch": 1.1, + "learning_rate": 0.00045430089597367664, + "loss": 0.1013, + "step": 25000 + }, + { + "epoch": 1.1, + "eval_cer": 0.49968442062805807, + "eval_loss": 0.08791407942771912, + "eval_runtime": 83.8318, + "eval_samples_per_second": 119.215, + "eval_steps_per_second": 7.455, + "step": 25000 + }, + { + "epoch": 1.1, + "learning_rate": 0.00045407856999933306, + "loss": 0.102, + "step": 25010 + }, + { + "epoch": 1.1, + "learning_rate": 0.00045385624402498944, + "loss": 0.1043, + "step": 25020 + }, + { + "epoch": 1.1, + "learning_rate": 0.00045363391805064586, + "loss": 0.0992, + "step": 25030 + }, + { + "epoch": 1.1, + "learning_rate": 0.0004534115920763023, + "loss": 0.1012, + "step": 25040 + }, + { + "epoch": 1.1, + "learning_rate": 0.0004531892661019587, + "loss": 0.1017, + "step": 25050 + }, + { + "epoch": 1.1, + "learning_rate": 0.00045296694012761515, + "loss": 0.1018, + "step": 25060 + }, + { + "epoch": 1.1, + "learning_rate": 0.0004527446141532716, + "loss": 0.0989, + "step": 25070 + }, + { + "epoch": 1.1, + "learning_rate": 0.00045252228817892795, + "loss": 0.105, + "step": 25080 + }, + { + "epoch": 1.1, + "learning_rate": 0.0004522999622045844, + "loss": 0.1021, + "step": 25090 + }, + { + "epoch": 1.1, + "learning_rate": 0.0004520776362302408, + "loss": 0.1011, + "step": 25100 + }, + { + "epoch": 1.11, + "learning_rate": 0.00045185531025589723, + "loss": 0.1051, + "step": 25110 + }, + { + "epoch": 1.11, + "learning_rate": 0.00045163298428155366, + "loss": 0.1019, + "step": 25120 + }, + { + "epoch": 1.11, + "learning_rate": 0.0004514106583072101, + "loss": 0.098, + "step": 25130 + }, + { + "epoch": 1.11, + "learning_rate": 0.00045118833233286646, + "loss": 0.0991, + "step": 25140 + }, + { + "epoch": 1.11, + "learning_rate": 0.0004509660063585229, + "loss": 0.1044, + "step": 25150 + }, + { + "epoch": 1.11, + "learning_rate": 0.0004507436803841793, + "loss": 0.1001, + "step": 25160 + }, + { + "epoch": 1.11, + "learning_rate": 0.00045052135440983574, + "loss": 0.1019, + "step": 25170 + }, + { + "epoch": 1.11, + "learning_rate": 0.00045029902843549217, + "loss": 0.1046, + "step": 25180 + }, + { + "epoch": 1.11, + "learning_rate": 0.0004500767024611486, + "loss": 0.1017, + "step": 25190 + }, + { + "epoch": 1.11, + "learning_rate": 0.0004498543764868049, + "loss": 0.1035, + "step": 25200 + }, + { + "epoch": 1.11, + "learning_rate": 0.00044963205051246134, + "loss": 0.1055, + "step": 25210 + }, + { + "epoch": 1.11, + "learning_rate": 0.00044940972453811777, + "loss": 0.1023, + "step": 25220 + }, + { + "epoch": 1.11, + "learning_rate": 0.0004491873985637742, + "loss": 0.0983, + "step": 25230 + }, + { + "epoch": 1.11, + "learning_rate": 0.0004489650725894306, + "loss": 0.1038, + "step": 25240 + }, + { + "epoch": 1.11, + "learning_rate": 0.00044874274661508705, + "loss": 0.1015, + "step": 25250 + }, + { + "epoch": 1.11, + "learning_rate": 0.0004485204206407434, + "loss": 0.1009, + "step": 25260 + }, + { + "epoch": 1.11, + "learning_rate": 0.00044829809466639985, + "loss": 0.1009, + "step": 25270 + }, + { + "epoch": 1.11, + "learning_rate": 0.0004480757686920563, + "loss": 0.1038, + "step": 25280 + }, + { + "epoch": 1.11, + "learning_rate": 0.0004478534427177127, + "loss": 0.1036, + "step": 25290 + }, + { + "epoch": 1.11, + "learning_rate": 0.00044763111674336913, + "loss": 0.1019, + "step": 25300 + }, + { + "epoch": 1.11, + "learning_rate": 0.00044740879076902556, + "loss": 0.1055, + "step": 25310 + }, + { + "epoch": 1.11, + "learning_rate": 0.00044718646479468193, + "loss": 0.1025, + "step": 25320 + }, + { + "epoch": 1.11, + "learning_rate": 0.00044696413882033836, + "loss": 0.0995, + "step": 25330 + }, + { + "epoch": 1.12, + "learning_rate": 0.0004467418128459948, + "loss": 0.0981, + "step": 25340 + }, + { + "epoch": 1.12, + "learning_rate": 0.0004465194868716512, + "loss": 0.1016, + "step": 25350 + }, + { + "epoch": 1.12, + "learning_rate": 0.00044629716089730764, + "loss": 0.1028, + "step": 25360 + }, + { + "epoch": 1.12, + "learning_rate": 0.00044607483492296407, + "loss": 0.1073, + "step": 25370 + }, + { + "epoch": 1.12, + "learning_rate": 0.00044585250894862044, + "loss": 0.1018, + "step": 25380 + }, + { + "epoch": 1.12, + "learning_rate": 0.00044563018297427687, + "loss": 0.1044, + "step": 25390 + }, + { + "epoch": 1.12, + "learning_rate": 0.0004454078569999333, + "loss": 0.1006, + "step": 25400 + }, + { + "epoch": 1.12, + "learning_rate": 0.0004451855310255897, + "loss": 0.1076, + "step": 25410 + }, + { + "epoch": 1.12, + "learning_rate": 0.00044496320505124615, + "loss": 0.0993, + "step": 25420 + }, + { + "epoch": 1.12, + "learning_rate": 0.0004447408790769026, + "loss": 0.1046, + "step": 25430 + }, + { + "epoch": 1.12, + "learning_rate": 0.00044451855310255895, + "loss": 0.0996, + "step": 25440 + }, + { + "epoch": 1.12, + "learning_rate": 0.0004442962271282154, + "loss": 0.1005, + "step": 25450 + }, + { + "epoch": 1.12, + "learning_rate": 0.0004440739011538718, + "loss": 0.1052, + "step": 25460 + }, + { + "epoch": 1.12, + "learning_rate": 0.00044385157517952824, + "loss": 0.0975, + "step": 25470 + }, + { + "epoch": 1.12, + "learning_rate": 0.00044362924920518466, + "loss": 0.0996, + "step": 25480 + }, + { + "epoch": 1.12, + "learning_rate": 0.0004434069232308411, + "loss": 0.1024, + "step": 25490 + }, + { + "epoch": 1.12, + "learning_rate": 0.0004431845972564975, + "loss": 0.1, + "step": 25500 + }, + { + "epoch": 1.12, + "eval_cer": 0.4995371718769156, + "eval_loss": 0.08730906993150711, + "eval_runtime": 83.8596, + "eval_samples_per_second": 119.175, + "eval_steps_per_second": 7.453, + "step": 25500 + }, + { + "epoch": 1.12, + "learning_rate": 0.0004429622712821539, + "loss": 0.1011, + "step": 25510 + }, + { + "epoch": 1.12, + "learning_rate": 0.0004427399453078103, + "loss": 0.1034, + "step": 25520 + }, + { + "epoch": 1.12, + "learning_rate": 0.00044251761933346675, + "loss": 0.1034, + "step": 25530 + }, + { + "epoch": 1.12, + "learning_rate": 0.0004422952933591232, + "loss": 0.1014, + "step": 25540 + }, + { + "epoch": 1.12, + "learning_rate": 0.0004420729673847796, + "loss": 0.1036, + "step": 25550 + }, + { + "epoch": 1.13, + "learning_rate": 0.00044185064141043603, + "loss": 0.1029, + "step": 25560 + }, + { + "epoch": 1.13, + "learning_rate": 0.0004416283154360924, + "loss": 0.1011, + "step": 25570 + }, + { + "epoch": 1.13, + "learning_rate": 0.00044140598946174883, + "loss": 0.0998, + "step": 25580 + }, + { + "epoch": 1.13, + "learning_rate": 0.00044118366348740526, + "loss": 0.0999, + "step": 25590 + }, + { + "epoch": 1.13, + "learning_rate": 0.0004409613375130617, + "loss": 0.1073, + "step": 25600 + }, + { + "epoch": 1.13, + "learning_rate": 0.0004407390115387181, + "loss": 0.0995, + "step": 25610 + }, + { + "epoch": 1.13, + "learning_rate": 0.00044051668556437454, + "loss": 0.1027, + "step": 25620 + }, + { + "epoch": 1.13, + "learning_rate": 0.0004402943595900309, + "loss": 0.0967, + "step": 25630 + }, + { + "epoch": 1.13, + "learning_rate": 0.00044007203361568734, + "loss": 0.1033, + "step": 25640 + }, + { + "epoch": 1.13, + "learning_rate": 0.00043984970764134377, + "loss": 0.1023, + "step": 25650 + }, + { + "epoch": 1.13, + "learning_rate": 0.0004396273816670002, + "loss": 0.1021, + "step": 25660 + }, + { + "epoch": 1.13, + "learning_rate": 0.0004394050556926566, + "loss": 0.0972, + "step": 25670 + }, + { + "epoch": 1.13, + "learning_rate": 0.00043918272971831305, + "loss": 0.1012, + "step": 25680 + }, + { + "epoch": 1.13, + "learning_rate": 0.00043896040374396937, + "loss": 0.103, + "step": 25690 + }, + { + "epoch": 1.13, + "learning_rate": 0.0004387380777696258, + "loss": 0.0996, + "step": 25700 + }, + { + "epoch": 1.13, + "learning_rate": 0.0004385157517952822, + "loss": 0.1011, + "step": 25710 + }, + { + "epoch": 1.13, + "learning_rate": 0.00043829342582093865, + "loss": 0.1092, + "step": 25720 + }, + { + "epoch": 1.13, + "learning_rate": 0.0004380710998465951, + "loss": 0.0992, + "step": 25730 + }, + { + "epoch": 1.13, + "learning_rate": 0.0004378487738722515, + "loss": 0.1068, + "step": 25740 + }, + { + "epoch": 1.13, + "learning_rate": 0.0004376264478979079, + "loss": 0.1013, + "step": 25750 + }, + { + "epoch": 1.13, + "learning_rate": 0.0004374041219235643, + "loss": 0.1027, + "step": 25760 + }, + { + "epoch": 1.13, + "learning_rate": 0.00043718179594922073, + "loss": 0.1011, + "step": 25770 + }, + { + "epoch": 1.13, + "learning_rate": 0.00043695946997487716, + "loss": 0.1022, + "step": 25780 + }, + { + "epoch": 1.14, + "learning_rate": 0.0004367371440005336, + "loss": 0.1035, + "step": 25790 + }, + { + "epoch": 1.14, + "learning_rate": 0.00043651481802619, + "loss": 0.109, + "step": 25800 + }, + { + "epoch": 1.14, + "learning_rate": 0.0004362924920518464, + "loss": 0.1029, + "step": 25810 + }, + { + "epoch": 1.14, + "learning_rate": 0.0004360701660775028, + "loss": 0.0985, + "step": 25820 + }, + { + "epoch": 1.14, + "learning_rate": 0.00043584784010315924, + "loss": 0.1077, + "step": 25830 + }, + { + "epoch": 1.14, + "learning_rate": 0.00043562551412881567, + "loss": 0.1004, + "step": 25840 + }, + { + "epoch": 1.14, + "learning_rate": 0.0004354031881544721, + "loss": 0.0984, + "step": 25850 + }, + { + "epoch": 1.14, + "learning_rate": 0.0004351808621801285, + "loss": 0.1054, + "step": 25860 + }, + { + "epoch": 1.14, + "learning_rate": 0.0004349585362057849, + "loss": 0.1022, + "step": 25870 + }, + { + "epoch": 1.14, + "learning_rate": 0.0004347362102314413, + "loss": 0.1034, + "step": 25880 + }, + { + "epoch": 1.14, + "learning_rate": 0.00043451388425709775, + "loss": 0.1058, + "step": 25890 + }, + { + "epoch": 1.14, + "learning_rate": 0.0004342915582827542, + "loss": 0.099, + "step": 25900 + }, + { + "epoch": 1.14, + "learning_rate": 0.0004340692323084106, + "loss": 0.0995, + "step": 25910 + }, + { + "epoch": 1.14, + "learning_rate": 0.00043384690633406703, + "loss": 0.1033, + "step": 25920 + }, + { + "epoch": 1.14, + "learning_rate": 0.0004336245803597234, + "loss": 0.1018, + "step": 25930 + }, + { + "epoch": 1.14, + "learning_rate": 0.00043340225438537984, + "loss": 0.1013, + "step": 25940 + }, + { + "epoch": 1.14, + "learning_rate": 0.00043317992841103626, + "loss": 0.1015, + "step": 25950 + }, + { + "epoch": 1.14, + "learning_rate": 0.0004329576024366927, + "loss": 0.1015, + "step": 25960 + }, + { + "epoch": 1.14, + "learning_rate": 0.0004327352764623491, + "loss": 0.1019, + "step": 25970 + }, + { + "epoch": 1.14, + "learning_rate": 0.00043251295048800554, + "loss": 0.0932, + "step": 25980 + }, + { + "epoch": 1.14, + "learning_rate": 0.00043229062451366197, + "loss": 0.1032, + "step": 25990 + }, + { + "epoch": 1.14, + "learning_rate": 0.00043206829853931835, + "loss": 0.103, + "step": 26000 + }, + { + "epoch": 1.14, + "eval_cer": 0.4998037764439621, + "eval_loss": 0.08562103658914566, + "eval_runtime": 83.8891, + "eval_samples_per_second": 119.134, + "eval_steps_per_second": 7.45, + "step": 26000 + }, + { + "epoch": 1.14, + "learning_rate": 0.00043184597256497477, + "loss": 0.0992, + "step": 26010 + }, + { + "epoch": 1.15, + "learning_rate": 0.0004316236465906312, + "loss": 0.1044, + "step": 26020 + }, + { + "epoch": 1.15, + "learning_rate": 0.00043140132061628763, + "loss": 0.1032, + "step": 26030 + }, + { + "epoch": 1.15, + "learning_rate": 0.00043117899464194405, + "loss": 0.1029, + "step": 26040 + }, + { + "epoch": 1.15, + "learning_rate": 0.0004309566686676005, + "loss": 0.1051, + "step": 26050 + }, + { + "epoch": 1.15, + "learning_rate": 0.00043073434269325686, + "loss": 0.1054, + "step": 26060 + }, + { + "epoch": 1.15, + "learning_rate": 0.0004305120167189133, + "loss": 0.104, + "step": 26070 + }, + { + "epoch": 1.15, + "learning_rate": 0.0004302896907445697, + "loss": 0.1014, + "step": 26080 + }, + { + "epoch": 1.15, + "learning_rate": 0.00043006736477022614, + "loss": 0.1041, + "step": 26090 + }, + { + "epoch": 1.15, + "learning_rate": 0.00042984503879588256, + "loss": 0.0986, + "step": 26100 + }, + { + "epoch": 1.15, + "learning_rate": 0.000429622712821539, + "loss": 0.1034, + "step": 26110 + }, + { + "epoch": 1.15, + "learning_rate": 0.00042940038684719537, + "loss": 0.0998, + "step": 26120 + }, + { + "epoch": 1.15, + "learning_rate": 0.0004291780608728518, + "loss": 0.1038, + "step": 26130 + }, + { + "epoch": 1.15, + "learning_rate": 0.0004289557348985082, + "loss": 0.0975, + "step": 26140 + }, + { + "epoch": 1.15, + "learning_rate": 0.00042873340892416465, + "loss": 0.1044, + "step": 26150 + }, + { + "epoch": 1.15, + "learning_rate": 0.0004285110829498211, + "loss": 0.0984, + "step": 26160 + }, + { + "epoch": 1.15, + "learning_rate": 0.0004282887569754775, + "loss": 0.1023, + "step": 26170 + }, + { + "epoch": 1.15, + "learning_rate": 0.0004280664310011339, + "loss": 0.1035, + "step": 26180 + }, + { + "epoch": 1.15, + "learning_rate": 0.0004278441050267903, + "loss": 0.0992, + "step": 26190 + }, + { + "epoch": 1.15, + "learning_rate": 0.0004276217790524467, + "loss": 0.0986, + "step": 26200 + }, + { + "epoch": 1.15, + "learning_rate": 0.0004273994530781031, + "loss": 0.0999, + "step": 26210 + }, + { + "epoch": 1.15, + "learning_rate": 0.00042717712710375953, + "loss": 0.0996, + "step": 26220 + }, + { + "epoch": 1.15, + "learning_rate": 0.00042695480112941596, + "loss": 0.0995, + "step": 26230 + }, + { + "epoch": 1.16, + "learning_rate": 0.00042673247515507233, + "loss": 0.0998, + "step": 26240 + }, + { + "epoch": 1.16, + "learning_rate": 0.00042651014918072876, + "loss": 0.098, + "step": 26250 + }, + { + "epoch": 1.16, + "learning_rate": 0.0004262878232063852, + "loss": 0.1074, + "step": 26260 + }, + { + "epoch": 1.16, + "learning_rate": 0.0004260654972320416, + "loss": 0.1031, + "step": 26270 + }, + { + "epoch": 1.16, + "learning_rate": 0.00042584317125769804, + "loss": 0.1008, + "step": 26280 + }, + { + "epoch": 1.16, + "learning_rate": 0.00042562084528335447, + "loss": 0.102, + "step": 26290 + }, + { + "epoch": 1.16, + "learning_rate": 0.00042539851930901084, + "loss": 0.097, + "step": 26300 + }, + { + "epoch": 1.16, + "learning_rate": 0.00042517619333466727, + "loss": 0.0983, + "step": 26310 + }, + { + "epoch": 1.16, + "learning_rate": 0.0004249538673603237, + "loss": 0.0954, + "step": 26320 + }, + { + "epoch": 1.16, + "learning_rate": 0.0004247315413859801, + "loss": 0.1066, + "step": 26330 + }, + { + "epoch": 1.16, + "learning_rate": 0.00042450921541163655, + "loss": 0.1003, + "step": 26340 + }, + { + "epoch": 1.16, + "learning_rate": 0.000424286889437293, + "loss": 0.1025, + "step": 26350 + }, + { + "epoch": 1.16, + "learning_rate": 0.00042406456346294935, + "loss": 0.1063, + "step": 26360 + }, + { + "epoch": 1.16, + "learning_rate": 0.0004238422374886058, + "loss": 0.1054, + "step": 26370 + }, + { + "epoch": 1.16, + "learning_rate": 0.0004236199115142622, + "loss": 0.1044, + "step": 26380 + }, + { + "epoch": 1.16, + "learning_rate": 0.00042339758553991863, + "loss": 0.105, + "step": 26390 + }, + { + "epoch": 1.16, + "learning_rate": 0.00042317525956557506, + "loss": 0.1031, + "step": 26400 + }, + { + "epoch": 1.16, + "learning_rate": 0.0004229529335912315, + "loss": 0.0989, + "step": 26410 + }, + { + "epoch": 1.16, + "learning_rate": 0.00042273060761688786, + "loss": 0.1025, + "step": 26420 + }, + { + "epoch": 1.16, + "learning_rate": 0.0004225082816425443, + "loss": 0.1039, + "step": 26430 + }, + { + "epoch": 1.16, + "learning_rate": 0.0004222859556682007, + "loss": 0.1011, + "step": 26440 + }, + { + "epoch": 1.16, + "learning_rate": 0.00042206362969385714, + "loss": 0.0942, + "step": 26450 + }, + { + "epoch": 1.16, + "learning_rate": 0.00042184130371951357, + "loss": 0.1023, + "step": 26460 + }, + { + "epoch": 1.17, + "learning_rate": 0.00042161897774517, + "loss": 0.102, + "step": 26470 + }, + { + "epoch": 1.17, + "learning_rate": 0.00042139665177082637, + "loss": 0.0984, + "step": 26480 + }, + { + "epoch": 1.17, + "learning_rate": 0.0004211743257964828, + "loss": 0.1045, + "step": 26490 + }, + { + "epoch": 1.17, + "learning_rate": 0.0004209519998221392, + "loss": 0.0971, + "step": 26500 + }, + { + "epoch": 1.17, + "eval_cer": 0.4994301408463495, + "eval_loss": 0.085819311439991, + "eval_runtime": 83.9721, + "eval_samples_per_second": 119.016, + "eval_steps_per_second": 7.443, + "step": 26500 + }, + { + "epoch": 1.17, + "learning_rate": 0.00042072967384779565, + "loss": 0.0997, + "step": 26510 + }, + { + "epoch": 1.17, + "learning_rate": 0.0004205073478734521, + "loss": 0.1008, + "step": 26520 + }, + { + "epoch": 1.17, + "learning_rate": 0.0004202850218991085, + "loss": 0.1006, + "step": 26530 + }, + { + "epoch": 1.17, + "learning_rate": 0.00042006269592476494, + "loss": 0.0999, + "step": 26540 + }, + { + "epoch": 1.17, + "learning_rate": 0.0004198403699504213, + "loss": 0.1069, + "step": 26550 + }, + { + "epoch": 1.17, + "learning_rate": 0.00041961804397607774, + "loss": 0.1001, + "step": 26560 + }, + { + "epoch": 1.17, + "learning_rate": 0.00041939571800173416, + "loss": 0.1012, + "step": 26570 + }, + { + "epoch": 1.17, + "learning_rate": 0.0004191733920273906, + "loss": 0.1014, + "step": 26580 + }, + { + "epoch": 1.17, + "learning_rate": 0.000418951066053047, + "loss": 0.0991, + "step": 26590 + }, + { + "epoch": 1.17, + "learning_rate": 0.00041872874007870345, + "loss": 0.1044, + "step": 26600 + }, + { + "epoch": 1.17, + "learning_rate": 0.0004185064141043598, + "loss": 0.1001, + "step": 26610 + }, + { + "epoch": 1.17, + "learning_rate": 0.00041828408813001625, + "loss": 0.093, + "step": 26620 + }, + { + "epoch": 1.17, + "learning_rate": 0.0004180617621556727, + "loss": 0.1016, + "step": 26630 + }, + { + "epoch": 1.17, + "learning_rate": 0.0004178394361813291, + "loss": 0.0997, + "step": 26640 + }, + { + "epoch": 1.17, + "learning_rate": 0.00041761711020698553, + "loss": 0.1006, + "step": 26650 + }, + { + "epoch": 1.17, + "learning_rate": 0.00041739478423264196, + "loss": 0.1069, + "step": 26660 + }, + { + "epoch": 1.17, + "learning_rate": 0.00041717245825829833, + "loss": 0.1024, + "step": 26670 + }, + { + "epoch": 1.17, + "learning_rate": 0.00041695013228395476, + "loss": 0.1015, + "step": 26680 + }, + { + "epoch": 1.17, + "learning_rate": 0.0004167278063096112, + "loss": 0.1024, + "step": 26690 + }, + { + "epoch": 1.18, + "learning_rate": 0.00041650548033526756, + "loss": 0.1104, + "step": 26700 + }, + { + "epoch": 1.18, + "learning_rate": 0.000416283154360924, + "loss": 0.1025, + "step": 26710 + }, + { + "epoch": 1.18, + "learning_rate": 0.0004160608283865804, + "loss": 0.0991, + "step": 26720 + }, + { + "epoch": 1.18, + "learning_rate": 0.0004158385024122368, + "loss": 0.0969, + "step": 26730 + }, + { + "epoch": 1.18, + "learning_rate": 0.0004156161764378932, + "loss": 0.0992, + "step": 26740 + }, + { + "epoch": 1.18, + "learning_rate": 0.00041539385046354964, + "loss": 0.0998, + "step": 26750 + }, + { + "epoch": 1.18, + "learning_rate": 0.00041517152448920607, + "loss": 0.1037, + "step": 26760 + }, + { + "epoch": 1.18, + "learning_rate": 0.0004149491985148625, + "loss": 0.1006, + "step": 26770 + }, + { + "epoch": 1.18, + "learning_rate": 0.0004147268725405189, + "loss": 0.1005, + "step": 26780 + }, + { + "epoch": 1.18, + "learning_rate": 0.0004145045465661753, + "loss": 0.1013, + "step": 26790 + }, + { + "epoch": 1.18, + "learning_rate": 0.0004142822205918317, + "loss": 0.1032, + "step": 26800 + }, + { + "epoch": 1.18, + "learning_rate": 0.00041405989461748815, + "loss": 0.0989, + "step": 26810 + }, + { + "epoch": 1.18, + "learning_rate": 0.0004138375686431446, + "loss": 0.1017, + "step": 26820 + }, + { + "epoch": 1.18, + "learning_rate": 0.000413615242668801, + "loss": 0.1024, + "step": 26830 + }, + { + "epoch": 1.18, + "learning_rate": 0.00041339291669445743, + "loss": 0.1004, + "step": 26840 + }, + { + "epoch": 1.18, + "learning_rate": 0.0004131705907201138, + "loss": 0.0997, + "step": 26850 + }, + { + "epoch": 1.18, + "learning_rate": 0.00041294826474577023, + "loss": 0.1026, + "step": 26860 + }, + { + "epoch": 1.18, + "learning_rate": 0.00041272593877142666, + "loss": 0.1014, + "step": 26870 + }, + { + "epoch": 1.18, + "learning_rate": 0.0004125036127970831, + "loss": 0.1015, + "step": 26880 + }, + { + "epoch": 1.18, + "learning_rate": 0.0004122812868227395, + "loss": 0.0997, + "step": 26890 + }, + { + "epoch": 1.18, + "learning_rate": 0.00041205896084839594, + "loss": 0.1008, + "step": 26900 + }, + { + "epoch": 1.18, + "learning_rate": 0.0004118366348740523, + "loss": 0.1015, + "step": 26910 + }, + { + "epoch": 1.18, + "learning_rate": 0.00041161430889970874, + "loss": 0.1027, + "step": 26920 + }, + { + "epoch": 1.19, + "learning_rate": 0.00041139198292536517, + "loss": 0.098, + "step": 26930 + }, + { + "epoch": 1.19, + "learning_rate": 0.0004111696569510216, + "loss": 0.0972, + "step": 26940 + }, + { + "epoch": 1.19, + "learning_rate": 0.000410947330976678, + "loss": 0.1047, + "step": 26950 + }, + { + "epoch": 1.19, + "learning_rate": 0.00041072500500233445, + "loss": 0.1014, + "step": 26960 + }, + { + "epoch": 1.19, + "learning_rate": 0.0004105026790279908, + "loss": 0.104, + "step": 26970 + }, + { + "epoch": 1.19, + "learning_rate": 0.00041028035305364725, + "loss": 0.1022, + "step": 26980 + }, + { + "epoch": 1.19, + "learning_rate": 0.0004100580270793037, + "loss": 0.103, + "step": 26990 + }, + { + "epoch": 1.19, + "learning_rate": 0.0004098357011049601, + "loss": 0.101, + "step": 27000 + }, + { + "epoch": 1.19, + "eval_cer": 0.49954819931642847, + "eval_loss": 0.08485907316207886, + "eval_runtime": 83.7713, + "eval_samples_per_second": 119.301, + "eval_steps_per_second": 7.461, + "step": 27000 + }, + { + "epoch": 1.19, + "learning_rate": 0.00040961337513061653, + "loss": 0.0989, + "step": 27010 + }, + { + "epoch": 1.19, + "learning_rate": 0.00040939104915627296, + "loss": 0.1024, + "step": 27020 + }, + { + "epoch": 1.19, + "learning_rate": 0.0004091687231819294, + "loss": 0.097, + "step": 27030 + }, + { + "epoch": 1.19, + "learning_rate": 0.00040894639720758576, + "loss": 0.0955, + "step": 27040 + }, + { + "epoch": 1.19, + "learning_rate": 0.0004087240712332422, + "loss": 0.0984, + "step": 27050 + }, + { + "epoch": 1.19, + "learning_rate": 0.0004085017452588986, + "loss": 0.1025, + "step": 27060 + }, + { + "epoch": 1.19, + "learning_rate": 0.00040827941928455505, + "loss": 0.1023, + "step": 27070 + }, + { + "epoch": 1.19, + "learning_rate": 0.00040805709331021147, + "loss": 0.1043, + "step": 27080 + }, + { + "epoch": 1.19, + "learning_rate": 0.0004078347673358679, + "loss": 0.1016, + "step": 27090 + }, + { + "epoch": 1.19, + "learning_rate": 0.0004076124413615243, + "loss": 0.0965, + "step": 27100 + }, + { + "epoch": 1.19, + "learning_rate": 0.0004073901153871807, + "loss": 0.0978, + "step": 27110 + }, + { + "epoch": 1.19, + "learning_rate": 0.00040716778941283713, + "loss": 0.0994, + "step": 27120 + }, + { + "epoch": 1.19, + "learning_rate": 0.00040694546343849356, + "loss": 0.0956, + "step": 27130 + }, + { + "epoch": 1.19, + "learning_rate": 0.00040672313746415, + "loss": 0.0975, + "step": 27140 + }, + { + "epoch": 1.2, + "learning_rate": 0.0004065008114898064, + "loss": 0.0978, + "step": 27150 + }, + { + "epoch": 1.2, + "learning_rate": 0.0004062784855154628, + "loss": 0.1002, + "step": 27160 + }, + { + "epoch": 1.2, + "learning_rate": 0.0004060561595411192, + "loss": 0.0989, + "step": 27170 + }, + { + "epoch": 1.2, + "learning_rate": 0.00040583383356677564, + "loss": 0.1005, + "step": 27180 + }, + { + "epoch": 1.2, + "learning_rate": 0.00040561150759243207, + "loss": 0.0974, + "step": 27190 + }, + { + "epoch": 1.2, + "learning_rate": 0.00040538918161808844, + "loss": 0.0974, + "step": 27200 + }, + { + "epoch": 1.2, + "learning_rate": 0.00040516685564374487, + "loss": 0.1048, + "step": 27210 + }, + { + "epoch": 1.2, + "learning_rate": 0.00040494452966940124, + "loss": 0.0965, + "step": 27220 + }, + { + "epoch": 1.2, + "learning_rate": 0.00040472220369505767, + "loss": 0.0998, + "step": 27230 + }, + { + "epoch": 1.2, + "learning_rate": 0.0004044998777207141, + "loss": 0.0976, + "step": 27240 + }, + { + "epoch": 1.2, + "learning_rate": 0.0004042775517463705, + "loss": 0.1027, + "step": 27250 + }, + { + "epoch": 1.2, + "learning_rate": 0.00040405522577202695, + "loss": 0.0993, + "step": 27260 + }, + { + "epoch": 1.2, + "learning_rate": 0.0004038328997976834, + "loss": 0.0999, + "step": 27270 + }, + { + "epoch": 1.2, + "learning_rate": 0.00040361057382333975, + "loss": 0.1066, + "step": 27280 + }, + { + "epoch": 1.2, + "learning_rate": 0.0004033882478489962, + "loss": 0.1051, + "step": 27290 + }, + { + "epoch": 1.2, + "learning_rate": 0.0004031659218746526, + "loss": 0.097, + "step": 27300 + }, + { + "epoch": 1.2, + "learning_rate": 0.00040294359590030903, + "loss": 0.0953, + "step": 27310 + }, + { + "epoch": 1.2, + "learning_rate": 0.00040272126992596546, + "loss": 0.108, + "step": 27320 + }, + { + "epoch": 1.2, + "learning_rate": 0.0004024989439516219, + "loss": 0.0962, + "step": 27330 + }, + { + "epoch": 1.2, + "learning_rate": 0.00040227661797727826, + "loss": 0.0998, + "step": 27340 + }, + { + "epoch": 1.2, + "learning_rate": 0.0004020542920029347, + "loss": 0.1029, + "step": 27350 + }, + { + "epoch": 1.2, + "learning_rate": 0.0004018319660285911, + "loss": 0.105, + "step": 27360 + }, + { + "epoch": 1.2, + "learning_rate": 0.00040160964005424754, + "loss": 0.101, + "step": 27370 + }, + { + "epoch": 1.21, + "learning_rate": 0.00040138731407990397, + "loss": 0.1004, + "step": 27380 + }, + { + "epoch": 1.21, + "learning_rate": 0.0004011649881055604, + "loss": 0.0978, + "step": 27390 + }, + { + "epoch": 1.21, + "learning_rate": 0.00040094266213121677, + "loss": 0.1039, + "step": 27400 + }, + { + "epoch": 1.21, + "learning_rate": 0.0004007203361568732, + "loss": 0.1017, + "step": 27410 + }, + { + "epoch": 1.21, + "learning_rate": 0.0004004980101825296, + "loss": 0.1028, + "step": 27420 + }, + { + "epoch": 1.21, + "learning_rate": 0.00040027568420818605, + "loss": 0.1006, + "step": 27430 + }, + { + "epoch": 1.21, + "learning_rate": 0.0004000533582338425, + "loss": 0.1031, + "step": 27440 + }, + { + "epoch": 1.21, + "learning_rate": 0.0003998310322594989, + "loss": 0.1035, + "step": 27450 + }, + { + "epoch": 1.21, + "learning_rate": 0.0003996087062851553, + "loss": 0.1039, + "step": 27460 + }, + { + "epoch": 1.21, + "learning_rate": 0.0003993863803108117, + "loss": 0.0991, + "step": 27470 + }, + { + "epoch": 1.21, + "learning_rate": 0.00039916405433646813, + "loss": 0.0943, + "step": 27480 + }, + { + "epoch": 1.21, + "learning_rate": 0.00039894172836212456, + "loss": 0.1001, + "step": 27490 + }, + { + "epoch": 1.21, + "learning_rate": 0.000398719402387781, + "loss": 0.1034, + "step": 27500 + }, + { + "epoch": 1.21, + "eval_cer": 0.4996584737115572, + "eval_loss": 0.08409909158945084, + "eval_runtime": 83.873, + "eval_samples_per_second": 119.156, + "eval_steps_per_second": 7.452, + "step": 27500 + }, + { + "epoch": 1.21, + "learning_rate": 0.0003984970764134374, + "loss": 0.102, + "step": 27510 + }, + { + "epoch": 1.21, + "learning_rate": 0.0003982747504390938, + "loss": 0.0949, + "step": 27520 + }, + { + "epoch": 1.21, + "learning_rate": 0.0003980524244647502, + "loss": 0.0988, + "step": 27530 + }, + { + "epoch": 1.21, + "learning_rate": 0.00039783009849040664, + "loss": 0.1053, + "step": 27540 + }, + { + "epoch": 1.21, + "learning_rate": 0.00039760777251606307, + "loss": 0.1029, + "step": 27550 + }, + { + "epoch": 1.21, + "learning_rate": 0.0003973854465417195, + "loss": 0.1011, + "step": 27560 + }, + { + "epoch": 1.21, + "learning_rate": 0.0003971631205673759, + "loss": 0.0998, + "step": 27570 + }, + { + "epoch": 1.21, + "learning_rate": 0.00039694079459303235, + "loss": 0.1011, + "step": 27580 + }, + { + "epoch": 1.21, + "learning_rate": 0.0003967184686186887, + "loss": 0.1037, + "step": 27590 + }, + { + "epoch": 1.21, + "learning_rate": 0.00039649614264434515, + "loss": 0.0976, + "step": 27600 + }, + { + "epoch": 1.22, + "learning_rate": 0.0003962738166700016, + "loss": 0.099, + "step": 27610 + }, + { + "epoch": 1.22, + "learning_rate": 0.000396051490695658, + "loss": 0.1016, + "step": 27620 + }, + { + "epoch": 1.22, + "learning_rate": 0.00039582916472131444, + "loss": 0.1003, + "step": 27630 + }, + { + "epoch": 1.22, + "learning_rate": 0.00039560683874697086, + "loss": 0.0943, + "step": 27640 + }, + { + "epoch": 1.22, + "learning_rate": 0.00039538451277262724, + "loss": 0.0974, + "step": 27650 + }, + { + "epoch": 1.22, + "learning_rate": 0.00039516218679828366, + "loss": 0.1016, + "step": 27660 + }, + { + "epoch": 1.22, + "learning_rate": 0.0003949398608239401, + "loss": 0.1011, + "step": 27670 + }, + { + "epoch": 1.22, + "learning_rate": 0.0003947175348495965, + "loss": 0.0974, + "step": 27680 + }, + { + "epoch": 1.22, + "learning_rate": 0.00039449520887525295, + "loss": 0.0973, + "step": 27690 + }, + { + "epoch": 1.22, + "learning_rate": 0.0003942728829009093, + "loss": 0.0994, + "step": 27700 + }, + { + "epoch": 1.22, + "learning_rate": 0.0003940505569265657, + "loss": 0.0977, + "step": 27710 + }, + { + "epoch": 1.22, + "learning_rate": 0.0003938282309522221, + "loss": 0.0976, + "step": 27720 + }, + { + "epoch": 1.22, + "learning_rate": 0.00039360590497787855, + "loss": 0.103, + "step": 27730 + }, + { + "epoch": 1.22, + "learning_rate": 0.000393383579003535, + "loss": 0.0995, + "step": 27740 + }, + { + "epoch": 1.22, + "learning_rate": 0.0003931612530291914, + "loss": 0.0974, + "step": 27750 + }, + { + "epoch": 1.22, + "learning_rate": 0.00039293892705484783, + "loss": 0.1012, + "step": 27760 + }, + { + "epoch": 1.22, + "learning_rate": 0.0003927166010805042, + "loss": 0.1051, + "step": 27770 + }, + { + "epoch": 1.22, + "learning_rate": 0.00039249427510616063, + "loss": 0.1021, + "step": 27780 + }, + { + "epoch": 1.22, + "learning_rate": 0.00039227194913181706, + "loss": 0.0993, + "step": 27790 + }, + { + "epoch": 1.22, + "learning_rate": 0.0003920496231574735, + "loss": 0.102, + "step": 27800 + }, + { + "epoch": 1.22, + "learning_rate": 0.0003918272971831299, + "loss": 0.1014, + "step": 27810 + }, + { + "epoch": 1.22, + "learning_rate": 0.00039160497120878634, + "loss": 0.1085, + "step": 27820 + }, + { + "epoch": 1.23, + "learning_rate": 0.0003913826452344427, + "loss": 0.1024, + "step": 27830 + }, + { + "epoch": 1.23, + "learning_rate": 0.00039116031926009914, + "loss": 0.0977, + "step": 27840 + }, + { + "epoch": 1.23, + "learning_rate": 0.00039093799328575557, + "loss": 0.096, + "step": 27850 + }, + { + "epoch": 1.23, + "learning_rate": 0.000390715667311412, + "loss": 0.1012, + "step": 27860 + }, + { + "epoch": 1.23, + "learning_rate": 0.0003904933413370684, + "loss": 0.1024, + "step": 27870 + }, + { + "epoch": 1.23, + "learning_rate": 0.00039027101536272485, + "loss": 0.0999, + "step": 27880 + }, + { + "epoch": 1.23, + "learning_rate": 0.0003900486893883812, + "loss": 0.1033, + "step": 27890 + }, + { + "epoch": 1.23, + "learning_rate": 0.00038982636341403765, + "loss": 0.1007, + "step": 27900 + }, + { + "epoch": 1.23, + "learning_rate": 0.0003896040374396941, + "loss": 0.0954, + "step": 27910 + }, + { + "epoch": 1.23, + "learning_rate": 0.0003893817114653505, + "loss": 0.0994, + "step": 27920 + }, + { + "epoch": 1.23, + "learning_rate": 0.00038915938549100693, + "loss": 0.0993, + "step": 27930 + }, + { + "epoch": 1.23, + "learning_rate": 0.00038893705951666336, + "loss": 0.1027, + "step": 27940 + }, + { + "epoch": 1.23, + "learning_rate": 0.00038871473354231973, + "loss": 0.0972, + "step": 27950 + }, + { + "epoch": 1.23, + "learning_rate": 0.00038849240756797616, + "loss": 0.0982, + "step": 27960 + }, + { + "epoch": 1.23, + "learning_rate": 0.0003882700815936326, + "loss": 0.1002, + "step": 27970 + }, + { + "epoch": 1.23, + "learning_rate": 0.000388047755619289, + "loss": 0.0975, + "step": 27980 + }, + { + "epoch": 1.23, + "learning_rate": 0.00038782542964494544, + "loss": 0.0974, + "step": 27990 + }, + { + "epoch": 1.23, + "learning_rate": 0.00038760310367060187, + "loss": 0.0967, + "step": 28000 + }, + { + "epoch": 1.23, + "eval_cer": 0.49939965321946095, + "eval_loss": 0.08435171097517014, + "eval_runtime": 83.6514, + "eval_samples_per_second": 119.472, + "eval_steps_per_second": 7.471, + "step": 28000 + }, + { + "epoch": 1.23, + "learning_rate": 0.00038738077769625824, + "loss": 0.0986, + "step": 28010 + }, + { + "epoch": 1.23, + "learning_rate": 0.00038715845172191467, + "loss": 0.0983, + "step": 28020 + }, + { + "epoch": 1.23, + "learning_rate": 0.0003869361257475711, + "loss": 0.1006, + "step": 28030 + }, + { + "epoch": 1.23, + "learning_rate": 0.0003867137997732275, + "loss": 0.0976, + "step": 28040 + }, + { + "epoch": 1.23, + "learning_rate": 0.00038649147379888395, + "loss": 0.0973, + "step": 28050 + }, + { + "epoch": 1.24, + "learning_rate": 0.0003862691478245404, + "loss": 0.0999, + "step": 28060 + }, + { + "epoch": 1.24, + "learning_rate": 0.00038604682185019675, + "loss": 0.104, + "step": 28070 + }, + { + "epoch": 1.24, + "learning_rate": 0.0003858244958758532, + "loss": 0.0938, + "step": 28080 + }, + { + "epoch": 1.24, + "learning_rate": 0.0003856021699015096, + "loss": 0.0987, + "step": 28090 + }, + { + "epoch": 1.24, + "learning_rate": 0.00038537984392716604, + "loss": 0.1014, + "step": 28100 + }, + { + "epoch": 1.24, + "learning_rate": 0.00038515751795282246, + "loss": 0.0967, + "step": 28110 + }, + { + "epoch": 1.24, + "learning_rate": 0.0003849351919784789, + "loss": 0.0989, + "step": 28120 + }, + { + "epoch": 1.24, + "learning_rate": 0.0003847128660041353, + "loss": 0.0973, + "step": 28130 + }, + { + "epoch": 1.24, + "learning_rate": 0.0003844905400297917, + "loss": 0.1004, + "step": 28140 + }, + { + "epoch": 1.24, + "learning_rate": 0.0003842682140554481, + "loss": 0.0941, + "step": 28150 + }, + { + "epoch": 1.24, + "learning_rate": 0.00038404588808110455, + "loss": 0.102, + "step": 28160 + }, + { + "epoch": 1.24, + "learning_rate": 0.000383823562106761, + "loss": 0.1001, + "step": 28170 + }, + { + "epoch": 1.24, + "learning_rate": 0.0003836012361324174, + "loss": 0.0984, + "step": 28180 + }, + { + "epoch": 1.24, + "learning_rate": 0.00038337891015807383, + "loss": 0.0969, + "step": 28190 + }, + { + "epoch": 1.24, + "learning_rate": 0.00038315658418373015, + "loss": 0.0954, + "step": 28200 + }, + { + "epoch": 1.24, + "learning_rate": 0.0003829342582093866, + "loss": 0.1012, + "step": 28210 + }, + { + "epoch": 1.24, + "learning_rate": 0.000382711932235043, + "loss": 0.1002, + "step": 28220 + }, + { + "epoch": 1.24, + "learning_rate": 0.00038248960626069943, + "loss": 0.0981, + "step": 28230 + }, + { + "epoch": 1.24, + "learning_rate": 0.00038226728028635586, + "loss": 0.0975, + "step": 28240 + }, + { + "epoch": 1.24, + "learning_rate": 0.0003820449543120123, + "loss": 0.0969, + "step": 28250 + }, + { + "epoch": 1.24, + "learning_rate": 0.00038182262833766866, + "loss": 0.1022, + "step": 28260 + }, + { + "epoch": 1.24, + "learning_rate": 0.0003816003023633251, + "loss": 0.1032, + "step": 28270 + }, + { + "epoch": 1.24, + "learning_rate": 0.0003813779763889815, + "loss": 0.0973, + "step": 28280 + }, + { + "epoch": 1.25, + "learning_rate": 0.00038115565041463794, + "loss": 0.0962, + "step": 28290 + }, + { + "epoch": 1.25, + "learning_rate": 0.00038093332444029437, + "loss": 0.0954, + "step": 28300 + }, + { + "epoch": 1.25, + "learning_rate": 0.0003807109984659508, + "loss": 0.0974, + "step": 28310 + }, + { + "epoch": 1.25, + "learning_rate": 0.00038048867249160717, + "loss": 0.0989, + "step": 28320 + }, + { + "epoch": 1.25, + "learning_rate": 0.0003802663465172636, + "loss": 0.1035, + "step": 28330 + }, + { + "epoch": 1.25, + "learning_rate": 0.00038004402054292, + "loss": 0.1023, + "step": 28340 + }, + { + "epoch": 1.25, + "learning_rate": 0.00037982169456857645, + "loss": 0.1, + "step": 28350 + }, + { + "epoch": 1.25, + "learning_rate": 0.0003795993685942329, + "loss": 0.1009, + "step": 28360 + }, + { + "epoch": 1.25, + "learning_rate": 0.0003793770426198893, + "loss": 0.0994, + "step": 28370 + }, + { + "epoch": 1.25, + "learning_rate": 0.0003791547166455457, + "loss": 0.099, + "step": 28380 + }, + { + "epoch": 1.25, + "learning_rate": 0.0003789323906712021, + "loss": 0.0989, + "step": 28390 + }, + { + "epoch": 1.25, + "learning_rate": 0.00037871006469685853, + "loss": 0.0985, + "step": 28400 + }, + { + "epoch": 1.25, + "learning_rate": 0.00037848773872251496, + "loss": 0.1003, + "step": 28410 + }, + { + "epoch": 1.25, + "learning_rate": 0.0003782654127481714, + "loss": 0.1031, + "step": 28420 + }, + { + "epoch": 1.25, + "learning_rate": 0.0003780430867738278, + "loss": 0.1024, + "step": 28430 + }, + { + "epoch": 1.25, + "learning_rate": 0.0003778207607994842, + "loss": 0.1058, + "step": 28440 + }, + { + "epoch": 1.25, + "learning_rate": 0.0003775984348251406, + "loss": 0.0983, + "step": 28450 + }, + { + "epoch": 1.25, + "learning_rate": 0.00037737610885079704, + "loss": 0.0948, + "step": 28460 + }, + { + "epoch": 1.25, + "learning_rate": 0.00037715378287645347, + "loss": 0.0981, + "step": 28470 + }, + { + "epoch": 1.25, + "learning_rate": 0.0003769314569021099, + "loss": 0.0992, + "step": 28480 + }, + { + "epoch": 1.25, + "learning_rate": 0.0003767091309277663, + "loss": 0.0975, + "step": 28490 + }, + { + "epoch": 1.25, + "learning_rate": 0.0003764868049534227, + "loss": 0.0979, + "step": 28500 + }, + { + "epoch": 1.25, + "eval_cer": 0.49911229111921374, + "eval_loss": 0.08381886035203934, + "eval_runtime": 83.8853, + "eval_samples_per_second": 119.139, + "eval_steps_per_second": 7.451, + "step": 28500 + }, + { + "epoch": 1.25, + "learning_rate": 0.0003762644789790791, + "loss": 0.1039, + "step": 28510 + }, + { + "epoch": 1.26, + "learning_rate": 0.00037604215300473555, + "loss": 0.0979, + "step": 28520 + }, + { + "epoch": 1.26, + "learning_rate": 0.000375819827030392, + "loss": 0.1009, + "step": 28530 + }, + { + "epoch": 1.26, + "learning_rate": 0.0003755975010560484, + "loss": 0.0981, + "step": 28540 + }, + { + "epoch": 1.26, + "learning_rate": 0.00037537517508170483, + "loss": 0.0979, + "step": 28550 + }, + { + "epoch": 1.26, + "learning_rate": 0.0003751528491073612, + "loss": 0.1005, + "step": 28560 + }, + { + "epoch": 1.26, + "learning_rate": 0.00037493052313301763, + "loss": 0.0995, + "step": 28570 + }, + { + "epoch": 1.26, + "learning_rate": 0.00037470819715867406, + "loss": 0.0976, + "step": 28580 + }, + { + "epoch": 1.26, + "learning_rate": 0.0003744858711843305, + "loss": 0.0965, + "step": 28590 + }, + { + "epoch": 1.26, + "learning_rate": 0.0003742635452099869, + "loss": 0.097, + "step": 28600 + }, + { + "epoch": 1.26, + "learning_rate": 0.00037404121923564334, + "loss": 0.0997, + "step": 28610 + }, + { + "epoch": 1.26, + "learning_rate": 0.00037381889326129977, + "loss": 0.0927, + "step": 28620 + }, + { + "epoch": 1.26, + "learning_rate": 0.00037359656728695614, + "loss": 0.1045, + "step": 28630 + }, + { + "epoch": 1.26, + "learning_rate": 0.00037337424131261257, + "loss": 0.0973, + "step": 28640 + }, + { + "epoch": 1.26, + "learning_rate": 0.000373151915338269, + "loss": 0.1006, + "step": 28650 + }, + { + "epoch": 1.26, + "learning_rate": 0.0003729295893639254, + "loss": 0.0975, + "step": 28660 + }, + { + "epoch": 1.26, + "learning_rate": 0.00037270726338958185, + "loss": 0.097, + "step": 28670 + }, + { + "epoch": 1.26, + "learning_rate": 0.0003724849374152383, + "loss": 0.1015, + "step": 28680 + }, + { + "epoch": 1.26, + "learning_rate": 0.0003722626114408946, + "loss": 0.1007, + "step": 28690 + }, + { + "epoch": 1.26, + "learning_rate": 0.00037204028546655103, + "loss": 0.1009, + "step": 28700 + }, + { + "epoch": 1.26, + "learning_rate": 0.00037181795949220746, + "loss": 0.0943, + "step": 28710 + }, + { + "epoch": 1.26, + "learning_rate": 0.0003715956335178639, + "loss": 0.0997, + "step": 28720 + }, + { + "epoch": 1.26, + "learning_rate": 0.0003713733075435203, + "loss": 0.1012, + "step": 28730 + }, + { + "epoch": 1.27, + "learning_rate": 0.00037115098156917674, + "loss": 0.0964, + "step": 28740 + }, + { + "epoch": 1.27, + "learning_rate": 0.0003709286555948331, + "loss": 0.1023, + "step": 28750 + }, + { + "epoch": 1.27, + "learning_rate": 0.00037070632962048954, + "loss": 0.1008, + "step": 28760 + }, + { + "epoch": 1.27, + "learning_rate": 0.00037048400364614597, + "loss": 0.0971, + "step": 28770 + }, + { + "epoch": 1.27, + "learning_rate": 0.0003702616776718024, + "loss": 0.0959, + "step": 28780 + }, + { + "epoch": 1.27, + "learning_rate": 0.0003700393516974588, + "loss": 0.0992, + "step": 28790 + }, + { + "epoch": 1.27, + "learning_rate": 0.00036981702572311525, + "loss": 0.096, + "step": 28800 + }, + { + "epoch": 1.27, + "learning_rate": 0.0003695946997487716, + "loss": 0.0952, + "step": 28810 + }, + { + "epoch": 1.27, + "learning_rate": 0.00036937237377442805, + "loss": 0.0983, + "step": 28820 + }, + { + "epoch": 1.27, + "learning_rate": 0.0003691500478000845, + "loss": 0.0967, + "step": 28830 + }, + { + "epoch": 1.27, + "learning_rate": 0.0003689277218257409, + "loss": 0.0981, + "step": 28840 + }, + { + "epoch": 1.27, + "learning_rate": 0.00036870539585139733, + "loss": 0.0978, + "step": 28850 + }, + { + "epoch": 1.27, + "learning_rate": 0.00036848306987705376, + "loss": 0.1012, + "step": 28860 + }, + { + "epoch": 1.27, + "learning_rate": 0.00036826074390271013, + "loss": 0.102, + "step": 28870 + }, + { + "epoch": 1.27, + "learning_rate": 0.00036803841792836656, + "loss": 0.1039, + "step": 28880 + }, + { + "epoch": 1.27, + "learning_rate": 0.000367816091954023, + "loss": 0.1037, + "step": 28890 + }, + { + "epoch": 1.27, + "learning_rate": 0.0003675937659796794, + "loss": 0.0977, + "step": 28900 + }, + { + "epoch": 1.27, + "learning_rate": 0.00036737144000533584, + "loss": 0.0964, + "step": 28910 + }, + { + "epoch": 1.27, + "learning_rate": 0.00036714911403099227, + "loss": 0.0983, + "step": 28920 + }, + { + "epoch": 1.27, + "learning_rate": 0.00036692678805664864, + "loss": 0.1004, + "step": 28930 + }, + { + "epoch": 1.27, + "learning_rate": 0.00036670446208230507, + "loss": 0.1013, + "step": 28940 + }, + { + "epoch": 1.27, + "learning_rate": 0.0003664821361079615, + "loss": 0.0968, + "step": 28950 + }, + { + "epoch": 1.27, + "learning_rate": 0.0003662598101336179, + "loss": 0.1004, + "step": 28960 + }, + { + "epoch": 1.28, + "learning_rate": 0.00036603748415927435, + "loss": 0.0998, + "step": 28970 + }, + { + "epoch": 1.28, + "learning_rate": 0.0003658151581849308, + "loss": 0.0971, + "step": 28980 + }, + { + "epoch": 1.28, + "learning_rate": 0.00036559283221058715, + "loss": 0.0969, + "step": 28990 + }, + { + "epoch": 1.28, + "learning_rate": 0.0003653705062362436, + "loss": 0.0962, + "step": 29000 + }, + { + "epoch": 1.28, + "eval_cer": 0.49958841703700485, + "eval_loss": 0.08326420187950134, + "eval_runtime": 83.8446, + "eval_samples_per_second": 119.197, + "eval_steps_per_second": 7.454, + "step": 29000 + }, + { + "epoch": 1.28, + "learning_rate": 0.0003651481802619, + "loss": 0.0944, + "step": 29010 + }, + { + "epoch": 1.28, + "learning_rate": 0.00036492585428755643, + "loss": 0.102, + "step": 29020 + }, + { + "epoch": 1.28, + "learning_rate": 0.00036470352831321286, + "loss": 0.1062, + "step": 29030 + }, + { + "epoch": 1.28, + "learning_rate": 0.0003644812023388693, + "loss": 0.0982, + "step": 29040 + }, + { + "epoch": 1.28, + "learning_rate": 0.00036425887636452566, + "loss": 0.0992, + "step": 29050 + }, + { + "epoch": 1.28, + "learning_rate": 0.0003640365503901821, + "loss": 0.0972, + "step": 29060 + }, + { + "epoch": 1.28, + "learning_rate": 0.0003638142244158385, + "loss": 0.1043, + "step": 29070 + }, + { + "epoch": 1.28, + "learning_rate": 0.00036359189844149494, + "loss": 0.0996, + "step": 29080 + }, + { + "epoch": 1.28, + "learning_rate": 0.00036336957246715137, + "loss": 0.0945, + "step": 29090 + }, + { + "epoch": 1.28, + "learning_rate": 0.0003631472464928078, + "loss": 0.1036, + "step": 29100 + }, + { + "epoch": 1.28, + "learning_rate": 0.00036292492051846417, + "loss": 0.0986, + "step": 29110 + }, + { + "epoch": 1.28, + "learning_rate": 0.0003627025945441206, + "loss": 0.0978, + "step": 29120 + }, + { + "epoch": 1.28, + "learning_rate": 0.000362480268569777, + "loss": 0.0998, + "step": 29130 + }, + { + "epoch": 1.28, + "learning_rate": 0.00036225794259543345, + "loss": 0.0994, + "step": 29140 + }, + { + "epoch": 1.28, + "learning_rate": 0.0003620356166210899, + "loss": 0.1005, + "step": 29150 + }, + { + "epoch": 1.28, + "learning_rate": 0.0003618132906467463, + "loss": 0.0984, + "step": 29160 + }, + { + "epoch": 1.28, + "learning_rate": 0.00036159096467240274, + "loss": 0.1009, + "step": 29170 + }, + { + "epoch": 1.28, + "learning_rate": 0.0003613686386980591, + "loss": 0.1002, + "step": 29180 + }, + { + "epoch": 1.28, + "learning_rate": 0.0003611463127237155, + "loss": 0.0957, + "step": 29190 + }, + { + "epoch": 1.29, + "learning_rate": 0.0003609239867493719, + "loss": 0.1021, + "step": 29200 + }, + { + "epoch": 1.29, + "learning_rate": 0.00036070166077502834, + "loss": 0.0936, + "step": 29210 + }, + { + "epoch": 1.29, + "learning_rate": 0.00036047933480068476, + "loss": 0.1027, + "step": 29220 + }, + { + "epoch": 1.29, + "learning_rate": 0.0003602570088263412, + "loss": 0.0975, + "step": 29230 + }, + { + "epoch": 1.29, + "learning_rate": 0.00036003468285199756, + "loss": 0.0975, + "step": 29240 + }, + { + "epoch": 1.29, + "learning_rate": 0.000359812356877654, + "loss": 0.0974, + "step": 29250 + }, + { + "epoch": 1.29, + "learning_rate": 0.0003595900309033104, + "loss": 0.1015, + "step": 29260 + }, + { + "epoch": 1.29, + "learning_rate": 0.00035936770492896685, + "loss": 0.0944, + "step": 29270 + }, + { + "epoch": 1.29, + "learning_rate": 0.0003591453789546233, + "loss": 0.1003, + "step": 29280 + }, + { + "epoch": 1.29, + "learning_rate": 0.0003589230529802797, + "loss": 0.1016, + "step": 29290 + }, + { + "epoch": 1.29, + "learning_rate": 0.0003587007270059361, + "loss": 0.0973, + "step": 29300 + }, + { + "epoch": 1.29, + "learning_rate": 0.0003584784010315925, + "loss": 0.0993, + "step": 29310 + }, + { + "epoch": 1.29, + "learning_rate": 0.00035825607505724893, + "loss": 0.0945, + "step": 29320 + }, + { + "epoch": 1.29, + "learning_rate": 0.00035803374908290536, + "loss": 0.1016, + "step": 29330 + }, + { + "epoch": 1.29, + "learning_rate": 0.0003578114231085618, + "loss": 0.1032, + "step": 29340 + }, + { + "epoch": 1.29, + "learning_rate": 0.0003575890971342182, + "loss": 0.1016, + "step": 29350 + }, + { + "epoch": 1.29, + "learning_rate": 0.0003573667711598746, + "loss": 0.1004, + "step": 29360 + }, + { + "epoch": 1.29, + "learning_rate": 0.000357144445185531, + "loss": 0.1064, + "step": 29370 + }, + { + "epoch": 1.29, + "learning_rate": 0.00035692211921118744, + "loss": 0.104, + "step": 29380 + }, + { + "epoch": 1.29, + "learning_rate": 0.00035669979323684387, + "loss": 0.0991, + "step": 29390 + }, + { + "epoch": 1.29, + "learning_rate": 0.0003564774672625003, + "loss": 0.0919, + "step": 29400 + }, + { + "epoch": 1.29, + "learning_rate": 0.0003562551412881567, + "loss": 0.098, + "step": 29410 + }, + { + "epoch": 1.3, + "learning_rate": 0.0003560328153138131, + "loss": 0.0986, + "step": 29420 + }, + { + "epoch": 1.3, + "learning_rate": 0.0003558104893394695, + "loss": 0.0933, + "step": 29430 + }, + { + "epoch": 1.3, + "learning_rate": 0.00035558816336512595, + "loss": 0.0989, + "step": 29440 + }, + { + "epoch": 1.3, + "learning_rate": 0.0003553658373907824, + "loss": 0.0979, + "step": 29450 + }, + { + "epoch": 1.3, + "learning_rate": 0.0003551435114164388, + "loss": 0.0956, + "step": 29460 + }, + { + "epoch": 1.3, + "learning_rate": 0.00035492118544209523, + "loss": 0.0965, + "step": 29470 + }, + { + "epoch": 1.3, + "learning_rate": 0.0003546988594677516, + "loss": 0.0961, + "step": 29480 + }, + { + "epoch": 1.3, + "learning_rate": 0.00035447653349340803, + "loss": 0.0972, + "step": 29490 + }, + { + "epoch": 1.3, + "learning_rate": 0.00035425420751906446, + "loss": 0.0948, + "step": 29500 + }, + { + "epoch": 1.3, + "eval_cer": 0.49944700634207506, + "eval_loss": 0.08346997946500778, + "eval_runtime": 83.9537, + "eval_samples_per_second": 119.042, + "eval_steps_per_second": 7.445, + "step": 29500 + }, + { + "epoch": 1.3, + "learning_rate": 0.0003540318815447209, + "loss": 0.1035, + "step": 29510 + }, + { + "epoch": 1.3, + "learning_rate": 0.0003538095555703773, + "loss": 0.0921, + "step": 29520 + }, + { + "epoch": 1.3, + "learning_rate": 0.00035358722959603374, + "loss": 0.0965, + "step": 29530 + }, + { + "epoch": 1.3, + "learning_rate": 0.0003533649036216901, + "loss": 0.1001, + "step": 29540 + }, + { + "epoch": 1.3, + "learning_rate": 0.00035314257764734654, + "loss": 0.0994, + "step": 29550 + }, + { + "epoch": 1.3, + "learning_rate": 0.00035292025167300297, + "loss": 0.0997, + "step": 29560 + }, + { + "epoch": 1.3, + "learning_rate": 0.0003526979256986594, + "loss": 0.0972, + "step": 29570 + }, + { + "epoch": 1.3, + "learning_rate": 0.0003524755997243158, + "loss": 0.0992, + "step": 29580 + }, + { + "epoch": 1.3, + "learning_rate": 0.00035225327374997225, + "loss": 0.0981, + "step": 29590 + }, + { + "epoch": 1.3, + "learning_rate": 0.0003520309477756286, + "loss": 0.1011, + "step": 29600 + }, + { + "epoch": 1.3, + "learning_rate": 0.00035180862180128505, + "loss": 0.1012, + "step": 29610 + }, + { + "epoch": 1.3, + "learning_rate": 0.0003515862958269415, + "loss": 0.0972, + "step": 29620 + }, + { + "epoch": 1.3, + "learning_rate": 0.0003513639698525979, + "loss": 0.0968, + "step": 29630 + }, + { + "epoch": 1.3, + "learning_rate": 0.00035114164387825433, + "loss": 0.0965, + "step": 29640 + }, + { + "epoch": 1.31, + "learning_rate": 0.00035091931790391076, + "loss": 0.1016, + "step": 29650 + }, + { + "epoch": 1.31, + "learning_rate": 0.0003506969919295672, + "loss": 0.0944, + "step": 29660 + }, + { + "epoch": 1.31, + "learning_rate": 0.00035047466595522356, + "loss": 0.1006, + "step": 29670 + }, + { + "epoch": 1.31, + "learning_rate": 0.00035025233998088, + "loss": 0.0968, + "step": 29680 + }, + { + "epoch": 1.31, + "learning_rate": 0.0003500300140065364, + "loss": 0.0961, + "step": 29690 + }, + { + "epoch": 1.31, + "learning_rate": 0.0003498076880321928, + "loss": 0.0974, + "step": 29700 + }, + { + "epoch": 1.31, + "learning_rate": 0.0003495853620578492, + "loss": 0.0983, + "step": 29710 + }, + { + "epoch": 1.31, + "learning_rate": 0.00034936303608350565, + "loss": 0.104, + "step": 29720 + }, + { + "epoch": 1.31, + "learning_rate": 0.000349140710109162, + "loss": 0.0989, + "step": 29730 + }, + { + "epoch": 1.31, + "learning_rate": 0.00034891838413481845, + "loss": 0.1002, + "step": 29740 + }, + { + "epoch": 1.31, + "learning_rate": 0.0003486960581604749, + "loss": 0.0978, + "step": 29750 + }, + { + "epoch": 1.31, + "learning_rate": 0.0003484737321861313, + "loss": 0.1018, + "step": 29760 + }, + { + "epoch": 1.31, + "learning_rate": 0.00034825140621178773, + "loss": 0.0965, + "step": 29770 + }, + { + "epoch": 1.31, + "learning_rate": 0.00034802908023744416, + "loss": 0.0979, + "step": 29780 + }, + { + "epoch": 1.31, + "learning_rate": 0.00034780675426310053, + "loss": 0.0984, + "step": 29790 + }, + { + "epoch": 1.31, + "learning_rate": 0.00034758442828875696, + "loss": 0.0972, + "step": 29800 + }, + { + "epoch": 1.31, + "learning_rate": 0.0003473621023144134, + "loss": 0.0948, + "step": 29810 + }, + { + "epoch": 1.31, + "learning_rate": 0.0003471397763400698, + "loss": 0.0986, + "step": 29820 + }, + { + "epoch": 1.31, + "learning_rate": 0.00034691745036572624, + "loss": 0.0962, + "step": 29830 + }, + { + "epoch": 1.31, + "learning_rate": 0.00034669512439138267, + "loss": 0.0976, + "step": 29840 + }, + { + "epoch": 1.31, + "learning_rate": 0.00034647279841703904, + "loss": 0.097, + "step": 29850 + }, + { + "epoch": 1.31, + "learning_rate": 0.00034625047244269547, + "loss": 0.0967, + "step": 29860 + }, + { + "epoch": 1.31, + "learning_rate": 0.0003460281464683519, + "loss": 0.0998, + "step": 29870 + }, + { + "epoch": 1.32, + "learning_rate": 0.0003458058204940083, + "loss": 0.0999, + "step": 29880 + }, + { + "epoch": 1.32, + "learning_rate": 0.00034558349451966475, + "loss": 0.0975, + "step": 29890 + }, + { + "epoch": 1.32, + "learning_rate": 0.0003453611685453212, + "loss": 0.097, + "step": 29900 + }, + { + "epoch": 1.32, + "learning_rate": 0.00034513884257097755, + "loss": 0.099, + "step": 29910 + }, + { + "epoch": 1.32, + "learning_rate": 0.000344916516596634, + "loss": 0.0959, + "step": 29920 + }, + { + "epoch": 1.32, + "learning_rate": 0.0003446941906222904, + "loss": 0.096, + "step": 29930 + }, + { + "epoch": 1.32, + "learning_rate": 0.00034447186464794683, + "loss": 0.0977, + "step": 29940 + }, + { + "epoch": 1.32, + "learning_rate": 0.00034424953867360326, + "loss": 0.0987, + "step": 29950 + }, + { + "epoch": 1.32, + "learning_rate": 0.0003440272126992597, + "loss": 0.1009, + "step": 29960 + }, + { + "epoch": 1.32, + "learning_rate": 0.00034380488672491606, + "loss": 0.1006, + "step": 29970 + }, + { + "epoch": 1.32, + "learning_rate": 0.0003435825607505725, + "loss": 0.0955, + "step": 29980 + }, + { + "epoch": 1.32, + "learning_rate": 0.0003433602347762289, + "loss": 0.0967, + "step": 29990 + }, + { + "epoch": 1.32, + "learning_rate": 0.00034313790880188534, + "loss": 0.1009, + "step": 30000 + }, + { + "epoch": 1.32, + "eval_cer": 0.4992290522434677, + "eval_loss": 0.08234353363513947, + "eval_runtime": 83.7488, + "eval_samples_per_second": 119.333, + "eval_steps_per_second": 7.463, + "step": 30000 + }, + { + "epoch": 1.32, + "learning_rate": 0.00034291558282754177, + "loss": 0.0986, + "step": 30010 + }, + { + "epoch": 1.32, + "learning_rate": 0.0003426932568531982, + "loss": 0.0985, + "step": 30020 + }, + { + "epoch": 1.32, + "learning_rate": 0.00034247093087885457, + "loss": 0.0989, + "step": 30030 + }, + { + "epoch": 1.32, + "learning_rate": 0.000342248604904511, + "loss": 0.0955, + "step": 30040 + }, + { + "epoch": 1.32, + "learning_rate": 0.0003420262789301674, + "loss": 0.0991, + "step": 30050 + }, + { + "epoch": 1.32, + "learning_rate": 0.00034180395295582385, + "loss": 0.0968, + "step": 30060 + }, + { + "epoch": 1.32, + "learning_rate": 0.0003415816269814803, + "loss": 0.0973, + "step": 30070 + }, + { + "epoch": 1.32, + "learning_rate": 0.0003413593010071367, + "loss": 0.0971, + "step": 30080 + }, + { + "epoch": 1.32, + "learning_rate": 0.0003411369750327931, + "loss": 0.0916, + "step": 30090 + }, + { + "epoch": 1.32, + "learning_rate": 0.0003409146490584495, + "loss": 0.0986, + "step": 30100 + }, + { + "epoch": 1.33, + "learning_rate": 0.00034069232308410593, + "loss": 0.0943, + "step": 30110 + }, + { + "epoch": 1.33, + "learning_rate": 0.00034046999710976236, + "loss": 0.0951, + "step": 30120 + }, + { + "epoch": 1.33, + "learning_rate": 0.0003402476711354188, + "loss": 0.0969, + "step": 30130 + }, + { + "epoch": 1.33, + "learning_rate": 0.0003400253451610752, + "loss": 0.0981, + "step": 30140 + }, + { + "epoch": 1.33, + "learning_rate": 0.0003398030191867316, + "loss": 0.0992, + "step": 30150 + }, + { + "epoch": 1.33, + "learning_rate": 0.000339580693212388, + "loss": 0.0978, + "step": 30160 + }, + { + "epoch": 1.33, + "learning_rate": 0.00033935836723804444, + "loss": 0.0971, + "step": 30170 + }, + { + "epoch": 1.33, + "learning_rate": 0.00033913604126370087, + "loss": 0.093, + "step": 30180 + }, + { + "epoch": 1.33, + "learning_rate": 0.0003389137152893573, + "loss": 0.0967, + "step": 30190 + }, + { + "epoch": 1.33, + "learning_rate": 0.00033869138931501367, + "loss": 0.0987, + "step": 30200 + }, + { + "epoch": 1.33, + "learning_rate": 0.0003384690633406701, + "loss": 0.0975, + "step": 30210 + }, + { + "epoch": 1.33, + "learning_rate": 0.00033824673736632647, + "loss": 0.0969, + "step": 30220 + }, + { + "epoch": 1.33, + "learning_rate": 0.0003380244113919829, + "loss": 0.0958, + "step": 30230 + }, + { + "epoch": 1.33, + "learning_rate": 0.0003378020854176393, + "loss": 0.0919, + "step": 30240 + }, + { + "epoch": 1.33, + "learning_rate": 0.00033757975944329575, + "loss": 0.0973, + "step": 30250 + }, + { + "epoch": 1.33, + "learning_rate": 0.0003373574334689522, + "loss": 0.0943, + "step": 30260 + }, + { + "epoch": 1.33, + "learning_rate": 0.0003371351074946086, + "loss": 0.0981, + "step": 30270 + }, + { + "epoch": 1.33, + "learning_rate": 0.000336912781520265, + "loss": 0.0982, + "step": 30280 + }, + { + "epoch": 1.33, + "learning_rate": 0.0003366904555459214, + "loss": 0.0958, + "step": 30290 + }, + { + "epoch": 1.33, + "learning_rate": 0.00033646812957157784, + "loss": 0.0999, + "step": 30300 + }, + { + "epoch": 1.33, + "learning_rate": 0.00033624580359723426, + "loss": 0.0983, + "step": 30310 + }, + { + "epoch": 1.33, + "learning_rate": 0.0003360234776228907, + "loss": 0.0964, + "step": 30320 + }, + { + "epoch": 1.34, + "learning_rate": 0.0003358011516485471, + "loss": 0.0938, + "step": 30330 + }, + { + "epoch": 1.34, + "learning_rate": 0.0003355788256742035, + "loss": 0.0984, + "step": 30340 + }, + { + "epoch": 1.34, + "learning_rate": 0.0003353564996998599, + "loss": 0.1001, + "step": 30350 + }, + { + "epoch": 1.34, + "learning_rate": 0.00033513417372551635, + "loss": 0.098, + "step": 30360 + }, + { + "epoch": 1.34, + "learning_rate": 0.0003349118477511728, + "loss": 0.102, + "step": 30370 + }, + { + "epoch": 1.34, + "learning_rate": 0.0003346895217768292, + "loss": 0.0943, + "step": 30380 + }, + { + "epoch": 1.34, + "learning_rate": 0.00033446719580248563, + "loss": 0.0973, + "step": 30390 + }, + { + "epoch": 1.34, + "learning_rate": 0.000334244869828142, + "loss": 0.0951, + "step": 30400 + }, + { + "epoch": 1.34, + "learning_rate": 0.00033402254385379843, + "loss": 0.0977, + "step": 30410 + }, + { + "epoch": 1.34, + "learning_rate": 0.00033380021787945486, + "loss": 0.0953, + "step": 30420 + }, + { + "epoch": 1.34, + "learning_rate": 0.0003335778919051113, + "loss": 0.0961, + "step": 30430 + }, + { + "epoch": 1.34, + "learning_rate": 0.0003333555659307677, + "loss": 0.1025, + "step": 30440 + }, + { + "epoch": 1.34, + "learning_rate": 0.00033313323995642414, + "loss": 0.0972, + "step": 30450 + }, + { + "epoch": 1.34, + "learning_rate": 0.0003329109139820805, + "loss": 0.093, + "step": 30460 + }, + { + "epoch": 1.34, + "learning_rate": 0.00033268858800773694, + "loss": 0.094, + "step": 30470 + }, + { + "epoch": 1.34, + "learning_rate": 0.00033246626203339337, + "loss": 0.0996, + "step": 30480 + }, + { + "epoch": 1.34, + "learning_rate": 0.0003322439360590498, + "loss": 0.0999, + "step": 30490 + }, + { + "epoch": 1.34, + "learning_rate": 0.0003320216100847062, + "loss": 0.0977, + "step": 30500 + }, + { + "epoch": 1.34, + "eval_cer": 0.4994781426418761, + "eval_loss": 0.08195621520280838, + "eval_runtime": 83.7847, + "eval_samples_per_second": 119.282, + "eval_steps_per_second": 7.46, + "step": 30500 + }, + { + "epoch": 1.34, + "learning_rate": 0.00033179928411036265, + "loss": 0.0937, + "step": 30510 + }, + { + "epoch": 1.34, + "learning_rate": 0.000331576958136019, + "loss": 0.0999, + "step": 30520 + }, + { + "epoch": 1.34, + "learning_rate": 0.00033135463216167545, + "loss": 0.0959, + "step": 30530 + }, + { + "epoch": 1.34, + "learning_rate": 0.0003311323061873319, + "loss": 0.0977, + "step": 30540 + }, + { + "epoch": 1.34, + "learning_rate": 0.0003309099802129883, + "loss": 0.1007, + "step": 30550 + }, + { + "epoch": 1.35, + "learning_rate": 0.00033068765423864473, + "loss": 0.0954, + "step": 30560 + }, + { + "epoch": 1.35, + "learning_rate": 0.00033046532826430116, + "loss": 0.0985, + "step": 30570 + }, + { + "epoch": 1.35, + "learning_rate": 0.00033024300228995753, + "loss": 0.0979, + "step": 30580 + }, + { + "epoch": 1.35, + "learning_rate": 0.00033002067631561396, + "loss": 0.0934, + "step": 30590 + }, + { + "epoch": 1.35, + "learning_rate": 0.0003297983503412704, + "loss": 0.099, + "step": 30600 + }, + { + "epoch": 1.35, + "learning_rate": 0.0003295760243669268, + "loss": 0.1013, + "step": 30610 + }, + { + "epoch": 1.35, + "learning_rate": 0.00032935369839258324, + "loss": 0.0924, + "step": 30620 + }, + { + "epoch": 1.35, + "learning_rate": 0.00032913137241823967, + "loss": 0.1005, + "step": 30630 + }, + { + "epoch": 1.35, + "learning_rate": 0.00032890904644389604, + "loss": 0.0915, + "step": 30640 + }, + { + "epoch": 1.35, + "learning_rate": 0.00032868672046955247, + "loss": 0.0988, + "step": 30650 + }, + { + "epoch": 1.35, + "learning_rate": 0.0003284643944952089, + "loss": 0.0944, + "step": 30660 + }, + { + "epoch": 1.35, + "learning_rate": 0.0003282420685208653, + "loss": 0.1015, + "step": 30670 + }, + { + "epoch": 1.35, + "learning_rate": 0.00032801974254652175, + "loss": 0.1015, + "step": 30680 + }, + { + "epoch": 1.35, + "learning_rate": 0.0003277974165721782, + "loss": 0.1, + "step": 30690 + }, + { + "epoch": 1.35, + "learning_rate": 0.0003275750905978345, + "loss": 0.0943, + "step": 30700 + }, + { + "epoch": 1.35, + "learning_rate": 0.0003273527646234909, + "loss": 0.0936, + "step": 30710 + }, + { + "epoch": 1.35, + "learning_rate": 0.00032713043864914735, + "loss": 0.0927, + "step": 30720 + }, + { + "epoch": 1.35, + "learning_rate": 0.0003269081126748038, + "loss": 0.0966, + "step": 30730 + }, + { + "epoch": 1.35, + "learning_rate": 0.0003266857867004602, + "loss": 0.0994, + "step": 30740 + }, + { + "epoch": 1.35, + "learning_rate": 0.00032646346072611664, + "loss": 0.0933, + "step": 30750 + }, + { + "epoch": 1.35, + "learning_rate": 0.00032624113475177306, + "loss": 0.0988, + "step": 30760 + }, + { + "epoch": 1.35, + "learning_rate": 0.00032601880877742944, + "loss": 0.0963, + "step": 30770 + }, + { + "epoch": 1.35, + "learning_rate": 0.00032579648280308586, + "loss": 0.0971, + "step": 30780 + }, + { + "epoch": 1.36, + "learning_rate": 0.0003255741568287423, + "loss": 0.0905, + "step": 30790 + }, + { + "epoch": 1.36, + "learning_rate": 0.0003253518308543987, + "loss": 0.1016, + "step": 30800 + }, + { + "epoch": 1.36, + "learning_rate": 0.00032512950488005515, + "loss": 0.0978, + "step": 30810 + }, + { + "epoch": 1.36, + "learning_rate": 0.0003249071789057116, + "loss": 0.0928, + "step": 30820 + }, + { + "epoch": 1.36, + "learning_rate": 0.00032468485293136795, + "loss": 0.1008, + "step": 30830 + }, + { + "epoch": 1.36, + "learning_rate": 0.0003244625269570244, + "loss": 0.0974, + "step": 30840 + }, + { + "epoch": 1.36, + "learning_rate": 0.0003242402009826808, + "loss": 0.0995, + "step": 30850 + }, + { + "epoch": 1.36, + "learning_rate": 0.00032401787500833723, + "loss": 0.0962, + "step": 30860 + }, + { + "epoch": 1.36, + "learning_rate": 0.00032379554903399366, + "loss": 0.0978, + "step": 30870 + }, + { + "epoch": 1.36, + "learning_rate": 0.0003235732230596501, + "loss": 0.0972, + "step": 30880 + }, + { + "epoch": 1.36, + "learning_rate": 0.00032335089708530646, + "loss": 0.0969, + "step": 30890 + }, + { + "epoch": 1.36, + "learning_rate": 0.0003231285711109629, + "loss": 0.0975, + "step": 30900 + }, + { + "epoch": 1.36, + "learning_rate": 0.0003229062451366193, + "loss": 0.0983, + "step": 30910 + }, + { + "epoch": 1.36, + "learning_rate": 0.00032268391916227574, + "loss": 0.0939, + "step": 30920 + }, + { + "epoch": 1.36, + "learning_rate": 0.00032246159318793217, + "loss": 0.0946, + "step": 30930 + }, + { + "epoch": 1.36, + "learning_rate": 0.0003222392672135886, + "loss": 0.0955, + "step": 30940 + }, + { + "epoch": 1.36, + "learning_rate": 0.00032201694123924497, + "loss": 0.0946, + "step": 30950 + }, + { + "epoch": 1.36, + "learning_rate": 0.0003217946152649014, + "loss": 0.0957, + "step": 30960 + }, + { + "epoch": 1.36, + "learning_rate": 0.0003215722892905578, + "loss": 0.0912, + "step": 30970 + }, + { + "epoch": 1.36, + "learning_rate": 0.00032134996331621425, + "loss": 0.1, + "step": 30980 + }, + { + "epoch": 1.36, + "learning_rate": 0.0003211276373418707, + "loss": 0.0975, + "step": 30990 + }, + { + "epoch": 1.36, + "learning_rate": 0.0003209053113675271, + "loss": 0.0938, + "step": 31000 + }, + { + "epoch": 1.36, + "eval_cer": 0.49915445485852766, + "eval_loss": 0.08223430067300797, + "eval_runtime": 83.7546, + "eval_samples_per_second": 119.325, + "eval_steps_per_second": 7.462, + "step": 31000 + }, + { + "epoch": 1.37, + "learning_rate": 0.0003206829853931835, + "loss": 0.1009, + "step": 31010 + }, + { + "epoch": 1.37, + "learning_rate": 0.0003204606594188399, + "loss": 0.0986, + "step": 31020 + }, + { + "epoch": 1.37, + "learning_rate": 0.00032023833344449633, + "loss": 0.0967, + "step": 31030 + }, + { + "epoch": 1.37, + "learning_rate": 0.00032001600747015276, + "loss": 0.0955, + "step": 31040 + }, + { + "epoch": 1.37, + "learning_rate": 0.0003197936814958092, + "loss": 0.0935, + "step": 31050 + }, + { + "epoch": 1.37, + "learning_rate": 0.0003195713555214656, + "loss": 0.0961, + "step": 31060 + }, + { + "epoch": 1.37, + "learning_rate": 0.000319349029547122, + "loss": 0.095, + "step": 31070 + }, + { + "epoch": 1.37, + "learning_rate": 0.0003191267035727784, + "loss": 0.0958, + "step": 31080 + }, + { + "epoch": 1.37, + "learning_rate": 0.00031890437759843484, + "loss": 0.0969, + "step": 31090 + }, + { + "epoch": 1.37, + "learning_rate": 0.00031868205162409127, + "loss": 0.0976, + "step": 31100 + }, + { + "epoch": 1.37, + "learning_rate": 0.0003184597256497477, + "loss": 0.0965, + "step": 31110 + }, + { + "epoch": 1.37, + "learning_rate": 0.0003182373996754041, + "loss": 0.098, + "step": 31120 + }, + { + "epoch": 1.37, + "learning_rate": 0.0003180150737010605, + "loss": 0.1003, + "step": 31130 + }, + { + "epoch": 1.37, + "learning_rate": 0.0003177927477267169, + "loss": 0.097, + "step": 31140 + }, + { + "epoch": 1.37, + "learning_rate": 0.00031757042175237335, + "loss": 0.0997, + "step": 31150 + }, + { + "epoch": 1.37, + "learning_rate": 0.0003173480957780298, + "loss": 0.0989, + "step": 31160 + }, + { + "epoch": 1.37, + "learning_rate": 0.0003171257698036862, + "loss": 0.0941, + "step": 31170 + }, + { + "epoch": 1.37, + "learning_rate": 0.00031690344382934263, + "loss": 0.0929, + "step": 31180 + }, + { + "epoch": 1.37, + "learning_rate": 0.00031668111785499895, + "loss": 0.0979, + "step": 31190 + }, + { + "epoch": 1.37, + "learning_rate": 0.0003164587918806554, + "loss": 0.094, + "step": 31200 + }, + { + "epoch": 1.37, + "learning_rate": 0.0003162364659063118, + "loss": 0.0938, + "step": 31210 + }, + { + "epoch": 1.37, + "learning_rate": 0.00031601413993196823, + "loss": 0.0981, + "step": 31220 + }, + { + "epoch": 1.37, + "learning_rate": 0.00031579181395762466, + "loss": 0.0998, + "step": 31230 + }, + { + "epoch": 1.38, + "learning_rate": 0.0003155694879832811, + "loss": 0.0929, + "step": 31240 + }, + { + "epoch": 1.38, + "learning_rate": 0.0003153471620089375, + "loss": 0.1005, + "step": 31250 + }, + { + "epoch": 1.38, + "learning_rate": 0.0003151248360345939, + "loss": 0.0962, + "step": 31260 + }, + { + "epoch": 1.38, + "learning_rate": 0.0003149025100602503, + "loss": 0.0988, + "step": 31270 + }, + { + "epoch": 1.38, + "learning_rate": 0.00031468018408590674, + "loss": 0.0946, + "step": 31280 + }, + { + "epoch": 1.38, + "learning_rate": 0.00031445785811156317, + "loss": 0.0977, + "step": 31290 + }, + { + "epoch": 1.38, + "learning_rate": 0.0003142355321372196, + "loss": 0.0978, + "step": 31300 + }, + { + "epoch": 1.38, + "learning_rate": 0.000314013206162876, + "loss": 0.0981, + "step": 31310 + }, + { + "epoch": 1.38, + "learning_rate": 0.0003137908801885324, + "loss": 0.0927, + "step": 31320 + }, + { + "epoch": 1.38, + "learning_rate": 0.00031356855421418883, + "loss": 0.0955, + "step": 31330 + }, + { + "epoch": 1.38, + "learning_rate": 0.00031334622823984525, + "loss": 0.0936, + "step": 31340 + }, + { + "epoch": 1.38, + "learning_rate": 0.0003131239022655017, + "loss": 0.0915, + "step": 31350 + }, + { + "epoch": 1.38, + "learning_rate": 0.0003129015762911581, + "loss": 0.0944, + "step": 31360 + }, + { + "epoch": 1.38, + "learning_rate": 0.00031267925031681454, + "loss": 0.0999, + "step": 31370 + }, + { + "epoch": 1.38, + "learning_rate": 0.0003124569243424709, + "loss": 0.0985, + "step": 31380 + }, + { + "epoch": 1.38, + "learning_rate": 0.00031223459836812734, + "loss": 0.0967, + "step": 31390 + }, + { + "epoch": 1.38, + "learning_rate": 0.00031201227239378376, + "loss": 0.0907, + "step": 31400 + }, + { + "epoch": 1.38, + "learning_rate": 0.0003117899464194402, + "loss": 0.0958, + "step": 31410 + }, + { + "epoch": 1.38, + "learning_rate": 0.0003115676204450966, + "loss": 0.093, + "step": 31420 + }, + { + "epoch": 1.38, + "learning_rate": 0.00031134529447075305, + "loss": 0.0928, + "step": 31430 + }, + { + "epoch": 1.38, + "learning_rate": 0.0003111229684964094, + "loss": 0.0958, + "step": 31440 + }, + { + "epoch": 1.38, + "learning_rate": 0.00031090064252206585, + "loss": 0.0992, + "step": 31450 + }, + { + "epoch": 1.38, + "learning_rate": 0.0003106783165477223, + "loss": 0.0925, + "step": 31460 + }, + { + "epoch": 1.39, + "learning_rate": 0.0003104559905733787, + "loss": 0.0955, + "step": 31470 + }, + { + "epoch": 1.39, + "learning_rate": 0.00031023366459903513, + "loss": 0.0929, + "step": 31480 + }, + { + "epoch": 1.39, + "learning_rate": 0.00031001133862469156, + "loss": 0.0978, + "step": 31490 + }, + { + "epoch": 1.39, + "learning_rate": 0.00030978901265034793, + "loss": 0.0935, + "step": 31500 + }, + { + "epoch": 1.39, + "eval_cer": 0.49902861231349843, + "eval_loss": 0.0814196914434433, + "eval_runtime": 83.7896, + "eval_samples_per_second": 119.275, + "eval_steps_per_second": 7.459, + "step": 31500 + }, + { + "epoch": 1.39, + "learning_rate": 0.00030956668667600436, + "loss": 0.0972, + "step": 31510 + }, + { + "epoch": 1.39, + "learning_rate": 0.0003093443607016608, + "loss": 0.0953, + "step": 31520 + }, + { + "epoch": 1.39, + "learning_rate": 0.0003091220347273172, + "loss": 0.0965, + "step": 31530 + }, + { + "epoch": 1.39, + "learning_rate": 0.00030889970875297364, + "loss": 0.0984, + "step": 31540 + }, + { + "epoch": 1.39, + "learning_rate": 0.00030867738277863007, + "loss": 0.0969, + "step": 31550 + }, + { + "epoch": 1.39, + "learning_rate": 0.00030845505680428644, + "loss": 0.0954, + "step": 31560 + }, + { + "epoch": 1.39, + "learning_rate": 0.00030823273082994287, + "loss": 0.095, + "step": 31570 + }, + { + "epoch": 1.39, + "learning_rate": 0.0003080104048555993, + "loss": 0.0925, + "step": 31580 + }, + { + "epoch": 1.39, + "learning_rate": 0.0003077880788812557, + "loss": 0.0966, + "step": 31590 + }, + { + "epoch": 1.39, + "learning_rate": 0.00030756575290691215, + "loss": 0.0978, + "step": 31600 + }, + { + "epoch": 1.39, + "learning_rate": 0.0003073434269325686, + "loss": 0.0958, + "step": 31610 + }, + { + "epoch": 1.39, + "learning_rate": 0.00030712110095822495, + "loss": 0.0972, + "step": 31620 + }, + { + "epoch": 1.39, + "learning_rate": 0.0003068987749838814, + "loss": 0.092, + "step": 31630 + }, + { + "epoch": 1.39, + "learning_rate": 0.0003066764490095378, + "loss": 0.0952, + "step": 31640 + }, + { + "epoch": 1.39, + "learning_rate": 0.00030645412303519423, + "loss": 0.097, + "step": 31650 + }, + { + "epoch": 1.39, + "learning_rate": 0.00030623179706085066, + "loss": 0.0983, + "step": 31660 + }, + { + "epoch": 1.39, + "learning_rate": 0.0003060094710865071, + "loss": 0.096, + "step": 31670 + }, + { + "epoch": 1.39, + "learning_rate": 0.00030578714511216346, + "loss": 0.1014, + "step": 31680 + }, + { + "epoch": 1.39, + "learning_rate": 0.00030556481913781983, + "loss": 0.095, + "step": 31690 + }, + { + "epoch": 1.4, + "learning_rate": 0.00030534249316347626, + "loss": 0.0989, + "step": 31700 + }, + { + "epoch": 1.4, + "learning_rate": 0.0003051201671891327, + "loss": 0.0933, + "step": 31710 + }, + { + "epoch": 1.4, + "learning_rate": 0.0003048978412147891, + "loss": 0.0951, + "step": 31720 + }, + { + "epoch": 1.4, + "learning_rate": 0.00030467551524044554, + "loss": 0.0958, + "step": 31730 + }, + { + "epoch": 1.4, + "learning_rate": 0.0003044531892661019, + "loss": 0.0969, + "step": 31740 + }, + { + "epoch": 1.4, + "learning_rate": 0.00030423086329175834, + "loss": 0.0976, + "step": 31750 + }, + { + "epoch": 1.4, + "learning_rate": 0.00030400853731741477, + "loss": 0.0976, + "step": 31760 + }, + { + "epoch": 1.4, + "learning_rate": 0.0003037862113430712, + "loss": 0.0934, + "step": 31770 + }, + { + "epoch": 1.4, + "learning_rate": 0.0003035638853687276, + "loss": 0.0993, + "step": 31780 + }, + { + "epoch": 1.4, + "learning_rate": 0.00030334155939438405, + "loss": 0.0994, + "step": 31790 + }, + { + "epoch": 1.4, + "learning_rate": 0.0003031192334200405, + "loss": 0.094, + "step": 31800 + }, + { + "epoch": 1.4, + "learning_rate": 0.00030289690744569685, + "loss": 0.0979, + "step": 31810 + }, + { + "epoch": 1.4, + "learning_rate": 0.0003026745814713533, + "loss": 0.0979, + "step": 31820 + }, + { + "epoch": 1.4, + "learning_rate": 0.0003024522554970097, + "loss": 0.1006, + "step": 31830 + }, + { + "epoch": 1.4, + "learning_rate": 0.00030222992952266614, + "loss": 0.0948, + "step": 31840 + }, + { + "epoch": 1.4, + "learning_rate": 0.00030200760354832256, + "loss": 0.0942, + "step": 31850 + }, + { + "epoch": 1.4, + "learning_rate": 0.000301785277573979, + "loss": 0.1016, + "step": 31860 + }, + { + "epoch": 1.4, + "learning_rate": 0.00030156295159963536, + "loss": 0.097, + "step": 31870 + }, + { + "epoch": 1.4, + "learning_rate": 0.0003013406256252918, + "loss": 0.0973, + "step": 31880 + }, + { + "epoch": 1.4, + "learning_rate": 0.0003011182996509482, + "loss": 0.0962, + "step": 31890 + }, + { + "epoch": 1.4, + "learning_rate": 0.00030089597367660465, + "loss": 0.094, + "step": 31900 + }, + { + "epoch": 1.4, + "learning_rate": 0.0003006736477022611, + "loss": 0.0923, + "step": 31910 + }, + { + "epoch": 1.41, + "learning_rate": 0.0003004513217279175, + "loss": 0.094, + "step": 31920 + }, + { + "epoch": 1.41, + "learning_rate": 0.0003002289957535739, + "loss": 0.0991, + "step": 31930 + }, + { + "epoch": 1.41, + "learning_rate": 0.0003000066697792303, + "loss": 0.1012, + "step": 31940 + }, + { + "epoch": 1.41, + "learning_rate": 0.00029978434380488673, + "loss": 0.0999, + "step": 31950 + }, + { + "epoch": 1.41, + "learning_rate": 0.00029956201783054316, + "loss": 0.0995, + "step": 31960 + }, + { + "epoch": 1.41, + "learning_rate": 0.0002993396918561996, + "loss": 0.0942, + "step": 31970 + }, + { + "epoch": 1.41, + "learning_rate": 0.000299117365881856, + "loss": 0.0975, + "step": 31980 + }, + { + "epoch": 1.41, + "learning_rate": 0.0002988950399075124, + "loss": 0.0952, + "step": 31990 + }, + { + "epoch": 1.41, + "learning_rate": 0.0002986727139331688, + "loss": 0.0961, + "step": 32000 + }, + { + "epoch": 1.41, + "eval_cer": 0.49931921777830823, + "eval_loss": 0.08108332753181458, + "eval_runtime": 84.2512, + "eval_samples_per_second": 118.621, + "eval_steps_per_second": 7.418, + "step": 32000 + }, + { + "epoch": 1.41, + "learning_rate": 0.00029845038795882524, + "loss": 0.0963, + "step": 32010 + }, + { + "epoch": 1.41, + "learning_rate": 0.00029822806198448167, + "loss": 0.1016, + "step": 32020 + }, + { + "epoch": 1.41, + "learning_rate": 0.0002980057360101381, + "loss": 0.0936, + "step": 32030 + }, + { + "epoch": 1.41, + "learning_rate": 0.0002977834100357945, + "loss": 0.095, + "step": 32040 + }, + { + "epoch": 1.41, + "learning_rate": 0.0002975610840614509, + "loss": 0.0984, + "step": 32050 + }, + { + "epoch": 1.41, + "learning_rate": 0.0002973387580871073, + "loss": 0.0991, + "step": 32060 + }, + { + "epoch": 1.41, + "learning_rate": 0.00029711643211276375, + "loss": 0.0938, + "step": 32070 + }, + { + "epoch": 1.41, + "learning_rate": 0.0002968941061384202, + "loss": 0.0942, + "step": 32080 + }, + { + "epoch": 1.41, + "learning_rate": 0.0002966717801640766, + "loss": 0.0922, + "step": 32090 + }, + { + "epoch": 1.41, + "learning_rate": 0.00029644945418973303, + "loss": 0.0963, + "step": 32100 + }, + { + "epoch": 1.41, + "learning_rate": 0.0002962271282153894, + "loss": 0.1015, + "step": 32110 + }, + { + "epoch": 1.41, + "learning_rate": 0.00029600480224104583, + "loss": 0.0977, + "step": 32120 + }, + { + "epoch": 1.41, + "learning_rate": 0.00029578247626670226, + "loss": 0.0966, + "step": 32130 + }, + { + "epoch": 1.41, + "learning_rate": 0.0002955601502923587, + "loss": 0.0993, + "step": 32140 + }, + { + "epoch": 1.42, + "learning_rate": 0.0002953378243180151, + "loss": 0.0929, + "step": 32150 + }, + { + "epoch": 1.42, + "learning_rate": 0.00029511549834367154, + "loss": 0.0991, + "step": 32160 + }, + { + "epoch": 1.42, + "learning_rate": 0.0002948931723693279, + "loss": 0.0969, + "step": 32170 + }, + { + "epoch": 1.42, + "learning_rate": 0.00029467084639498434, + "loss": 0.1002, + "step": 32180 + }, + { + "epoch": 1.42, + "learning_rate": 0.0002944485204206407, + "loss": 0.0981, + "step": 32190 + }, + { + "epoch": 1.42, + "learning_rate": 0.00029422619444629714, + "loss": 0.0907, + "step": 32200 + }, + { + "epoch": 1.42, + "learning_rate": 0.00029400386847195357, + "loss": 0.0941, + "step": 32210 + }, + { + "epoch": 1.42, + "learning_rate": 0.00029378154249761, + "loss": 0.0901, + "step": 32220 + }, + { + "epoch": 1.42, + "learning_rate": 0.00029355921652326637, + "loss": 0.0986, + "step": 32230 + }, + { + "epoch": 1.42, + "learning_rate": 0.0002933368905489228, + "loss": 0.0982, + "step": 32240 + }, + { + "epoch": 1.42, + "learning_rate": 0.0002931145645745792, + "loss": 0.0963, + "step": 32250 + }, + { + "epoch": 1.42, + "learning_rate": 0.00029289223860023565, + "loss": 0.0996, + "step": 32260 + }, + { + "epoch": 1.42, + "learning_rate": 0.0002926699126258921, + "loss": 0.0917, + "step": 32270 + }, + { + "epoch": 1.42, + "learning_rate": 0.0002924475866515485, + "loss": 0.0954, + "step": 32280 + }, + { + "epoch": 1.42, + "learning_rate": 0.00029222526067720493, + "loss": 0.0982, + "step": 32290 + }, + { + "epoch": 1.42, + "learning_rate": 0.0002920029347028613, + "loss": 0.0978, + "step": 32300 + }, + { + "epoch": 1.42, + "learning_rate": 0.00029178060872851774, + "loss": 0.0961, + "step": 32310 + }, + { + "epoch": 1.42, + "learning_rate": 0.00029155828275417416, + "loss": 0.0962, + "step": 32320 + }, + { + "epoch": 1.42, + "learning_rate": 0.0002913359567798306, + "loss": 0.0937, + "step": 32330 + }, + { + "epoch": 1.42, + "learning_rate": 0.000291113630805487, + "loss": 0.0986, + "step": 32340 + }, + { + "epoch": 1.42, + "learning_rate": 0.00029089130483114344, + "loss": 0.0985, + "step": 32350 + }, + { + "epoch": 1.42, + "learning_rate": 0.0002906689788567998, + "loss": 0.0946, + "step": 32360 + }, + { + "epoch": 1.42, + "learning_rate": 0.00029044665288245625, + "loss": 0.095, + "step": 32370 + }, + { + "epoch": 1.43, + "learning_rate": 0.00029022432690811267, + "loss": 0.097, + "step": 32380 + }, + { + "epoch": 1.43, + "learning_rate": 0.0002900020009337691, + "loss": 0.0999, + "step": 32390 + }, + { + "epoch": 1.43, + "learning_rate": 0.00028977967495942553, + "loss": 0.0944, + "step": 32400 + }, + { + "epoch": 1.43, + "learning_rate": 0.00028955734898508195, + "loss": 0.0925, + "step": 32410 + }, + { + "epoch": 1.43, + "learning_rate": 0.00028933502301073833, + "loss": 0.0958, + "step": 32420 + }, + { + "epoch": 1.43, + "learning_rate": 0.00028911269703639476, + "loss": 0.0935, + "step": 32430 + }, + { + "epoch": 1.43, + "learning_rate": 0.0002888903710620512, + "loss": 0.0956, + "step": 32440 + }, + { + "epoch": 1.43, + "learning_rate": 0.0002886680450877076, + "loss": 0.0992, + "step": 32450 + }, + { + "epoch": 1.43, + "learning_rate": 0.00028844571911336404, + "loss": 0.0993, + "step": 32460 + }, + { + "epoch": 1.43, + "learning_rate": 0.00028822339313902046, + "loss": 0.0957, + "step": 32470 + }, + { + "epoch": 1.43, + "learning_rate": 0.00028800106716467684, + "loss": 0.0966, + "step": 32480 + }, + { + "epoch": 1.43, + "learning_rate": 0.00028777874119033327, + "loss": 0.0956, + "step": 32490 + }, + { + "epoch": 1.43, + "learning_rate": 0.0002875564152159897, + "loss": 0.0977, + "step": 32500 + }, + { + "epoch": 1.43, + "eval_cer": 0.49893196004953266, + "eval_loss": 0.08076103031635284, + "eval_runtime": 83.7148, + "eval_samples_per_second": 119.382, + "eval_steps_per_second": 7.466, + "step": 32500 + }, + { + "epoch": 1.43, + "learning_rate": 0.0002873340892416461, + "loss": 0.0917, + "step": 32510 + }, + { + "epoch": 1.43, + "learning_rate": 0.00028711176326730255, + "loss": 0.0966, + "step": 32520 + }, + { + "epoch": 1.43, + "learning_rate": 0.000286889437292959, + "loss": 0.092, + "step": 32530 + }, + { + "epoch": 1.43, + "learning_rate": 0.00028666711131861535, + "loss": 0.099, + "step": 32540 + }, + { + "epoch": 1.43, + "learning_rate": 0.0002864447853442718, + "loss": 0.0981, + "step": 32550 + }, + { + "epoch": 1.43, + "learning_rate": 0.0002862224593699282, + "loss": 0.0966, + "step": 32560 + }, + { + "epoch": 1.43, + "learning_rate": 0.00028600013339558463, + "loss": 0.0959, + "step": 32570 + }, + { + "epoch": 1.43, + "learning_rate": 0.00028577780742124106, + "loss": 0.0936, + "step": 32580 + }, + { + "epoch": 1.43, + "learning_rate": 0.0002855554814468975, + "loss": 0.0967, + "step": 32590 + }, + { + "epoch": 1.43, + "learning_rate": 0.00028533315547255386, + "loss": 0.0962, + "step": 32600 + }, + { + "epoch": 1.44, + "learning_rate": 0.0002851108294982103, + "loss": 0.0902, + "step": 32610 + }, + { + "epoch": 1.44, + "learning_rate": 0.0002848885035238667, + "loss": 0.0962, + "step": 32620 + }, + { + "epoch": 1.44, + "learning_rate": 0.00028466617754952314, + "loss": 0.0971, + "step": 32630 + }, + { + "epoch": 1.44, + "learning_rate": 0.00028444385157517957, + "loss": 0.0958, + "step": 32640 + }, + { + "epoch": 1.44, + "learning_rate": 0.000284221525600836, + "loss": 0.0986, + "step": 32650 + }, + { + "epoch": 1.44, + "learning_rate": 0.00028399919962649237, + "loss": 0.0974, + "step": 32660 + }, + { + "epoch": 1.44, + "learning_rate": 0.0002837768736521488, + "loss": 0.0945, + "step": 32670 + }, + { + "epoch": 1.44, + "learning_rate": 0.0002835545476778052, + "loss": 0.0941, + "step": 32680 + }, + { + "epoch": 1.44, + "learning_rate": 0.00028333222170346165, + "loss": 0.0938, + "step": 32690 + }, + { + "epoch": 1.44, + "learning_rate": 0.000283109895729118, + "loss": 0.0926, + "step": 32700 + }, + { + "epoch": 1.44, + "learning_rate": 0.00028288756975477445, + "loss": 0.0959, + "step": 32710 + }, + { + "epoch": 1.44, + "learning_rate": 0.0002826652437804308, + "loss": 0.093, + "step": 32720 + }, + { + "epoch": 1.44, + "learning_rate": 0.00028244291780608725, + "loss": 0.0951, + "step": 32730 + }, + { + "epoch": 1.44, + "learning_rate": 0.0002822205918317437, + "loss": 0.0973, + "step": 32740 + }, + { + "epoch": 1.44, + "learning_rate": 0.0002819982658574001, + "loss": 0.0936, + "step": 32750 + }, + { + "epoch": 1.44, + "learning_rate": 0.00028177593988305653, + "loss": 0.0978, + "step": 32760 + }, + { + "epoch": 1.44, + "learning_rate": 0.00028155361390871296, + "loss": 0.0958, + "step": 32770 + }, + { + "epoch": 1.44, + "learning_rate": 0.00028133128793436933, + "loss": 0.0947, + "step": 32780 + }, + { + "epoch": 1.44, + "learning_rate": 0.00028110896196002576, + "loss": 0.0939, + "step": 32790 + }, + { + "epoch": 1.44, + "learning_rate": 0.0002808866359856822, + "loss": 0.0948, + "step": 32800 + }, + { + "epoch": 1.44, + "learning_rate": 0.0002806643100113386, + "loss": 0.0931, + "step": 32810 + }, + { + "epoch": 1.44, + "learning_rate": 0.00028044198403699504, + "loss": 0.0975, + "step": 32820 + }, + { + "epoch": 1.45, + "learning_rate": 0.00028021965806265147, + "loss": 0.0956, + "step": 32830 + }, + { + "epoch": 1.45, + "learning_rate": 0.0002799973320883079, + "loss": 0.0943, + "step": 32840 + }, + { + "epoch": 1.45, + "learning_rate": 0.00027977500611396427, + "loss": 0.0934, + "step": 32850 + }, + { + "epoch": 1.45, + "learning_rate": 0.0002795526801396207, + "loss": 0.095, + "step": 32860 + }, + { + "epoch": 1.45, + "learning_rate": 0.0002793303541652771, + "loss": 0.0939, + "step": 32870 + }, + { + "epoch": 1.45, + "learning_rate": 0.00027910802819093355, + "loss": 0.0954, + "step": 32880 + }, + { + "epoch": 1.45, + "learning_rate": 0.00027888570221659, + "loss": 0.0954, + "step": 32890 + }, + { + "epoch": 1.45, + "learning_rate": 0.0002786633762422464, + "loss": 0.0915, + "step": 32900 + }, + { + "epoch": 1.45, + "learning_rate": 0.0002784410502679028, + "loss": 0.0936, + "step": 32910 + }, + { + "epoch": 1.45, + "learning_rate": 0.0002782187242935592, + "loss": 0.0949, + "step": 32920 + }, + { + "epoch": 1.45, + "learning_rate": 0.00027799639831921564, + "loss": 0.0981, + "step": 32930 + }, + { + "epoch": 1.45, + "learning_rate": 0.00027777407234487206, + "loss": 0.0921, + "step": 32940 + }, + { + "epoch": 1.45, + "learning_rate": 0.0002775517463705285, + "loss": 0.0929, + "step": 32950 + }, + { + "epoch": 1.45, + "learning_rate": 0.0002773294203961849, + "loss": 0.0948, + "step": 32960 + }, + { + "epoch": 1.45, + "learning_rate": 0.0002771070944218413, + "loss": 0.0955, + "step": 32970 + }, + { + "epoch": 1.45, + "learning_rate": 0.0002768847684474977, + "loss": 0.0936, + "step": 32980 + }, + { + "epoch": 1.45, + "learning_rate": 0.00027666244247315415, + "loss": 0.0956, + "step": 32990 + }, + { + "epoch": 1.45, + "learning_rate": 0.0002764401164988106, + "loss": 0.0979, + "step": 33000 + }, + { + "epoch": 1.45, + "eval_cer": 0.49929846024510754, + "eval_loss": 0.08017853647470474, + "eval_runtime": 84.2308, + "eval_samples_per_second": 118.65, + "eval_steps_per_second": 7.42, + "step": 33000 + }, + { + "epoch": 1.45, + "learning_rate": 0.000276217790524467, + "loss": 0.0949, + "step": 33010 + }, + { + "epoch": 1.45, + "learning_rate": 0.00027599546455012343, + "loss": 0.0928, + "step": 33020 + }, + { + "epoch": 1.45, + "learning_rate": 0.0002757731385757798, + "loss": 0.0921, + "step": 33030 + }, + { + "epoch": 1.45, + "learning_rate": 0.00027555081260143623, + "loss": 0.0955, + "step": 33040 + }, + { + "epoch": 1.45, + "learning_rate": 0.00027532848662709266, + "loss": 0.0918, + "step": 33050 + }, + { + "epoch": 1.46, + "learning_rate": 0.0002751061606527491, + "loss": 0.0998, + "step": 33060 + }, + { + "epoch": 1.46, + "learning_rate": 0.0002748838346784055, + "loss": 0.0975, + "step": 33070 + }, + { + "epoch": 1.46, + "learning_rate": 0.00027466150870406194, + "loss": 0.0988, + "step": 33080 + }, + { + "epoch": 1.46, + "learning_rate": 0.0002744391827297183, + "loss": 0.0952, + "step": 33090 + }, + { + "epoch": 1.46, + "learning_rate": 0.00027421685675537474, + "loss": 0.0987, + "step": 33100 + }, + { + "epoch": 1.46, + "learning_rate": 0.00027399453078103117, + "loss": 0.0907, + "step": 33110 + }, + { + "epoch": 1.46, + "learning_rate": 0.0002737722048066876, + "loss": 0.0954, + "step": 33120 + }, + { + "epoch": 1.46, + "learning_rate": 0.000273549878832344, + "loss": 0.0921, + "step": 33130 + }, + { + "epoch": 1.46, + "learning_rate": 0.00027332755285800045, + "loss": 0.0913, + "step": 33140 + }, + { + "epoch": 1.46, + "learning_rate": 0.0002731052268836568, + "loss": 0.0963, + "step": 33150 + }, + { + "epoch": 1.46, + "learning_rate": 0.00027288290090931325, + "loss": 0.0951, + "step": 33160 + }, + { + "epoch": 1.46, + "learning_rate": 0.0002726605749349697, + "loss": 0.0978, + "step": 33170 + }, + { + "epoch": 1.46, + "learning_rate": 0.0002724382489606261, + "loss": 0.0929, + "step": 33180 + }, + { + "epoch": 1.46, + "learning_rate": 0.00027221592298628253, + "loss": 0.0947, + "step": 33190 + }, + { + "epoch": 1.46, + "learning_rate": 0.0002719935970119389, + "loss": 0.0932, + "step": 33200 + }, + { + "epoch": 1.46, + "learning_rate": 0.0002717712710375953, + "loss": 0.0943, + "step": 33210 + }, + { + "epoch": 1.46, + "learning_rate": 0.0002715489450632517, + "loss": 0.0946, + "step": 33220 + }, + { + "epoch": 1.46, + "learning_rate": 0.00027132661908890813, + "loss": 0.0915, + "step": 33230 + }, + { + "epoch": 1.46, + "learning_rate": 0.00027110429311456456, + "loss": 0.0945, + "step": 33240 + }, + { + "epoch": 1.46, + "learning_rate": 0.000270881967140221, + "loss": 0.094, + "step": 33250 + }, + { + "epoch": 1.46, + "learning_rate": 0.0002706596411658774, + "loss": 0.1009, + "step": 33260 + }, + { + "epoch": 1.46, + "learning_rate": 0.0002704373151915338, + "loss": 0.0969, + "step": 33270 + }, + { + "epoch": 1.46, + "learning_rate": 0.0002702149892171902, + "loss": 0.0947, + "step": 33280 + }, + { + "epoch": 1.47, + "learning_rate": 0.00026999266324284664, + "loss": 0.0929, + "step": 33290 + }, + { + "epoch": 1.47, + "learning_rate": 0.00026977033726850307, + "loss": 0.0964, + "step": 33300 + }, + { + "epoch": 1.47, + "learning_rate": 0.0002695480112941595, + "loss": 0.0922, + "step": 33310 + }, + { + "epoch": 1.47, + "learning_rate": 0.0002693256853198159, + "loss": 0.0927, + "step": 33320 + }, + { + "epoch": 1.47, + "learning_rate": 0.0002691033593454723, + "loss": 0.096, + "step": 33330 + }, + { + "epoch": 1.47, + "learning_rate": 0.0002688810333711287, + "loss": 0.0954, + "step": 33340 + }, + { + "epoch": 1.47, + "learning_rate": 0.00026865870739678515, + "loss": 0.0953, + "step": 33350 + }, + { + "epoch": 1.47, + "learning_rate": 0.0002684363814224416, + "loss": 0.0918, + "step": 33360 + }, + { + "epoch": 1.47, + "learning_rate": 0.000268214055448098, + "loss": 0.0925, + "step": 33370 + }, + { + "epoch": 1.47, + "learning_rate": 0.00026799172947375444, + "loss": 0.0926, + "step": 33380 + }, + { + "epoch": 1.47, + "learning_rate": 0.00026776940349941086, + "loss": 0.0923, + "step": 33390 + }, + { + "epoch": 1.47, + "learning_rate": 0.00026754707752506724, + "loss": 0.0932, + "step": 33400 + }, + { + "epoch": 1.47, + "learning_rate": 0.00026732475155072366, + "loss": 0.0965, + "step": 33410 + }, + { + "epoch": 1.47, + "learning_rate": 0.0002671024255763801, + "loss": 0.0922, + "step": 33420 + }, + { + "epoch": 1.47, + "learning_rate": 0.0002668800996020365, + "loss": 0.0952, + "step": 33430 + }, + { + "epoch": 1.47, + "learning_rate": 0.00026665777362769295, + "loss": 0.0965, + "step": 33440 + }, + { + "epoch": 1.47, + "learning_rate": 0.00026643544765334937, + "loss": 0.092, + "step": 33450 + }, + { + "epoch": 1.47, + "learning_rate": 0.00026621312167900575, + "loss": 0.0921, + "step": 33460 + }, + { + "epoch": 1.47, + "learning_rate": 0.0002659907957046622, + "loss": 0.0964, + "step": 33470 + }, + { + "epoch": 1.47, + "learning_rate": 0.0002657684697303186, + "loss": 0.0938, + "step": 33480 + }, + { + "epoch": 1.47, + "learning_rate": 0.00026554614375597503, + "loss": 0.0916, + "step": 33490 + }, + { + "epoch": 1.47, + "learning_rate": 0.00026532381778163146, + "loss": 0.0941, + "step": 33500 + }, + { + "epoch": 1.47, + "eval_cer": 0.49920440267279187, + "eval_loss": 0.08023788779973984, + "eval_runtime": 84.3212, + "eval_samples_per_second": 118.523, + "eval_steps_per_second": 7.412, + "step": 33500 + }, + { + "epoch": 1.48, + "learning_rate": 0.0002651014918072879, + "loss": 0.0931, + "step": 33510 + }, + { + "epoch": 1.48, + "learning_rate": 0.00026487916583294426, + "loss": 0.0952, + "step": 33520 + }, + { + "epoch": 1.48, + "learning_rate": 0.0002646568398586007, + "loss": 0.0903, + "step": 33530 + }, + { + "epoch": 1.48, + "learning_rate": 0.0002644345138842571, + "loss": 0.0926, + "step": 33540 + }, + { + "epoch": 1.48, + "learning_rate": 0.00026421218790991354, + "loss": 0.0946, + "step": 33550 + }, + { + "epoch": 1.48, + "learning_rate": 0.00026398986193556997, + "loss": 0.0908, + "step": 33560 + }, + { + "epoch": 1.48, + "learning_rate": 0.0002637675359612264, + "loss": 0.0951, + "step": 33570 + }, + { + "epoch": 1.48, + "learning_rate": 0.00026354520998688277, + "loss": 0.0945, + "step": 33580 + }, + { + "epoch": 1.48, + "learning_rate": 0.0002633228840125392, + "loss": 0.0944, + "step": 33590 + }, + { + "epoch": 1.48, + "learning_rate": 0.0002631005580381956, + "loss": 0.0965, + "step": 33600 + }, + { + "epoch": 1.48, + "learning_rate": 0.00026287823206385205, + "loss": 0.0924, + "step": 33610 + }, + { + "epoch": 1.48, + "learning_rate": 0.0002626559060895085, + "loss": 0.0975, + "step": 33620 + }, + { + "epoch": 1.48, + "learning_rate": 0.0002624335801151649, + "loss": 0.0961, + "step": 33630 + }, + { + "epoch": 1.48, + "learning_rate": 0.0002622112541408213, + "loss": 0.098, + "step": 33640 + }, + { + "epoch": 1.48, + "learning_rate": 0.0002619889281664777, + "loss": 0.0967, + "step": 33650 + }, + { + "epoch": 1.48, + "learning_rate": 0.00026176660219213413, + "loss": 0.0937, + "step": 33660 + }, + { + "epoch": 1.48, + "learning_rate": 0.00026154427621779056, + "loss": 0.0976, + "step": 33670 + }, + { + "epoch": 1.48, + "learning_rate": 0.000261321950243447, + "loss": 0.0954, + "step": 33680 + }, + { + "epoch": 1.48, + "learning_rate": 0.0002610996242691034, + "loss": 0.0917, + "step": 33690 + }, + { + "epoch": 1.48, + "learning_rate": 0.00026087729829475973, + "loss": 0.0913, + "step": 33700 + }, + { + "epoch": 1.48, + "learning_rate": 0.00026065497232041616, + "loss": 0.094, + "step": 33710 + }, + { + "epoch": 1.48, + "learning_rate": 0.0002604326463460726, + "loss": 0.093, + "step": 33720 + }, + { + "epoch": 1.48, + "learning_rate": 0.000260210320371729, + "loss": 0.0922, + "step": 33730 + }, + { + "epoch": 1.49, + "learning_rate": 0.00025998799439738544, + "loss": 0.0941, + "step": 33740 + }, + { + "epoch": 1.49, + "learning_rate": 0.00025976566842304187, + "loss": 0.0933, + "step": 33750 + }, + { + "epoch": 1.49, + "learning_rate": 0.00025954334244869824, + "loss": 0.0952, + "step": 33760 + }, + { + "epoch": 1.49, + "learning_rate": 0.00025932101647435467, + "loss": 0.092, + "step": 33770 + }, + { + "epoch": 1.49, + "learning_rate": 0.0002590986905000111, + "loss": 0.0964, + "step": 33780 + }, + { + "epoch": 1.49, + "learning_rate": 0.0002588763645256675, + "loss": 0.0945, + "step": 33790 + }, + { + "epoch": 1.49, + "learning_rate": 0.00025865403855132395, + "loss": 0.093, + "step": 33800 + }, + { + "epoch": 1.49, + "learning_rate": 0.0002584317125769804, + "loss": 0.0894, + "step": 33810 + }, + { + "epoch": 1.49, + "learning_rate": 0.00025820938660263675, + "loss": 0.0906, + "step": 33820 + }, + { + "epoch": 1.49, + "learning_rate": 0.0002579870606282932, + "loss": 0.1003, + "step": 33830 + }, + { + "epoch": 1.49, + "learning_rate": 0.0002577647346539496, + "loss": 0.0991, + "step": 33840 + }, + { + "epoch": 1.49, + "learning_rate": 0.00025754240867960603, + "loss": 0.0968, + "step": 33850 + }, + { + "epoch": 1.49, + "learning_rate": 0.00025732008270526246, + "loss": 0.0915, + "step": 33860 + }, + { + "epoch": 1.49, + "learning_rate": 0.0002570977567309189, + "loss": 0.0915, + "step": 33870 + }, + { + "epoch": 1.49, + "learning_rate": 0.0002568754307565753, + "loss": 0.0931, + "step": 33880 + }, + { + "epoch": 1.49, + "learning_rate": 0.0002566531047822317, + "loss": 0.0928, + "step": 33890 + }, + { + "epoch": 1.49, + "learning_rate": 0.0002564307788078881, + "loss": 0.0959, + "step": 33900 + }, + { + "epoch": 1.49, + "learning_rate": 0.00025620845283354454, + "loss": 0.0941, + "step": 33910 + }, + { + "epoch": 1.49, + "learning_rate": 0.00025598612685920097, + "loss": 0.0999, + "step": 33920 + }, + { + "epoch": 1.49, + "learning_rate": 0.0002557638008848574, + "loss": 0.0955, + "step": 33930 + }, + { + "epoch": 1.49, + "learning_rate": 0.0002555414749105138, + "loss": 0.0953, + "step": 33940 + }, + { + "epoch": 1.49, + "learning_rate": 0.0002553191489361702, + "loss": 0.0906, + "step": 33950 + }, + { + "epoch": 1.49, + "learning_rate": 0.0002550968229618266, + "loss": 0.0916, + "step": 33960 + }, + { + "epoch": 1.5, + "learning_rate": 0.00025487449698748305, + "loss": 0.0948, + "step": 33970 + }, + { + "epoch": 1.5, + "learning_rate": 0.0002546521710131395, + "loss": 0.0976, + "step": 33980 + }, + { + "epoch": 1.5, + "learning_rate": 0.0002544298450387959, + "loss": 0.0939, + "step": 33990 + }, + { + "epoch": 1.5, + "learning_rate": 0.00025420751906445234, + "loss": 0.0923, + "step": 34000 + }, + { + "epoch": 1.5, + "eval_cer": 0.4990312070051485, + "eval_loss": 0.07928372919559479, + "eval_runtime": 84.2302, + "eval_samples_per_second": 118.651, + "eval_steps_per_second": 7.42, + "step": 34000 + }, + { + "epoch": 1.5, + "learning_rate": 0.0002539851930901087, + "loss": 0.0901, + "step": 34010 + }, + { + "epoch": 1.5, + "learning_rate": 0.00025376286711576514, + "loss": 0.0913, + "step": 34020 + }, + { + "epoch": 1.5, + "learning_rate": 0.00025354054114142156, + "loss": 0.0928, + "step": 34030 + }, + { + "epoch": 1.5, + "learning_rate": 0.000253318215167078, + "loss": 0.0936, + "step": 34040 + }, + { + "epoch": 1.5, + "learning_rate": 0.0002530958891927344, + "loss": 0.0921, + "step": 34050 + }, + { + "epoch": 1.5, + "learning_rate": 0.00025287356321839085, + "loss": 0.0948, + "step": 34060 + }, + { + "epoch": 1.5, + "learning_rate": 0.0002526512372440472, + "loss": 0.0911, + "step": 34070 + }, + { + "epoch": 1.5, + "learning_rate": 0.00025242891126970365, + "loss": 0.097, + "step": 34080 + }, + { + "epoch": 1.5, + "learning_rate": 0.0002522065852953601, + "loss": 0.0962, + "step": 34090 + }, + { + "epoch": 1.5, + "learning_rate": 0.0002519842593210165, + "loss": 0.09, + "step": 34100 + }, + { + "epoch": 1.5, + "learning_rate": 0.00025176193334667293, + "loss": 0.0972, + "step": 34110 + }, + { + "epoch": 1.5, + "learning_rate": 0.00025153960737232936, + "loss": 0.0961, + "step": 34120 + }, + { + "epoch": 1.5, + "learning_rate": 0.00025131728139798573, + "loss": 0.0911, + "step": 34130 + }, + { + "epoch": 1.5, + "learning_rate": 0.00025109495542364216, + "loss": 0.0944, + "step": 34140 + }, + { + "epoch": 1.5, + "learning_rate": 0.0002508726294492986, + "loss": 0.0977, + "step": 34150 + }, + { + "epoch": 1.5, + "learning_rate": 0.000250650303474955, + "loss": 0.0969, + "step": 34160 + }, + { + "epoch": 1.5, + "learning_rate": 0.00025042797750061144, + "loss": 0.0978, + "step": 34170 + }, + { + "epoch": 1.5, + "learning_rate": 0.00025020565152626787, + "loss": 0.0914, + "step": 34180 + }, + { + "epoch": 1.5, + "learning_rate": 0.00024998332555192424, + "loss": 0.0944, + "step": 34190 + }, + { + "epoch": 1.51, + "learning_rate": 0.00024976099957758067, + "loss": 0.0944, + "step": 34200 + }, + { + "epoch": 1.51, + "learning_rate": 0.00024953867360323704, + "loss": 0.0922, + "step": 34210 + }, + { + "epoch": 1.51, + "learning_rate": 0.00024931634762889347, + "loss": 0.0927, + "step": 34220 + }, + { + "epoch": 1.51, + "learning_rate": 0.0002490940216545499, + "loss": 0.0937, + "step": 34230 + }, + { + "epoch": 1.51, + "learning_rate": 0.0002488716956802063, + "loss": 0.0854, + "step": 34240 + }, + { + "epoch": 1.51, + "learning_rate": 0.00024864936970586275, + "loss": 0.0945, + "step": 34250 + }, + { + "epoch": 1.51, + "learning_rate": 0.0002484270437315192, + "loss": 0.0959, + "step": 34260 + }, + { + "epoch": 1.51, + "learning_rate": 0.00024820471775717555, + "loss": 0.0943, + "step": 34270 + }, + { + "epoch": 1.51, + "learning_rate": 0.000247982391782832, + "loss": 0.0923, + "step": 34280 + }, + { + "epoch": 1.51, + "learning_rate": 0.0002477600658084884, + "loss": 0.0926, + "step": 34290 + }, + { + "epoch": 1.51, + "learning_rate": 0.00024753773983414483, + "loss": 0.0905, + "step": 34300 + }, + { + "epoch": 1.51, + "learning_rate": 0.00024731541385980126, + "loss": 0.0953, + "step": 34310 + }, + { + "epoch": 1.51, + "learning_rate": 0.0002470930878854577, + "loss": 0.0915, + "step": 34320 + }, + { + "epoch": 1.51, + "learning_rate": 0.00024687076191111406, + "loss": 0.0912, + "step": 34330 + }, + { + "epoch": 1.51, + "learning_rate": 0.0002466484359367705, + "loss": 0.0983, + "step": 34340 + }, + { + "epoch": 1.51, + "learning_rate": 0.0002464261099624269, + "loss": 0.0936, + "step": 34350 + }, + { + "epoch": 1.51, + "learning_rate": 0.00024620378398808334, + "loss": 0.0887, + "step": 34360 + }, + { + "epoch": 1.51, + "learning_rate": 0.00024598145801373977, + "loss": 0.0929, + "step": 34370 + }, + { + "epoch": 1.51, + "learning_rate": 0.0002457591320393962, + "loss": 0.0984, + "step": 34380 + }, + { + "epoch": 1.51, + "learning_rate": 0.00024553680606505257, + "loss": 0.0944, + "step": 34390 + }, + { + "epoch": 1.51, + "learning_rate": 0.000245314480090709, + "loss": 0.0931, + "step": 34400 + }, + { + "epoch": 1.51, + "learning_rate": 0.0002450921541163654, + "loss": 0.092, + "step": 34410 + }, + { + "epoch": 1.52, + "learning_rate": 0.00024486982814202185, + "loss": 0.0985, + "step": 34420 + }, + { + "epoch": 1.52, + "learning_rate": 0.0002446475021676783, + "loss": 0.0962, + "step": 34430 + }, + { + "epoch": 1.52, + "learning_rate": 0.0002444251761933347, + "loss": 0.092, + "step": 34440 + }, + { + "epoch": 1.52, + "learning_rate": 0.0002442028502189911, + "loss": 0.0926, + "step": 34450 + }, + { + "epoch": 1.52, + "learning_rate": 0.0002439805242446475, + "loss": 0.0898, + "step": 34460 + }, + { + "epoch": 1.52, + "learning_rate": 0.00024375819827030394, + "loss": 0.0974, + "step": 34470 + }, + { + "epoch": 1.52, + "learning_rate": 0.00024353587229596034, + "loss": 0.0947, + "step": 34480 + }, + { + "epoch": 1.52, + "learning_rate": 0.00024331354632161676, + "loss": 0.0986, + "step": 34490 + }, + { + "epoch": 1.52, + "learning_rate": 0.0002430912203472732, + "loss": 0.0905, + "step": 34500 + }, + { + "epoch": 1.52, + "eval_cer": 0.498808063523241, + "eval_loss": 0.07936431467533112, + "eval_runtime": 84.1313, + "eval_samples_per_second": 118.791, + "eval_steps_per_second": 7.429, + "step": 34500 + }, + { + "epoch": 1.52, + "learning_rate": 0.0002428688943729296, + "loss": 0.0946, + "step": 34510 + }, + { + "epoch": 1.52, + "learning_rate": 0.00024264656839858602, + "loss": 0.0938, + "step": 34520 + }, + { + "epoch": 1.52, + "learning_rate": 0.00024242424242424245, + "loss": 0.0922, + "step": 34530 + }, + { + "epoch": 1.52, + "learning_rate": 0.00024220191644989885, + "loss": 0.0897, + "step": 34540 + }, + { + "epoch": 1.52, + "learning_rate": 0.00024197959047555527, + "loss": 0.097, + "step": 34550 + }, + { + "epoch": 1.52, + "learning_rate": 0.0002417572645012117, + "loss": 0.0892, + "step": 34560 + }, + { + "epoch": 1.52, + "learning_rate": 0.00024153493852686807, + "loss": 0.1004, + "step": 34570 + }, + { + "epoch": 1.52, + "learning_rate": 0.0002413126125525245, + "loss": 0.0944, + "step": 34580 + }, + { + "epoch": 1.52, + "learning_rate": 0.00024109028657818093, + "loss": 0.093, + "step": 34590 + }, + { + "epoch": 1.52, + "learning_rate": 0.00024086796060383733, + "loss": 0.093, + "step": 34600 + }, + { + "epoch": 1.52, + "learning_rate": 0.00024064563462949376, + "loss": 0.0957, + "step": 34610 + }, + { + "epoch": 1.52, + "learning_rate": 0.00024042330865515018, + "loss": 0.0955, + "step": 34620 + }, + { + "epoch": 1.52, + "learning_rate": 0.00024020098268080658, + "loss": 0.0925, + "step": 34630 + }, + { + "epoch": 1.52, + "learning_rate": 0.000239978656706463, + "loss": 0.0932, + "step": 34640 + }, + { + "epoch": 1.53, + "learning_rate": 0.00023975633073211944, + "loss": 0.0946, + "step": 34650 + }, + { + "epoch": 1.53, + "learning_rate": 0.00023953400475777587, + "loss": 0.0934, + "step": 34660 + }, + { + "epoch": 1.53, + "learning_rate": 0.00023931167878343227, + "loss": 0.0934, + "step": 34670 + }, + { + "epoch": 1.53, + "learning_rate": 0.0002390893528090887, + "loss": 0.0972, + "step": 34680 + }, + { + "epoch": 1.53, + "learning_rate": 0.00023886702683474512, + "loss": 0.0963, + "step": 34690 + }, + { + "epoch": 1.53, + "learning_rate": 0.00023864470086040152, + "loss": 0.093, + "step": 34700 + }, + { + "epoch": 1.53, + "learning_rate": 0.00023842237488605795, + "loss": 0.0904, + "step": 34710 + }, + { + "epoch": 1.53, + "learning_rate": 0.00023820004891171438, + "loss": 0.0956, + "step": 34720 + }, + { + "epoch": 1.53, + "learning_rate": 0.00023797772293737078, + "loss": 0.0911, + "step": 34730 + }, + { + "epoch": 1.53, + "learning_rate": 0.0002377553969630272, + "loss": 0.0949, + "step": 34740 + }, + { + "epoch": 1.53, + "learning_rate": 0.00023753307098868363, + "loss": 0.0949, + "step": 34750 + }, + { + "epoch": 1.53, + "learning_rate": 0.00023731074501434003, + "loss": 0.0948, + "step": 34760 + }, + { + "epoch": 1.53, + "learning_rate": 0.00023708841903999646, + "loss": 0.0926, + "step": 34770 + }, + { + "epoch": 1.53, + "learning_rate": 0.00023686609306565289, + "loss": 0.09, + "step": 34780 + }, + { + "epoch": 1.53, + "learning_rate": 0.0002366437670913093, + "loss": 0.0911, + "step": 34790 + }, + { + "epoch": 1.53, + "learning_rate": 0.00023642144111696571, + "loss": 0.093, + "step": 34800 + }, + { + "epoch": 1.53, + "learning_rate": 0.00023619911514262214, + "loss": 0.0956, + "step": 34810 + }, + { + "epoch": 1.53, + "learning_rate": 0.00023597678916827851, + "loss": 0.0917, + "step": 34820 + }, + { + "epoch": 1.53, + "learning_rate": 0.00023575446319393494, + "loss": 0.0934, + "step": 34830 + }, + { + "epoch": 1.53, + "learning_rate": 0.00023553213721959137, + "loss": 0.0936, + "step": 34840 + }, + { + "epoch": 1.53, + "learning_rate": 0.00023530981124524777, + "loss": 0.091, + "step": 34850 + }, + { + "epoch": 1.53, + "learning_rate": 0.0002350874852709042, + "loss": 0.0947, + "step": 34860 + }, + { + "epoch": 1.53, + "learning_rate": 0.00023486515929656062, + "loss": 0.0889, + "step": 34870 + }, + { + "epoch": 1.54, + "learning_rate": 0.00023464283332221702, + "loss": 0.0926, + "step": 34880 + }, + { + "epoch": 1.54, + "learning_rate": 0.00023442050734787345, + "loss": 0.0922, + "step": 34890 + }, + { + "epoch": 1.54, + "learning_rate": 0.00023419818137352988, + "loss": 0.0943, + "step": 34900 + }, + { + "epoch": 1.54, + "learning_rate": 0.00023397585539918628, + "loss": 0.0927, + "step": 34910 + }, + { + "epoch": 1.54, + "learning_rate": 0.0002337535294248427, + "loss": 0.0943, + "step": 34920 + }, + { + "epoch": 1.54, + "learning_rate": 0.00023353120345049913, + "loss": 0.0916, + "step": 34930 + }, + { + "epoch": 1.54, + "learning_rate": 0.00023330887747615553, + "loss": 0.0939, + "step": 34940 + }, + { + "epoch": 1.54, + "learning_rate": 0.00023308655150181196, + "loss": 0.0921, + "step": 34950 + }, + { + "epoch": 1.54, + "learning_rate": 0.0002328642255274684, + "loss": 0.0941, + "step": 34960 + }, + { + "epoch": 1.54, + "learning_rate": 0.0002326418995531248, + "loss": 0.0917, + "step": 34970 + }, + { + "epoch": 1.54, + "learning_rate": 0.00023241957357878122, + "loss": 0.092, + "step": 34980 + }, + { + "epoch": 1.54, + "learning_rate": 0.00023219724760443764, + "loss": 0.092, + "step": 34990 + }, + { + "epoch": 1.54, + "learning_rate": 0.00023197492163009404, + "loss": 0.0885, + "step": 35000 + }, + { + "epoch": 1.54, + "eval_cer": 0.49906753268824977, + "eval_loss": 0.07893533259630203, + "eval_runtime": 84.2504, + "eval_samples_per_second": 118.623, + "eval_steps_per_second": 7.418, + "step": 35000 + }, + { + "epoch": 1.54, + "learning_rate": 0.00023175259565575047, + "loss": 0.095, + "step": 35010 + }, + { + "epoch": 1.54, + "learning_rate": 0.0002315302696814069, + "loss": 0.0929, + "step": 35020 + }, + { + "epoch": 1.54, + "learning_rate": 0.0002313079437070633, + "loss": 0.0916, + "step": 35030 + }, + { + "epoch": 1.54, + "learning_rate": 0.00023108561773271973, + "loss": 0.0939, + "step": 35040 + }, + { + "epoch": 1.54, + "learning_rate": 0.00023086329175837615, + "loss": 0.0964, + "step": 35050 + }, + { + "epoch": 1.54, + "learning_rate": 0.00023064096578403255, + "loss": 0.0907, + "step": 35060 + }, + { + "epoch": 1.54, + "learning_rate": 0.00023041863980968896, + "loss": 0.0915, + "step": 35070 + }, + { + "epoch": 1.54, + "learning_rate": 0.00023019631383534538, + "loss": 0.0912, + "step": 35080 + }, + { + "epoch": 1.54, + "learning_rate": 0.00022997398786100178, + "loss": 0.0945, + "step": 35090 + }, + { + "epoch": 1.55, + "learning_rate": 0.0002297516618866582, + "loss": 0.0946, + "step": 35100 + }, + { + "epoch": 1.55, + "learning_rate": 0.00022952933591231464, + "loss": 0.0937, + "step": 35110 + }, + { + "epoch": 1.55, + "learning_rate": 0.00022930700993797104, + "loss": 0.0904, + "step": 35120 + }, + { + "epoch": 1.55, + "learning_rate": 0.00022908468396362747, + "loss": 0.0964, + "step": 35130 + }, + { + "epoch": 1.55, + "learning_rate": 0.0002288623579892839, + "loss": 0.0963, + "step": 35140 + }, + { + "epoch": 1.55, + "learning_rate": 0.0002286400320149403, + "loss": 0.0939, + "step": 35150 + }, + { + "epoch": 1.55, + "learning_rate": 0.00022841770604059672, + "loss": 0.0894, + "step": 35160 + }, + { + "epoch": 1.55, + "learning_rate": 0.00022819538006625315, + "loss": 0.09, + "step": 35170 + }, + { + "epoch": 1.55, + "learning_rate": 0.00022797305409190958, + "loss": 0.0944, + "step": 35180 + }, + { + "epoch": 1.55, + "learning_rate": 0.00022775072811756598, + "loss": 0.0918, + "step": 35190 + }, + { + "epoch": 1.55, + "learning_rate": 0.0002275284021432224, + "loss": 0.0952, + "step": 35200 + }, + { + "epoch": 1.55, + "learning_rate": 0.00022730607616887883, + "loss": 0.0947, + "step": 35210 + }, + { + "epoch": 1.55, + "learning_rate": 0.00022708375019453523, + "loss": 0.0976, + "step": 35220 + }, + { + "epoch": 1.55, + "learning_rate": 0.00022686142422019166, + "loss": 0.0928, + "step": 35230 + }, + { + "epoch": 1.55, + "learning_rate": 0.00022663909824584809, + "loss": 0.0936, + "step": 35240 + }, + { + "epoch": 1.55, + "learning_rate": 0.00022641677227150449, + "loss": 0.09, + "step": 35250 + }, + { + "epoch": 1.55, + "learning_rate": 0.0002261944462971609, + "loss": 0.0933, + "step": 35260 + }, + { + "epoch": 1.55, + "learning_rate": 0.00022597212032281734, + "loss": 0.089, + "step": 35270 + }, + { + "epoch": 1.55, + "learning_rate": 0.00022574979434847374, + "loss": 0.0927, + "step": 35280 + }, + { + "epoch": 1.55, + "learning_rate": 0.00022552746837413017, + "loss": 0.095, + "step": 35290 + }, + { + "epoch": 1.55, + "learning_rate": 0.0002253051423997866, + "loss": 0.0923, + "step": 35300 + }, + { + "epoch": 1.55, + "learning_rate": 0.000225082816425443, + "loss": 0.0942, + "step": 35310 + }, + { + "epoch": 1.55, + "learning_rate": 0.0002248604904510994, + "loss": 0.0937, + "step": 35320 + }, + { + "epoch": 1.56, + "learning_rate": 0.00022463816447675582, + "loss": 0.09, + "step": 35330 + }, + { + "epoch": 1.56, + "learning_rate": 0.00022441583850241222, + "loss": 0.0916, + "step": 35340 + }, + { + "epoch": 1.56, + "learning_rate": 0.00022419351252806865, + "loss": 0.0899, + "step": 35350 + }, + { + "epoch": 1.56, + "learning_rate": 0.00022397118655372508, + "loss": 0.0921, + "step": 35360 + }, + { + "epoch": 1.56, + "learning_rate": 0.00022374886057938148, + "loss": 0.0928, + "step": 35370 + }, + { + "epoch": 1.56, + "learning_rate": 0.0002235265346050379, + "loss": 0.0967, + "step": 35380 + }, + { + "epoch": 1.56, + "learning_rate": 0.00022330420863069433, + "loss": 0.095, + "step": 35390 + }, + { + "epoch": 1.56, + "learning_rate": 0.00022308188265635073, + "loss": 0.0962, + "step": 35400 + }, + { + "epoch": 1.56, + "learning_rate": 0.00022285955668200716, + "loss": 0.0967, + "step": 35410 + }, + { + "epoch": 1.56, + "learning_rate": 0.0002226372307076636, + "loss": 0.0945, + "step": 35420 + }, + { + "epoch": 1.56, + "learning_rate": 0.00022241490473332, + "loss": 0.0902, + "step": 35430 + }, + { + "epoch": 1.56, + "learning_rate": 0.00022219257875897642, + "loss": 0.0964, + "step": 35440 + }, + { + "epoch": 1.56, + "learning_rate": 0.00022197025278463284, + "loss": 0.0919, + "step": 35450 + }, + { + "epoch": 1.56, + "learning_rate": 0.00022174792681028924, + "loss": 0.0905, + "step": 35460 + }, + { + "epoch": 1.56, + "learning_rate": 0.00022152560083594567, + "loss": 0.0948, + "step": 35470 + }, + { + "epoch": 1.56, + "learning_rate": 0.0002213032748616021, + "loss": 0.0911, + "step": 35480 + }, + { + "epoch": 1.56, + "learning_rate": 0.0002210809488872585, + "loss": 0.0887, + "step": 35490 + }, + { + "epoch": 1.56, + "learning_rate": 0.00022085862291291493, + "loss": 0.0941, + "step": 35500 + }, + { + "epoch": 1.56, + "eval_cer": 0.49907077605281236, + "eval_loss": 0.07846389710903168, + "eval_runtime": 84.3242, + "eval_samples_per_second": 118.519, + "eval_steps_per_second": 7.412, + "step": 35500 + }, + { + "epoch": 1.56, + "learning_rate": 0.00022063629693857135, + "loss": 0.0955, + "step": 35510 + }, + { + "epoch": 1.56, + "learning_rate": 0.00022041397096422775, + "loss": 0.0933, + "step": 35520 + }, + { + "epoch": 1.56, + "learning_rate": 0.00022019164498988418, + "loss": 0.0924, + "step": 35530 + }, + { + "epoch": 1.56, + "learning_rate": 0.0002199693190155406, + "loss": 0.0917, + "step": 35540 + }, + { + "epoch": 1.56, + "learning_rate": 0.000219746993041197, + "loss": 0.0963, + "step": 35550 + }, + { + "epoch": 1.57, + "learning_rate": 0.00021952466706685344, + "loss": 0.093, + "step": 35560 + }, + { + "epoch": 1.57, + "learning_rate": 0.00021930234109250984, + "loss": 0.0911, + "step": 35570 + }, + { + "epoch": 1.57, + "learning_rate": 0.00021908001511816624, + "loss": 0.0974, + "step": 35580 + }, + { + "epoch": 1.57, + "learning_rate": 0.00021885768914382266, + "loss": 0.093, + "step": 35590 + }, + { + "epoch": 1.57, + "learning_rate": 0.0002186353631694791, + "loss": 0.0916, + "step": 35600 + }, + { + "epoch": 1.57, + "learning_rate": 0.0002184130371951355, + "loss": 0.0938, + "step": 35610 + }, + { + "epoch": 1.57, + "learning_rate": 0.00021819071122079192, + "loss": 0.0919, + "step": 35620 + }, + { + "epoch": 1.57, + "learning_rate": 0.00021796838524644835, + "loss": 0.0935, + "step": 35630 + }, + { + "epoch": 1.57, + "learning_rate": 0.00021774605927210475, + "loss": 0.0961, + "step": 35640 + }, + { + "epoch": 1.57, + "learning_rate": 0.00021752373329776117, + "loss": 0.0913, + "step": 35650 + }, + { + "epoch": 1.57, + "learning_rate": 0.0002173014073234176, + "loss": 0.0946, + "step": 35660 + }, + { + "epoch": 1.57, + "learning_rate": 0.000217079081349074, + "loss": 0.0971, + "step": 35670 + }, + { + "epoch": 1.57, + "learning_rate": 0.00021685675537473043, + "loss": 0.0934, + "step": 35680 + }, + { + "epoch": 1.57, + "learning_rate": 0.00021663442940038686, + "loss": 0.0937, + "step": 35690 + }, + { + "epoch": 1.57, + "learning_rate": 0.00021641210342604326, + "loss": 0.0926, + "step": 35700 + }, + { + "epoch": 1.57, + "learning_rate": 0.00021618977745169968, + "loss": 0.0923, + "step": 35710 + }, + { + "epoch": 1.57, + "learning_rate": 0.0002159674514773561, + "loss": 0.0899, + "step": 35720 + }, + { + "epoch": 1.57, + "learning_rate": 0.00021574512550301254, + "loss": 0.0878, + "step": 35730 + }, + { + "epoch": 1.57, + "learning_rate": 0.00021552279952866894, + "loss": 0.0912, + "step": 35740 + }, + { + "epoch": 1.57, + "learning_rate": 0.00021530047355432537, + "loss": 0.0927, + "step": 35750 + }, + { + "epoch": 1.57, + "learning_rate": 0.0002150781475799818, + "loss": 0.0894, + "step": 35760 + }, + { + "epoch": 1.57, + "learning_rate": 0.0002148558216056382, + "loss": 0.0893, + "step": 35770 + }, + { + "epoch": 1.57, + "learning_rate": 0.00021463349563129462, + "loss": 0.0931, + "step": 35780 + }, + { + "epoch": 1.58, + "learning_rate": 0.00021441116965695105, + "loss": 0.0901, + "step": 35790 + }, + { + "epoch": 1.58, + "learning_rate": 0.00021418884368260745, + "loss": 0.0916, + "step": 35800 + }, + { + "epoch": 1.58, + "learning_rate": 0.00021396651770826388, + "loss": 0.0923, + "step": 35810 + }, + { + "epoch": 1.58, + "learning_rate": 0.00021374419173392028, + "loss": 0.0948, + "step": 35820 + }, + { + "epoch": 1.58, + "learning_rate": 0.00021352186575957668, + "loss": 0.0974, + "step": 35830 + }, + { + "epoch": 1.58, + "learning_rate": 0.0002132995397852331, + "loss": 0.0907, + "step": 35840 + }, + { + "epoch": 1.58, + "learning_rate": 0.00021307721381088953, + "loss": 0.0925, + "step": 35850 + }, + { + "epoch": 1.58, + "learning_rate": 0.00021285488783654593, + "loss": 0.0896, + "step": 35860 + }, + { + "epoch": 1.58, + "learning_rate": 0.00021263256186220236, + "loss": 0.0908, + "step": 35870 + }, + { + "epoch": 1.58, + "learning_rate": 0.0002124102358878588, + "loss": 0.0914, + "step": 35880 + }, + { + "epoch": 1.58, + "learning_rate": 0.0002121879099135152, + "loss": 0.0954, + "step": 35890 + }, + { + "epoch": 1.58, + "learning_rate": 0.00021196558393917161, + "loss": 0.093, + "step": 35900 + }, + { + "epoch": 1.58, + "learning_rate": 0.00021174325796482804, + "loss": 0.092, + "step": 35910 + }, + { + "epoch": 1.58, + "learning_rate": 0.00021152093199048444, + "loss": 0.0945, + "step": 35920 + }, + { + "epoch": 1.58, + "learning_rate": 0.00021129860601614087, + "loss": 0.0974, + "step": 35930 + }, + { + "epoch": 1.58, + "learning_rate": 0.0002110762800417973, + "loss": 0.0885, + "step": 35940 + }, + { + "epoch": 1.58, + "learning_rate": 0.0002108539540674537, + "loss": 0.088, + "step": 35950 + }, + { + "epoch": 1.58, + "learning_rate": 0.00021063162809311012, + "loss": 0.0919, + "step": 35960 + }, + { + "epoch": 1.58, + "learning_rate": 0.00021040930211876655, + "loss": 0.0952, + "step": 35970 + }, + { + "epoch": 1.58, + "learning_rate": 0.00021018697614442295, + "loss": 0.0932, + "step": 35980 + }, + { + "epoch": 1.58, + "learning_rate": 0.00020996465017007938, + "loss": 0.0942, + "step": 35990 + }, + { + "epoch": 1.58, + "learning_rate": 0.0002097423241957358, + "loss": 0.0966, + "step": 36000 + }, + { + "epoch": 1.58, + "eval_cer": 0.4989390954515704, + "eval_loss": 0.07804570347070694, + "eval_runtime": 84.2907, + "eval_samples_per_second": 118.566, + "eval_steps_per_second": 7.415, + "step": 36000 + }, + { + "epoch": 1.59, + "learning_rate": 0.0002095199982213922, + "loss": 0.0926, + "step": 36010 + }, + { + "epoch": 1.59, + "learning_rate": 0.00020929767224704863, + "loss": 0.0915, + "step": 36020 + }, + { + "epoch": 1.59, + "learning_rate": 0.00020907534627270506, + "loss": 0.0921, + "step": 36030 + }, + { + "epoch": 1.59, + "learning_rate": 0.00020885302029836146, + "loss": 0.0907, + "step": 36040 + }, + { + "epoch": 1.59, + "learning_rate": 0.0002086306943240179, + "loss": 0.0872, + "step": 36050 + }, + { + "epoch": 1.59, + "learning_rate": 0.00020840836834967432, + "loss": 0.0901, + "step": 36060 + }, + { + "epoch": 1.59, + "learning_rate": 0.0002081860423753307, + "loss": 0.0955, + "step": 36070 + }, + { + "epoch": 1.59, + "learning_rate": 0.00020796371640098712, + "loss": 0.0917, + "step": 36080 + }, + { + "epoch": 1.59, + "learning_rate": 0.00020774139042664355, + "loss": 0.0907, + "step": 36090 + }, + { + "epoch": 1.59, + "learning_rate": 0.00020751906445229995, + "loss": 0.0908, + "step": 36100 + }, + { + "epoch": 1.59, + "learning_rate": 0.00020729673847795637, + "loss": 0.0897, + "step": 36110 + }, + { + "epoch": 1.59, + "learning_rate": 0.0002070744125036128, + "loss": 0.0881, + "step": 36120 + }, + { + "epoch": 1.59, + "learning_rate": 0.0002068520865292692, + "loss": 0.0939, + "step": 36130 + }, + { + "epoch": 1.59, + "learning_rate": 0.00020662976055492563, + "loss": 0.0929, + "step": 36140 + }, + { + "epoch": 1.59, + "learning_rate": 0.00020640743458058206, + "loss": 0.0937, + "step": 36150 + }, + { + "epoch": 1.59, + "learning_rate": 0.00020618510860623846, + "loss": 0.0886, + "step": 36160 + }, + { + "epoch": 1.59, + "learning_rate": 0.00020596278263189488, + "loss": 0.0922, + "step": 36170 + }, + { + "epoch": 1.59, + "learning_rate": 0.0002057404566575513, + "loss": 0.0917, + "step": 36180 + }, + { + "epoch": 1.59, + "learning_rate": 0.0002055181306832077, + "loss": 0.0925, + "step": 36190 + }, + { + "epoch": 1.59, + "learning_rate": 0.00020529580470886414, + "loss": 0.0902, + "step": 36200 + }, + { + "epoch": 1.59, + "learning_rate": 0.00020507347873452057, + "loss": 0.0913, + "step": 36210 + }, + { + "epoch": 1.59, + "learning_rate": 0.00020485115276017697, + "loss": 0.0895, + "step": 36220 + }, + { + "epoch": 1.59, + "learning_rate": 0.0002046288267858334, + "loss": 0.0931, + "step": 36230 + }, + { + "epoch": 1.6, + "learning_rate": 0.00020440650081148982, + "loss": 0.0973, + "step": 36240 + }, + { + "epoch": 1.6, + "learning_rate": 0.00020418417483714625, + "loss": 0.0905, + "step": 36250 + }, + { + "epoch": 1.6, + "learning_rate": 0.00020396184886280265, + "loss": 0.0913, + "step": 36260 + }, + { + "epoch": 1.6, + "learning_rate": 0.00020373952288845908, + "loss": 0.0903, + "step": 36270 + }, + { + "epoch": 1.6, + "learning_rate": 0.0002035171969141155, + "loss": 0.095, + "step": 36280 + }, + { + "epoch": 1.6, + "learning_rate": 0.0002032948709397719, + "loss": 0.0937, + "step": 36290 + }, + { + "epoch": 1.6, + "learning_rate": 0.00020307254496542833, + "loss": 0.0901, + "step": 36300 + }, + { + "epoch": 1.6, + "learning_rate": 0.00020285021899108476, + "loss": 0.0899, + "step": 36310 + }, + { + "epoch": 1.6, + "learning_rate": 0.00020262789301674113, + "loss": 0.0914, + "step": 36320 + }, + { + "epoch": 1.6, + "learning_rate": 0.00020240556704239756, + "loss": 0.0892, + "step": 36330 + }, + { + "epoch": 1.6, + "learning_rate": 0.00020218324106805399, + "loss": 0.0908, + "step": 36340 + }, + { + "epoch": 1.6, + "learning_rate": 0.00020196091509371039, + "loss": 0.0893, + "step": 36350 + }, + { + "epoch": 1.6, + "learning_rate": 0.00020173858911936681, + "loss": 0.0887, + "step": 36360 + }, + { + "epoch": 1.6, + "learning_rate": 0.00020151626314502324, + "loss": 0.0955, + "step": 36370 + }, + { + "epoch": 1.6, + "learning_rate": 0.00020129393717067964, + "loss": 0.0928, + "step": 36380 + }, + { + "epoch": 1.6, + "learning_rate": 0.00020107161119633607, + "loss": 0.0923, + "step": 36390 + }, + { + "epoch": 1.6, + "learning_rate": 0.0002008492852219925, + "loss": 0.0887, + "step": 36400 + }, + { + "epoch": 1.6, + "learning_rate": 0.0002006269592476489, + "loss": 0.0937, + "step": 36410 + }, + { + "epoch": 1.6, + "learning_rate": 0.00020040463327330532, + "loss": 0.0927, + "step": 36420 + }, + { + "epoch": 1.6, + "learning_rate": 0.00020018230729896175, + "loss": 0.0852, + "step": 36430 + }, + { + "epoch": 1.6, + "learning_rate": 0.00019995998132461815, + "loss": 0.0964, + "step": 36440 + }, + { + "epoch": 1.6, + "learning_rate": 0.00019973765535027458, + "loss": 0.0912, + "step": 36450 + }, + { + "epoch": 1.6, + "learning_rate": 0.000199515329375931, + "loss": 0.0929, + "step": 36460 + }, + { + "epoch": 1.61, + "learning_rate": 0.0001992930034015874, + "loss": 0.0882, + "step": 36470 + }, + { + "epoch": 1.61, + "learning_rate": 0.00019907067742724383, + "loss": 0.0891, + "step": 36480 + }, + { + "epoch": 1.61, + "learning_rate": 0.00019884835145290026, + "loss": 0.0928, + "step": 36490 + }, + { + "epoch": 1.61, + "learning_rate": 0.00019862602547855666, + "loss": 0.0891, + "step": 36500 + }, + { + "epoch": 1.61, + "eval_cer": 0.49908829022145046, + "eval_loss": 0.07801735401153564, + "eval_runtime": 84.4038, + "eval_samples_per_second": 118.407, + "eval_steps_per_second": 7.405, + "step": 36500 + }, + { + "epoch": 1.61, + "learning_rate": 0.0001984036995042131, + "loss": 0.095, + "step": 36510 + }, + { + "epoch": 1.61, + "learning_rate": 0.00019818137352986952, + "loss": 0.0954, + "step": 36520 + }, + { + "epoch": 1.61, + "learning_rate": 0.00019795904755552592, + "loss": 0.0944, + "step": 36530 + }, + { + "epoch": 1.61, + "learning_rate": 0.00019773672158118234, + "loss": 0.0908, + "step": 36540 + }, + { + "epoch": 1.61, + "learning_rate": 0.00019751439560683877, + "loss": 0.0912, + "step": 36550 + }, + { + "epoch": 1.61, + "learning_rate": 0.00019729206963249517, + "loss": 0.0875, + "step": 36560 + }, + { + "epoch": 1.61, + "learning_rate": 0.00019706974365815157, + "loss": 0.0928, + "step": 36570 + }, + { + "epoch": 1.61, + "learning_rate": 0.000196847417683808, + "loss": 0.0902, + "step": 36580 + }, + { + "epoch": 1.61, + "learning_rate": 0.0001966250917094644, + "loss": 0.0981, + "step": 36590 + }, + { + "epoch": 1.61, + "learning_rate": 0.00019640276573512083, + "loss": 0.0953, + "step": 36600 + }, + { + "epoch": 1.61, + "learning_rate": 0.00019618043976077725, + "loss": 0.0923, + "step": 36610 + }, + { + "epoch": 1.61, + "learning_rate": 0.00019595811378643365, + "loss": 0.0916, + "step": 36620 + }, + { + "epoch": 1.61, + "learning_rate": 0.00019573578781209008, + "loss": 0.0881, + "step": 36630 + }, + { + "epoch": 1.61, + "learning_rate": 0.0001955134618377465, + "loss": 0.089, + "step": 36640 + }, + { + "epoch": 1.61, + "learning_rate": 0.0001952911358634029, + "loss": 0.0909, + "step": 36650 + }, + { + "epoch": 1.61, + "learning_rate": 0.00019506880988905934, + "loss": 0.0931, + "step": 36660 + }, + { + "epoch": 1.61, + "learning_rate": 0.00019484648391471576, + "loss": 0.0898, + "step": 36670 + }, + { + "epoch": 1.61, + "learning_rate": 0.00019462415794037216, + "loss": 0.096, + "step": 36680 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001944018319660286, + "loss": 0.0914, + "step": 36690 + }, + { + "epoch": 1.62, + "learning_rate": 0.00019417950599168502, + "loss": 0.0916, + "step": 36700 + }, + { + "epoch": 1.62, + "learning_rate": 0.00019395718001734142, + "loss": 0.0943, + "step": 36710 + }, + { + "epoch": 1.62, + "learning_rate": 0.00019373485404299785, + "loss": 0.0908, + "step": 36720 + }, + { + "epoch": 1.62, + "learning_rate": 0.00019351252806865427, + "loss": 0.0912, + "step": 36730 + }, + { + "epoch": 1.62, + "learning_rate": 0.00019329020209431067, + "loss": 0.094, + "step": 36740 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001930678761199671, + "loss": 0.0937, + "step": 36750 + }, + { + "epoch": 1.62, + "learning_rate": 0.00019284555014562353, + "loss": 0.0909, + "step": 36760 + }, + { + "epoch": 1.62, + "learning_rate": 0.00019262322417127996, + "loss": 0.0887, + "step": 36770 + }, + { + "epoch": 1.62, + "learning_rate": 0.00019240089819693636, + "loss": 0.0869, + "step": 36780 + }, + { + "epoch": 1.62, + "learning_rate": 0.00019217857222259278, + "loss": 0.0919, + "step": 36790 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001919562462482492, + "loss": 0.0918, + "step": 36800 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001917339202739056, + "loss": 0.0905, + "step": 36810 + }, + { + "epoch": 1.62, + "learning_rate": 0.000191511594299562, + "loss": 0.0895, + "step": 36820 + }, + { + "epoch": 1.62, + "learning_rate": 0.00019128926832521844, + "loss": 0.0929, + "step": 36830 + }, + { + "epoch": 1.62, + "learning_rate": 0.00019106694235087484, + "loss": 0.0922, + "step": 36840 + }, + { + "epoch": 1.62, + "learning_rate": 0.00019084461637653127, + "loss": 0.0883, + "step": 36850 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001906222904021877, + "loss": 0.091, + "step": 36860 + }, + { + "epoch": 1.62, + "learning_rate": 0.0001903999644278441, + "loss": 0.0922, + "step": 36870 + }, + { + "epoch": 1.62, + "learning_rate": 0.00019017763845350052, + "loss": 0.0911, + "step": 36880 + }, + { + "epoch": 1.62, + "learning_rate": 0.00018995531247915695, + "loss": 0.0882, + "step": 36890 + }, + { + "epoch": 1.62, + "learning_rate": 0.00018973298650481335, + "loss": 0.0892, + "step": 36900 + }, + { + "epoch": 1.62, + "learning_rate": 0.00018951066053046978, + "loss": 0.0926, + "step": 36910 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001892883345561262, + "loss": 0.0928, + "step": 36920 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001890660085817826, + "loss": 0.0946, + "step": 36930 + }, + { + "epoch": 1.63, + "learning_rate": 0.00018884368260743903, + "loss": 0.0944, + "step": 36940 + }, + { + "epoch": 1.63, + "learning_rate": 0.00018862135663309546, + "loss": 0.0913, + "step": 36950 + }, + { + "epoch": 1.63, + "learning_rate": 0.00018839903065875186, + "loss": 0.0963, + "step": 36960 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001881767046844083, + "loss": 0.0922, + "step": 36970 + }, + { + "epoch": 1.63, + "learning_rate": 0.00018795437871006471, + "loss": 0.0948, + "step": 36980 + }, + { + "epoch": 1.63, + "learning_rate": 0.00018773205273572112, + "loss": 0.0903, + "step": 36990 + }, + { + "epoch": 1.63, + "learning_rate": 0.00018750972676137754, + "loss": 0.0973, + "step": 37000 + }, + { + "epoch": 1.63, + "eval_cer": 0.4991849424854162, + "eval_loss": 0.07733184844255447, + "eval_runtime": 84.1491, + "eval_samples_per_second": 118.765, + "eval_steps_per_second": 7.427, + "step": 37000 + }, + { + "epoch": 1.63, + "learning_rate": 0.00018728740078703397, + "loss": 0.0883, + "step": 37010 + }, + { + "epoch": 1.63, + "learning_rate": 0.00018706507481269037, + "loss": 0.0893, + "step": 37020 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001868427488383468, + "loss": 0.0936, + "step": 37030 + }, + { + "epoch": 1.63, + "learning_rate": 0.00018662042286400323, + "loss": 0.0926, + "step": 37040 + }, + { + "epoch": 1.63, + "learning_rate": 0.00018639809688965963, + "loss": 0.0941, + "step": 37050 + }, + { + "epoch": 1.63, + "learning_rate": 0.00018617577091531605, + "loss": 0.0943, + "step": 37060 + }, + { + "epoch": 1.63, + "learning_rate": 0.00018595344494097245, + "loss": 0.0931, + "step": 37070 + }, + { + "epoch": 1.63, + "learning_rate": 0.00018573111896662885, + "loss": 0.0955, + "step": 37080 + }, + { + "epoch": 1.63, + "learning_rate": 0.00018550879299228528, + "loss": 0.0942, + "step": 37090 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001852864670179417, + "loss": 0.0913, + "step": 37100 + }, + { + "epoch": 1.63, + "learning_rate": 0.0001850641410435981, + "loss": 0.0894, + "step": 37110 + }, + { + "epoch": 1.63, + "learning_rate": 0.00018484181506925454, + "loss": 0.0917, + "step": 37120 + }, + { + "epoch": 1.63, + "learning_rate": 0.00018461948909491096, + "loss": 0.0877, + "step": 37130 + }, + { + "epoch": 1.63, + "learning_rate": 0.00018439716312056736, + "loss": 0.0902, + "step": 37140 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001841748371462238, + "loss": 0.0927, + "step": 37150 + }, + { + "epoch": 1.64, + "learning_rate": 0.00018395251117188022, + "loss": 0.0926, + "step": 37160 + }, + { + "epoch": 1.64, + "learning_rate": 0.00018373018519753662, + "loss": 0.0934, + "step": 37170 + }, + { + "epoch": 1.64, + "learning_rate": 0.00018350785922319305, + "loss": 0.0901, + "step": 37180 + }, + { + "epoch": 1.64, + "learning_rate": 0.00018328553324884947, + "loss": 0.0894, + "step": 37190 + }, + { + "epoch": 1.64, + "learning_rate": 0.00018306320727450587, + "loss": 0.0896, + "step": 37200 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001828408813001623, + "loss": 0.0946, + "step": 37210 + }, + { + "epoch": 1.64, + "learning_rate": 0.00018261855532581873, + "loss": 0.0922, + "step": 37220 + }, + { + "epoch": 1.64, + "learning_rate": 0.00018239622935147513, + "loss": 0.0916, + "step": 37230 + }, + { + "epoch": 1.64, + "learning_rate": 0.00018217390337713156, + "loss": 0.0921, + "step": 37240 + }, + { + "epoch": 1.64, + "learning_rate": 0.00018195157740278798, + "loss": 0.0932, + "step": 37250 + }, + { + "epoch": 1.64, + "learning_rate": 0.00018172925142844438, + "loss": 0.0936, + "step": 37260 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001815069254541008, + "loss": 0.0863, + "step": 37270 + }, + { + "epoch": 1.64, + "learning_rate": 0.00018128459947975724, + "loss": 0.0913, + "step": 37280 + }, + { + "epoch": 1.64, + "learning_rate": 0.00018106227350541367, + "loss": 0.0853, + "step": 37290 + }, + { + "epoch": 1.64, + "learning_rate": 0.00018083994753107007, + "loss": 0.0969, + "step": 37300 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001806176215567265, + "loss": 0.0933, + "step": 37310 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001803952955823829, + "loss": 0.0891, + "step": 37320 + }, + { + "epoch": 1.64, + "learning_rate": 0.0001801729696080393, + "loss": 0.0889, + "step": 37330 + }, + { + "epoch": 1.64, + "learning_rate": 0.00017995064363369572, + "loss": 0.0927, + "step": 37340 + }, + { + "epoch": 1.64, + "learning_rate": 0.00017972831765935215, + "loss": 0.0931, + "step": 37350 + }, + { + "epoch": 1.64, + "learning_rate": 0.00017950599168500855, + "loss": 0.0905, + "step": 37360 + }, + { + "epoch": 1.64, + "learning_rate": 0.00017928366571066498, + "loss": 0.088, + "step": 37370 + }, + { + "epoch": 1.65, + "learning_rate": 0.0001790613397363214, + "loss": 0.0932, + "step": 37380 + }, + { + "epoch": 1.65, + "learning_rate": 0.0001788390137619778, + "loss": 0.0855, + "step": 37390 + }, + { + "epoch": 1.65, + "learning_rate": 0.00017861668778763423, + "loss": 0.0922, + "step": 37400 + }, + { + "epoch": 1.65, + "learning_rate": 0.00017839436181329066, + "loss": 0.09, + "step": 37410 + }, + { + "epoch": 1.65, + "learning_rate": 0.00017817203583894706, + "loss": 0.0906, + "step": 37420 + }, + { + "epoch": 1.65, + "learning_rate": 0.00017794970986460349, + "loss": 0.0928, + "step": 37430 + }, + { + "epoch": 1.65, + "learning_rate": 0.00017772738389025991, + "loss": 0.0958, + "step": 37440 + }, + { + "epoch": 1.65, + "learning_rate": 0.00017750505791591631, + "loss": 0.091, + "step": 37450 + }, + { + "epoch": 1.65, + "learning_rate": 0.00017728273194157274, + "loss": 0.0854, + "step": 37460 + }, + { + "epoch": 1.65, + "learning_rate": 0.00017706040596722917, + "loss": 0.0886, + "step": 37470 + }, + { + "epoch": 1.65, + "learning_rate": 0.00017683807999288557, + "loss": 0.0881, + "step": 37480 + }, + { + "epoch": 1.65, + "learning_rate": 0.000176615754018542, + "loss": 0.0926, + "step": 37490 + }, + { + "epoch": 1.65, + "learning_rate": 0.00017639342804419842, + "loss": 0.0914, + "step": 37500 + }, + { + "epoch": 1.65, + "eval_cer": 0.4993626788634472, + "eval_loss": 0.07686267048120499, + "eval_runtime": 84.2554, + "eval_samples_per_second": 118.616, + "eval_steps_per_second": 7.418, + "step": 37500 + }, + { + "epoch": 1.65, + "learning_rate": 0.00017617110206985482, + "loss": 0.0938, + "step": 37510 + }, + { + "epoch": 1.65, + "learning_rate": 0.00017594877609551125, + "loss": 0.0892, + "step": 37520 + }, + { + "epoch": 1.65, + "learning_rate": 0.00017572645012116768, + "loss": 0.0898, + "step": 37530 + }, + { + "epoch": 1.65, + "learning_rate": 0.00017550412414682408, + "loss": 0.0888, + "step": 37540 + }, + { + "epoch": 1.65, + "learning_rate": 0.0001752817981724805, + "loss": 0.0869, + "step": 37550 + }, + { + "epoch": 1.65, + "learning_rate": 0.00017505947219813693, + "loss": 0.0876, + "step": 37560 + }, + { + "epoch": 1.65, + "learning_rate": 0.0001748371462237933, + "loss": 0.0892, + "step": 37570 + }, + { + "epoch": 1.65, + "learning_rate": 0.00017461482024944973, + "loss": 0.0889, + "step": 37580 + }, + { + "epoch": 1.65, + "learning_rate": 0.00017439249427510616, + "loss": 0.0909, + "step": 37590 + }, + { + "epoch": 1.66, + "learning_rate": 0.00017417016830076256, + "loss": 0.0931, + "step": 37600 + }, + { + "epoch": 1.66, + "learning_rate": 0.000173947842326419, + "loss": 0.0909, + "step": 37610 + }, + { + "epoch": 1.66, + "learning_rate": 0.00017372551635207542, + "loss": 0.0917, + "step": 37620 + }, + { + "epoch": 1.66, + "learning_rate": 0.00017350319037773182, + "loss": 0.0927, + "step": 37630 + }, + { + "epoch": 1.66, + "learning_rate": 0.00017328086440338824, + "loss": 0.0922, + "step": 37640 + }, + { + "epoch": 1.66, + "learning_rate": 0.00017305853842904467, + "loss": 0.0927, + "step": 37650 + }, + { + "epoch": 1.66, + "learning_rate": 0.00017283621245470107, + "loss": 0.0967, + "step": 37660 + }, + { + "epoch": 1.66, + "learning_rate": 0.0001726138864803575, + "loss": 0.0895, + "step": 37670 + }, + { + "epoch": 1.66, + "learning_rate": 0.00017239156050601393, + "loss": 0.0871, + "step": 37680 + }, + { + "epoch": 1.66, + "learning_rate": 0.00017216923453167033, + "loss": 0.093, + "step": 37690 + }, + { + "epoch": 1.66, + "learning_rate": 0.00017194690855732675, + "loss": 0.0945, + "step": 37700 + }, + { + "epoch": 1.66, + "learning_rate": 0.00017172458258298318, + "loss": 0.0907, + "step": 37710 + }, + { + "epoch": 1.66, + "learning_rate": 0.00017150225660863958, + "loss": 0.0894, + "step": 37720 + }, + { + "epoch": 1.66, + "learning_rate": 0.000171279930634296, + "loss": 0.0903, + "step": 37730 + }, + { + "epoch": 1.66, + "learning_rate": 0.00017105760465995244, + "loss": 0.0909, + "step": 37740 + }, + { + "epoch": 1.66, + "learning_rate": 0.00017083527868560884, + "loss": 0.0916, + "step": 37750 + }, + { + "epoch": 1.66, + "learning_rate": 0.00017061295271126526, + "loss": 0.0939, + "step": 37760 + }, + { + "epoch": 1.66, + "learning_rate": 0.0001703906267369217, + "loss": 0.0887, + "step": 37770 + }, + { + "epoch": 1.66, + "learning_rate": 0.0001701683007625781, + "loss": 0.0924, + "step": 37780 + }, + { + "epoch": 1.66, + "learning_rate": 0.00016994597478823452, + "loss": 0.0901, + "step": 37790 + }, + { + "epoch": 1.66, + "learning_rate": 0.00016972364881389095, + "loss": 0.0885, + "step": 37800 + }, + { + "epoch": 1.66, + "learning_rate": 0.00016950132283954737, + "loss": 0.0959, + "step": 37810 + }, + { + "epoch": 1.66, + "learning_rate": 0.00016927899686520375, + "loss": 0.0909, + "step": 37820 + }, + { + "epoch": 1.67, + "learning_rate": 0.00016905667089086018, + "loss": 0.0892, + "step": 37830 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001688343449165166, + "loss": 0.092, + "step": 37840 + }, + { + "epoch": 1.67, + "learning_rate": 0.000168612018942173, + "loss": 0.0914, + "step": 37850 + }, + { + "epoch": 1.67, + "learning_rate": 0.00016838969296782943, + "loss": 0.0914, + "step": 37860 + }, + { + "epoch": 1.67, + "learning_rate": 0.00016816736699348586, + "loss": 0.0909, + "step": 37870 + }, + { + "epoch": 1.67, + "learning_rate": 0.00016794504101914226, + "loss": 0.0849, + "step": 37880 + }, + { + "epoch": 1.67, + "learning_rate": 0.00016772271504479869, + "loss": 0.0944, + "step": 37890 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001675003890704551, + "loss": 0.0874, + "step": 37900 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001672780630961115, + "loss": 0.0947, + "step": 37910 + }, + { + "epoch": 1.67, + "learning_rate": 0.00016705573712176794, + "loss": 0.0938, + "step": 37920 + }, + { + "epoch": 1.67, + "learning_rate": 0.00016683341114742437, + "loss": 0.0963, + "step": 37930 + }, + { + "epoch": 1.67, + "learning_rate": 0.00016661108517308077, + "loss": 0.0905, + "step": 37940 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001663887591987372, + "loss": 0.0967, + "step": 37950 + }, + { + "epoch": 1.67, + "learning_rate": 0.00016616643322439362, + "loss": 0.0875, + "step": 37960 + }, + { + "epoch": 1.67, + "learning_rate": 0.00016594410725005002, + "loss": 0.0933, + "step": 37970 + }, + { + "epoch": 1.67, + "learning_rate": 0.00016572178127570645, + "loss": 0.0948, + "step": 37980 + }, + { + "epoch": 1.67, + "learning_rate": 0.00016549945530136288, + "loss": 0.0909, + "step": 37990 + }, + { + "epoch": 1.67, + "learning_rate": 0.00016527712932701928, + "loss": 0.0971, + "step": 38000 + }, + { + "epoch": 1.67, + "eval_cer": 0.499148616802315, + "eval_loss": 0.07678454369306564, + "eval_runtime": 84.2037, + "eval_samples_per_second": 118.688, + "eval_steps_per_second": 7.422, + "step": 38000 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001650548033526757, + "loss": 0.0876, + "step": 38010 + }, + { + "epoch": 1.67, + "learning_rate": 0.00016483247737833213, + "loss": 0.0898, + "step": 38020 + }, + { + "epoch": 1.67, + "learning_rate": 0.00016461015140398853, + "loss": 0.0884, + "step": 38030 + }, + { + "epoch": 1.67, + "learning_rate": 0.00016438782542964496, + "loss": 0.088, + "step": 38040 + }, + { + "epoch": 1.67, + "learning_rate": 0.0001641654994553014, + "loss": 0.0886, + "step": 38050 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001639431734809578, + "loss": 0.09, + "step": 38060 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001637208475066142, + "loss": 0.0858, + "step": 38070 + }, + { + "epoch": 1.68, + "learning_rate": 0.00016349852153227062, + "loss": 0.0928, + "step": 38080 + }, + { + "epoch": 1.68, + "learning_rate": 0.00016327619555792702, + "loss": 0.0886, + "step": 38090 + }, + { + "epoch": 1.68, + "learning_rate": 0.00016305386958358344, + "loss": 0.0884, + "step": 38100 + }, + { + "epoch": 1.68, + "learning_rate": 0.00016283154360923987, + "loss": 0.0927, + "step": 38110 + }, + { + "epoch": 1.68, + "learning_rate": 0.00016260921763489627, + "loss": 0.0922, + "step": 38120 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001623868916605527, + "loss": 0.0924, + "step": 38130 + }, + { + "epoch": 1.68, + "learning_rate": 0.00016216456568620913, + "loss": 0.0925, + "step": 38140 + }, + { + "epoch": 1.68, + "learning_rate": 0.00016194223971186553, + "loss": 0.0894, + "step": 38150 + }, + { + "epoch": 1.68, + "learning_rate": 0.00016171991373752195, + "loss": 0.0876, + "step": 38160 + }, + { + "epoch": 1.68, + "learning_rate": 0.00016149758776317838, + "loss": 0.0919, + "step": 38170 + }, + { + "epoch": 1.68, + "learning_rate": 0.00016127526178883478, + "loss": 0.0864, + "step": 38180 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001610529358144912, + "loss": 0.0875, + "step": 38190 + }, + { + "epoch": 1.68, + "learning_rate": 0.00016083060984014764, + "loss": 0.0902, + "step": 38200 + }, + { + "epoch": 1.68, + "learning_rate": 0.00016060828386580404, + "loss": 0.0898, + "step": 38210 + }, + { + "epoch": 1.68, + "learning_rate": 0.00016038595789146046, + "loss": 0.0915, + "step": 38220 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001601636319171169, + "loss": 0.0936, + "step": 38230 + }, + { + "epoch": 1.68, + "learning_rate": 0.0001599413059427733, + "loss": 0.0944, + "step": 38240 + }, + { + "epoch": 1.68, + "learning_rate": 0.00015971897996842972, + "loss": 0.0904, + "step": 38250 + }, + { + "epoch": 1.68, + "learning_rate": 0.00015949665399408615, + "loss": 0.0915, + "step": 38260 + }, + { + "epoch": 1.68, + "learning_rate": 0.00015927432801974255, + "loss": 0.0888, + "step": 38270 + }, + { + "epoch": 1.69, + "learning_rate": 0.00015905200204539897, + "loss": 0.0887, + "step": 38280 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001588296760710554, + "loss": 0.0919, + "step": 38290 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001586073500967118, + "loss": 0.0923, + "step": 38300 + }, + { + "epoch": 1.69, + "learning_rate": 0.00015838502412236823, + "loss": 0.0908, + "step": 38310 + }, + { + "epoch": 1.69, + "learning_rate": 0.00015816269814802463, + "loss": 0.0876, + "step": 38320 + }, + { + "epoch": 1.69, + "learning_rate": 0.00015794037217368103, + "loss": 0.0922, + "step": 38330 + }, + { + "epoch": 1.69, + "learning_rate": 0.00015771804619933746, + "loss": 0.0919, + "step": 38340 + }, + { + "epoch": 1.69, + "learning_rate": 0.00015749572022499388, + "loss": 0.0958, + "step": 38350 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001572733942506503, + "loss": 0.0938, + "step": 38360 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001570510682763067, + "loss": 0.0923, + "step": 38370 + }, + { + "epoch": 1.69, + "learning_rate": 0.00015682874230196314, + "loss": 0.0858, + "step": 38380 + }, + { + "epoch": 1.69, + "learning_rate": 0.00015660641632761957, + "loss": 0.09, + "step": 38390 + }, + { + "epoch": 1.69, + "learning_rate": 0.00015638409035327597, + "loss": 0.0876, + "step": 38400 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001561617643789324, + "loss": 0.0912, + "step": 38410 + }, + { + "epoch": 1.69, + "learning_rate": 0.00015593943840458882, + "loss": 0.0862, + "step": 38420 + }, + { + "epoch": 1.69, + "learning_rate": 0.00015571711243024522, + "loss": 0.096, + "step": 38430 + }, + { + "epoch": 1.69, + "learning_rate": 0.00015549478645590165, + "loss": 0.0905, + "step": 38440 + }, + { + "epoch": 1.69, + "learning_rate": 0.00015527246048155808, + "loss": 0.0906, + "step": 38450 + }, + { + "epoch": 1.69, + "learning_rate": 0.00015505013450721448, + "loss": 0.0939, + "step": 38460 + }, + { + "epoch": 1.69, + "learning_rate": 0.0001548278085328709, + "loss": 0.0902, + "step": 38470 + }, + { + "epoch": 1.69, + "learning_rate": 0.00015460548255852733, + "loss": 0.0931, + "step": 38480 + }, + { + "epoch": 1.69, + "learning_rate": 0.00015438315658418373, + "loss": 0.0897, + "step": 38490 + }, + { + "epoch": 1.69, + "learning_rate": 0.00015416083060984016, + "loss": 0.0933, + "step": 38500 + }, + { + "epoch": 1.69, + "eval_cer": 0.49899552999495983, + "eval_loss": 0.07686174660921097, + "eval_runtime": 84.2098, + "eval_samples_per_second": 118.68, + "eval_steps_per_second": 7.422, + "step": 38500 + }, + { + "epoch": 1.7, + "learning_rate": 0.0001539385046354966, + "loss": 0.0922, + "step": 38510 + }, + { + "epoch": 1.7, + "learning_rate": 0.000153716178661153, + "loss": 0.0912, + "step": 38520 + }, + { + "epoch": 1.7, + "learning_rate": 0.00015349385268680941, + "loss": 0.0903, + "step": 38530 + }, + { + "epoch": 1.7, + "learning_rate": 0.00015327152671246584, + "loss": 0.0887, + "step": 38540 + }, + { + "epoch": 1.7, + "learning_rate": 0.00015304920073812224, + "loss": 0.0927, + "step": 38550 + }, + { + "epoch": 1.7, + "learning_rate": 0.00015282687476377867, + "loss": 0.0919, + "step": 38560 + }, + { + "epoch": 1.7, + "learning_rate": 0.00015260454878943507, + "loss": 0.0925, + "step": 38570 + }, + { + "epoch": 1.7, + "learning_rate": 0.00015238222281509147, + "loss": 0.0873, + "step": 38580 + }, + { + "epoch": 1.7, + "learning_rate": 0.0001521598968407479, + "loss": 0.0932, + "step": 38590 + }, + { + "epoch": 1.7, + "learning_rate": 0.00015193757086640432, + "loss": 0.0895, + "step": 38600 + }, + { + "epoch": 1.7, + "learning_rate": 0.00015171524489206072, + "loss": 0.0921, + "step": 38610 + }, + { + "epoch": 1.7, + "learning_rate": 0.00015149291891771715, + "loss": 0.0913, + "step": 38620 + }, + { + "epoch": 1.7, + "learning_rate": 0.00015127059294337358, + "loss": 0.0894, + "step": 38630 + }, + { + "epoch": 1.7, + "learning_rate": 0.00015104826696902998, + "loss": 0.0929, + "step": 38640 + }, + { + "epoch": 1.7, + "learning_rate": 0.0001508259409946864, + "loss": 0.0936, + "step": 38650 + }, + { + "epoch": 1.7, + "learning_rate": 0.00015060361502034283, + "loss": 0.0887, + "step": 38660 + }, + { + "epoch": 1.7, + "learning_rate": 0.00015038128904599924, + "loss": 0.0892, + "step": 38670 + }, + { + "epoch": 1.7, + "learning_rate": 0.00015015896307165566, + "loss": 0.0916, + "step": 38680 + }, + { + "epoch": 1.7, + "learning_rate": 0.0001499366370973121, + "loss": 0.0908, + "step": 38690 + }, + { + "epoch": 1.7, + "learning_rate": 0.0001497143111229685, + "loss": 0.0879, + "step": 38700 + }, + { + "epoch": 1.7, + "learning_rate": 0.00014949198514862492, + "loss": 0.0861, + "step": 38710 + }, + { + "epoch": 1.7, + "learning_rate": 0.00014926965917428134, + "loss": 0.0906, + "step": 38720 + }, + { + "epoch": 1.7, + "learning_rate": 0.00014904733319993775, + "loss": 0.0916, + "step": 38730 + }, + { + "epoch": 1.71, + "learning_rate": 0.00014882500722559417, + "loss": 0.0892, + "step": 38740 + }, + { + "epoch": 1.71, + "learning_rate": 0.0001486026812512506, + "loss": 0.0891, + "step": 38750 + }, + { + "epoch": 1.71, + "learning_rate": 0.000148380355276907, + "loss": 0.0899, + "step": 38760 + }, + { + "epoch": 1.71, + "learning_rate": 0.00014815802930256343, + "loss": 0.0906, + "step": 38770 + }, + { + "epoch": 1.71, + "learning_rate": 0.00014793570332821985, + "loss": 0.0885, + "step": 38780 + }, + { + "epoch": 1.71, + "learning_rate": 0.00014771337735387626, + "loss": 0.0923, + "step": 38790 + }, + { + "epoch": 1.71, + "learning_rate": 0.00014749105137953268, + "loss": 0.0907, + "step": 38800 + }, + { + "epoch": 1.71, + "learning_rate": 0.0001472687254051891, + "loss": 0.093, + "step": 38810 + }, + { + "epoch": 1.71, + "learning_rate": 0.00014704639943084548, + "loss": 0.0888, + "step": 38820 + }, + { + "epoch": 1.71, + "learning_rate": 0.0001468240734565019, + "loss": 0.0895, + "step": 38830 + }, + { + "epoch": 1.71, + "learning_rate": 0.00014660174748215834, + "loss": 0.0877, + "step": 38840 + }, + { + "epoch": 1.71, + "learning_rate": 0.00014637942150781474, + "loss": 0.0927, + "step": 38850 + }, + { + "epoch": 1.71, + "learning_rate": 0.00014615709553347117, + "loss": 0.0898, + "step": 38860 + }, + { + "epoch": 1.71, + "learning_rate": 0.0001459347695591276, + "loss": 0.0938, + "step": 38870 + }, + { + "epoch": 1.71, + "learning_rate": 0.00014571244358478402, + "loss": 0.0904, + "step": 38880 + }, + { + "epoch": 1.71, + "learning_rate": 0.00014549011761044042, + "loss": 0.0908, + "step": 38890 + }, + { + "epoch": 1.71, + "learning_rate": 0.00014526779163609685, + "loss": 0.0931, + "step": 38900 + }, + { + "epoch": 1.71, + "learning_rate": 0.00014504546566175328, + "loss": 0.0915, + "step": 38910 + }, + { + "epoch": 1.71, + "learning_rate": 0.00014482313968740968, + "loss": 0.0921, + "step": 38920 + }, + { + "epoch": 1.71, + "learning_rate": 0.0001446008137130661, + "loss": 0.0886, + "step": 38930 + }, + { + "epoch": 1.71, + "learning_rate": 0.00014437848773872253, + "loss": 0.0912, + "step": 38940 + }, + { + "epoch": 1.71, + "learning_rate": 0.00014415616176437893, + "loss": 0.0911, + "step": 38950 + }, + { + "epoch": 1.71, + "learning_rate": 0.00014393383579003536, + "loss": 0.0922, + "step": 38960 + }, + { + "epoch": 1.72, + "learning_rate": 0.00014371150981569179, + "loss": 0.0923, + "step": 38970 + }, + { + "epoch": 1.72, + "learning_rate": 0.00014348918384134819, + "loss": 0.0869, + "step": 38980 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001432668578670046, + "loss": 0.0925, + "step": 38990 + }, + { + "epoch": 1.72, + "learning_rate": 0.00014304453189266104, + "loss": 0.0912, + "step": 39000 + }, + { + "epoch": 1.72, + "eval_cer": 0.49903574771553616, + "eval_loss": 0.07630728930234909, + "eval_runtime": 83.8297, + "eval_samples_per_second": 119.218, + "eval_steps_per_second": 7.456, + "step": 39000 + }, + { + "epoch": 1.72, + "learning_rate": 0.00014282220591831744, + "loss": 0.091, + "step": 39010 + }, + { + "epoch": 1.72, + "learning_rate": 0.00014259987994397387, + "loss": 0.0903, + "step": 39020 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001423775539696303, + "loss": 0.0906, + "step": 39030 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001421552279952867, + "loss": 0.0864, + "step": 39040 + }, + { + "epoch": 1.72, + "learning_rate": 0.00014193290202094312, + "loss": 0.0886, + "step": 39050 + }, + { + "epoch": 1.72, + "learning_rate": 0.00014171057604659955, + "loss": 0.0869, + "step": 39060 + }, + { + "epoch": 1.72, + "learning_rate": 0.00014148825007225592, + "loss": 0.0876, + "step": 39070 + }, + { + "epoch": 1.72, + "learning_rate": 0.00014126592409791235, + "loss": 0.09, + "step": 39080 + }, + { + "epoch": 1.72, + "learning_rate": 0.00014104359812356878, + "loss": 0.0915, + "step": 39090 + }, + { + "epoch": 1.72, + "learning_rate": 0.00014082127214922518, + "loss": 0.0891, + "step": 39100 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001405989461748816, + "loss": 0.0896, + "step": 39110 + }, + { + "epoch": 1.72, + "learning_rate": 0.00014037662020053803, + "loss": 0.0914, + "step": 39120 + }, + { + "epoch": 1.72, + "learning_rate": 0.00014015429422619443, + "loss": 0.0884, + "step": 39130 + }, + { + "epoch": 1.72, + "learning_rate": 0.00013993196825185086, + "loss": 0.0906, + "step": 39140 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001397096422775073, + "loss": 0.0892, + "step": 39150 + }, + { + "epoch": 1.72, + "learning_rate": 0.0001394873163031637, + "loss": 0.0881, + "step": 39160 + }, + { + "epoch": 1.72, + "learning_rate": 0.00013926499032882012, + "loss": 0.0911, + "step": 39170 + }, + { + "epoch": 1.72, + "learning_rate": 0.00013904266435447654, + "loss": 0.0925, + "step": 39180 + }, + { + "epoch": 1.73, + "learning_rate": 0.00013882033838013294, + "loss": 0.0893, + "step": 39190 + }, + { + "epoch": 1.73, + "learning_rate": 0.00013859801240578937, + "loss": 0.0859, + "step": 39200 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001383756864314458, + "loss": 0.0885, + "step": 39210 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001381533604571022, + "loss": 0.0903, + "step": 39220 + }, + { + "epoch": 1.73, + "learning_rate": 0.00013793103448275863, + "loss": 0.0925, + "step": 39230 + }, + { + "epoch": 1.73, + "learning_rate": 0.00013770870850841505, + "loss": 0.0876, + "step": 39240 + }, + { + "epoch": 1.73, + "learning_rate": 0.00013748638253407145, + "loss": 0.0936, + "step": 39250 + }, + { + "epoch": 1.73, + "learning_rate": 0.00013726405655972788, + "loss": 0.0889, + "step": 39260 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001370417305853843, + "loss": 0.0906, + "step": 39270 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001368194046110407, + "loss": 0.0892, + "step": 39280 + }, + { + "epoch": 1.73, + "learning_rate": 0.00013659707863669714, + "loss": 0.0885, + "step": 39290 + }, + { + "epoch": 1.73, + "learning_rate": 0.00013637475266235356, + "loss": 0.0882, + "step": 39300 + }, + { + "epoch": 1.73, + "learning_rate": 0.00013615242668800994, + "loss": 0.0919, + "step": 39310 + }, + { + "epoch": 1.73, + "learning_rate": 0.00013593010071366636, + "loss": 0.0854, + "step": 39320 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001357077747393228, + "loss": 0.0908, + "step": 39330 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001354854487649792, + "loss": 0.0919, + "step": 39340 + }, + { + "epoch": 1.73, + "learning_rate": 0.00013526312279063562, + "loss": 0.0887, + "step": 39350 + }, + { + "epoch": 1.73, + "learning_rate": 0.00013504079681629205, + "loss": 0.0898, + "step": 39360 + }, + { + "epoch": 1.73, + "learning_rate": 0.00013481847084194845, + "loss": 0.0896, + "step": 39370 + }, + { + "epoch": 1.73, + "learning_rate": 0.00013459614486760487, + "loss": 0.0898, + "step": 39380 + }, + { + "epoch": 1.73, + "learning_rate": 0.0001343738188932613, + "loss": 0.0889, + "step": 39390 + }, + { + "epoch": 1.73, + "learning_rate": 0.00013415149291891773, + "loss": 0.0934, + "step": 39400 + }, + { + "epoch": 1.73, + "learning_rate": 0.00013392916694457413, + "loss": 0.0908, + "step": 39410 + }, + { + "epoch": 1.74, + "learning_rate": 0.00013370684097023056, + "loss": 0.0907, + "step": 39420 + }, + { + "epoch": 1.74, + "learning_rate": 0.00013348451499588698, + "loss": 0.0933, + "step": 39430 + }, + { + "epoch": 1.74, + "learning_rate": 0.00013326218902154338, + "loss": 0.0917, + "step": 39440 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001330398630471998, + "loss": 0.0849, + "step": 39450 + }, + { + "epoch": 1.74, + "learning_rate": 0.00013281753707285624, + "loss": 0.0928, + "step": 39460 + }, + { + "epoch": 1.74, + "learning_rate": 0.00013259521109851264, + "loss": 0.0891, + "step": 39470 + }, + { + "epoch": 1.74, + "learning_rate": 0.00013237288512416907, + "loss": 0.0923, + "step": 39480 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001321505591498255, + "loss": 0.0894, + "step": 39490 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001319282331754819, + "loss": 0.0921, + "step": 39500 + }, + { + "epoch": 1.74, + "eval_cer": 0.4988781201977933, + "eval_loss": 0.07647676020860672, + "eval_runtime": 84.1161, + "eval_samples_per_second": 118.812, + "eval_steps_per_second": 7.43, + "step": 39500 + }, + { + "epoch": 1.74, + "learning_rate": 0.00013170590720113832, + "loss": 0.0895, + "step": 39510 + }, + { + "epoch": 1.74, + "learning_rate": 0.00013148358122679475, + "loss": 0.0919, + "step": 39520 + }, + { + "epoch": 1.74, + "learning_rate": 0.00013126125525245115, + "loss": 0.0936, + "step": 39530 + }, + { + "epoch": 1.74, + "learning_rate": 0.00013103892927810758, + "loss": 0.0907, + "step": 39540 + }, + { + "epoch": 1.74, + "learning_rate": 0.000130816603303764, + "loss": 0.0912, + "step": 39550 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001305942773294204, + "loss": 0.0905, + "step": 39560 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001303719513550768, + "loss": 0.093, + "step": 39570 + }, + { + "epoch": 1.74, + "learning_rate": 0.00013014962538073323, + "loss": 0.0911, + "step": 39580 + }, + { + "epoch": 1.74, + "learning_rate": 0.00012992729940638963, + "loss": 0.0884, + "step": 39590 + }, + { + "epoch": 1.74, + "learning_rate": 0.00012970497343204606, + "loss": 0.0879, + "step": 39600 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001294826474577025, + "loss": 0.0901, + "step": 39610 + }, + { + "epoch": 1.74, + "learning_rate": 0.0001292603214833589, + "loss": 0.0885, + "step": 39620 + }, + { + "epoch": 1.74, + "learning_rate": 0.00012903799550901532, + "loss": 0.0948, + "step": 39630 + }, + { + "epoch": 1.74, + "learning_rate": 0.00012881566953467174, + "loss": 0.0886, + "step": 39640 + }, + { + "epoch": 1.75, + "learning_rate": 0.00012859334356032814, + "loss": 0.0875, + "step": 39650 + }, + { + "epoch": 1.75, + "learning_rate": 0.00012837101758598457, + "loss": 0.0917, + "step": 39660 + }, + { + "epoch": 1.75, + "learning_rate": 0.000128148691611641, + "loss": 0.0879, + "step": 39670 + }, + { + "epoch": 1.75, + "learning_rate": 0.0001279263656372974, + "loss": 0.0896, + "step": 39680 + }, + { + "epoch": 1.75, + "learning_rate": 0.00012770403966295383, + "loss": 0.0904, + "step": 39690 + }, + { + "epoch": 1.75, + "learning_rate": 0.00012748171368861025, + "loss": 0.0874, + "step": 39700 + }, + { + "epoch": 1.75, + "learning_rate": 0.00012725938771426665, + "loss": 0.0883, + "step": 39710 + }, + { + "epoch": 1.75, + "learning_rate": 0.00012703706173992308, + "loss": 0.0914, + "step": 39720 + }, + { + "epoch": 1.75, + "learning_rate": 0.0001268147357655795, + "loss": 0.0899, + "step": 39730 + }, + { + "epoch": 1.75, + "learning_rate": 0.0001265924097912359, + "loss": 0.0947, + "step": 39740 + }, + { + "epoch": 1.75, + "learning_rate": 0.00012637008381689234, + "loss": 0.0888, + "step": 39750 + }, + { + "epoch": 1.75, + "learning_rate": 0.00012614775784254876, + "loss": 0.0892, + "step": 39760 + }, + { + "epoch": 1.75, + "learning_rate": 0.00012592543186820516, + "loss": 0.087, + "step": 39770 + }, + { + "epoch": 1.75, + "learning_rate": 0.0001257031058938616, + "loss": 0.0862, + "step": 39780 + }, + { + "epoch": 1.75, + "learning_rate": 0.00012548077991951802, + "loss": 0.092, + "step": 39790 + }, + { + "epoch": 1.75, + "learning_rate": 0.00012525845394517442, + "loss": 0.0922, + "step": 39800 + }, + { + "epoch": 1.75, + "learning_rate": 0.00012503612797083085, + "loss": 0.0885, + "step": 39810 + }, + { + "epoch": 1.75, + "learning_rate": 0.00012481380199648725, + "loss": 0.0882, + "step": 39820 + }, + { + "epoch": 1.75, + "learning_rate": 0.00012459147602214367, + "loss": 0.0917, + "step": 39830 + }, + { + "epoch": 1.75, + "learning_rate": 0.00012436915004780007, + "loss": 0.0888, + "step": 39840 + }, + { + "epoch": 1.75, + "learning_rate": 0.0001241468240734565, + "loss": 0.0921, + "step": 39850 + }, + { + "epoch": 1.75, + "learning_rate": 0.00012392449809911293, + "loss": 0.0879, + "step": 39860 + }, + { + "epoch": 1.76, + "learning_rate": 0.00012370217212476933, + "loss": 0.0924, + "step": 39870 + }, + { + "epoch": 1.76, + "learning_rate": 0.00012347984615042576, + "loss": 0.0893, + "step": 39880 + }, + { + "epoch": 1.76, + "learning_rate": 0.00012325752017608218, + "loss": 0.0891, + "step": 39890 + }, + { + "epoch": 1.76, + "learning_rate": 0.00012303519420173858, + "loss": 0.0902, + "step": 39900 + }, + { + "epoch": 1.76, + "learning_rate": 0.000122812868227395, + "loss": 0.0886, + "step": 39910 + }, + { + "epoch": 1.76, + "learning_rate": 0.00012259054225305144, + "loss": 0.0894, + "step": 39920 + }, + { + "epoch": 1.76, + "learning_rate": 0.00012236821627870784, + "loss": 0.0903, + "step": 39930 + }, + { + "epoch": 1.76, + "learning_rate": 0.00012214589030436427, + "loss": 0.0936, + "step": 39940 + }, + { + "epoch": 1.76, + "learning_rate": 0.00012192356433002068, + "loss": 0.0911, + "step": 39950 + }, + { + "epoch": 1.76, + "learning_rate": 0.0001217012383556771, + "loss": 0.0953, + "step": 39960 + }, + { + "epoch": 1.76, + "learning_rate": 0.00012147891238133352, + "loss": 0.088, + "step": 39970 + }, + { + "epoch": 1.76, + "learning_rate": 0.00012125658640698993, + "loss": 0.0859, + "step": 39980 + }, + { + "epoch": 1.76, + "learning_rate": 0.00012103426043264635, + "loss": 0.0923, + "step": 39990 + }, + { + "epoch": 1.76, + "learning_rate": 0.00012081193445830276, + "loss": 0.0879, + "step": 40000 + }, + { + "epoch": 1.76, + "eval_cer": 0.4989462308536081, + "eval_loss": 0.07592933624982834, + "eval_runtime": 83.5394, + "eval_samples_per_second": 119.632, + "eval_steps_per_second": 7.481, + "step": 40000 + }, + { + "epoch": 1.76, + "learning_rate": 0.00012058960848395918, + "loss": 0.0933, + "step": 40010 + }, + { + "epoch": 1.76, + "learning_rate": 0.00012036728250961559, + "loss": 0.0839, + "step": 40020 + }, + { + "epoch": 1.76, + "learning_rate": 0.00012014495653527202, + "loss": 0.091, + "step": 40030 + }, + { + "epoch": 1.76, + "learning_rate": 0.00011992263056092843, + "loss": 0.0913, + "step": 40040 + }, + { + "epoch": 1.76, + "learning_rate": 0.00011970030458658486, + "loss": 0.0839, + "step": 40050 + }, + { + "epoch": 1.76, + "learning_rate": 0.00011947797861224127, + "loss": 0.092, + "step": 40060 + }, + { + "epoch": 1.76, + "learning_rate": 0.00011925565263789769, + "loss": 0.0864, + "step": 40070 + }, + { + "epoch": 1.76, + "learning_rate": 0.00011903332666355411, + "loss": 0.0902, + "step": 40080 + }, + { + "epoch": 1.76, + "learning_rate": 0.00011881100068921053, + "loss": 0.0896, + "step": 40090 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011858867471486694, + "loss": 0.0894, + "step": 40100 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011836634874052337, + "loss": 0.0886, + "step": 40110 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011814402276617978, + "loss": 0.0866, + "step": 40120 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011792169679183618, + "loss": 0.0892, + "step": 40130 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011769937081749261, + "loss": 0.0881, + "step": 40140 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011747704484314902, + "loss": 0.0887, + "step": 40150 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011725471886880544, + "loss": 0.0907, + "step": 40160 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011703239289446187, + "loss": 0.0908, + "step": 40170 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011681006692011828, + "loss": 0.0934, + "step": 40180 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011658774094577469, + "loss": 0.0904, + "step": 40190 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011636541497143112, + "loss": 0.0891, + "step": 40200 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011614308899708753, + "loss": 0.0895, + "step": 40210 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011592076302274395, + "loss": 0.0833, + "step": 40220 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011569843704840038, + "loss": 0.0902, + "step": 40230 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011547611107405679, + "loss": 0.0874, + "step": 40240 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011525378509971319, + "loss": 0.09, + "step": 40250 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011503145912536962, + "loss": 0.0866, + "step": 40260 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011480913315102603, + "loss": 0.0937, + "step": 40270 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011458680717668244, + "loss": 0.0885, + "step": 40280 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011436448120233887, + "loss": 0.089, + "step": 40290 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011414215522799529, + "loss": 0.0872, + "step": 40300 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011391982925365171, + "loss": 0.088, + "step": 40310 + }, + { + "epoch": 1.77, + "learning_rate": 0.00011369750327930813, + "loss": 0.092, + "step": 40320 + }, + { + "epoch": 1.78, + "learning_rate": 0.00011347517730496454, + "loss": 0.0878, + "step": 40330 + }, + { + "epoch": 1.78, + "learning_rate": 0.00011325285133062097, + "loss": 0.0901, + "step": 40340 + }, + { + "epoch": 1.78, + "learning_rate": 0.00011303052535627738, + "loss": 0.0903, + "step": 40350 + }, + { + "epoch": 1.78, + "learning_rate": 0.0001128081993819338, + "loss": 0.0873, + "step": 40360 + }, + { + "epoch": 1.78, + "learning_rate": 0.00011258587340759022, + "loss": 0.0856, + "step": 40370 + }, + { + "epoch": 1.78, + "learning_rate": 0.00011236354743324662, + "loss": 0.0864, + "step": 40380 + }, + { + "epoch": 1.78, + "learning_rate": 0.00011214122145890304, + "loss": 0.0911, + "step": 40390 + }, + { + "epoch": 1.78, + "learning_rate": 0.00011191889548455946, + "loss": 0.0908, + "step": 40400 + }, + { + "epoch": 1.78, + "learning_rate": 0.00011169656951021588, + "loss": 0.0848, + "step": 40410 + }, + { + "epoch": 1.78, + "learning_rate": 0.00011147424353587229, + "loss": 0.0909, + "step": 40420 + }, + { + "epoch": 1.78, + "learning_rate": 0.00011125191756152872, + "loss": 0.0879, + "step": 40430 + }, + { + "epoch": 1.78, + "learning_rate": 0.00011102959158718513, + "loss": 0.0884, + "step": 40440 + }, + { + "epoch": 1.78, + "learning_rate": 0.00011080726561284155, + "loss": 0.0871, + "step": 40450 + }, + { + "epoch": 1.78, + "learning_rate": 0.00011058493963849797, + "loss": 0.0943, + "step": 40460 + }, + { + "epoch": 1.78, + "learning_rate": 0.00011036261366415439, + "loss": 0.0916, + "step": 40470 + }, + { + "epoch": 1.78, + "learning_rate": 0.0001101402876898108, + "loss": 0.0884, + "step": 40480 + }, + { + "epoch": 1.78, + "learning_rate": 0.00010991796171546723, + "loss": 0.0887, + "step": 40490 + }, + { + "epoch": 1.78, + "learning_rate": 0.00010969563574112363, + "loss": 0.0877, + "step": 40500 + }, + { + "epoch": 1.78, + "eval_cer": 0.4990214769114607, + "eval_loss": 0.07558812946081161, + "eval_runtime": 83.6646, + "eval_samples_per_second": 119.453, + "eval_steps_per_second": 7.47, + "step": 40500 + }, + { + "epoch": 1.78, + "learning_rate": 0.00010947330976678004, + "loss": 0.0923, + "step": 40510 + }, + { + "epoch": 1.78, + "learning_rate": 0.00010925098379243647, + "loss": 0.0939, + "step": 40520 + }, + { + "epoch": 1.78, + "learning_rate": 0.00010902865781809289, + "loss": 0.086, + "step": 40530 + }, + { + "epoch": 1.78, + "learning_rate": 0.0001088063318437493, + "loss": 0.0906, + "step": 40540 + }, + { + "epoch": 1.78, + "learning_rate": 0.00010858400586940573, + "loss": 0.0908, + "step": 40550 + }, + { + "epoch": 1.79, + "learning_rate": 0.00010836167989506214, + "loss": 0.0874, + "step": 40560 + }, + { + "epoch": 1.79, + "learning_rate": 0.00010813935392071857, + "loss": 0.0924, + "step": 40570 + }, + { + "epoch": 1.79, + "learning_rate": 0.00010791702794637498, + "loss": 0.0932, + "step": 40580 + }, + { + "epoch": 1.79, + "learning_rate": 0.0001076947019720314, + "loss": 0.0896, + "step": 40590 + }, + { + "epoch": 1.79, + "learning_rate": 0.00010747237599768782, + "loss": 0.0896, + "step": 40600 + }, + { + "epoch": 1.79, + "learning_rate": 0.00010725005002334424, + "loss": 0.0917, + "step": 40610 + }, + { + "epoch": 1.79, + "learning_rate": 0.00010702772404900065, + "loss": 0.0899, + "step": 40620 + }, + { + "epoch": 1.79, + "learning_rate": 0.00010680539807465706, + "loss": 0.0877, + "step": 40630 + }, + { + "epoch": 1.79, + "learning_rate": 0.00010658307210031348, + "loss": 0.0893, + "step": 40640 + }, + { + "epoch": 1.79, + "learning_rate": 0.00010636074612596989, + "loss": 0.0905, + "step": 40650 + }, + { + "epoch": 1.79, + "learning_rate": 0.00010613842015162632, + "loss": 0.0914, + "step": 40660 + }, + { + "epoch": 1.79, + "learning_rate": 0.00010591609417728273, + "loss": 0.0865, + "step": 40670 + }, + { + "epoch": 1.79, + "learning_rate": 0.00010569376820293915, + "loss": 0.0889, + "step": 40680 + }, + { + "epoch": 1.79, + "learning_rate": 0.00010547144222859557, + "loss": 0.0928, + "step": 40690 + }, + { + "epoch": 1.79, + "learning_rate": 0.00010524911625425199, + "loss": 0.0928, + "step": 40700 + }, + { + "epoch": 1.79, + "learning_rate": 0.0001050267902799084, + "loss": 0.0911, + "step": 40710 + }, + { + "epoch": 1.79, + "learning_rate": 0.00010480446430556483, + "loss": 0.0903, + "step": 40720 + }, + { + "epoch": 1.79, + "learning_rate": 0.00010458213833122124, + "loss": 0.0836, + "step": 40730 + }, + { + "epoch": 1.79, + "learning_rate": 0.00010435981235687766, + "loss": 0.0902, + "step": 40740 + }, + { + "epoch": 1.79, + "learning_rate": 0.00010413748638253407, + "loss": 0.0903, + "step": 40750 + }, + { + "epoch": 1.79, + "learning_rate": 0.00010391516040819048, + "loss": 0.0853, + "step": 40760 + }, + { + "epoch": 1.79, + "learning_rate": 0.0001036928344338469, + "loss": 0.0907, + "step": 40770 + }, + { + "epoch": 1.8, + "learning_rate": 0.00010347050845950333, + "loss": 0.0867, + "step": 40780 + }, + { + "epoch": 1.8, + "learning_rate": 0.00010324818248515974, + "loss": 0.0891, + "step": 40790 + }, + { + "epoch": 1.8, + "learning_rate": 0.00010302585651081615, + "loss": 0.0872, + "step": 40800 + }, + { + "epoch": 1.8, + "learning_rate": 0.00010280353053647258, + "loss": 0.0886, + "step": 40810 + }, + { + "epoch": 1.8, + "learning_rate": 0.000102581204562129, + "loss": 0.089, + "step": 40820 + }, + { + "epoch": 1.8, + "learning_rate": 0.00010235887858778541, + "loss": 0.0928, + "step": 40830 + }, + { + "epoch": 1.8, + "learning_rate": 0.00010213655261344184, + "loss": 0.0859, + "step": 40840 + }, + { + "epoch": 1.8, + "learning_rate": 0.00010191422663909825, + "loss": 0.0876, + "step": 40850 + }, + { + "epoch": 1.8, + "learning_rate": 0.00010169190066475468, + "loss": 0.0847, + "step": 40860 + }, + { + "epoch": 1.8, + "learning_rate": 0.00010146957469041109, + "loss": 0.0918, + "step": 40870 + }, + { + "epoch": 1.8, + "learning_rate": 0.00010124724871606749, + "loss": 0.0853, + "step": 40880 + }, + { + "epoch": 1.8, + "learning_rate": 0.00010102492274172392, + "loss": 0.0904, + "step": 40890 + }, + { + "epoch": 1.8, + "learning_rate": 0.00010080259676738033, + "loss": 0.0851, + "step": 40900 + }, + { + "epoch": 1.8, + "learning_rate": 0.00010058027079303675, + "loss": 0.0881, + "step": 40910 + }, + { + "epoch": 1.8, + "learning_rate": 0.00010035794481869317, + "loss": 0.0908, + "step": 40920 + }, + { + "epoch": 1.8, + "learning_rate": 0.00010013561884434959, + "loss": 0.092, + "step": 40930 + }, + { + "epoch": 1.8, + "learning_rate": 9.9913292870006e-05, + "loss": 0.0918, + "step": 40940 + }, + { + "epoch": 1.8, + "learning_rate": 9.969096689566243e-05, + "loss": 0.0898, + "step": 40950 + }, + { + "epoch": 1.8, + "learning_rate": 9.946864092131884e-05, + "loss": 0.0902, + "step": 40960 + }, + { + "epoch": 1.8, + "learning_rate": 9.924631494697526e-05, + "loss": 0.0838, + "step": 40970 + }, + { + "epoch": 1.8, + "learning_rate": 9.902398897263168e-05, + "loss": 0.0881, + "step": 40980 + }, + { + "epoch": 1.8, + "learning_rate": 9.88016629982881e-05, + "loss": 0.0846, + "step": 40990 + }, + { + "epoch": 1.8, + "learning_rate": 9.85793370239445e-05, + "loss": 0.0847, + "step": 41000 + }, + { + "epoch": 1.8, + "eval_cer": 0.49894039279739544, + "eval_loss": 0.07562576234340668, + "eval_runtime": 83.7761, + "eval_samples_per_second": 119.294, + "eval_steps_per_second": 7.46, + "step": 41000 + }, + { + "epoch": 1.81, + "learning_rate": 9.835701104960093e-05, + "loss": 0.0874, + "step": 41010 + }, + { + "epoch": 1.81, + "learning_rate": 9.813468507525734e-05, + "loss": 0.0908, + "step": 41020 + }, + { + "epoch": 1.81, + "learning_rate": 9.791235910091375e-05, + "loss": 0.0902, + "step": 41030 + }, + { + "epoch": 1.81, + "learning_rate": 9.769003312657018e-05, + "loss": 0.0894, + "step": 41040 + }, + { + "epoch": 1.81, + "learning_rate": 9.74677071522266e-05, + "loss": 0.0875, + "step": 41050 + }, + { + "epoch": 1.81, + "learning_rate": 9.724538117788301e-05, + "loss": 0.0883, + "step": 41060 + }, + { + "epoch": 1.81, + "learning_rate": 9.702305520353944e-05, + "loss": 0.0874, + "step": 41070 + }, + { + "epoch": 1.81, + "learning_rate": 9.680072922919585e-05, + "loss": 0.0855, + "step": 41080 + }, + { + "epoch": 1.81, + "learning_rate": 9.657840325485226e-05, + "loss": 0.0936, + "step": 41090 + }, + { + "epoch": 1.81, + "learning_rate": 9.635607728050869e-05, + "loss": 0.0909, + "step": 41100 + }, + { + "epoch": 1.81, + "learning_rate": 9.61337513061651e-05, + "loss": 0.0868, + "step": 41110 + }, + { + "epoch": 1.81, + "learning_rate": 9.591142533182153e-05, + "loss": 0.0861, + "step": 41120 + }, + { + "epoch": 1.81, + "learning_rate": 9.568909935747793e-05, + "loss": 0.0907, + "step": 41130 + }, + { + "epoch": 1.81, + "learning_rate": 9.546677338313435e-05, + "loss": 0.0915, + "step": 41140 + }, + { + "epoch": 1.81, + "learning_rate": 9.524444740879077e-05, + "loss": 0.0933, + "step": 41150 + }, + { + "epoch": 1.81, + "learning_rate": 9.502212143444719e-05, + "loss": 0.0877, + "step": 41160 + }, + { + "epoch": 1.81, + "learning_rate": 9.47997954601036e-05, + "loss": 0.0928, + "step": 41170 + }, + { + "epoch": 1.81, + "learning_rate": 9.457746948576003e-05, + "loss": 0.0857, + "step": 41180 + }, + { + "epoch": 1.81, + "learning_rate": 9.435514351141644e-05, + "loss": 0.0883, + "step": 41190 + }, + { + "epoch": 1.81, + "learning_rate": 9.413281753707286e-05, + "loss": 0.0872, + "step": 41200 + }, + { + "epoch": 1.81, + "learning_rate": 9.391049156272928e-05, + "loss": 0.0876, + "step": 41210 + }, + { + "epoch": 1.81, + "learning_rate": 9.36881655883857e-05, + "loss": 0.0908, + "step": 41220 + }, + { + "epoch": 1.81, + "learning_rate": 9.346583961404211e-05, + "loss": 0.0906, + "step": 41230 + }, + { + "epoch": 1.82, + "learning_rate": 9.324351363969854e-05, + "loss": 0.0882, + "step": 41240 + }, + { + "epoch": 1.82, + "learning_rate": 9.302118766535494e-05, + "loss": 0.0899, + "step": 41250 + }, + { + "epoch": 1.82, + "learning_rate": 9.279886169101135e-05, + "loss": 0.0899, + "step": 41260 + }, + { + "epoch": 1.82, + "learning_rate": 9.257653571666778e-05, + "loss": 0.0899, + "step": 41270 + }, + { + "epoch": 1.82, + "learning_rate": 9.23542097423242e-05, + "loss": 0.0881, + "step": 41280 + }, + { + "epoch": 1.82, + "learning_rate": 9.213188376798061e-05, + "loss": 0.0906, + "step": 41290 + }, + { + "epoch": 1.82, + "learning_rate": 9.190955779363703e-05, + "loss": 0.0926, + "step": 41300 + }, + { + "epoch": 1.82, + "learning_rate": 9.168723181929345e-05, + "loss": 0.0906, + "step": 41310 + }, + { + "epoch": 1.82, + "learning_rate": 9.146490584494986e-05, + "loss": 0.0892, + "step": 41320 + }, + { + "epoch": 1.82, + "learning_rate": 9.124257987060629e-05, + "loss": 0.0927, + "step": 41330 + }, + { + "epoch": 1.82, + "learning_rate": 9.10202538962627e-05, + "loss": 0.091, + "step": 41340 + }, + { + "epoch": 1.82, + "learning_rate": 9.079792792191912e-05, + "loss": 0.0875, + "step": 41350 + }, + { + "epoch": 1.82, + "learning_rate": 9.057560194757554e-05, + "loss": 0.0878, + "step": 41360 + }, + { + "epoch": 1.82, + "learning_rate": 9.035327597323196e-05, + "loss": 0.0877, + "step": 41370 + }, + { + "epoch": 1.82, + "learning_rate": 9.013094999888837e-05, + "loss": 0.0901, + "step": 41380 + }, + { + "epoch": 1.82, + "learning_rate": 8.990862402454479e-05, + "loss": 0.0888, + "step": 41390 + }, + { + "epoch": 1.82, + "learning_rate": 8.96862980502012e-05, + "loss": 0.0901, + "step": 41400 + }, + { + "epoch": 1.82, + "learning_rate": 8.946397207585763e-05, + "loss": 0.0883, + "step": 41410 + }, + { + "epoch": 1.82, + "learning_rate": 8.924164610151404e-05, + "loss": 0.0892, + "step": 41420 + }, + { + "epoch": 1.82, + "learning_rate": 8.901932012717046e-05, + "loss": 0.0881, + "step": 41430 + }, + { + "epoch": 1.82, + "learning_rate": 8.879699415282688e-05, + "loss": 0.0917, + "step": 41440 + }, + { + "epoch": 1.82, + "learning_rate": 8.85746681784833e-05, + "loss": 0.0899, + "step": 41450 + }, + { + "epoch": 1.82, + "learning_rate": 8.835234220413971e-05, + "loss": 0.0887, + "step": 41460 + }, + { + "epoch": 1.83, + "learning_rate": 8.813001622979614e-05, + "loss": 0.0907, + "step": 41470 + }, + { + "epoch": 1.83, + "learning_rate": 8.790769025545255e-05, + "loss": 0.0882, + "step": 41480 + }, + { + "epoch": 1.83, + "learning_rate": 8.768536428110897e-05, + "loss": 0.0881, + "step": 41490 + }, + { + "epoch": 1.83, + "learning_rate": 8.746303830676538e-05, + "loss": 0.0889, + "step": 41500 + }, + { + "epoch": 1.83, + "eval_cer": 0.49899034061165964, + "eval_loss": 0.07531926035881042, + "eval_runtime": 83.7634, + "eval_samples_per_second": 119.312, + "eval_steps_per_second": 7.461, + "step": 41500 + }, + { + "epoch": 1.83, + "learning_rate": 8.724071233242179e-05, + "loss": 0.09, + "step": 41510 + }, + { + "epoch": 1.83, + "learning_rate": 8.70183863580782e-05, + "loss": 0.09, + "step": 41520 + }, + { + "epoch": 1.83, + "learning_rate": 8.679606038373463e-05, + "loss": 0.0859, + "step": 41530 + }, + { + "epoch": 1.83, + "learning_rate": 8.657373440939105e-05, + "loss": 0.0928, + "step": 41540 + }, + { + "epoch": 1.83, + "learning_rate": 8.635140843504746e-05, + "loss": 0.0867, + "step": 41550 + }, + { + "epoch": 1.83, + "learning_rate": 8.612908246070389e-05, + "loss": 0.093, + "step": 41560 + }, + { + "epoch": 1.83, + "learning_rate": 8.59067564863603e-05, + "loss": 0.0848, + "step": 41570 + }, + { + "epoch": 1.83, + "learning_rate": 8.568443051201672e-05, + "loss": 0.0893, + "step": 41580 + }, + { + "epoch": 1.83, + "learning_rate": 8.546210453767314e-05, + "loss": 0.0885, + "step": 41590 + }, + { + "epoch": 1.83, + "learning_rate": 8.523977856332956e-05, + "loss": 0.0867, + "step": 41600 + }, + { + "epoch": 1.83, + "learning_rate": 8.501745258898597e-05, + "loss": 0.0903, + "step": 41610 + }, + { + "epoch": 1.83, + "learning_rate": 8.47951266146424e-05, + "loss": 0.0879, + "step": 41620 + }, + { + "epoch": 1.83, + "learning_rate": 8.45728006402988e-05, + "loss": 0.0838, + "step": 41630 + }, + { + "epoch": 1.83, + "learning_rate": 8.435047466595523e-05, + "loss": 0.0909, + "step": 41640 + }, + { + "epoch": 1.83, + "learning_rate": 8.412814869161164e-05, + "loss": 0.0835, + "step": 41650 + }, + { + "epoch": 1.83, + "learning_rate": 8.390582271726805e-05, + "loss": 0.0846, + "step": 41660 + }, + { + "epoch": 1.83, + "learning_rate": 8.368349674292448e-05, + "loss": 0.0855, + "step": 41670 + }, + { + "epoch": 1.83, + "learning_rate": 8.34611707685809e-05, + "loss": 0.0858, + "step": 41680 + }, + { + "epoch": 1.84, + "learning_rate": 8.323884479423731e-05, + "loss": 0.0903, + "step": 41690 + }, + { + "epoch": 1.84, + "learning_rate": 8.301651881989374e-05, + "loss": 0.0853, + "step": 41700 + }, + { + "epoch": 1.84, + "learning_rate": 8.279419284555015e-05, + "loss": 0.0821, + "step": 41710 + }, + { + "epoch": 1.84, + "learning_rate": 8.257186687120656e-05, + "loss": 0.0882, + "step": 41720 + }, + { + "epoch": 1.84, + "learning_rate": 8.234954089686299e-05, + "loss": 0.0862, + "step": 41730 + }, + { + "epoch": 1.84, + "learning_rate": 8.21272149225194e-05, + "loss": 0.086, + "step": 41740 + }, + { + "epoch": 1.84, + "learning_rate": 8.19048889481758e-05, + "loss": 0.0872, + "step": 41750 + }, + { + "epoch": 1.84, + "learning_rate": 8.168256297383223e-05, + "loss": 0.0925, + "step": 41760 + }, + { + "epoch": 1.84, + "learning_rate": 8.146023699948865e-05, + "loss": 0.0888, + "step": 41770 + }, + { + "epoch": 1.84, + "learning_rate": 8.123791102514506e-05, + "loss": 0.0898, + "step": 41780 + }, + { + "epoch": 1.84, + "learning_rate": 8.101558505080149e-05, + "loss": 0.0917, + "step": 41790 + }, + { + "epoch": 1.84, + "learning_rate": 8.07932590764579e-05, + "loss": 0.0868, + "step": 41800 + }, + { + "epoch": 1.84, + "learning_rate": 8.057093310211432e-05, + "loss": 0.0896, + "step": 41810 + }, + { + "epoch": 1.84, + "learning_rate": 8.034860712777074e-05, + "loss": 0.0923, + "step": 41820 + }, + { + "epoch": 1.84, + "learning_rate": 8.012628115342716e-05, + "loss": 0.0892, + "step": 41830 + }, + { + "epoch": 1.84, + "learning_rate": 7.990395517908357e-05, + "loss": 0.0896, + "step": 41840 + }, + { + "epoch": 1.84, + "learning_rate": 7.968162920474e-05, + "loss": 0.09, + "step": 41850 + }, + { + "epoch": 1.84, + "learning_rate": 7.945930323039641e-05, + "loss": 0.0868, + "step": 41860 + }, + { + "epoch": 1.84, + "learning_rate": 7.923697725605283e-05, + "loss": 0.0926, + "step": 41870 + }, + { + "epoch": 1.84, + "learning_rate": 7.901465128170924e-05, + "loss": 0.0876, + "step": 41880 + }, + { + "epoch": 1.84, + "learning_rate": 7.879232530736565e-05, + "loss": 0.087, + "step": 41890 + }, + { + "epoch": 1.84, + "learning_rate": 7.856999933302208e-05, + "loss": 0.0905, + "step": 41900 + }, + { + "epoch": 1.84, + "learning_rate": 7.83476733586785e-05, + "loss": 0.0854, + "step": 41910 + }, + { + "epoch": 1.85, + "learning_rate": 7.812534738433491e-05, + "loss": 0.0883, + "step": 41920 + }, + { + "epoch": 1.85, + "learning_rate": 7.790302140999134e-05, + "loss": 0.0908, + "step": 41930 + }, + { + "epoch": 1.85, + "learning_rate": 7.768069543564775e-05, + "loss": 0.0905, + "step": 41940 + }, + { + "epoch": 1.85, + "learning_rate": 7.745836946130416e-05, + "loss": 0.0852, + "step": 41950 + }, + { + "epoch": 1.85, + "learning_rate": 7.723604348696059e-05, + "loss": 0.0893, + "step": 41960 + }, + { + "epoch": 1.85, + "learning_rate": 7.7013717512617e-05, + "loss": 0.0904, + "step": 41970 + }, + { + "epoch": 1.85, + "learning_rate": 7.679139153827342e-05, + "loss": 0.0891, + "step": 41980 + }, + { + "epoch": 1.85, + "learning_rate": 7.656906556392985e-05, + "loss": 0.0907, + "step": 41990 + }, + { + "epoch": 1.85, + "learning_rate": 7.634673958958625e-05, + "loss": 0.0875, + "step": 42000 + }, + { + "epoch": 1.85, + "eval_cer": 0.4991862398312413, + "eval_loss": 0.07488802075386047, + "eval_runtime": 84.361, + "eval_samples_per_second": 118.467, + "eval_steps_per_second": 7.409, + "step": 42000 + }, + { + "epoch": 1.85, + "learning_rate": 7.612441361524266e-05, + "loss": 0.0866, + "step": 42010 + }, + { + "epoch": 1.85, + "learning_rate": 7.590208764089909e-05, + "loss": 0.0879, + "step": 42020 + }, + { + "epoch": 1.85, + "learning_rate": 7.56797616665555e-05, + "loss": 0.0877, + "step": 42030 + }, + { + "epoch": 1.85, + "learning_rate": 7.545743569221192e-05, + "loss": 0.089, + "step": 42040 + }, + { + "epoch": 1.85, + "learning_rate": 7.523510971786834e-05, + "loss": 0.0914, + "step": 42050 + }, + { + "epoch": 1.85, + "learning_rate": 7.501278374352476e-05, + "loss": 0.0829, + "step": 42060 + }, + { + "epoch": 1.85, + "learning_rate": 7.479045776918117e-05, + "loss": 0.0869, + "step": 42070 + }, + { + "epoch": 1.85, + "learning_rate": 7.45681317948376e-05, + "loss": 0.088, + "step": 42080 + }, + { + "epoch": 1.85, + "learning_rate": 7.434580582049401e-05, + "loss": 0.087, + "step": 42090 + }, + { + "epoch": 1.85, + "learning_rate": 7.412347984615043e-05, + "loss": 0.0831, + "step": 42100 + }, + { + "epoch": 1.85, + "learning_rate": 7.390115387180685e-05, + "loss": 0.0871, + "step": 42110 + }, + { + "epoch": 1.85, + "learning_rate": 7.367882789746327e-05, + "loss": 0.0908, + "step": 42120 + }, + { + "epoch": 1.85, + "learning_rate": 7.345650192311967e-05, + "loss": 0.0857, + "step": 42130 + }, + { + "epoch": 1.85, + "learning_rate": 7.32341759487761e-05, + "loss": 0.0861, + "step": 42140 + }, + { + "epoch": 1.86, + "learning_rate": 7.301184997443251e-05, + "loss": 0.0895, + "step": 42150 + }, + { + "epoch": 1.86, + "learning_rate": 7.278952400008894e-05, + "loss": 0.086, + "step": 42160 + }, + { + "epoch": 1.86, + "learning_rate": 7.256719802574535e-05, + "loss": 0.0857, + "step": 42170 + }, + { + "epoch": 1.86, + "learning_rate": 7.234487205140176e-05, + "loss": 0.0886, + "step": 42180 + }, + { + "epoch": 1.86, + "learning_rate": 7.212254607705819e-05, + "loss": 0.0885, + "step": 42190 + }, + { + "epoch": 1.86, + "learning_rate": 7.19002201027146e-05, + "loss": 0.091, + "step": 42200 + }, + { + "epoch": 1.86, + "learning_rate": 7.167789412837102e-05, + "loss": 0.0865, + "step": 42210 + }, + { + "epoch": 1.86, + "learning_rate": 7.145556815402745e-05, + "loss": 0.089, + "step": 42220 + }, + { + "epoch": 1.86, + "learning_rate": 7.123324217968386e-05, + "loss": 0.0864, + "step": 42230 + }, + { + "epoch": 1.86, + "learning_rate": 7.101091620534027e-05, + "loss": 0.0907, + "step": 42240 + }, + { + "epoch": 1.86, + "learning_rate": 7.078859023099669e-05, + "loss": 0.0866, + "step": 42250 + }, + { + "epoch": 1.86, + "learning_rate": 7.05662642566531e-05, + "loss": 0.0888, + "step": 42260 + }, + { + "epoch": 1.86, + "learning_rate": 7.034393828230951e-05, + "loss": 0.0877, + "step": 42270 + }, + { + "epoch": 1.86, + "learning_rate": 7.012161230796594e-05, + "loss": 0.0934, + "step": 42280 + }, + { + "epoch": 1.86, + "learning_rate": 6.989928633362236e-05, + "loss": 0.0855, + "step": 42290 + }, + { + "epoch": 1.86, + "learning_rate": 6.967696035927877e-05, + "loss": 0.0861, + "step": 42300 + }, + { + "epoch": 1.86, + "learning_rate": 6.94546343849352e-05, + "loss": 0.0894, + "step": 42310 + }, + { + "epoch": 1.86, + "learning_rate": 6.923230841059161e-05, + "loss": 0.0881, + "step": 42320 + }, + { + "epoch": 1.86, + "learning_rate": 6.900998243624802e-05, + "loss": 0.0886, + "step": 42330 + }, + { + "epoch": 1.86, + "learning_rate": 6.878765646190445e-05, + "loss": 0.0879, + "step": 42340 + }, + { + "epoch": 1.86, + "learning_rate": 6.856533048756087e-05, + "loss": 0.0875, + "step": 42350 + }, + { + "epoch": 1.86, + "learning_rate": 6.834300451321728e-05, + "loss": 0.0874, + "step": 42360 + }, + { + "epoch": 1.87, + "learning_rate": 6.812067853887371e-05, + "loss": 0.0865, + "step": 42370 + }, + { + "epoch": 1.87, + "learning_rate": 6.789835256453011e-05, + "loss": 0.0895, + "step": 42380 + }, + { + "epoch": 1.87, + "learning_rate": 6.767602659018652e-05, + "loss": 0.091, + "step": 42390 + }, + { + "epoch": 1.87, + "learning_rate": 6.745370061584295e-05, + "loss": 0.0892, + "step": 42400 + }, + { + "epoch": 1.87, + "learning_rate": 6.723137464149936e-05, + "loss": 0.0875, + "step": 42410 + }, + { + "epoch": 1.87, + "learning_rate": 6.700904866715579e-05, + "loss": 0.0905, + "step": 42420 + }, + { + "epoch": 1.87, + "learning_rate": 6.67867226928122e-05, + "loss": 0.0891, + "step": 42430 + }, + { + "epoch": 1.87, + "learning_rate": 6.656439671846862e-05, + "loss": 0.0882, + "step": 42440 + }, + { + "epoch": 1.87, + "learning_rate": 6.634207074412505e-05, + "loss": 0.0846, + "step": 42450 + }, + { + "epoch": 1.87, + "learning_rate": 6.611974476978146e-05, + "loss": 0.0864, + "step": 42460 + }, + { + "epoch": 1.87, + "learning_rate": 6.589741879543787e-05, + "loss": 0.0916, + "step": 42470 + }, + { + "epoch": 1.87, + "learning_rate": 6.56750928210943e-05, + "loss": 0.0909, + "step": 42480 + }, + { + "epoch": 1.87, + "learning_rate": 6.545276684675071e-05, + "loss": 0.0869, + "step": 42490 + }, + { + "epoch": 1.87, + "learning_rate": 6.523044087240711e-05, + "loss": 0.0875, + "step": 42500 + }, + { + "epoch": 1.87, + "eval_cer": 0.4990214769114607, + "eval_loss": 0.0748269185423851, + "eval_runtime": 84.306, + "eval_samples_per_second": 118.544, + "eval_steps_per_second": 7.413, + "step": 42500 + }, + { + "epoch": 1.87, + "learning_rate": 6.500811489806354e-05, + "loss": 0.0859, + "step": 42510 + }, + { + "epoch": 1.87, + "learning_rate": 6.478578892371996e-05, + "loss": 0.0854, + "step": 42520 + }, + { + "epoch": 1.87, + "learning_rate": 6.456346294937637e-05, + "loss": 0.0914, + "step": 42530 + }, + { + "epoch": 1.87, + "learning_rate": 6.43411369750328e-05, + "loss": 0.0884, + "step": 42540 + }, + { + "epoch": 1.87, + "learning_rate": 6.411881100068921e-05, + "loss": 0.0857, + "step": 42550 + }, + { + "epoch": 1.87, + "learning_rate": 6.389648502634562e-05, + "loss": 0.0894, + "step": 42560 + }, + { + "epoch": 1.87, + "learning_rate": 6.367415905200205e-05, + "loss": 0.0887, + "step": 42570 + }, + { + "epoch": 1.87, + "learning_rate": 6.345183307765847e-05, + "loss": 0.0865, + "step": 42580 + }, + { + "epoch": 1.87, + "learning_rate": 6.322950710331488e-05, + "loss": 0.0871, + "step": 42590 + }, + { + "epoch": 1.88, + "learning_rate": 6.300718112897131e-05, + "loss": 0.0871, + "step": 42600 + }, + { + "epoch": 1.88, + "learning_rate": 6.278485515462772e-05, + "loss": 0.0868, + "step": 42610 + }, + { + "epoch": 1.88, + "learning_rate": 6.256252918028413e-05, + "loss": 0.0908, + "step": 42620 + }, + { + "epoch": 1.88, + "learning_rate": 6.234020320594055e-05, + "loss": 0.0853, + "step": 42630 + }, + { + "epoch": 1.88, + "learning_rate": 6.211787723159696e-05, + "loss": 0.0902, + "step": 42640 + }, + { + "epoch": 1.88, + "learning_rate": 6.189555125725339e-05, + "loss": 0.0887, + "step": 42650 + }, + { + "epoch": 1.88, + "learning_rate": 6.16732252829098e-05, + "loss": 0.0888, + "step": 42660 + }, + { + "epoch": 1.88, + "learning_rate": 6.145089930856622e-05, + "loss": 0.0878, + "step": 42670 + }, + { + "epoch": 1.88, + "learning_rate": 6.122857333422264e-05, + "loss": 0.0883, + "step": 42680 + }, + { + "epoch": 1.88, + "learning_rate": 6.100624735987906e-05, + "loss": 0.088, + "step": 42690 + }, + { + "epoch": 1.88, + "learning_rate": 6.078392138553547e-05, + "loss": 0.0914, + "step": 42700 + }, + { + "epoch": 1.88, + "learning_rate": 6.056159541119189e-05, + "loss": 0.0895, + "step": 42710 + }, + { + "epoch": 1.88, + "learning_rate": 6.033926943684831e-05, + "loss": 0.0909, + "step": 42720 + }, + { + "epoch": 1.88, + "learning_rate": 6.011694346250472e-05, + "loss": 0.0927, + "step": 42730 + }, + { + "epoch": 1.88, + "learning_rate": 5.989461748816114e-05, + "loss": 0.0876, + "step": 42740 + }, + { + "epoch": 1.88, + "learning_rate": 5.967229151381756e-05, + "loss": 0.0888, + "step": 42750 + }, + { + "epoch": 1.88, + "learning_rate": 5.944996553947398e-05, + "loss": 0.0905, + "step": 42760 + }, + { + "epoch": 1.88, + "learning_rate": 5.9227639565130396e-05, + "loss": 0.0902, + "step": 42770 + }, + { + "epoch": 1.88, + "learning_rate": 5.900531359078681e-05, + "loss": 0.088, + "step": 42780 + }, + { + "epoch": 1.88, + "learning_rate": 5.878298761644323e-05, + "loss": 0.0854, + "step": 42790 + }, + { + "epoch": 1.88, + "learning_rate": 5.8560661642099644e-05, + "loss": 0.088, + "step": 42800 + }, + { + "epoch": 1.88, + "learning_rate": 5.8338335667756065e-05, + "loss": 0.0905, + "step": 42810 + }, + { + "epoch": 1.88, + "learning_rate": 5.8116009693412486e-05, + "loss": 0.0893, + "step": 42820 + }, + { + "epoch": 1.89, + "learning_rate": 5.78936837190689e-05, + "loss": 0.0886, + "step": 42830 + }, + { + "epoch": 1.89, + "learning_rate": 5.767135774472532e-05, + "loss": 0.0848, + "step": 42840 + }, + { + "epoch": 1.89, + "learning_rate": 5.7449031770381734e-05, + "loss": 0.0875, + "step": 42850 + }, + { + "epoch": 1.89, + "learning_rate": 5.722670579603815e-05, + "loss": 0.0907, + "step": 42860 + }, + { + "epoch": 1.89, + "learning_rate": 5.700437982169457e-05, + "loss": 0.0824, + "step": 42870 + }, + { + "epoch": 1.89, + "learning_rate": 5.678205384735099e-05, + "loss": 0.0844, + "step": 42880 + }, + { + "epoch": 1.89, + "learning_rate": 5.655972787300741e-05, + "loss": 0.0873, + "step": 42890 + }, + { + "epoch": 1.89, + "learning_rate": 5.633740189866382e-05, + "loss": 0.0883, + "step": 42900 + }, + { + "epoch": 1.89, + "learning_rate": 5.611507592432024e-05, + "loss": 0.0847, + "step": 42910 + }, + { + "epoch": 1.89, + "learning_rate": 5.589274994997666e-05, + "loss": 0.089, + "step": 42920 + }, + { + "epoch": 1.89, + "learning_rate": 5.567042397563307e-05, + "loss": 0.089, + "step": 42930 + }, + { + "epoch": 1.89, + "learning_rate": 5.544809800128949e-05, + "loss": 0.0889, + "step": 42940 + }, + { + "epoch": 1.89, + "learning_rate": 5.522577202694591e-05, + "loss": 0.0884, + "step": 42950 + }, + { + "epoch": 1.89, + "learning_rate": 5.500344605260233e-05, + "loss": 0.0871, + "step": 42960 + }, + { + "epoch": 1.89, + "learning_rate": 5.478112007825874e-05, + "loss": 0.0899, + "step": 42970 + }, + { + "epoch": 1.89, + "learning_rate": 5.455879410391516e-05, + "loss": 0.0877, + "step": 42980 + }, + { + "epoch": 1.89, + "learning_rate": 5.4336468129571575e-05, + "loss": 0.0881, + "step": 42990 + }, + { + "epoch": 1.89, + "learning_rate": 5.4114142155227996e-05, + "loss": 0.0953, + "step": 43000 + }, + { + "epoch": 1.89, + "eval_cer": 0.4989300140307951, + "eval_loss": 0.07458040863275528, + "eval_runtime": 84.0307, + "eval_samples_per_second": 118.933, + "eval_steps_per_second": 7.438, + "step": 43000 + }, + { + "epoch": 1.89, + "learning_rate": 5.3891816180884416e-05, + "loss": 0.0895, + "step": 43010 + }, + { + "epoch": 1.89, + "learning_rate": 5.366949020654084e-05, + "loss": 0.0891, + "step": 43020 + }, + { + "epoch": 1.89, + "learning_rate": 5.3447164232197244e-05, + "loss": 0.0855, + "step": 43030 + }, + { + "epoch": 1.89, + "learning_rate": 5.3224838257853664e-05, + "loss": 0.0913, + "step": 43040 + }, + { + "epoch": 1.89, + "learning_rate": 5.3002512283510085e-05, + "loss": 0.0902, + "step": 43050 + }, + { + "epoch": 1.9, + "learning_rate": 5.27801863091665e-05, + "loss": 0.0893, + "step": 43060 + }, + { + "epoch": 1.9, + "learning_rate": 5.255786033482292e-05, + "loss": 0.0896, + "step": 43070 + }, + { + "epoch": 1.9, + "learning_rate": 5.233553436047934e-05, + "loss": 0.0869, + "step": 43080 + }, + { + "epoch": 1.9, + "learning_rate": 5.2113208386135754e-05, + "loss": 0.0877, + "step": 43090 + }, + { + "epoch": 1.9, + "learning_rate": 5.189088241179217e-05, + "loss": 0.087, + "step": 43100 + }, + { + "epoch": 1.9, + "learning_rate": 5.166855643744859e-05, + "loss": 0.0901, + "step": 43110 + }, + { + "epoch": 1.9, + "learning_rate": 5.1446230463105e-05, + "loss": 0.0872, + "step": 43120 + }, + { + "epoch": 1.9, + "learning_rate": 5.122390448876142e-05, + "loss": 0.0863, + "step": 43130 + }, + { + "epoch": 1.9, + "learning_rate": 5.100157851441784e-05, + "loss": 0.0862, + "step": 43140 + }, + { + "epoch": 1.9, + "learning_rate": 5.0779252540074264e-05, + "loss": 0.0895, + "step": 43150 + }, + { + "epoch": 1.9, + "learning_rate": 5.055692656573067e-05, + "loss": 0.0869, + "step": 43160 + }, + { + "epoch": 1.9, + "learning_rate": 5.033460059138709e-05, + "loss": 0.0887, + "step": 43170 + }, + { + "epoch": 1.9, + "learning_rate": 5.011227461704351e-05, + "loss": 0.0875, + "step": 43180 + }, + { + "epoch": 1.9, + "learning_rate": 4.9889948642699926e-05, + "loss": 0.0867, + "step": 43190 + }, + { + "epoch": 1.9, + "learning_rate": 4.966762266835635e-05, + "loss": 0.088, + "step": 43200 + }, + { + "epoch": 1.9, + "learning_rate": 4.944529669401277e-05, + "loss": 0.0832, + "step": 43210 + }, + { + "epoch": 1.9, + "learning_rate": 4.9222970719669174e-05, + "loss": 0.089, + "step": 43220 + }, + { + "epoch": 1.9, + "learning_rate": 4.9000644745325595e-05, + "loss": 0.0891, + "step": 43230 + }, + { + "epoch": 1.9, + "learning_rate": 4.8778318770982016e-05, + "loss": 0.0885, + "step": 43240 + }, + { + "epoch": 1.9, + "learning_rate": 4.855599279663843e-05, + "loss": 0.0879, + "step": 43250 + }, + { + "epoch": 1.9, + "learning_rate": 4.833366682229485e-05, + "loss": 0.0881, + "step": 43260 + }, + { + "epoch": 1.9, + "learning_rate": 4.811134084795127e-05, + "loss": 0.089, + "step": 43270 + }, + { + "epoch": 1.91, + "learning_rate": 4.7889014873607684e-05, + "loss": 0.0868, + "step": 43280 + }, + { + "epoch": 1.91, + "learning_rate": 4.76666888992641e-05, + "loss": 0.0891, + "step": 43290 + }, + { + "epoch": 1.91, + "learning_rate": 4.744436292492052e-05, + "loss": 0.0871, + "step": 43300 + }, + { + "epoch": 1.91, + "learning_rate": 4.722203695057694e-05, + "loss": 0.0873, + "step": 43310 + }, + { + "epoch": 1.91, + "learning_rate": 4.699971097623335e-05, + "loss": 0.0911, + "step": 43320 + }, + { + "epoch": 1.91, + "learning_rate": 4.6777385001889774e-05, + "loss": 0.0873, + "step": 43330 + }, + { + "epoch": 1.91, + "learning_rate": 4.6555059027546195e-05, + "loss": 0.087, + "step": 43340 + }, + { + "epoch": 1.91, + "learning_rate": 4.63327330532026e-05, + "loss": 0.0888, + "step": 43350 + }, + { + "epoch": 1.91, + "learning_rate": 4.611040707885902e-05, + "loss": 0.085, + "step": 43360 + }, + { + "epoch": 1.91, + "learning_rate": 4.588808110451544e-05, + "loss": 0.0847, + "step": 43370 + }, + { + "epoch": 1.91, + "learning_rate": 4.566575513017186e-05, + "loss": 0.0874, + "step": 43380 + }, + { + "epoch": 1.91, + "learning_rate": 4.544342915582828e-05, + "loss": 0.0865, + "step": 43390 + }, + { + "epoch": 1.91, + "learning_rate": 4.52211031814847e-05, + "loss": 0.0888, + "step": 43400 + }, + { + "epoch": 1.91, + "learning_rate": 4.499877720714111e-05, + "loss": 0.085, + "step": 43410 + }, + { + "epoch": 1.91, + "learning_rate": 4.4776451232797525e-05, + "loss": 0.0923, + "step": 43420 + }, + { + "epoch": 1.91, + "learning_rate": 4.4554125258453946e-05, + "loss": 0.089, + "step": 43430 + }, + { + "epoch": 1.91, + "learning_rate": 4.433179928411037e-05, + "loss": 0.0887, + "step": 43440 + }, + { + "epoch": 1.91, + "learning_rate": 4.410947330976678e-05, + "loss": 0.0857, + "step": 43450 + }, + { + "epoch": 1.91, + "learning_rate": 4.38871473354232e-05, + "loss": 0.0892, + "step": 43460 + }, + { + "epoch": 1.91, + "learning_rate": 4.3664821361079615e-05, + "loss": 0.0878, + "step": 43470 + }, + { + "epoch": 1.91, + "learning_rate": 4.344249538673603e-05, + "loss": 0.089, + "step": 43480 + }, + { + "epoch": 1.91, + "learning_rate": 4.322016941239245e-05, + "loss": 0.086, + "step": 43490 + }, + { + "epoch": 1.91, + "learning_rate": 4.299784343804887e-05, + "loss": 0.0861, + "step": 43500 + }, + { + "epoch": 1.91, + "eval_cer": 0.4990655866695122, + "eval_loss": 0.07449883967638016, + "eval_runtime": 84.1406, + "eval_samples_per_second": 118.777, + "eval_steps_per_second": 7.428, + "step": 43500 + }, + { + "epoch": 1.92, + "learning_rate": 4.2775517463705284e-05, + "loss": 0.0855, + "step": 43510 + }, + { + "epoch": 1.92, + "learning_rate": 4.2553191489361704e-05, + "loss": 0.086, + "step": 43520 + }, + { + "epoch": 1.92, + "learning_rate": 4.233086551501812e-05, + "loss": 0.0865, + "step": 43530 + }, + { + "epoch": 1.92, + "learning_rate": 4.210853954067454e-05, + "loss": 0.0884, + "step": 43540 + }, + { + "epoch": 1.92, + "learning_rate": 4.188621356633095e-05, + "loss": 0.0905, + "step": 43550 + }, + { + "epoch": 1.92, + "learning_rate": 4.166388759198737e-05, + "loss": 0.0929, + "step": 43560 + }, + { + "epoch": 1.92, + "learning_rate": 4.1441561617643794e-05, + "loss": 0.0904, + "step": 43570 + }, + { + "epoch": 1.92, + "learning_rate": 4.121923564330021e-05, + "loss": 0.0892, + "step": 43580 + }, + { + "epoch": 1.92, + "learning_rate": 4.099690966895663e-05, + "loss": 0.0858, + "step": 43590 + }, + { + "epoch": 1.92, + "learning_rate": 4.077458369461304e-05, + "loss": 0.0891, + "step": 43600 + }, + { + "epoch": 1.92, + "learning_rate": 4.0552257720269456e-05, + "loss": 0.0839, + "step": 43610 + }, + { + "epoch": 1.92, + "learning_rate": 4.032993174592588e-05, + "loss": 0.0892, + "step": 43620 + }, + { + "epoch": 1.92, + "learning_rate": 4.01076057715823e-05, + "loss": 0.0844, + "step": 43630 + }, + { + "epoch": 1.92, + "learning_rate": 3.988527979723871e-05, + "loss": 0.0847, + "step": 43640 + }, + { + "epoch": 1.92, + "learning_rate": 3.966295382289513e-05, + "loss": 0.0895, + "step": 43650 + }, + { + "epoch": 1.92, + "learning_rate": 3.9440627848551546e-05, + "loss": 0.093, + "step": 43660 + }, + { + "epoch": 1.92, + "learning_rate": 3.9218301874207966e-05, + "loss": 0.0934, + "step": 43670 + }, + { + "epoch": 1.92, + "learning_rate": 3.899597589986438e-05, + "loss": 0.0865, + "step": 43680 + }, + { + "epoch": 1.92, + "learning_rate": 3.87736499255208e-05, + "loss": 0.0891, + "step": 43690 + }, + { + "epoch": 1.92, + "learning_rate": 3.855132395117722e-05, + "loss": 0.086, + "step": 43700 + }, + { + "epoch": 1.92, + "learning_rate": 3.8328997976833635e-05, + "loss": 0.0867, + "step": 43710 + }, + { + "epoch": 1.92, + "learning_rate": 3.810667200249005e-05, + "loss": 0.0869, + "step": 43720 + }, + { + "epoch": 1.92, + "learning_rate": 3.788434602814647e-05, + "loss": 0.089, + "step": 43730 + }, + { + "epoch": 1.93, + "learning_rate": 3.766202005380288e-05, + "loss": 0.0913, + "step": 43740 + }, + { + "epoch": 1.93, + "learning_rate": 3.7439694079459304e-05, + "loss": 0.085, + "step": 43750 + }, + { + "epoch": 1.93, + "learning_rate": 3.7217368105115724e-05, + "loss": 0.0822, + "step": 43760 + }, + { + "epoch": 1.93, + "learning_rate": 3.699504213077214e-05, + "loss": 0.0888, + "step": 43770 + }, + { + "epoch": 1.93, + "learning_rate": 3.677271615642855e-05, + "loss": 0.0907, + "step": 43780 + }, + { + "epoch": 1.93, + "learning_rate": 3.655039018208497e-05, + "loss": 0.0857, + "step": 43790 + }, + { + "epoch": 1.93, + "learning_rate": 3.632806420774139e-05, + "loss": 0.0877, + "step": 43800 + }, + { + "epoch": 1.93, + "learning_rate": 3.610573823339781e-05, + "loss": 0.0866, + "step": 43810 + }, + { + "epoch": 1.93, + "learning_rate": 3.588341225905423e-05, + "loss": 0.0899, + "step": 43820 + }, + { + "epoch": 1.93, + "learning_rate": 3.566108628471065e-05, + "loss": 0.0881, + "step": 43830 + }, + { + "epoch": 1.93, + "learning_rate": 3.543876031036706e-05, + "loss": 0.0898, + "step": 43840 + }, + { + "epoch": 1.93, + "learning_rate": 3.5216434336023476e-05, + "loss": 0.0905, + "step": 43850 + }, + { + "epoch": 1.93, + "learning_rate": 3.49941083616799e-05, + "loss": 0.0883, + "step": 43860 + }, + { + "epoch": 1.93, + "learning_rate": 3.477178238733631e-05, + "loss": 0.0884, + "step": 43870 + }, + { + "epoch": 1.93, + "learning_rate": 3.454945641299273e-05, + "loss": 0.0886, + "step": 43880 + }, + { + "epoch": 1.93, + "learning_rate": 3.432713043864915e-05, + "loss": 0.087, + "step": 43890 + }, + { + "epoch": 1.93, + "learning_rate": 3.4104804464305566e-05, + "loss": 0.0846, + "step": 43900 + }, + { + "epoch": 1.93, + "learning_rate": 3.388247848996198e-05, + "loss": 0.0917, + "step": 43910 + }, + { + "epoch": 1.93, + "learning_rate": 3.36601525156184e-05, + "loss": 0.0869, + "step": 43920 + }, + { + "epoch": 1.93, + "learning_rate": 3.343782654127482e-05, + "loss": 0.089, + "step": 43930 + }, + { + "epoch": 1.93, + "learning_rate": 3.3215500566931234e-05, + "loss": 0.0871, + "step": 43940 + }, + { + "epoch": 1.93, + "learning_rate": 3.2993174592587655e-05, + "loss": 0.0907, + "step": 43950 + }, + { + "epoch": 1.94, + "learning_rate": 3.2770848618244076e-05, + "loss": 0.086, + "step": 43960 + }, + { + "epoch": 1.94, + "learning_rate": 3.254852264390048e-05, + "loss": 0.0891, + "step": 43970 + }, + { + "epoch": 1.94, + "learning_rate": 3.23261966695569e-05, + "loss": 0.0847, + "step": 43980 + }, + { + "epoch": 1.94, + "learning_rate": 3.2103870695213324e-05, + "loss": 0.0867, + "step": 43990 + }, + { + "epoch": 1.94, + "learning_rate": 3.188154472086974e-05, + "loss": 0.0917, + "step": 44000 + }, + { + "epoch": 1.94, + "eval_cer": 0.49895531227438344, + "eval_loss": 0.07457119971513748, + "eval_runtime": 84.0182, + "eval_samples_per_second": 118.95, + "eval_steps_per_second": 7.439, + "step": 44000 + }, + { + "epoch": 1.94, + "learning_rate": 3.165921874652616e-05, + "loss": 0.0932, + "step": 44010 + }, + { + "epoch": 1.94, + "learning_rate": 3.143689277218258e-05, + "loss": 0.0906, + "step": 44020 + }, + { + "epoch": 1.94, + "learning_rate": 3.121456679783899e-05, + "loss": 0.0872, + "step": 44030 + }, + { + "epoch": 1.94, + "learning_rate": 3.0992240823495407e-05, + "loss": 0.0912, + "step": 44040 + }, + { + "epoch": 1.94, + "learning_rate": 3.076991484915183e-05, + "loss": 0.0832, + "step": 44050 + }, + { + "epoch": 1.94, + "learning_rate": 3.054758887480824e-05, + "loss": 0.0855, + "step": 44060 + }, + { + "epoch": 1.94, + "learning_rate": 3.032526290046466e-05, + "loss": 0.089, + "step": 44070 + }, + { + "epoch": 1.94, + "learning_rate": 3.010293692612108e-05, + "loss": 0.0875, + "step": 44080 + }, + { + "epoch": 1.94, + "learning_rate": 2.9880610951777496e-05, + "loss": 0.0871, + "step": 44090 + }, + { + "epoch": 1.94, + "learning_rate": 2.9658284977433917e-05, + "loss": 0.0869, + "step": 44100 + }, + { + "epoch": 1.94, + "learning_rate": 2.943595900309033e-05, + "loss": 0.0899, + "step": 44110 + }, + { + "epoch": 1.94, + "learning_rate": 2.9213633028746748e-05, + "loss": 0.0839, + "step": 44120 + }, + { + "epoch": 1.94, + "learning_rate": 2.899130705440317e-05, + "loss": 0.0874, + "step": 44130 + }, + { + "epoch": 1.94, + "learning_rate": 2.8768981080059582e-05, + "loss": 0.0907, + "step": 44140 + }, + { + "epoch": 1.94, + "learning_rate": 2.8546655105716e-05, + "loss": 0.0894, + "step": 44150 + }, + { + "epoch": 1.94, + "learning_rate": 2.832432913137242e-05, + "loss": 0.0845, + "step": 44160 + }, + { + "epoch": 1.94, + "learning_rate": 2.8102003157028834e-05, + "loss": 0.0857, + "step": 44170 + }, + { + "epoch": 1.94, + "learning_rate": 2.7879677182685254e-05, + "loss": 0.0898, + "step": 44180 + }, + { + "epoch": 1.95, + "learning_rate": 2.765735120834167e-05, + "loss": 0.0894, + "step": 44190 + }, + { + "epoch": 1.95, + "learning_rate": 2.7435025233998085e-05, + "loss": 0.0875, + "step": 44200 + }, + { + "epoch": 1.95, + "learning_rate": 2.7212699259654506e-05, + "loss": 0.0866, + "step": 44210 + }, + { + "epoch": 1.95, + "learning_rate": 2.6990373285310923e-05, + "loss": 0.0878, + "step": 44220 + }, + { + "epoch": 1.95, + "learning_rate": 2.6768047310967344e-05, + "loss": 0.0873, + "step": 44230 + }, + { + "epoch": 1.95, + "learning_rate": 2.6545721336623758e-05, + "loss": 0.0869, + "step": 44240 + }, + { + "epoch": 1.95, + "learning_rate": 2.6323395362280175e-05, + "loss": 0.093, + "step": 44250 + }, + { + "epoch": 1.95, + "learning_rate": 2.6101069387936596e-05, + "loss": 0.0911, + "step": 44260 + }, + { + "epoch": 1.95, + "learning_rate": 2.587874341359301e-05, + "loss": 0.0924, + "step": 44270 + }, + { + "epoch": 1.95, + "learning_rate": 2.5656417439249427e-05, + "loss": 0.0896, + "step": 44280 + }, + { + "epoch": 1.95, + "learning_rate": 2.5434091464905847e-05, + "loss": 0.0845, + "step": 44290 + }, + { + "epoch": 1.95, + "learning_rate": 2.521176549056226e-05, + "loss": 0.0863, + "step": 44300 + }, + { + "epoch": 1.95, + "learning_rate": 2.498943951621868e-05, + "loss": 0.0889, + "step": 44310 + }, + { + "epoch": 1.95, + "learning_rate": 2.47671135418751e-05, + "loss": 0.0898, + "step": 44320 + }, + { + "epoch": 1.95, + "learning_rate": 2.4544787567531513e-05, + "loss": 0.0892, + "step": 44330 + }, + { + "epoch": 1.95, + "learning_rate": 2.4322461593187933e-05, + "loss": 0.0907, + "step": 44340 + }, + { + "epoch": 1.95, + "learning_rate": 2.410013561884435e-05, + "loss": 0.0858, + "step": 44350 + }, + { + "epoch": 1.95, + "learning_rate": 2.3877809644500768e-05, + "loss": 0.0906, + "step": 44360 + }, + { + "epoch": 1.95, + "learning_rate": 2.3655483670157185e-05, + "loss": 0.0869, + "step": 44370 + }, + { + "epoch": 1.95, + "learning_rate": 2.3433157695813602e-05, + "loss": 0.0857, + "step": 44380 + }, + { + "epoch": 1.95, + "learning_rate": 2.321083172147002e-05, + "loss": 0.0852, + "step": 44390 + }, + { + "epoch": 1.95, + "learning_rate": 2.2988505747126437e-05, + "loss": 0.0878, + "step": 44400 + }, + { + "epoch": 1.95, + "learning_rate": 2.2766179772782854e-05, + "loss": 0.0867, + "step": 44410 + }, + { + "epoch": 1.96, + "learning_rate": 2.254385379843927e-05, + "loss": 0.0898, + "step": 44420 + }, + { + "epoch": 1.96, + "learning_rate": 2.2321527824095688e-05, + "loss": 0.091, + "step": 44430 + }, + { + "epoch": 1.96, + "learning_rate": 2.209920184975211e-05, + "loss": 0.089, + "step": 44440 + }, + { + "epoch": 1.96, + "learning_rate": 2.1876875875408526e-05, + "loss": 0.0878, + "step": 44450 + }, + { + "epoch": 1.96, + "learning_rate": 2.165454990106494e-05, + "loss": 0.088, + "step": 44460 + }, + { + "epoch": 1.96, + "learning_rate": 2.143222392672136e-05, + "loss": 0.0867, + "step": 44470 + }, + { + "epoch": 1.96, + "learning_rate": 2.1209897952377778e-05, + "loss": 0.0885, + "step": 44480 + }, + { + "epoch": 1.96, + "learning_rate": 2.0987571978034195e-05, + "loss": 0.0863, + "step": 44490 + }, + { + "epoch": 1.96, + "learning_rate": 2.0765246003690612e-05, + "loss": 0.0875, + "step": 44500 + }, + { + "epoch": 1.96, + "eval_cer": 0.498987097247097, + "eval_loss": 0.07432349026203156, + "eval_runtime": 84.2199, + "eval_samples_per_second": 118.666, + "eval_steps_per_second": 7.421, + "step": 44500 + }, + { + "epoch": 1.96, + "learning_rate": 2.054292002934703e-05, + "loss": 0.0871, + "step": 44510 + }, + { + "epoch": 1.96, + "learning_rate": 2.0320594055003447e-05, + "loss": 0.0861, + "step": 44520 + }, + { + "epoch": 1.96, + "learning_rate": 2.0098268080659864e-05, + "loss": 0.0774, + "step": 44530 + }, + { + "epoch": 1.96, + "learning_rate": 1.987594210631628e-05, + "loss": 0.0894, + "step": 44540 + }, + { + "epoch": 1.96, + "learning_rate": 1.9653616131972698e-05, + "loss": 0.0859, + "step": 44550 + }, + { + "epoch": 1.96, + "learning_rate": 1.9431290157629115e-05, + "loss": 0.0897, + "step": 44560 + }, + { + "epoch": 1.96, + "learning_rate": 1.9208964183285536e-05, + "loss": 0.0873, + "step": 44570 + }, + { + "epoch": 1.96, + "learning_rate": 1.898663820894195e-05, + "loss": 0.0916, + "step": 44580 + }, + { + "epoch": 1.96, + "learning_rate": 1.8764312234598367e-05, + "loss": 0.085, + "step": 44590 + }, + { + "epoch": 1.96, + "learning_rate": 1.8541986260254788e-05, + "loss": 0.0882, + "step": 44600 + }, + { + "epoch": 1.96, + "learning_rate": 1.83196602859112e-05, + "loss": 0.0881, + "step": 44610 + }, + { + "epoch": 1.96, + "learning_rate": 1.8097334311567622e-05, + "loss": 0.0863, + "step": 44620 + }, + { + "epoch": 1.96, + "learning_rate": 1.787500833722404e-05, + "loss": 0.0885, + "step": 44630 + }, + { + "epoch": 1.96, + "learning_rate": 1.7652682362880453e-05, + "loss": 0.0873, + "step": 44640 + }, + { + "epoch": 1.97, + "learning_rate": 1.7430356388536874e-05, + "loss": 0.0863, + "step": 44650 + }, + { + "epoch": 1.97, + "learning_rate": 1.720803041419329e-05, + "loss": 0.0923, + "step": 44660 + }, + { + "epoch": 1.97, + "learning_rate": 1.6985704439849705e-05, + "loss": 0.0866, + "step": 44670 + }, + { + "epoch": 1.97, + "learning_rate": 1.6763378465506125e-05, + "loss": 0.0852, + "step": 44680 + }, + { + "epoch": 1.97, + "learning_rate": 1.6541052491162543e-05, + "loss": 0.0897, + "step": 44690 + }, + { + "epoch": 1.97, + "learning_rate": 1.6318726516818963e-05, + "loss": 0.084, + "step": 44700 + }, + { + "epoch": 1.97, + "learning_rate": 1.6096400542475377e-05, + "loss": 0.0857, + "step": 44710 + }, + { + "epoch": 1.97, + "learning_rate": 1.5874074568131794e-05, + "loss": 0.0852, + "step": 44720 + }, + { + "epoch": 1.97, + "learning_rate": 1.5651748593788215e-05, + "loss": 0.0848, + "step": 44730 + }, + { + "epoch": 1.97, + "learning_rate": 1.542942261944463e-05, + "loss": 0.0906, + "step": 44740 + }, + { + "epoch": 1.97, + "learning_rate": 1.5207096645101048e-05, + "loss": 0.0876, + "step": 44750 + }, + { + "epoch": 1.97, + "learning_rate": 1.4984770670757465e-05, + "loss": 0.0868, + "step": 44760 + }, + { + "epoch": 1.97, + "learning_rate": 1.4762444696413884e-05, + "loss": 0.0906, + "step": 44770 + }, + { + "epoch": 1.97, + "learning_rate": 1.45401187220703e-05, + "loss": 0.0892, + "step": 44780 + }, + { + "epoch": 1.97, + "learning_rate": 1.4317792747726717e-05, + "loss": 0.0902, + "step": 44790 + }, + { + "epoch": 1.97, + "learning_rate": 1.4095466773383135e-05, + "loss": 0.0847, + "step": 44800 + }, + { + "epoch": 1.97, + "learning_rate": 1.3873140799039553e-05, + "loss": 0.0922, + "step": 44810 + }, + { + "epoch": 1.97, + "learning_rate": 1.3650814824695968e-05, + "loss": 0.0889, + "step": 44820 + }, + { + "epoch": 1.97, + "learning_rate": 1.3428488850352387e-05, + "loss": 0.0866, + "step": 44830 + }, + { + "epoch": 1.97, + "learning_rate": 1.3206162876008804e-05, + "loss": 0.0879, + "step": 44840 + }, + { + "epoch": 1.97, + "learning_rate": 1.2983836901665222e-05, + "loss": 0.0892, + "step": 44850 + }, + { + "epoch": 1.97, + "learning_rate": 1.276151092732164e-05, + "loss": 0.092, + "step": 44860 + }, + { + "epoch": 1.98, + "learning_rate": 1.2539184952978056e-05, + "loss": 0.0839, + "step": 44870 + }, + { + "epoch": 1.98, + "learning_rate": 1.2316858978634473e-05, + "loss": 0.0873, + "step": 44880 + }, + { + "epoch": 1.98, + "learning_rate": 1.2094533004290892e-05, + "loss": 0.0857, + "step": 44890 + }, + { + "epoch": 1.98, + "learning_rate": 1.187220702994731e-05, + "loss": 0.0869, + "step": 44900 + }, + { + "epoch": 1.98, + "learning_rate": 1.1649881055603727e-05, + "loss": 0.0893, + "step": 44910 + }, + { + "epoch": 1.98, + "learning_rate": 1.1427555081260144e-05, + "loss": 0.0869, + "step": 44920 + }, + { + "epoch": 1.98, + "learning_rate": 1.1205229106916561e-05, + "loss": 0.0906, + "step": 44930 + }, + { + "epoch": 1.98, + "learning_rate": 1.098290313257298e-05, + "loss": 0.0883, + "step": 44940 + }, + { + "epoch": 1.98, + "learning_rate": 1.0760577158229395e-05, + "loss": 0.0863, + "step": 44950 + }, + { + "epoch": 1.98, + "learning_rate": 1.0538251183885813e-05, + "loss": 0.0903, + "step": 44960 + }, + { + "epoch": 1.98, + "learning_rate": 1.0315925209542232e-05, + "loss": 0.0887, + "step": 44970 + }, + { + "epoch": 1.98, + "learning_rate": 1.0093599235198649e-05, + "loss": 0.0823, + "step": 44980 + }, + { + "epoch": 1.98, + "learning_rate": 9.871273260855064e-06, + "loss": 0.0859, + "step": 44990 + }, + { + "epoch": 1.98, + "learning_rate": 9.648947286511483e-06, + "loss": 0.0863, + "step": 45000 + }, + { + "epoch": 1.98, + "eval_cer": 0.49899034061165964, + "eval_loss": 0.07421446591615677, + "eval_runtime": 84.0432, + "eval_samples_per_second": 118.915, + "eval_steps_per_second": 7.437, + "step": 45000 + }, + { + "epoch": 1.98, + "learning_rate": 9.4266213121679e-06, + "loss": 0.0864, + "step": 45010 + }, + { + "epoch": 1.98, + "learning_rate": 9.20429533782432e-06, + "loss": 0.0879, + "step": 45020 + }, + { + "epoch": 1.98, + "learning_rate": 8.981969363480737e-06, + "loss": 0.089, + "step": 45030 + }, + { + "epoch": 1.98, + "learning_rate": 8.759643389137152e-06, + "loss": 0.0803, + "step": 45040 + }, + { + "epoch": 1.98, + "learning_rate": 8.537317414793571e-06, + "loss": 0.0901, + "step": 45050 + }, + { + "epoch": 1.98, + "learning_rate": 8.314991440449988e-06, + "loss": 0.086, + "step": 45060 + }, + { + "epoch": 1.98, + "learning_rate": 8.092665466106405e-06, + "loss": 0.0897, + "step": 45070 + }, + { + "epoch": 1.98, + "learning_rate": 7.870339491762823e-06, + "loss": 0.0918, + "step": 45080 + }, + { + "epoch": 1.98, + "learning_rate": 7.64801351741924e-06, + "loss": 0.0858, + "step": 45090 + }, + { + "epoch": 1.99, + "learning_rate": 7.425687543075658e-06, + "loss": 0.0863, + "step": 45100 + }, + { + "epoch": 1.99, + "learning_rate": 7.203361568732074e-06, + "loss": 0.086, + "step": 45110 + }, + { + "epoch": 1.99, + "learning_rate": 6.981035594388492e-06, + "loss": 0.086, + "step": 45120 + }, + { + "epoch": 1.99, + "learning_rate": 6.7587096200449105e-06, + "loss": 0.0871, + "step": 45130 + }, + { + "epoch": 1.99, + "learning_rate": 6.536383645701328e-06, + "loss": 0.0845, + "step": 45140 + }, + { + "epoch": 1.99, + "learning_rate": 6.314057671357745e-06, + "loss": 0.0897, + "step": 45150 + }, + { + "epoch": 1.99, + "learning_rate": 6.091731697014162e-06, + "loss": 0.0894, + "step": 45160 + }, + { + "epoch": 1.99, + "learning_rate": 5.86940572267058e-06, + "loss": 0.0855, + "step": 45170 + }, + { + "epoch": 1.99, + "learning_rate": 5.647079748326997e-06, + "loss": 0.0853, + "step": 45180 + }, + { + "epoch": 1.99, + "learning_rate": 5.424753773983415e-06, + "loss": 0.085, + "step": 45190 + }, + { + "epoch": 1.99, + "learning_rate": 5.202427799639832e-06, + "loss": 0.0875, + "step": 45200 + }, + { + "epoch": 1.99, + "learning_rate": 4.98010182529625e-06, + "loss": 0.0853, + "step": 45210 + }, + { + "epoch": 1.99, + "learning_rate": 4.757775850952666e-06, + "loss": 0.0857, + "step": 45220 + }, + { + "epoch": 1.99, + "learning_rate": 4.535449876609084e-06, + "loss": 0.0843, + "step": 45230 + }, + { + "epoch": 1.99, + "learning_rate": 4.3131239022655016e-06, + "loss": 0.0827, + "step": 45240 + }, + { + "epoch": 1.99, + "learning_rate": 4.09079792792192e-06, + "loss": 0.0841, + "step": 45250 + }, + { + "epoch": 1.99, + "learning_rate": 3.868471953578337e-06, + "loss": 0.0927, + "step": 45260 + }, + { + "epoch": 1.99, + "learning_rate": 3.646145979234754e-06, + "loss": 0.0881, + "step": 45270 + }, + { + "epoch": 1.99, + "learning_rate": 3.4238200048911713e-06, + "loss": 0.0893, + "step": 45280 + }, + { + "epoch": 1.99, + "learning_rate": 3.201494030547589e-06, + "loss": 0.0851, + "step": 45290 + }, + { + "epoch": 1.99, + "learning_rate": 2.979168056204006e-06, + "loss": 0.0854, + "step": 45300 + }, + { + "epoch": 1.99, + "learning_rate": 2.7568420818604238e-06, + "loss": 0.0887, + "step": 45310 + }, + { + "epoch": 1.99, + "learning_rate": 2.534516107516841e-06, + "loss": 0.0909, + "step": 45320 + }, + { + "epoch": 2.0, + "learning_rate": 2.3121901331732586e-06, + "loss": 0.0895, + "step": 45330 + }, + { + "epoch": 2.0, + "learning_rate": 2.0898641588296763e-06, + "loss": 0.0894, + "step": 45340 + }, + { + "epoch": 2.0, + "learning_rate": 1.8675381844860935e-06, + "loss": 0.0848, + "step": 45350 + }, + { + "epoch": 2.0, + "learning_rate": 1.645212210142511e-06, + "loss": 0.0867, + "step": 45360 + }, + { + "epoch": 2.0, + "learning_rate": 1.4228862357989284e-06, + "loss": 0.0844, + "step": 45370 + }, + { + "epoch": 2.0, + "learning_rate": 1.200560261455346e-06, + "loss": 0.0905, + "step": 45380 + }, + { + "epoch": 2.0, + "learning_rate": 9.782342871117634e-07, + "loss": 0.0872, + "step": 45390 + }, + { + "epoch": 2.0, + "learning_rate": 7.559083127681808e-07, + "loss": 0.0897, + "step": 45400 + }, + { + "epoch": 2.0, + "learning_rate": 5.335823384245982e-07, + "loss": 0.0865, + "step": 45410 + }, + { + "epoch": 2.0, + "learning_rate": 3.112563640810156e-07, + "loss": 0.0892, + "step": 45420 + }, + { + "epoch": 2.0, + "learning_rate": 8.893038973743302e-08, + "loss": 0.0864, + "step": 45430 + }, + { + "epoch": 2.0, + "step": 45434, + "total_flos": 1.7226765618429542e+18, + "train_loss": 0.13663575964982164, + "train_runtime": 49080.0234, + "train_samples_per_second": 236.991, + "train_steps_per_second": 0.926 + } + ], + "max_steps": 45434, + "num_train_epochs": 2, + "total_flos": 1.7226765618429542e+18, + "trial_name": null, + "trial_params": null +}