{ "best_metric": 0.45029670000076294, "best_model_checkpoint": "ai-light-dance_singing3_ft_wav2vec2-large-xlsr-53-v1-5gram/checkpoint-144", "epoch": 99.99653979238754, "global_step": 14400, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.07, "learning_rate": 4e-08, "loss": 0.3279, "step": 10 }, { "epoch": 0.14, "learning_rate": 8e-08, "loss": 0.2955, "step": 20 }, { "epoch": 0.21, "learning_rate": 1.1599999999999999e-07, "loss": 0.3391, "step": 30 }, { "epoch": 0.28, "learning_rate": 1.56e-07, "loss": 0.3143, "step": 40 }, { "epoch": 0.35, "learning_rate": 1.9199999999999997e-07, "loss": 0.2993, "step": 50 }, { "epoch": 0.42, "learning_rate": 2.3199999999999999e-07, "loss": 0.3166, "step": 60 }, { "epoch": 0.48, "learning_rate": 2.72e-07, "loss": 0.3098, "step": 70 }, { "epoch": 0.55, "learning_rate": 3.12e-07, "loss": 0.3187, "step": 80 }, { "epoch": 0.62, "learning_rate": 3.52e-07, "loss": 0.2942, "step": 90 }, { "epoch": 0.69, "learning_rate": 3.9199999999999996e-07, "loss": 0.3019, "step": 100 }, { "epoch": 0.76, "learning_rate": 3.9977622377622375e-07, "loss": 0.2931, "step": 110 }, { "epoch": 0.83, "learning_rate": 3.9949650349650347e-07, "loss": 0.2809, "step": 120 }, { "epoch": 0.9, "learning_rate": 3.992167832167832e-07, "loss": 0.2919, "step": 130 }, { "epoch": 0.97, "learning_rate": 3.9893706293706295e-07, "loss": 0.3376, "step": 140 }, { "epoch": 1.0, "eval_loss": 0.45029670000076294, "eval_runtime": 125.8866, "eval_samples_per_second": 4.226, "eval_steps_per_second": 0.532, "eval_wer": 0.2117374815195732, "step": 144 }, { "epoch": 1.04, "learning_rate": 3.9865734265734266e-07, "loss": 0.3174, "step": 150 }, { "epoch": 1.11, "learning_rate": 3.983776223776224e-07, "loss": 0.2944, "step": 160 }, { "epoch": 1.18, "learning_rate": 3.9809790209790204e-07, "loss": 0.3117, "step": 170 }, { "epoch": 1.25, "learning_rate": 3.978181818181818e-07, "loss": 0.2967, "step": 180 }, { "epoch": 1.32, "learning_rate": 3.975384615384615e-07, "loss": 0.2993, "step": 190 }, { "epoch": 1.39, "learning_rate": 3.9725874125874123e-07, "loss": 0.2847, "step": 200 }, { "epoch": 1.46, "learning_rate": 3.9697902097902095e-07, "loss": 0.31, "step": 210 }, { "epoch": 1.53, "learning_rate": 3.966993006993007e-07, "loss": 0.3074, "step": 220 }, { "epoch": 1.6, "learning_rate": 3.9641958041958043e-07, "loss": 0.3086, "step": 230 }, { "epoch": 1.66, "learning_rate": 3.9613986013986014e-07, "loss": 0.3035, "step": 240 }, { "epoch": 1.73, "learning_rate": 3.9586013986013986e-07, "loss": 0.3301, "step": 250 }, { "epoch": 1.8, "learning_rate": 3.9558041958041957e-07, "loss": 0.3173, "step": 260 }, { "epoch": 1.87, "learning_rate": 3.953006993006993e-07, "loss": 0.3089, "step": 270 }, { "epoch": 1.94, "learning_rate": 3.95020979020979e-07, "loss": 0.3064, "step": 280 }, { "epoch": 2.0, "eval_loss": 0.4526239335536957, "eval_runtime": 126.2511, "eval_samples_per_second": 4.214, "eval_steps_per_second": 0.531, "eval_wer": 0.2119303207559298, "step": 288 }, { "epoch": 2.01, "learning_rate": 3.947412587412587e-07, "loss": 0.3261, "step": 290 }, { "epoch": 2.08, "learning_rate": 3.944615384615384e-07, "loss": 0.293, "step": 300 }, { "epoch": 2.15, "learning_rate": 3.941818181818182e-07, "loss": 0.2931, "step": 310 }, { "epoch": 2.22, "learning_rate": 3.939020979020979e-07, "loss": 0.323, "step": 320 }, { "epoch": 2.29, "learning_rate": 3.936223776223776e-07, "loss": 0.3094, "step": 330 }, { "epoch": 2.36, "learning_rate": 3.9334265734265734e-07, "loss": 0.3025, "step": 340 }, { "epoch": 2.43, "learning_rate": 3.9306293706293705e-07, "loss": 0.3006, "step": 350 }, { "epoch": 2.5, "learning_rate": 3.9278321678321676e-07, "loss": 0.3051, "step": 360 }, { "epoch": 2.57, "learning_rate": 3.925034965034965e-07, "loss": 0.3071, "step": 370 }, { "epoch": 2.64, "learning_rate": 3.922237762237762e-07, "loss": 0.2814, "step": 380 }, { "epoch": 2.71, "learning_rate": 3.919440559440559e-07, "loss": 0.3263, "step": 390 }, { "epoch": 2.78, "learning_rate": 3.9166433566433567e-07, "loss": 0.2911, "step": 400 }, { "epoch": 2.84, "learning_rate": 3.913846153846154e-07, "loss": 0.299, "step": 410 }, { "epoch": 2.91, "learning_rate": 3.911048951048951e-07, "loss": 0.2921, "step": 420 }, { "epoch": 2.98, "learning_rate": 3.9082517482517476e-07, "loss": 0.3032, "step": 430 }, { "epoch": 3.0, "eval_loss": 0.45453566312789917, "eval_runtime": 126.716, "eval_samples_per_second": 4.198, "eval_steps_per_second": 0.529, "eval_wer": 0.21077328533779005, "step": 432 }, { "epoch": 3.06, "learning_rate": 3.9054545454545453e-07, "loss": 0.3096, "step": 440 }, { "epoch": 3.12, "learning_rate": 3.9026573426573424e-07, "loss": 0.2891, "step": 450 }, { "epoch": 3.19, "learning_rate": 3.8998601398601396e-07, "loss": 0.3079, "step": 460 }, { "epoch": 3.26, "learning_rate": 3.8970629370629367e-07, "loss": 0.3146, "step": 470 }, { "epoch": 3.33, "learning_rate": 3.8942657342657344e-07, "loss": 0.3078, "step": 480 }, { "epoch": 3.4, "learning_rate": 3.8914685314685315e-07, "loss": 0.2858, "step": 490 }, { "epoch": 3.47, "learning_rate": 3.8886713286713287e-07, "loss": 0.3078, "step": 500 }, { "epoch": 3.54, "learning_rate": 3.885874125874126e-07, "loss": 0.3041, "step": 510 }, { "epoch": 3.61, "learning_rate": 3.883076923076923e-07, "loss": 0.2828, "step": 520 }, { "epoch": 3.68, "learning_rate": 3.88027972027972e-07, "loss": 0.2955, "step": 530 }, { "epoch": 3.75, "learning_rate": 3.877482517482517e-07, "loss": 0.292, "step": 540 }, { "epoch": 3.82, "learning_rate": 3.8746853146853144e-07, "loss": 0.2835, "step": 550 }, { "epoch": 3.89, "learning_rate": 3.8718881118881115e-07, "loss": 0.2837, "step": 560 }, { "epoch": 3.96, "learning_rate": 3.869090909090909e-07, "loss": 0.3015, "step": 570 }, { "epoch": 4.0, "eval_loss": 0.45614418387413025, "eval_runtime": 126.2963, "eval_samples_per_second": 4.212, "eval_steps_per_second": 0.53, "eval_wer": 0.2112232435559555, "step": 576 }, { "epoch": 4.03, "learning_rate": 3.8662937062937063e-07, "loss": 0.316, "step": 580 }, { "epoch": 4.1, "learning_rate": 3.8634965034965035e-07, "loss": 0.2945, "step": 590 }, { "epoch": 4.17, "learning_rate": 3.8606993006993006e-07, "loss": 0.2913, "step": 600 }, { "epoch": 4.24, "learning_rate": 3.857902097902098e-07, "loss": 0.299, "step": 610 }, { "epoch": 4.3, "learning_rate": 3.855104895104895e-07, "loss": 0.29, "step": 620 }, { "epoch": 4.37, "learning_rate": 3.852307692307692e-07, "loss": 0.2924, "step": 630 }, { "epoch": 4.44, "learning_rate": 3.849510489510489e-07, "loss": 0.31, "step": 640 }, { "epoch": 4.51, "learning_rate": 3.8467132867132863e-07, "loss": 0.3074, "step": 650 }, { "epoch": 4.58, "learning_rate": 3.843916083916084e-07, "loss": 0.2956, "step": 660 }, { "epoch": 4.65, "learning_rate": 3.841118881118881e-07, "loss": 0.2937, "step": 670 }, { "epoch": 4.72, "learning_rate": 3.838321678321678e-07, "loss": 0.3293, "step": 680 }, { "epoch": 4.79, "learning_rate": 3.8355244755244754e-07, "loss": 0.2896, "step": 690 }, { "epoch": 4.86, "learning_rate": 3.8327272727272725e-07, "loss": 0.2871, "step": 700 }, { "epoch": 4.93, "learning_rate": 3.8299300699300697e-07, "loss": 0.2804, "step": 710 }, { "epoch": 5.0, "learning_rate": 3.827132867132867e-07, "loss": 0.3067, "step": 720 }, { "epoch": 5.0, "eval_loss": 0.455491304397583, "eval_runtime": 126.1611, "eval_samples_per_second": 4.217, "eval_steps_per_second": 0.531, "eval_wer": 0.21090184482869448, "step": 720 }, { "epoch": 5.07, "learning_rate": 3.824335664335664e-07, "loss": 0.3095, "step": 730 }, { "epoch": 5.14, "learning_rate": 3.8215384615384616e-07, "loss": 0.2745, "step": 740 }, { "epoch": 5.21, "learning_rate": 3.818741258741259e-07, "loss": 0.3084, "step": 750 }, { "epoch": 5.28, "learning_rate": 3.815944055944056e-07, "loss": 0.3066, "step": 760 }, { "epoch": 5.35, "learning_rate": 3.813146853146853e-07, "loss": 0.2947, "step": 770 }, { "epoch": 5.42, "learning_rate": 3.81034965034965e-07, "loss": 0.2923, "step": 780 }, { "epoch": 5.48, "learning_rate": 3.8075524475524473e-07, "loss": 0.2992, "step": 790 }, { "epoch": 5.55, "learning_rate": 3.8047552447552445e-07, "loss": 0.2936, "step": 800 }, { "epoch": 5.62, "learning_rate": 3.8019580419580416e-07, "loss": 0.3003, "step": 810 }, { "epoch": 5.69, "learning_rate": 3.799160839160839e-07, "loss": 0.3068, "step": 820 }, { "epoch": 5.76, "learning_rate": 3.7963636363636364e-07, "loss": 0.2928, "step": 830 }, { "epoch": 5.83, "learning_rate": 3.7935664335664336e-07, "loss": 0.2914, "step": 840 }, { "epoch": 5.9, "learning_rate": 3.7907692307692307e-07, "loss": 0.2897, "step": 850 }, { "epoch": 5.97, "learning_rate": 3.787972027972028e-07, "loss": 0.3001, "step": 860 }, { "epoch": 6.0, "eval_loss": 0.45580199360847473, "eval_runtime": 125.8438, "eval_samples_per_second": 4.227, "eval_steps_per_second": 0.532, "eval_wer": 0.21135180304685994, "step": 864 }, { "epoch": 6.04, "learning_rate": 3.785174825174825e-07, "loss": 0.3225, "step": 870 }, { "epoch": 6.11, "learning_rate": 3.782377622377622e-07, "loss": 0.2933, "step": 880 }, { "epoch": 6.18, "learning_rate": 3.7795804195804193e-07, "loss": 0.2816, "step": 890 }, { "epoch": 6.25, "learning_rate": 3.7767832167832164e-07, "loss": 0.2926, "step": 900 }, { "epoch": 6.32, "learning_rate": 3.7739860139860136e-07, "loss": 0.2839, "step": 910 }, { "epoch": 6.39, "learning_rate": 3.771188811188811e-07, "loss": 0.2865, "step": 920 }, { "epoch": 6.46, "learning_rate": 3.7683916083916084e-07, "loss": 0.3098, "step": 930 }, { "epoch": 6.53, "learning_rate": 3.7655944055944055e-07, "loss": 0.3185, "step": 940 }, { "epoch": 6.6, "learning_rate": 3.7627972027972026e-07, "loss": 0.3069, "step": 950 }, { "epoch": 6.66, "learning_rate": 3.76e-07, "loss": 0.3023, "step": 960 }, { "epoch": 6.73, "learning_rate": 3.757202797202797e-07, "loss": 0.2992, "step": 970 }, { "epoch": 6.8, "learning_rate": 3.754405594405594e-07, "loss": 0.2708, "step": 980 }, { "epoch": 6.87, "learning_rate": 3.751608391608391e-07, "loss": 0.2784, "step": 990 }, { "epoch": 6.94, "learning_rate": 3.748811188811189e-07, "loss": 0.2872, "step": 1000 }, { "epoch": 7.0, "eval_loss": 0.45643875002861023, "eval_runtime": 126.0308, "eval_samples_per_second": 4.221, "eval_steps_per_second": 0.532, "eval_wer": 0.21000192839236356, "step": 1008 }, { "epoch": 7.01, "learning_rate": 3.746013986013986e-07, "loss": 0.3204, "step": 1010 }, { "epoch": 7.08, "learning_rate": 3.743216783216783e-07, "loss": 0.3022, "step": 1020 }, { "epoch": 7.15, "learning_rate": 3.7404195804195803e-07, "loss": 0.2941, "step": 1030 }, { "epoch": 7.22, "learning_rate": 3.7376223776223774e-07, "loss": 0.3136, "step": 1040 }, { "epoch": 7.29, "learning_rate": 3.7348251748251746e-07, "loss": 0.2924, "step": 1050 }, { "epoch": 7.36, "learning_rate": 3.7320279720279717e-07, "loss": 0.3018, "step": 1060 }, { "epoch": 7.43, "learning_rate": 3.729230769230769e-07, "loss": 0.3014, "step": 1070 }, { "epoch": 7.5, "learning_rate": 3.726433566433566e-07, "loss": 0.3081, "step": 1080 }, { "epoch": 7.57, "learning_rate": 3.7236363636363637e-07, "loss": 0.2891, "step": 1090 }, { "epoch": 7.64, "learning_rate": 3.720839160839161e-07, "loss": 0.2847, "step": 1100 }, { "epoch": 7.71, "learning_rate": 3.718041958041958e-07, "loss": 0.297, "step": 1110 }, { "epoch": 7.78, "learning_rate": 3.715244755244755e-07, "loss": 0.2973, "step": 1120 }, { "epoch": 7.84, "learning_rate": 3.712447552447552e-07, "loss": 0.2964, "step": 1130 }, { "epoch": 7.91, "learning_rate": 3.7096503496503494e-07, "loss": 0.2964, "step": 1140 }, { "epoch": 7.98, "learning_rate": 3.7068531468531465e-07, "loss": 0.3082, "step": 1150 }, { "epoch": 8.0, "eval_loss": 0.4563812017440796, "eval_runtime": 126.312, "eval_samples_per_second": 4.212, "eval_steps_per_second": 0.53, "eval_wer": 0.20980908915600693, "step": 1152 }, { "epoch": 8.06, "learning_rate": 3.7040559440559437e-07, "loss": 0.3041, "step": 1160 }, { "epoch": 8.12, "learning_rate": 3.701258741258741e-07, "loss": 0.2912, "step": 1170 }, { "epoch": 8.19, "learning_rate": 3.6984615384615385e-07, "loss": 0.3023, "step": 1180 }, { "epoch": 8.26, "learning_rate": 3.6956643356643356e-07, "loss": 0.2903, "step": 1190 }, { "epoch": 8.33, "learning_rate": 3.692867132867133e-07, "loss": 0.2943, "step": 1200 }, { "epoch": 8.4, "learning_rate": 3.69006993006993e-07, "loss": 0.2872, "step": 1210 }, { "epoch": 8.47, "learning_rate": 3.6872727272727276e-07, "loss": 0.3175, "step": 1220 }, { "epoch": 8.54, "learning_rate": 3.684475524475524e-07, "loss": 0.2914, "step": 1230 }, { "epoch": 8.61, "learning_rate": 3.6816783216783213e-07, "loss": 0.2935, "step": 1240 }, { "epoch": 8.68, "learning_rate": 3.6788811188811185e-07, "loss": 0.2994, "step": 1250 }, { "epoch": 8.75, "learning_rate": 3.676083916083916e-07, "loss": 0.3011, "step": 1260 }, { "epoch": 8.82, "learning_rate": 3.6732867132867133e-07, "loss": 0.3058, "step": 1270 }, { "epoch": 8.89, "learning_rate": 3.6704895104895104e-07, "loss": 0.2797, "step": 1280 }, { "epoch": 8.96, "learning_rate": 3.6676923076923075e-07, "loss": 0.3032, "step": 1290 }, { "epoch": 9.0, "eval_loss": 0.4580211341381073, "eval_runtime": 126.451, "eval_samples_per_second": 4.207, "eval_steps_per_second": 0.53, "eval_wer": 0.21051616635598122, "step": 1296 }, { "epoch": 9.03, "learning_rate": 3.6648951048951047e-07, "loss": 0.3168, "step": 1300 }, { "epoch": 9.1, "learning_rate": 3.6620979020979024e-07, "loss": 0.2895, "step": 1310 }, { "epoch": 9.17, "learning_rate": 3.659300699300699e-07, "loss": 0.2926, "step": 1320 }, { "epoch": 9.24, "learning_rate": 3.656503496503496e-07, "loss": 0.3035, "step": 1330 }, { "epoch": 9.3, "learning_rate": 3.653706293706293e-07, "loss": 0.2975, "step": 1340 }, { "epoch": 9.37, "learning_rate": 3.650909090909091e-07, "loss": 0.2874, "step": 1350 }, { "epoch": 9.44, "learning_rate": 3.648111888111888e-07, "loss": 0.3158, "step": 1360 }, { "epoch": 9.51, "learning_rate": 3.645314685314685e-07, "loss": 0.2894, "step": 1370 }, { "epoch": 9.58, "learning_rate": 3.6425174825174823e-07, "loss": 0.2858, "step": 1380 }, { "epoch": 9.65, "learning_rate": 3.6397202797202795e-07, "loss": 0.2795, "step": 1390 }, { "epoch": 9.72, "learning_rate": 3.6369230769230766e-07, "loss": 0.3084, "step": 1400 }, { "epoch": 9.79, "learning_rate": 3.634125874125874e-07, "loss": 0.2874, "step": 1410 }, { "epoch": 9.86, "learning_rate": 3.631328671328671e-07, "loss": 0.2947, "step": 1420 }, { "epoch": 9.93, "learning_rate": 3.628531468531468e-07, "loss": 0.3126, "step": 1430 }, { "epoch": 10.0, "learning_rate": 3.6257342657342657e-07, "loss": 0.304, "step": 1440 }, { "epoch": 10.0, "eval_loss": 0.4562743306159973, "eval_runtime": 126.258, "eval_samples_per_second": 4.214, "eval_steps_per_second": 0.531, "eval_wer": 0.20923057144693707, "step": 1440 }, { "epoch": 10.07, "learning_rate": 3.622937062937063e-07, "loss": 0.3036, "step": 1450 }, { "epoch": 10.14, "learning_rate": 3.62013986013986e-07, "loss": 0.2625, "step": 1460 }, { "epoch": 10.21, "learning_rate": 3.617342657342657e-07, "loss": 0.3092, "step": 1470 }, { "epoch": 10.28, "learning_rate": 3.614545454545455e-07, "loss": 0.3034, "step": 1480 }, { "epoch": 10.35, "learning_rate": 3.6117482517482514e-07, "loss": 0.3057, "step": 1490 }, { "epoch": 10.42, "learning_rate": 3.6089510489510486e-07, "loss": 0.2996, "step": 1500 }, { "epoch": 10.48, "learning_rate": 3.6061538461538457e-07, "loss": 0.3187, "step": 1510 }, { "epoch": 10.55, "learning_rate": 3.6033566433566434e-07, "loss": 0.3052, "step": 1520 }, { "epoch": 10.62, "learning_rate": 3.6005594405594405e-07, "loss": 0.2923, "step": 1530 }, { "epoch": 10.69, "learning_rate": 3.5977622377622377e-07, "loss": 0.3025, "step": 1540 }, { "epoch": 10.76, "learning_rate": 3.594965034965035e-07, "loss": 0.2872, "step": 1550 }, { "epoch": 10.83, "learning_rate": 3.592167832167832e-07, "loss": 0.3026, "step": 1560 }, { "epoch": 10.9, "learning_rate": 3.5893706293706296e-07, "loss": 0.2842, "step": 1570 }, { "epoch": 10.97, "learning_rate": 3.586573426573426e-07, "loss": 0.3056, "step": 1580 }, { "epoch": 11.0, "eval_loss": 0.4582013785839081, "eval_runtime": 126.1153, "eval_samples_per_second": 4.218, "eval_steps_per_second": 0.531, "eval_wer": 0.20987336890145916, "step": 1584 }, { "epoch": 11.04, "learning_rate": 3.5837762237762234e-07, "loss": 0.3002, "step": 1590 }, { "epoch": 11.11, "learning_rate": 3.5809790209790205e-07, "loss": 0.2964, "step": 1600 }, { "epoch": 11.18, "learning_rate": 3.578181818181818e-07, "loss": 0.296, "step": 1610 }, { "epoch": 11.25, "learning_rate": 3.5753846153846153e-07, "loss": 0.3136, "step": 1620 }, { "epoch": 11.32, "learning_rate": 3.5725874125874125e-07, "loss": 0.2848, "step": 1630 }, { "epoch": 11.39, "learning_rate": 3.5697902097902096e-07, "loss": 0.2745, "step": 1640 }, { "epoch": 11.46, "learning_rate": 3.5669930069930067e-07, "loss": 0.3121, "step": 1650 }, { "epoch": 11.53, "learning_rate": 3.5641958041958044e-07, "loss": 0.3034, "step": 1660 }, { "epoch": 11.6, "learning_rate": 3.561398601398601e-07, "loss": 0.2818, "step": 1670 }, { "epoch": 11.66, "learning_rate": 3.558601398601398e-07, "loss": 0.2898, "step": 1680 }, { "epoch": 11.73, "learning_rate": 3.5558041958041953e-07, "loss": 0.3044, "step": 1690 }, { "epoch": 11.8, "learning_rate": 3.553006993006993e-07, "loss": 0.3081, "step": 1700 }, { "epoch": 11.87, "learning_rate": 3.55020979020979e-07, "loss": 0.2898, "step": 1710 }, { "epoch": 11.94, "learning_rate": 3.547412587412587e-07, "loss": 0.3142, "step": 1720 }, { "epoch": 12.0, "eval_loss": 0.45853883028030396, "eval_runtime": 126.2918, "eval_samples_per_second": 4.212, "eval_steps_per_second": 0.531, "eval_wer": 0.210130487883268, "step": 1728 }, { "epoch": 12.01, "learning_rate": 3.5446153846153844e-07, "loss": 0.2832, "step": 1730 }, { "epoch": 12.08, "learning_rate": 3.541818181818182e-07, "loss": 0.2875, "step": 1740 }, { "epoch": 12.15, "learning_rate": 3.539020979020979e-07, "loss": 0.2916, "step": 1750 }, { "epoch": 12.22, "learning_rate": 3.536223776223776e-07, "loss": 0.2865, "step": 1760 }, { "epoch": 12.29, "learning_rate": 3.533426573426573e-07, "loss": 0.306, "step": 1770 }, { "epoch": 12.36, "learning_rate": 3.5309090909090904e-07, "loss": 0.2971, "step": 1780 }, { "epoch": 12.43, "learning_rate": 3.528111888111888e-07, "loss": 0.3009, "step": 1790 }, { "epoch": 12.5, "learning_rate": 3.525314685314685e-07, "loss": 0.3043, "step": 1800 }, { "epoch": 12.57, "learning_rate": 3.5225174825174823e-07, "loss": 0.2726, "step": 1810 }, { "epoch": 12.64, "learning_rate": 3.5197202797202795e-07, "loss": 0.2725, "step": 1820 }, { "epoch": 12.71, "learning_rate": 3.516923076923077e-07, "loss": 0.3035, "step": 1830 }, { "epoch": 12.78, "learning_rate": 3.5141258741258743e-07, "loss": 0.2942, "step": 1840 }, { "epoch": 12.84, "learning_rate": 3.5113286713286714e-07, "loss": 0.2861, "step": 1850 }, { "epoch": 12.91, "learning_rate": 3.508531468531468e-07, "loss": 0.293, "step": 1860 }, { "epoch": 12.98, "learning_rate": 3.5057342657342657e-07, "loss": 0.3012, "step": 1870 }, { "epoch": 13.0, "eval_loss": 0.4594975411891937, "eval_runtime": 126.1025, "eval_samples_per_second": 4.219, "eval_steps_per_second": 0.531, "eval_wer": 0.20974480941055473, "step": 1872 }, { "epoch": 13.06, "learning_rate": 3.502937062937063e-07, "loss": 0.3152, "step": 1880 }, { "epoch": 13.12, "learning_rate": 3.50013986013986e-07, "loss": 0.2786, "step": 1890 }, { "epoch": 13.19, "learning_rate": 3.497342657342657e-07, "loss": 0.2972, "step": 1900 }, { "epoch": 13.26, "learning_rate": 3.4945454545454543e-07, "loss": 0.3007, "step": 1910 }, { "epoch": 13.33, "learning_rate": 3.491748251748252e-07, "loss": 0.3023, "step": 1920 }, { "epoch": 13.4, "learning_rate": 3.488951048951049e-07, "loss": 0.2807, "step": 1930 }, { "epoch": 13.47, "learning_rate": 3.4861538461538457e-07, "loss": 0.3088, "step": 1940 }, { "epoch": 13.54, "learning_rate": 3.483356643356643e-07, "loss": 0.2778, "step": 1950 }, { "epoch": 13.61, "learning_rate": 3.4805594405594405e-07, "loss": 0.2845, "step": 1960 }, { "epoch": 13.68, "learning_rate": 3.4777622377622376e-07, "loss": 0.3127, "step": 1970 }, { "epoch": 13.75, "learning_rate": 3.474965034965035e-07, "loss": 0.301, "step": 1980 }, { "epoch": 13.82, "learning_rate": 3.472167832167832e-07, "loss": 0.2911, "step": 1990 }, { "epoch": 13.89, "learning_rate": 3.469370629370629e-07, "loss": 0.299, "step": 2000 }, { "epoch": 13.96, "learning_rate": 3.4665734265734267e-07, "loss": 0.2886, "step": 2010 }, { "epoch": 14.0, "eval_loss": 0.4589780569076538, "eval_runtime": 126.6498, "eval_samples_per_second": 4.201, "eval_steps_per_second": 0.529, "eval_wer": 0.21032332711962462, "step": 2016 }, { "epoch": 14.03, "learning_rate": 3.463776223776224e-07, "loss": 0.3151, "step": 2020 }, { "epoch": 14.1, "learning_rate": 3.4609790209790205e-07, "loss": 0.2859, "step": 2030 }, { "epoch": 14.17, "learning_rate": 3.4581818181818176e-07, "loss": 0.2918, "step": 2040 }, { "epoch": 14.24, "learning_rate": 3.4553846153846153e-07, "loss": 0.3003, "step": 2050 }, { "epoch": 14.3, "learning_rate": 3.4525874125874124e-07, "loss": 0.302, "step": 2060 }, { "epoch": 14.37, "learning_rate": 3.4497902097902096e-07, "loss": 0.2772, "step": 2070 }, { "epoch": 14.44, "learning_rate": 3.4469930069930067e-07, "loss": 0.2938, "step": 2080 }, { "epoch": 14.51, "learning_rate": 3.4441958041958044e-07, "loss": 0.2947, "step": 2090 }, { "epoch": 14.58, "learning_rate": 3.4413986013986015e-07, "loss": 0.3019, "step": 2100 }, { "epoch": 14.65, "learning_rate": 3.4386013986013987e-07, "loss": 0.3013, "step": 2110 }, { "epoch": 14.72, "learning_rate": 3.4358041958041953e-07, "loss": 0.3019, "step": 2120 }, { "epoch": 14.79, "learning_rate": 3.433006993006993e-07, "loss": 0.2972, "step": 2130 }, { "epoch": 14.86, "learning_rate": 3.43020979020979e-07, "loss": 0.2683, "step": 2140 }, { "epoch": 14.93, "learning_rate": 3.427412587412587e-07, "loss": 0.2969, "step": 2150 }, { "epoch": 15.0, "learning_rate": 3.4246153846153844e-07, "loss": 0.3071, "step": 2160 }, { "epoch": 15.0, "eval_loss": 0.45672807097435, "eval_runtime": 126.2984, "eval_samples_per_second": 4.212, "eval_steps_per_second": 0.53, "eval_wer": 0.20935913093784148, "step": 2160 }, { "epoch": 15.07, "learning_rate": 3.4218181818181815e-07, "loss": 0.3242, "step": 2170 }, { "epoch": 15.14, "learning_rate": 3.419020979020979e-07, "loss": 0.2801, "step": 2180 }, { "epoch": 15.21, "learning_rate": 3.4162237762237763e-07, "loss": 0.312, "step": 2190 }, { "epoch": 15.28, "learning_rate": 3.4134265734265735e-07, "loss": 0.3056, "step": 2200 }, { "epoch": 15.35, "learning_rate": 3.41062937062937e-07, "loss": 0.2897, "step": 2210 }, { "epoch": 15.42, "learning_rate": 3.407832167832168e-07, "loss": 0.2915, "step": 2220 }, { "epoch": 15.48, "learning_rate": 3.405034965034965e-07, "loss": 0.3079, "step": 2230 }, { "epoch": 15.55, "learning_rate": 3.402237762237762e-07, "loss": 0.2779, "step": 2240 }, { "epoch": 15.62, "learning_rate": 3.399440559440559e-07, "loss": 0.2793, "step": 2250 }, { "epoch": 15.69, "learning_rate": 3.3966433566433563e-07, "loss": 0.2898, "step": 2260 }, { "epoch": 15.76, "learning_rate": 3.393846153846154e-07, "loss": 0.2942, "step": 2270 }, { "epoch": 15.83, "learning_rate": 3.391048951048951e-07, "loss": 0.3064, "step": 2280 }, { "epoch": 15.9, "learning_rate": 3.388251748251748e-07, "loss": 0.2887, "step": 2290 }, { "epoch": 15.97, "learning_rate": 3.385454545454545e-07, "loss": 0.3234, "step": 2300 }, { "epoch": 16.0, "eval_loss": 0.4554351270198822, "eval_runtime": 126.4726, "eval_samples_per_second": 4.206, "eval_steps_per_second": 0.53, "eval_wer": 0.20910201195603265, "step": 2304 }, { "epoch": 16.04, "learning_rate": 3.3826573426573425e-07, "loss": 0.3007, "step": 2310 }, { "epoch": 16.11, "learning_rate": 3.3798601398601397e-07, "loss": 0.2902, "step": 2320 }, { "epoch": 16.18, "learning_rate": 3.377062937062937e-07, "loss": 0.2987, "step": 2330 }, { "epoch": 16.25, "learning_rate": 3.374265734265734e-07, "loss": 0.2981, "step": 2340 }, { "epoch": 16.32, "learning_rate": 3.3714685314685316e-07, "loss": 0.277, "step": 2350 }, { "epoch": 16.39, "learning_rate": 3.368671328671329e-07, "loss": 0.2905, "step": 2360 }, { "epoch": 16.46, "learning_rate": 3.365874125874126e-07, "loss": 0.2997, "step": 2370 }, { "epoch": 16.53, "learning_rate": 3.363076923076923e-07, "loss": 0.312, "step": 2380 }, { "epoch": 16.6, "learning_rate": 3.36027972027972e-07, "loss": 0.3083, "step": 2390 }, { "epoch": 16.66, "learning_rate": 3.3574825174825173e-07, "loss": 0.2986, "step": 2400 }, { "epoch": 16.73, "learning_rate": 3.3546853146853145e-07, "loss": 0.2869, "step": 2410 }, { "epoch": 16.8, "learning_rate": 3.3518881118881116e-07, "loss": 0.2812, "step": 2420 }, { "epoch": 16.87, "learning_rate": 3.349090909090909e-07, "loss": 0.2856, "step": 2430 }, { "epoch": 16.94, "learning_rate": 3.3462937062937064e-07, "loss": 0.2973, "step": 2440 }, { "epoch": 17.0, "eval_loss": 0.4580243229866028, "eval_runtime": 126.6356, "eval_samples_per_second": 4.201, "eval_steps_per_second": 0.529, "eval_wer": 0.210130487883268, "step": 2448 }, { "epoch": 17.01, "learning_rate": 3.3434965034965036e-07, "loss": 0.3165, "step": 2450 }, { "epoch": 17.08, "learning_rate": 3.3406993006993007e-07, "loss": 0.3053, "step": 2460 }, { "epoch": 17.15, "learning_rate": 3.338181818181818e-07, "loss": 0.2819, "step": 2470 }, { "epoch": 17.22, "learning_rate": 3.3353846153846153e-07, "loss": 0.2878, "step": 2480 }, { "epoch": 17.29, "learning_rate": 3.3325874125874124e-07, "loss": 0.2884, "step": 2490 }, { "epoch": 17.36, "learning_rate": 3.3297902097902096e-07, "loss": 0.2955, "step": 2500 }, { "epoch": 17.43, "learning_rate": 3.3269930069930067e-07, "loss": 0.3029, "step": 2510 }, { "epoch": 17.5, "learning_rate": 3.324195804195804e-07, "loss": 0.3119, "step": 2520 }, { "epoch": 17.57, "learning_rate": 3.3213986013986015e-07, "loss": 0.2833, "step": 2530 }, { "epoch": 17.64, "learning_rate": 3.3186013986013987e-07, "loss": 0.2678, "step": 2540 }, { "epoch": 17.71, "learning_rate": 3.315804195804196e-07, "loss": 0.3054, "step": 2550 }, { "epoch": 17.78, "learning_rate": 3.313006993006993e-07, "loss": 0.3051, "step": 2560 }, { "epoch": 17.84, "learning_rate": 3.31020979020979e-07, "loss": 0.2812, "step": 2570 }, { "epoch": 17.91, "learning_rate": 3.307412587412587e-07, "loss": 0.2964, "step": 2580 }, { "epoch": 17.98, "learning_rate": 3.3046153846153844e-07, "loss": 0.3126, "step": 2590 }, { "epoch": 18.0, "eval_loss": 0.460181325674057, "eval_runtime": 126.3776, "eval_samples_per_second": 4.21, "eval_steps_per_second": 0.53, "eval_wer": 0.2085234942469628, "step": 2592 }, { "epoch": 18.06, "learning_rate": 3.3018181818181815e-07, "loss": 0.3038, "step": 2600 }, { "epoch": 18.12, "learning_rate": 3.2990209790209786e-07, "loss": 0.2856, "step": 2610 }, { "epoch": 18.19, "learning_rate": 3.2962237762237763e-07, "loss": 0.3128, "step": 2620 }, { "epoch": 18.26, "learning_rate": 3.2934265734265734e-07, "loss": 0.2964, "step": 2630 }, { "epoch": 18.33, "learning_rate": 3.2906293706293706e-07, "loss": 0.2939, "step": 2640 }, { "epoch": 18.4, "learning_rate": 3.2878321678321677e-07, "loss": 0.2784, "step": 2650 }, { "epoch": 18.47, "learning_rate": 3.285034965034965e-07, "loss": 0.3157, "step": 2660 }, { "epoch": 18.54, "learning_rate": 3.282237762237762e-07, "loss": 0.2807, "step": 2670 }, { "epoch": 18.61, "learning_rate": 3.279440559440559e-07, "loss": 0.2956, "step": 2680 }, { "epoch": 18.68, "learning_rate": 3.2766433566433563e-07, "loss": 0.2814, "step": 2690 }, { "epoch": 18.75, "learning_rate": 3.273846153846154e-07, "loss": 0.2973, "step": 2700 }, { "epoch": 18.82, "learning_rate": 3.271048951048951e-07, "loss": 0.2943, "step": 2710 }, { "epoch": 18.89, "learning_rate": 3.268251748251748e-07, "loss": 0.2875, "step": 2720 }, { "epoch": 18.96, "learning_rate": 3.2654545454545454e-07, "loss": 0.3027, "step": 2730 }, { "epoch": 19.0, "eval_loss": 0.46277734637260437, "eval_runtime": 126.8019, "eval_samples_per_second": 4.196, "eval_steps_per_second": 0.528, "eval_wer": 0.20916629170148487, "step": 2736 }, { "epoch": 19.03, "learning_rate": 3.2626573426573425e-07, "loss": 0.2928, "step": 2740 }, { "epoch": 19.1, "learning_rate": 3.2598601398601397e-07, "loss": 0.2847, "step": 2750 }, { "epoch": 19.17, "learning_rate": 3.257062937062937e-07, "loss": 0.2852, "step": 2760 }, { "epoch": 19.24, "learning_rate": 3.254265734265734e-07, "loss": 0.3102, "step": 2770 }, { "epoch": 19.3, "learning_rate": 3.251468531468531e-07, "loss": 0.2926, "step": 2780 }, { "epoch": 19.37, "learning_rate": 3.248671328671329e-07, "loss": 0.2848, "step": 2790 }, { "epoch": 19.44, "learning_rate": 3.245874125874126e-07, "loss": 0.2856, "step": 2800 }, { "epoch": 19.51, "learning_rate": 3.243076923076923e-07, "loss": 0.2912, "step": 2810 }, { "epoch": 19.58, "learning_rate": 3.24027972027972e-07, "loss": 0.2953, "step": 2820 }, { "epoch": 19.65, "learning_rate": 3.2374825174825173e-07, "loss": 0.2886, "step": 2830 }, { "epoch": 19.72, "learning_rate": 3.2346853146853145e-07, "loss": 0.3151, "step": 2840 }, { "epoch": 19.79, "learning_rate": 3.2318881118881116e-07, "loss": 0.2947, "step": 2850 }, { "epoch": 19.86, "learning_rate": 3.229090909090909e-07, "loss": 0.2835, "step": 2860 }, { "epoch": 19.93, "learning_rate": 3.226293706293706e-07, "loss": 0.2952, "step": 2870 }, { "epoch": 20.0, "learning_rate": 3.2234965034965036e-07, "loss": 0.2837, "step": 2880 }, { "epoch": 20.0, "eval_loss": 0.465202271938324, "eval_runtime": 126.7842, "eval_samples_per_second": 4.196, "eval_steps_per_second": 0.528, "eval_wer": 0.2110304043195989, "step": 2880 }, { "epoch": 20.07, "learning_rate": 3.2206993006993007e-07, "loss": 0.3054, "step": 2890 }, { "epoch": 20.14, "learning_rate": 3.217902097902098e-07, "loss": 0.2852, "step": 2900 }, { "epoch": 20.21, "learning_rate": 3.215104895104895e-07, "loss": 0.2852, "step": 2910 }, { "epoch": 20.28, "learning_rate": 3.212307692307692e-07, "loss": 0.2838, "step": 2920 }, { "epoch": 20.35, "learning_rate": 3.209510489510489e-07, "loss": 0.2799, "step": 2930 }, { "epoch": 20.42, "learning_rate": 3.2067132867132864e-07, "loss": 0.2907, "step": 2940 }, { "epoch": 20.48, "learning_rate": 3.2039160839160835e-07, "loss": 0.3084, "step": 2950 }, { "epoch": 20.55, "learning_rate": 3.201118881118881e-07, "loss": 0.2843, "step": 2960 }, { "epoch": 20.62, "learning_rate": 3.1983216783216784e-07, "loss": 0.2889, "step": 2970 }, { "epoch": 20.69, "learning_rate": 3.1955244755244755e-07, "loss": 0.2921, "step": 2980 }, { "epoch": 20.76, "learning_rate": 3.1927272727272726e-07, "loss": 0.3003, "step": 2990 }, { "epoch": 20.83, "learning_rate": 3.18993006993007e-07, "loss": 0.3006, "step": 3000 }, { "epoch": 20.9, "learning_rate": 3.187132867132867e-07, "loss": 0.2818, "step": 3010 }, { "epoch": 20.97, "learning_rate": 3.184335664335664e-07, "loss": 0.3145, "step": 3020 }, { "epoch": 21.0, "eval_loss": 0.4642258286476135, "eval_runtime": 127.0789, "eval_samples_per_second": 4.186, "eval_steps_per_second": 0.527, "eval_wer": 0.20987336890145916, "step": 3024 }, { "epoch": 21.04, "learning_rate": 3.181538461538461e-07, "loss": 0.3144, "step": 3030 }, { "epoch": 21.11, "learning_rate": 3.1787412587412583e-07, "loss": 0.2878, "step": 3040 }, { "epoch": 21.18, "learning_rate": 3.175944055944056e-07, "loss": 0.3117, "step": 3050 }, { "epoch": 21.25, "learning_rate": 3.173146853146853e-07, "loss": 0.307, "step": 3060 }, { "epoch": 21.32, "learning_rate": 3.1703496503496503e-07, "loss": 0.302, "step": 3070 }, { "epoch": 21.39, "learning_rate": 3.1675524475524474e-07, "loss": 0.3095, "step": 3080 }, { "epoch": 21.46, "learning_rate": 3.1647552447552446e-07, "loss": 0.315, "step": 3090 }, { "epoch": 21.53, "learning_rate": 3.1619580419580417e-07, "loss": 0.2889, "step": 3100 }, { "epoch": 21.6, "learning_rate": 3.159160839160839e-07, "loss": 0.2958, "step": 3110 }, { "epoch": 21.66, "learning_rate": 3.156363636363636e-07, "loss": 0.2688, "step": 3120 }, { "epoch": 21.73, "learning_rate": 3.153566433566433e-07, "loss": 0.2917, "step": 3130 }, { "epoch": 21.8, "learning_rate": 3.150769230769231e-07, "loss": 0.2888, "step": 3140 }, { "epoch": 21.87, "learning_rate": 3.147972027972028e-07, "loss": 0.3024, "step": 3150 }, { "epoch": 21.94, "learning_rate": 3.145174825174825e-07, "loss": 0.3133, "step": 3160 }, { "epoch": 22.0, "eval_loss": 0.45945289731025696, "eval_runtime": 127.5148, "eval_samples_per_second": 4.172, "eval_steps_per_second": 0.525, "eval_wer": 0.20929485119238928, "step": 3168 }, { "epoch": 22.01, "learning_rate": 3.142377622377622e-07, "loss": 0.3137, "step": 3170 }, { "epoch": 22.08, "learning_rate": 3.13958041958042e-07, "loss": 0.2997, "step": 3180 }, { "epoch": 22.15, "learning_rate": 3.1367832167832165e-07, "loss": 0.277, "step": 3190 }, { "epoch": 22.22, "learning_rate": 3.1339860139860136e-07, "loss": 0.3171, "step": 3200 }, { "epoch": 22.29, "learning_rate": 3.131188811188811e-07, "loss": 0.2843, "step": 3210 }, { "epoch": 22.36, "learning_rate": 3.1283916083916085e-07, "loss": 0.3043, "step": 3220 }, { "epoch": 22.43, "learning_rate": 3.1255944055944056e-07, "loss": 0.3131, "step": 3230 }, { "epoch": 22.5, "learning_rate": 3.122797202797203e-07, "loss": 0.311, "step": 3240 }, { "epoch": 22.57, "learning_rate": 3.12e-07, "loss": 0.3142, "step": 3250 }, { "epoch": 22.64, "learning_rate": 3.117202797202797e-07, "loss": 0.291, "step": 3260 }, { "epoch": 22.71, "learning_rate": 3.1144055944055947e-07, "loss": 0.2866, "step": 3270 }, { "epoch": 22.78, "learning_rate": 3.1116083916083913e-07, "loss": 0.3005, "step": 3280 }, { "epoch": 22.84, "learning_rate": 3.1088111888111884e-07, "loss": 0.2894, "step": 3290 }, { "epoch": 22.91, "learning_rate": 3.1060139860139856e-07, "loss": 0.2865, "step": 3300 }, { "epoch": 22.98, "learning_rate": 3.103216783216783e-07, "loss": 0.312, "step": 3310 }, { "epoch": 23.0, "eval_loss": 0.4610751271247864, "eval_runtime": 126.3842, "eval_samples_per_second": 4.209, "eval_steps_per_second": 0.53, "eval_wer": 0.20884489297422382, "step": 3312 }, { "epoch": 23.06, "learning_rate": 3.1004195804195804e-07, "loss": 0.2988, "step": 3320 }, { "epoch": 23.12, "learning_rate": 3.0976223776223775e-07, "loss": 0.3087, "step": 3330 }, { "epoch": 23.19, "learning_rate": 3.0948251748251747e-07, "loss": 0.3078, "step": 3340 }, { "epoch": 23.26, "learning_rate": 3.092027972027972e-07, "loss": 0.305, "step": 3350 }, { "epoch": 23.33, "learning_rate": 3.0892307692307695e-07, "loss": 0.2925, "step": 3360 }, { "epoch": 23.4, "learning_rate": 3.086433566433566e-07, "loss": 0.3025, "step": 3370 }, { "epoch": 23.47, "learning_rate": 3.083636363636363e-07, "loss": 0.3223, "step": 3380 }, { "epoch": 23.54, "learning_rate": 3.0808391608391604e-07, "loss": 0.3101, "step": 3390 }, { "epoch": 23.61, "learning_rate": 3.078041958041958e-07, "loss": 0.3041, "step": 3400 }, { "epoch": 23.68, "learning_rate": 3.075244755244755e-07, "loss": 0.2927, "step": 3410 }, { "epoch": 23.75, "learning_rate": 3.0724475524475523e-07, "loss": 0.3022, "step": 3420 }, { "epoch": 23.82, "learning_rate": 3.0696503496503495e-07, "loss": 0.2916, "step": 3430 }, { "epoch": 23.89, "learning_rate": 3.066853146853147e-07, "loss": 0.2737, "step": 3440 }, { "epoch": 23.96, "learning_rate": 3.064055944055944e-07, "loss": 0.2943, "step": 3450 }, { "epoch": 24.0, "eval_loss": 0.46063411235809326, "eval_runtime": 126.883, "eval_samples_per_second": 4.193, "eval_steps_per_second": 0.528, "eval_wer": 0.20903773221058045, "step": 3456 }, { "epoch": 24.03, "learning_rate": 3.061258741258741e-07, "loss": 0.2966, "step": 3460 }, { "epoch": 24.1, "learning_rate": 3.058461538461538e-07, "loss": 0.3, "step": 3470 }, { "epoch": 24.17, "learning_rate": 3.0556643356643357e-07, "loss": 0.3026, "step": 3480 }, { "epoch": 24.24, "learning_rate": 3.052867132867133e-07, "loss": 0.3146, "step": 3490 }, { "epoch": 24.3, "learning_rate": 3.05006993006993e-07, "loss": 0.2811, "step": 3500 }, { "epoch": 24.37, "learning_rate": 3.047272727272727e-07, "loss": 0.2874, "step": 3510 }, { "epoch": 24.44, "learning_rate": 3.0444755244755243e-07, "loss": 0.3108, "step": 3520 }, { "epoch": 24.51, "learning_rate": 3.041678321678322e-07, "loss": 0.3099, "step": 3530 }, { "epoch": 24.58, "learning_rate": 3.0388811188811185e-07, "loss": 0.301, "step": 3540 }, { "epoch": 24.65, "learning_rate": 3.0360839160839157e-07, "loss": 0.2939, "step": 3550 }, { "epoch": 24.72, "learning_rate": 3.033286713286713e-07, "loss": 0.3018, "step": 3560 }, { "epoch": 24.79, "learning_rate": 3.0304895104895105e-07, "loss": 0.3083, "step": 3570 }, { "epoch": 24.86, "learning_rate": 3.0276923076923076e-07, "loss": 0.2959, "step": 3580 }, { "epoch": 24.93, "learning_rate": 3.024895104895105e-07, "loss": 0.2838, "step": 3590 }, { "epoch": 25.0, "learning_rate": 3.022097902097902e-07, "loss": 0.3068, "step": 3600 }, { "epoch": 25.0, "eval_loss": 0.4597414433956146, "eval_runtime": 126.1713, "eval_samples_per_second": 4.216, "eval_steps_per_second": 0.531, "eval_wer": 0.20839493475605836, "step": 3600 }, { "epoch": 25.07, "learning_rate": 3.019300699300699e-07, "loss": 0.3236, "step": 3610 }, { "epoch": 25.14, "learning_rate": 3.0165034965034967e-07, "loss": 0.2917, "step": 3620 }, { "epoch": 25.21, "learning_rate": 3.0137062937062933e-07, "loss": 0.301, "step": 3630 }, { "epoch": 25.28, "learning_rate": 3.0109090909090905e-07, "loss": 0.2912, "step": 3640 }, { "epoch": 25.35, "learning_rate": 3.0081118881118876e-07, "loss": 0.2865, "step": 3650 }, { "epoch": 25.42, "learning_rate": 3.0053146853146853e-07, "loss": 0.3039, "step": 3660 }, { "epoch": 25.48, "learning_rate": 3.0025174825174824e-07, "loss": 0.308, "step": 3670 }, { "epoch": 25.55, "learning_rate": 2.9997202797202796e-07, "loss": 0.2881, "step": 3680 }, { "epoch": 25.62, "learning_rate": 2.9969230769230767e-07, "loss": 0.2999, "step": 3690 }, { "epoch": 25.69, "learning_rate": 2.9941258741258744e-07, "loss": 0.3059, "step": 3700 }, { "epoch": 25.76, "learning_rate": 2.9913286713286715e-07, "loss": 0.2872, "step": 3710 }, { "epoch": 25.83, "learning_rate": 2.988531468531468e-07, "loss": 0.3065, "step": 3720 }, { "epoch": 25.9, "learning_rate": 2.9857342657342653e-07, "loss": 0.2827, "step": 3730 }, { "epoch": 25.97, "learning_rate": 2.982937062937063e-07, "loss": 0.3184, "step": 3740 }, { "epoch": 26.0, "eval_loss": 0.46003487706184387, "eval_runtime": 127.1609, "eval_samples_per_second": 4.184, "eval_steps_per_second": 0.527, "eval_wer": 0.20890917271967602, "step": 3744 }, { "epoch": 26.04, "learning_rate": 2.98013986013986e-07, "loss": 0.3046, "step": 3750 }, { "epoch": 26.11, "learning_rate": 2.977342657342657e-07, "loss": 0.2901, "step": 3760 }, { "epoch": 26.18, "learning_rate": 2.9745454545454544e-07, "loss": 0.3013, "step": 3770 }, { "epoch": 26.25, "learning_rate": 2.9717482517482515e-07, "loss": 0.3061, "step": 3780 }, { "epoch": 26.32, "learning_rate": 2.968951048951049e-07, "loss": 0.2874, "step": 3790 }, { "epoch": 26.39, "learning_rate": 2.9661538461538463e-07, "loss": 0.2932, "step": 3800 }, { "epoch": 26.46, "learning_rate": 2.963356643356643e-07, "loss": 0.2944, "step": 3810 }, { "epoch": 26.53, "learning_rate": 2.96055944055944e-07, "loss": 0.295, "step": 3820 }, { "epoch": 26.6, "learning_rate": 2.957762237762238e-07, "loss": 0.2887, "step": 3830 }, { "epoch": 26.66, "learning_rate": 2.954965034965035e-07, "loss": 0.2922, "step": 3840 }, { "epoch": 26.73, "learning_rate": 2.952167832167832e-07, "loss": 0.3059, "step": 3850 }, { "epoch": 26.8, "learning_rate": 2.949370629370629e-07, "loss": 0.2966, "step": 3860 }, { "epoch": 26.87, "learning_rate": 2.9465734265734263e-07, "loss": 0.2792, "step": 3870 }, { "epoch": 26.94, "learning_rate": 2.943776223776224e-07, "loss": 0.3035, "step": 3880 }, { "epoch": 27.0, "eval_loss": 0.45960915088653564, "eval_runtime": 127.2463, "eval_samples_per_second": 4.181, "eval_steps_per_second": 0.527, "eval_wer": 0.2078164170469885, "step": 3888 }, { "epoch": 27.01, "learning_rate": 2.940979020979021e-07, "loss": 0.3181, "step": 3890 }, { "epoch": 27.08, "learning_rate": 2.9381818181818177e-07, "loss": 0.2892, "step": 3900 }, { "epoch": 27.15, "learning_rate": 2.935384615384615e-07, "loss": 0.2849, "step": 3910 }, { "epoch": 27.22, "learning_rate": 2.9325874125874125e-07, "loss": 0.3077, "step": 3920 }, { "epoch": 27.29, "learning_rate": 2.9297902097902097e-07, "loss": 0.2961, "step": 3930 }, { "epoch": 27.36, "learning_rate": 2.926993006993007e-07, "loss": 0.284, "step": 3940 }, { "epoch": 27.43, "learning_rate": 2.924195804195804e-07, "loss": 0.3077, "step": 3950 }, { "epoch": 27.5, "learning_rate": 2.9213986013986016e-07, "loss": 0.2836, "step": 3960 }, { "epoch": 27.57, "learning_rate": 2.918601398601399e-07, "loss": 0.2949, "step": 3970 }, { "epoch": 27.64, "learning_rate": 2.9158041958041954e-07, "loss": 0.2811, "step": 3980 }, { "epoch": 27.71, "learning_rate": 2.9130069930069925e-07, "loss": 0.3114, "step": 3990 }, { "epoch": 27.78, "learning_rate": 2.91020979020979e-07, "loss": 0.2982, "step": 4000 }, { "epoch": 27.84, "learning_rate": 2.9074125874125873e-07, "loss": 0.2915, "step": 4010 }, { "epoch": 27.91, "learning_rate": 2.9046153846153845e-07, "loss": 0.3045, "step": 4020 }, { "epoch": 27.98, "learning_rate": 2.9018181818181816e-07, "loss": 0.2956, "step": 4030 }, { "epoch": 28.0, "eval_loss": 0.4607580602169037, "eval_runtime": 126.0154, "eval_samples_per_second": 4.222, "eval_steps_per_second": 0.532, "eval_wer": 0.2085234942469628, "step": 4032 }, { "epoch": 28.06, "learning_rate": 2.899020979020979e-07, "loss": 0.312, "step": 4040 }, { "epoch": 28.12, "learning_rate": 2.8962237762237764e-07, "loss": 0.2945, "step": 4050 }, { "epoch": 28.19, "learning_rate": 2.8934265734265736e-07, "loss": 0.3022, "step": 4060 }, { "epoch": 28.26, "learning_rate": 2.89062937062937e-07, "loss": 0.3147, "step": 4070 }, { "epoch": 28.33, "learning_rate": 2.8878321678321673e-07, "loss": 0.2926, "step": 4080 }, { "epoch": 28.4, "learning_rate": 2.885034965034965e-07, "loss": 0.2961, "step": 4090 }, { "epoch": 28.47, "learning_rate": 2.882237762237762e-07, "loss": 0.2927, "step": 4100 }, { "epoch": 28.54, "learning_rate": 2.8794405594405593e-07, "loss": 0.3053, "step": 4110 }, { "epoch": 28.61, "learning_rate": 2.8766433566433564e-07, "loss": 0.296, "step": 4120 }, { "epoch": 28.68, "learning_rate": 2.8738461538461536e-07, "loss": 0.3038, "step": 4130 }, { "epoch": 28.75, "learning_rate": 2.871048951048951e-07, "loss": 0.3131, "step": 4140 }, { "epoch": 28.82, "learning_rate": 2.8682517482517484e-07, "loss": 0.2955, "step": 4150 }, { "epoch": 28.89, "learning_rate": 2.865454545454545e-07, "loss": 0.2809, "step": 4160 }, { "epoch": 28.96, "learning_rate": 2.862657342657342e-07, "loss": 0.3074, "step": 4170 }, { "epoch": 29.0, "eval_loss": 0.46335652470588684, "eval_runtime": 126.143, "eval_samples_per_second": 4.217, "eval_steps_per_second": 0.531, "eval_wer": 0.20820209551970173, "step": 4176 }, { "epoch": 29.03, "learning_rate": 2.85986013986014e-07, "loss": 0.283, "step": 4180 }, { "epoch": 29.1, "learning_rate": 2.857062937062937e-07, "loss": 0.2961, "step": 4190 }, { "epoch": 29.17, "learning_rate": 2.854265734265734e-07, "loss": 0.2871, "step": 4200 }, { "epoch": 29.24, "learning_rate": 2.851468531468531e-07, "loss": 0.3004, "step": 4210 }, { "epoch": 29.3, "learning_rate": 2.848671328671329e-07, "loss": 0.2916, "step": 4220 }, { "epoch": 29.37, "learning_rate": 2.845874125874126e-07, "loss": 0.2975, "step": 4230 }, { "epoch": 29.44, "learning_rate": 2.843076923076923e-07, "loss": 0.3051, "step": 4240 }, { "epoch": 29.51, "learning_rate": 2.84027972027972e-07, "loss": 0.3001, "step": 4250 }, { "epoch": 29.58, "learning_rate": 2.8374825174825174e-07, "loss": 0.2934, "step": 4260 }, { "epoch": 29.65, "learning_rate": 2.8346853146853146e-07, "loss": 0.289, "step": 4270 }, { "epoch": 29.72, "learning_rate": 2.8318881118881117e-07, "loss": 0.3126, "step": 4280 }, { "epoch": 29.79, "learning_rate": 2.829090909090909e-07, "loss": 0.3007, "step": 4290 }, { "epoch": 29.86, "learning_rate": 2.826293706293706e-07, "loss": 0.2952, "step": 4300 }, { "epoch": 29.93, "learning_rate": 2.8234965034965037e-07, "loss": 0.3051, "step": 4310 }, { "epoch": 30.0, "learning_rate": 2.820699300699301e-07, "loss": 0.309, "step": 4320 }, { "epoch": 30.0, "eval_loss": 0.46158549189567566, "eval_runtime": 126.5264, "eval_samples_per_second": 4.205, "eval_steps_per_second": 0.53, "eval_wer": 0.2072378993379186, "step": 4320 }, { "epoch": 30.07, "learning_rate": 2.817902097902098e-07, "loss": 0.3108, "step": 4330 }, { "epoch": 30.14, "learning_rate": 2.8151048951048946e-07, "loss": 0.2876, "step": 4340 }, { "epoch": 30.21, "learning_rate": 2.812307692307692e-07, "loss": 0.3085, "step": 4350 }, { "epoch": 30.28, "learning_rate": 2.8095104895104894e-07, "loss": 0.3017, "step": 4360 }, { "epoch": 30.35, "learning_rate": 2.8067132867132865e-07, "loss": 0.295, "step": 4370 }, { "epoch": 30.42, "learning_rate": 2.8039160839160837e-07, "loss": 0.2871, "step": 4380 }, { "epoch": 30.48, "learning_rate": 2.801118881118881e-07, "loss": 0.3074, "step": 4390 }, { "epoch": 30.55, "learning_rate": 2.7983216783216785e-07, "loss": 0.2963, "step": 4400 }, { "epoch": 30.62, "learning_rate": 2.7955244755244756e-07, "loss": 0.2838, "step": 4410 }, { "epoch": 30.69, "learning_rate": 2.792727272727273e-07, "loss": 0.3026, "step": 4420 }, { "epoch": 30.76, "learning_rate": 2.7899300699300694e-07, "loss": 0.3148, "step": 4430 }, { "epoch": 30.83, "learning_rate": 2.787132867132867e-07, "loss": 0.283, "step": 4440 }, { "epoch": 30.9, "learning_rate": 2.784335664335664e-07, "loss": 0.2913, "step": 4450 }, { "epoch": 30.97, "learning_rate": 2.7815384615384613e-07, "loss": 0.3048, "step": 4460 }, { "epoch": 31.0, "eval_loss": 0.4597069025039673, "eval_runtime": 126.9241, "eval_samples_per_second": 4.191, "eval_steps_per_second": 0.528, "eval_wer": 0.20691650061065758, "step": 4464 }, { "epoch": 31.04, "learning_rate": 2.7787412587412585e-07, "loss": 0.3093, "step": 4470 }, { "epoch": 31.11, "learning_rate": 2.775944055944056e-07, "loss": 0.2973, "step": 4480 }, { "epoch": 31.18, "learning_rate": 2.7731468531468533e-07, "loss": 0.2923, "step": 4490 }, { "epoch": 31.25, "learning_rate": 2.7703496503496504e-07, "loss": 0.3049, "step": 4500 }, { "epoch": 31.32, "learning_rate": 2.767552447552447e-07, "loss": 0.2851, "step": 4510 }, { "epoch": 31.39, "learning_rate": 2.7647552447552447e-07, "loss": 0.2647, "step": 4520 }, { "epoch": 31.46, "learning_rate": 2.761958041958042e-07, "loss": 0.2888, "step": 4530 }, { "epoch": 31.53, "learning_rate": 2.759160839160839e-07, "loss": 0.2903, "step": 4540 }, { "epoch": 31.6, "learning_rate": 2.756363636363636e-07, "loss": 0.2963, "step": 4550 }, { "epoch": 31.66, "learning_rate": 2.753566433566433e-07, "loss": 0.2874, "step": 4560 }, { "epoch": 31.73, "learning_rate": 2.750769230769231e-07, "loss": 0.3036, "step": 4570 }, { "epoch": 31.8, "learning_rate": 2.747972027972028e-07, "loss": 0.2977, "step": 4580 }, { "epoch": 31.87, "learning_rate": 2.745174825174825e-07, "loss": 0.2818, "step": 4590 }, { "epoch": 31.94, "learning_rate": 2.742377622377622e-07, "loss": 0.3146, "step": 4600 }, { "epoch": 32.0, "eval_loss": 0.46195778250694275, "eval_runtime": 126.7492, "eval_samples_per_second": 4.197, "eval_steps_per_second": 0.529, "eval_wer": 0.20691650061065758, "step": 4608 }, { "epoch": 32.01, "learning_rate": 2.7395804195804195e-07, "loss": 0.3153, "step": 4610 }, { "epoch": 32.08, "learning_rate": 2.7367832167832166e-07, "loss": 0.3078, "step": 4620 }, { "epoch": 32.15, "learning_rate": 2.733986013986014e-07, "loss": 0.2825, "step": 4630 }, { "epoch": 32.22, "learning_rate": 2.731188811188811e-07, "loss": 0.3108, "step": 4640 }, { "epoch": 32.29, "learning_rate": 2.728391608391608e-07, "loss": 0.3042, "step": 4650 }, { "epoch": 32.36, "learning_rate": 2.7255944055944057e-07, "loss": 0.2969, "step": 4660 }, { "epoch": 32.43, "learning_rate": 2.722797202797203e-07, "loss": 0.3099, "step": 4670 }, { "epoch": 32.5, "learning_rate": 2.72e-07, "loss": 0.2903, "step": 4680 }, { "epoch": 32.57, "learning_rate": 2.717202797202797e-07, "loss": 0.2808, "step": 4690 }, { "epoch": 32.64, "learning_rate": 2.7144055944055943e-07, "loss": 0.2856, "step": 4700 }, { "epoch": 32.71, "learning_rate": 2.7116083916083914e-07, "loss": 0.2983, "step": 4710 }, { "epoch": 32.78, "learning_rate": 2.7088111888111886e-07, "loss": 0.2949, "step": 4720 }, { "epoch": 32.84, "learning_rate": 2.7060139860139857e-07, "loss": 0.2881, "step": 4730 }, { "epoch": 32.91, "learning_rate": 2.7032167832167834e-07, "loss": 0.2735, "step": 4740 }, { "epoch": 32.98, "learning_rate": 2.7004195804195805e-07, "loss": 0.3122, "step": 4750 }, { "epoch": 33.0, "eval_loss": 0.46265074610710144, "eval_runtime": 126.7804, "eval_samples_per_second": 4.196, "eval_steps_per_second": 0.528, "eval_wer": 0.20698078035610978, "step": 4752 }, { "epoch": 33.06, "learning_rate": 2.6976223776223777e-07, "loss": 0.3178, "step": 4760 }, { "epoch": 33.12, "learning_rate": 2.694825174825175e-07, "loss": 0.2822, "step": 4770 }, { "epoch": 33.19, "learning_rate": 2.692027972027972e-07, "loss": 0.3019, "step": 4780 }, { "epoch": 33.26, "learning_rate": 2.689230769230769e-07, "loss": 0.3021, "step": 4790 }, { "epoch": 33.33, "learning_rate": 2.686433566433566e-07, "loss": 0.2948, "step": 4800 }, { "epoch": 33.4, "learning_rate": 2.6836363636363634e-07, "loss": 0.3007, "step": 4810 }, { "epoch": 33.47, "learning_rate": 2.6808391608391605e-07, "loss": 0.3161, "step": 4820 }, { "epoch": 33.54, "learning_rate": 2.678041958041958e-07, "loss": 0.3086, "step": 4830 }, { "epoch": 33.61, "learning_rate": 2.6752447552447553e-07, "loss": 0.2906, "step": 4840 }, { "epoch": 33.68, "learning_rate": 2.6724475524475525e-07, "loss": 0.2821, "step": 4850 }, { "epoch": 33.75, "learning_rate": 2.6696503496503496e-07, "loss": 0.3027, "step": 4860 }, { "epoch": 33.82, "learning_rate": 2.666853146853147e-07, "loss": 0.2884, "step": 4870 }, { "epoch": 33.89, "learning_rate": 2.664055944055944e-07, "loss": 0.2798, "step": 4880 }, { "epoch": 33.96, "learning_rate": 2.661258741258741e-07, "loss": 0.3014, "step": 4890 }, { "epoch": 34.0, "eval_loss": 0.46091532707214355, "eval_runtime": 126.6187, "eval_samples_per_second": 4.202, "eval_steps_per_second": 0.529, "eval_wer": 0.20672366137430095, "step": 4896 }, { "epoch": 34.03, "learning_rate": 2.658461538461538e-07, "loss": 0.3195, "step": 4900 }, { "epoch": 34.1, "learning_rate": 2.655664335664336e-07, "loss": 0.2988, "step": 4910 }, { "epoch": 34.17, "learning_rate": 2.652867132867133e-07, "loss": 0.2869, "step": 4920 }, { "epoch": 34.24, "learning_rate": 2.65006993006993e-07, "loss": 0.3116, "step": 4930 }, { "epoch": 34.3, "learning_rate": 2.647272727272727e-07, "loss": 0.2858, "step": 4940 }, { "epoch": 34.37, "learning_rate": 2.6444755244755244e-07, "loss": 0.2839, "step": 4950 }, { "epoch": 34.44, "learning_rate": 2.6416783216783215e-07, "loss": 0.3174, "step": 4960 }, { "epoch": 34.51, "learning_rate": 2.6388811188811187e-07, "loss": 0.2844, "step": 4970 }, { "epoch": 34.58, "learning_rate": 2.636083916083916e-07, "loss": 0.288, "step": 4980 }, { "epoch": 34.65, "learning_rate": 2.633286713286713e-07, "loss": 0.2767, "step": 4990 }, { "epoch": 34.72, "learning_rate": 2.6304895104895106e-07, "loss": 0.3234, "step": 5000 }, { "epoch": 34.79, "learning_rate": 2.627692307692308e-07, "loss": 0.3026, "step": 5010 }, { "epoch": 34.86, "learning_rate": 2.624895104895105e-07, "loss": 0.281, "step": 5020 }, { "epoch": 34.93, "learning_rate": 2.622097902097902e-07, "loss": 0.2852, "step": 5030 }, { "epoch": 35.0, "learning_rate": 2.619300699300699e-07, "loss": 0.3041, "step": 5040 }, { "epoch": 35.0, "eval_loss": 0.4620986878871918, "eval_runtime": 125.8849, "eval_samples_per_second": 4.226, "eval_steps_per_second": 0.532, "eval_wer": 0.2061451436652311, "step": 5040 }, { "epoch": 35.07, "learning_rate": 2.6165034965034963e-07, "loss": 0.3148, "step": 5050 }, { "epoch": 35.14, "learning_rate": 2.6137062937062935e-07, "loss": 0.2896, "step": 5060 }, { "epoch": 35.21, "learning_rate": 2.6109090909090906e-07, "loss": 0.2967, "step": 5070 }, { "epoch": 35.28, "learning_rate": 2.608111888111888e-07, "loss": 0.3021, "step": 5080 }, { "epoch": 35.35, "learning_rate": 2.6053146853146854e-07, "loss": 0.3001, "step": 5090 }, { "epoch": 35.42, "learning_rate": 2.6025174825174826e-07, "loss": 0.2915, "step": 5100 }, { "epoch": 35.48, "learning_rate": 2.5997202797202797e-07, "loss": 0.3184, "step": 5110 }, { "epoch": 35.55, "learning_rate": 2.596923076923077e-07, "loss": 0.2867, "step": 5120 }, { "epoch": 35.62, "learning_rate": 2.594125874125874e-07, "loss": 0.2849, "step": 5130 }, { "epoch": 35.69, "learning_rate": 2.591328671328671e-07, "loss": 0.3049, "step": 5140 }, { "epoch": 35.76, "learning_rate": 2.5885314685314683e-07, "loss": 0.2988, "step": 5150 }, { "epoch": 35.83, "learning_rate": 2.5857342657342654e-07, "loss": 0.2985, "step": 5160 }, { "epoch": 35.9, "learning_rate": 2.582937062937063e-07, "loss": 0.2941, "step": 5170 }, { "epoch": 35.97, "learning_rate": 2.58013986013986e-07, "loss": 0.295, "step": 5180 }, { "epoch": 36.0, "eval_loss": 0.4618644416332245, "eval_runtime": 127.2375, "eval_samples_per_second": 4.181, "eval_steps_per_second": 0.527, "eval_wer": 0.2061451436652311, "step": 5184 }, { "epoch": 36.04, "learning_rate": 2.5773426573426574e-07, "loss": 0.3072, "step": 5190 }, { "epoch": 36.11, "learning_rate": 2.5745454545454545e-07, "loss": 0.2962, "step": 5200 }, { "epoch": 36.18, "learning_rate": 2.5717482517482516e-07, "loss": 0.2939, "step": 5210 }, { "epoch": 36.25, "learning_rate": 2.568951048951049e-07, "loss": 0.2963, "step": 5220 }, { "epoch": 36.32, "learning_rate": 2.566153846153846e-07, "loss": 0.2956, "step": 5230 }, { "epoch": 36.39, "learning_rate": 2.563356643356643e-07, "loss": 0.2715, "step": 5240 }, { "epoch": 36.46, "learning_rate": 2.56055944055944e-07, "loss": 0.325, "step": 5250 }, { "epoch": 36.53, "learning_rate": 2.557762237762238e-07, "loss": 0.3027, "step": 5260 }, { "epoch": 36.6, "learning_rate": 2.554965034965035e-07, "loss": 0.2875, "step": 5270 }, { "epoch": 36.66, "learning_rate": 2.552167832167832e-07, "loss": 0.3137, "step": 5280 }, { "epoch": 36.73, "learning_rate": 2.5493706293706293e-07, "loss": 0.2969, "step": 5290 }, { "epoch": 36.8, "learning_rate": 2.5465734265734264e-07, "loss": 0.2929, "step": 5300 }, { "epoch": 36.87, "learning_rate": 2.5437762237762236e-07, "loss": 0.2954, "step": 5310 }, { "epoch": 36.94, "learning_rate": 2.5409790209790207e-07, "loss": 0.3068, "step": 5320 }, { "epoch": 37.0, "eval_loss": 0.4602077305316925, "eval_runtime": 126.5009, "eval_samples_per_second": 4.206, "eval_steps_per_second": 0.53, "eval_wer": 0.20601658417432667, "step": 5328 }, { "epoch": 37.01, "learning_rate": 2.538181818181818e-07, "loss": 0.317, "step": 5330 }, { "epoch": 37.08, "learning_rate": 2.535384615384615e-07, "loss": 0.3026, "step": 5340 }, { "epoch": 37.15, "learning_rate": 2.5325874125874127e-07, "loss": 0.2981, "step": 5350 }, { "epoch": 37.22, "learning_rate": 2.52979020979021e-07, "loss": 0.3199, "step": 5360 }, { "epoch": 37.29, "learning_rate": 2.526993006993007e-07, "loss": 0.282, "step": 5370 }, { "epoch": 37.36, "learning_rate": 2.524195804195804e-07, "loss": 0.303, "step": 5380 }, { "epoch": 37.43, "learning_rate": 2.521398601398602e-07, "loss": 0.299, "step": 5390 }, { "epoch": 37.5, "learning_rate": 2.5186013986013984e-07, "loss": 0.2958, "step": 5400 }, { "epoch": 37.57, "learning_rate": 2.5158041958041955e-07, "loss": 0.2786, "step": 5410 }, { "epoch": 37.64, "learning_rate": 2.5130069930069927e-07, "loss": 0.2844, "step": 5420 }, { "epoch": 37.71, "learning_rate": 2.5102097902097903e-07, "loss": 0.3002, "step": 5430 }, { "epoch": 37.78, "learning_rate": 2.5074125874125875e-07, "loss": 0.3062, "step": 5440 }, { "epoch": 37.84, "learning_rate": 2.5046153846153846e-07, "loss": 0.2871, "step": 5450 }, { "epoch": 37.91, "learning_rate": 2.501818181818182e-07, "loss": 0.2907, "step": 5460 }, { "epoch": 37.98, "learning_rate": 2.499020979020979e-07, "loss": 0.2993, "step": 5470 }, { "epoch": 38.0, "eval_loss": 0.4617382287979126, "eval_runtime": 126.4352, "eval_samples_per_second": 4.208, "eval_steps_per_second": 0.53, "eval_wer": 0.20685222086520538, "step": 5472 }, { "epoch": 38.06, "learning_rate": 2.496223776223776e-07, "loss": 0.3127, "step": 5480 }, { "epoch": 38.12, "learning_rate": 2.493426573426573e-07, "loss": 0.2939, "step": 5490 }, { "epoch": 38.19, "learning_rate": 2.4906293706293703e-07, "loss": 0.2969, "step": 5500 }, { "epoch": 38.26, "learning_rate": 2.4878321678321674e-07, "loss": 0.281, "step": 5510 }, { "epoch": 38.33, "learning_rate": 2.485034965034965e-07, "loss": 0.2799, "step": 5520 }, { "epoch": 38.4, "learning_rate": 2.482237762237762e-07, "loss": 0.2913, "step": 5530 }, { "epoch": 38.47, "learning_rate": 2.4794405594405594e-07, "loss": 0.3017, "step": 5540 }, { "epoch": 38.54, "learning_rate": 2.4766433566433565e-07, "loss": 0.2833, "step": 5550 }, { "epoch": 38.61, "learning_rate": 2.4738461538461537e-07, "loss": 0.2963, "step": 5560 }, { "epoch": 38.68, "learning_rate": 2.471048951048951e-07, "loss": 0.2998, "step": 5570 }, { "epoch": 38.75, "learning_rate": 2.468251748251748e-07, "loss": 0.3061, "step": 5580 }, { "epoch": 38.82, "learning_rate": 2.465454545454545e-07, "loss": 0.2901, "step": 5590 }, { "epoch": 38.89, "learning_rate": 2.462657342657342e-07, "loss": 0.2712, "step": 5600 }, { "epoch": 38.96, "learning_rate": 2.45986013986014e-07, "loss": 0.3065, "step": 5610 }, { "epoch": 39.0, "eval_loss": 0.4619607925415039, "eval_runtime": 126.1507, "eval_samples_per_second": 4.217, "eval_steps_per_second": 0.531, "eval_wer": 0.2071093398470142, "step": 5616 }, { "epoch": 39.03, "learning_rate": 2.457062937062937e-07, "loss": 0.3096, "step": 5620 }, { "epoch": 39.1, "learning_rate": 2.454265734265734e-07, "loss": 0.292, "step": 5630 }, { "epoch": 39.17, "learning_rate": 2.4514685314685313e-07, "loss": 0.3086, "step": 5640 }, { "epoch": 39.24, "learning_rate": 2.448671328671329e-07, "loss": 0.3047, "step": 5650 }, { "epoch": 39.3, "learning_rate": 2.4458741258741256e-07, "loss": 0.2885, "step": 5660 }, { "epoch": 39.37, "learning_rate": 2.443076923076923e-07, "loss": 0.2756, "step": 5670 }, { "epoch": 39.44, "learning_rate": 2.44027972027972e-07, "loss": 0.2899, "step": 5680 }, { "epoch": 39.51, "learning_rate": 2.4374825174825176e-07, "loss": 0.3047, "step": 5690 }, { "epoch": 39.58, "learning_rate": 2.4346853146853147e-07, "loss": 0.2833, "step": 5700 }, { "epoch": 39.65, "learning_rate": 2.431888111888112e-07, "loss": 0.3028, "step": 5710 }, { "epoch": 39.72, "learning_rate": 2.429090909090909e-07, "loss": 0.2982, "step": 5720 }, { "epoch": 39.79, "learning_rate": 2.426293706293706e-07, "loss": 0.2861, "step": 5730 }, { "epoch": 39.86, "learning_rate": 2.423496503496504e-07, "loss": 0.2889, "step": 5740 }, { "epoch": 39.93, "learning_rate": 2.4206993006993004e-07, "loss": 0.2993, "step": 5750 }, { "epoch": 40.0, "learning_rate": 2.4179020979020976e-07, "loss": 0.3102, "step": 5760 }, { "epoch": 40.0, "eval_loss": 0.4603091776371002, "eval_runtime": 125.9814, "eval_samples_per_second": 4.223, "eval_steps_per_second": 0.532, "eval_wer": 0.20653082213794433, "step": 5760 }, { "epoch": 40.07, "learning_rate": 2.4151048951048947e-07, "loss": 0.3043, "step": 5770 }, { "epoch": 40.14, "learning_rate": 2.4123076923076924e-07, "loss": 0.2954, "step": 5780 }, { "epoch": 40.21, "learning_rate": 2.4095104895104895e-07, "loss": 0.2937, "step": 5790 }, { "epoch": 40.28, "learning_rate": 2.4067132867132866e-07, "loss": 0.2927, "step": 5800 }, { "epoch": 40.35, "learning_rate": 2.403916083916084e-07, "loss": 0.2894, "step": 5810 }, { "epoch": 40.42, "learning_rate": 2.401118881118881e-07, "loss": 0.289, "step": 5820 }, { "epoch": 40.48, "learning_rate": 2.3983216783216786e-07, "loss": 0.3015, "step": 5830 }, { "epoch": 40.55, "learning_rate": 2.395524475524475e-07, "loss": 0.286, "step": 5840 }, { "epoch": 40.62, "learning_rate": 2.3927272727272724e-07, "loss": 0.284, "step": 5850 }, { "epoch": 40.69, "learning_rate": 2.3899300699300695e-07, "loss": 0.2925, "step": 5860 }, { "epoch": 40.76, "learning_rate": 2.387132867132867e-07, "loss": 0.3187, "step": 5870 }, { "epoch": 40.83, "learning_rate": 2.3843356643356643e-07, "loss": 0.2908, "step": 5880 }, { "epoch": 40.9, "learning_rate": 2.3815384615384614e-07, "loss": 0.2959, "step": 5890 }, { "epoch": 40.97, "learning_rate": 2.3787412587412588e-07, "loss": 0.3032, "step": 5900 }, { "epoch": 41.0, "eval_loss": 0.46081069111824036, "eval_runtime": 126.0299, "eval_samples_per_second": 4.221, "eval_steps_per_second": 0.532, "eval_wer": 0.2061451436652311, "step": 5904 }, { "epoch": 41.04, "learning_rate": 2.375944055944056e-07, "loss": 0.3126, "step": 5910 }, { "epoch": 41.11, "learning_rate": 2.373146853146853e-07, "loss": 0.305, "step": 5920 }, { "epoch": 41.18, "learning_rate": 2.37034965034965e-07, "loss": 0.2986, "step": 5930 }, { "epoch": 41.25, "learning_rate": 2.3675524475524474e-07, "loss": 0.2981, "step": 5940 }, { "epoch": 41.32, "learning_rate": 2.3647552447552446e-07, "loss": 0.2907, "step": 5950 }, { "epoch": 41.39, "learning_rate": 2.3619580419580417e-07, "loss": 0.2842, "step": 5960 }, { "epoch": 41.46, "learning_rate": 2.359160839160839e-07, "loss": 0.2852, "step": 5970 }, { "epoch": 41.53, "learning_rate": 2.3563636363636362e-07, "loss": 0.2901, "step": 5980 }, { "epoch": 41.6, "learning_rate": 2.3535664335664336e-07, "loss": 0.3025, "step": 5990 }, { "epoch": 41.66, "learning_rate": 2.3507692307692308e-07, "loss": 0.2994, "step": 6000 }, { "epoch": 41.73, "learning_rate": 2.3479720279720277e-07, "loss": 0.3084, "step": 6010 }, { "epoch": 41.8, "learning_rate": 2.3451748251748248e-07, "loss": 0.2922, "step": 6020 }, { "epoch": 41.87, "learning_rate": 2.3423776223776222e-07, "loss": 0.2717, "step": 6030 }, { "epoch": 41.94, "learning_rate": 2.3395804195804193e-07, "loss": 0.293, "step": 6040 }, { "epoch": 42.0, "eval_loss": 0.46132680773735046, "eval_runtime": 125.9829, "eval_samples_per_second": 4.223, "eval_steps_per_second": 0.532, "eval_wer": 0.20640226264703992, "step": 6048 }, { "epoch": 42.01, "learning_rate": 2.3367832167832168e-07, "loss": 0.3173, "step": 6050 }, { "epoch": 42.08, "learning_rate": 2.333986013986014e-07, "loss": 0.2966, "step": 6060 }, { "epoch": 42.15, "learning_rate": 2.331188811188811e-07, "loss": 0.2848, "step": 6070 }, { "epoch": 42.22, "learning_rate": 2.3283916083916084e-07, "loss": 0.2948, "step": 6080 }, { "epoch": 42.29, "learning_rate": 2.3255944055944056e-07, "loss": 0.292, "step": 6090 }, { "epoch": 42.36, "learning_rate": 2.3227972027972025e-07, "loss": 0.2944, "step": 6100 }, { "epoch": 42.43, "learning_rate": 2.3199999999999999e-07, "loss": 0.3107, "step": 6110 }, { "epoch": 42.5, "learning_rate": 2.317202797202797e-07, "loss": 0.3056, "step": 6120 }, { "epoch": 42.57, "learning_rate": 2.3144055944055941e-07, "loss": 0.2852, "step": 6130 }, { "epoch": 42.64, "learning_rate": 2.3116083916083916e-07, "loss": 0.2817, "step": 6140 }, { "epoch": 42.71, "learning_rate": 2.3088111888111887e-07, "loss": 0.3137, "step": 6150 }, { "epoch": 42.78, "learning_rate": 2.306013986013986e-07, "loss": 0.3, "step": 6160 }, { "epoch": 42.84, "learning_rate": 2.3032167832167832e-07, "loss": 0.2913, "step": 6170 }, { "epoch": 42.91, "learning_rate": 2.3004195804195804e-07, "loss": 0.2908, "step": 6180 }, { "epoch": 42.98, "learning_rate": 2.2976223776223773e-07, "loss": 0.305, "step": 6190 }, { "epoch": 43.0, "eval_loss": 0.46249884366989136, "eval_runtime": 126.0962, "eval_samples_per_second": 4.219, "eval_steps_per_second": 0.531, "eval_wer": 0.20601658417432667, "step": 6192 }, { "epoch": 43.06, "learning_rate": 2.2948251748251747e-07, "loss": 0.3029, "step": 6200 }, { "epoch": 43.12, "learning_rate": 2.2920279720279718e-07, "loss": 0.2848, "step": 6210 }, { "epoch": 43.19, "learning_rate": 2.2892307692307692e-07, "loss": 0.3024, "step": 6220 }, { "epoch": 43.26, "learning_rate": 2.2864335664335663e-07, "loss": 0.2899, "step": 6230 }, { "epoch": 43.33, "learning_rate": 2.2836363636363635e-07, "loss": 0.2946, "step": 6240 }, { "epoch": 43.4, "learning_rate": 2.280839160839161e-07, "loss": 0.2794, "step": 6250 }, { "epoch": 43.47, "learning_rate": 2.278041958041958e-07, "loss": 0.3074, "step": 6260 }, { "epoch": 43.54, "learning_rate": 2.2752447552447554e-07, "loss": 0.3048, "step": 6270 }, { "epoch": 43.61, "learning_rate": 2.272447552447552e-07, "loss": 0.2863, "step": 6280 }, { "epoch": 43.68, "learning_rate": 2.2696503496503495e-07, "loss": 0.2913, "step": 6290 }, { "epoch": 43.75, "learning_rate": 2.2668531468531466e-07, "loss": 0.2947, "step": 6300 }, { "epoch": 43.82, "learning_rate": 2.264055944055944e-07, "loss": 0.2863, "step": 6310 }, { "epoch": 43.89, "learning_rate": 2.2612587412587411e-07, "loss": 0.2909, "step": 6320 }, { "epoch": 43.96, "learning_rate": 2.2584615384615385e-07, "loss": 0.3108, "step": 6330 }, { "epoch": 44.0, "eval_loss": 0.4616779088973999, "eval_runtime": 126.2353, "eval_samples_per_second": 4.214, "eval_steps_per_second": 0.531, "eval_wer": 0.20428103104711706, "step": 6336 }, { "epoch": 44.03, "learning_rate": 2.2556643356643357e-07, "loss": 0.2984, "step": 6340 }, { "epoch": 44.1, "learning_rate": 2.2528671328671328e-07, "loss": 0.2872, "step": 6350 }, { "epoch": 44.17, "learning_rate": 2.2500699300699302e-07, "loss": 0.3015, "step": 6360 }, { "epoch": 44.24, "learning_rate": 2.247272727272727e-07, "loss": 0.3125, "step": 6370 }, { "epoch": 44.3, "learning_rate": 2.2444755244755243e-07, "loss": 0.2854, "step": 6380 }, { "epoch": 44.37, "learning_rate": 2.2416783216783214e-07, "loss": 0.2853, "step": 6390 }, { "epoch": 44.44, "learning_rate": 2.2388811188811188e-07, "loss": 0.2978, "step": 6400 }, { "epoch": 44.51, "learning_rate": 2.236083916083916e-07, "loss": 0.3037, "step": 6410 }, { "epoch": 44.58, "learning_rate": 2.2332867132867133e-07, "loss": 0.2876, "step": 6420 }, { "epoch": 44.65, "learning_rate": 2.2304895104895105e-07, "loss": 0.2853, "step": 6430 }, { "epoch": 44.72, "learning_rate": 2.227692307692308e-07, "loss": 0.2875, "step": 6440 }, { "epoch": 44.79, "learning_rate": 2.224895104895105e-07, "loss": 0.2884, "step": 6450 }, { "epoch": 44.86, "learning_rate": 2.222097902097902e-07, "loss": 0.2798, "step": 6460 }, { "epoch": 44.93, "learning_rate": 2.219300699300699e-07, "loss": 0.3013, "step": 6470 }, { "epoch": 45.0, "learning_rate": 2.2165034965034965e-07, "loss": 0.292, "step": 6480 }, { "epoch": 45.0, "eval_loss": 0.4616601765155792, "eval_runtime": 126.6152, "eval_samples_per_second": 4.202, "eval_steps_per_second": 0.529, "eval_wer": 0.20608086391977887, "step": 6480 }, { "epoch": 45.07, "learning_rate": 2.2137062937062936e-07, "loss": 0.2932, "step": 6490 }, { "epoch": 45.14, "learning_rate": 2.2109090909090907e-07, "loss": 0.2776, "step": 6500 }, { "epoch": 45.21, "learning_rate": 2.2081118881118881e-07, "loss": 0.2935, "step": 6510 }, { "epoch": 45.28, "learning_rate": 2.2053146853146853e-07, "loss": 0.2945, "step": 6520 }, { "epoch": 45.35, "learning_rate": 2.2025174825174827e-07, "loss": 0.2892, "step": 6530 }, { "epoch": 45.42, "learning_rate": 2.1997202797202793e-07, "loss": 0.2921, "step": 6540 }, { "epoch": 45.48, "learning_rate": 2.1969230769230767e-07, "loss": 0.3094, "step": 6550 }, { "epoch": 45.55, "learning_rate": 2.1941258741258738e-07, "loss": 0.294, "step": 6560 }, { "epoch": 45.62, "learning_rate": 2.1913286713286712e-07, "loss": 0.2897, "step": 6570 }, { "epoch": 45.69, "learning_rate": 2.1885314685314684e-07, "loss": 0.3052, "step": 6580 }, { "epoch": 45.76, "learning_rate": 2.1857342657342658e-07, "loss": 0.2988, "step": 6590 }, { "epoch": 45.83, "learning_rate": 2.182937062937063e-07, "loss": 0.295, "step": 6600 }, { "epoch": 45.9, "learning_rate": 2.18013986013986e-07, "loss": 0.2912, "step": 6610 }, { "epoch": 45.97, "learning_rate": 2.1773426573426575e-07, "loss": 0.3066, "step": 6620 }, { "epoch": 46.0, "eval_loss": 0.46197906136512756, "eval_runtime": 126.1594, "eval_samples_per_second": 4.217, "eval_steps_per_second": 0.531, "eval_wer": 0.20511666773799575, "step": 6624 }, { "epoch": 46.04, "learning_rate": 2.1745454545454544e-07, "loss": 0.3109, "step": 6630 }, { "epoch": 46.11, "learning_rate": 2.1717482517482515e-07, "loss": 0.3033, "step": 6640 }, { "epoch": 46.18, "learning_rate": 2.1689510489510486e-07, "loss": 0.2942, "step": 6650 }, { "epoch": 46.25, "learning_rate": 2.166153846153846e-07, "loss": 0.2951, "step": 6660 }, { "epoch": 46.32, "learning_rate": 2.1633566433566432e-07, "loss": 0.2996, "step": 6670 }, { "epoch": 46.39, "learning_rate": 2.1605594405594406e-07, "loss": 0.2852, "step": 6680 }, { "epoch": 46.46, "learning_rate": 2.1577622377622377e-07, "loss": 0.3073, "step": 6690 }, { "epoch": 46.53, "learning_rate": 2.1549650349650351e-07, "loss": 0.2848, "step": 6700 }, { "epoch": 46.6, "learning_rate": 2.1521678321678323e-07, "loss": 0.2856, "step": 6710 }, { "epoch": 46.66, "learning_rate": 2.1493706293706292e-07, "loss": 0.2985, "step": 6720 }, { "epoch": 46.73, "learning_rate": 2.1465734265734263e-07, "loss": 0.2962, "step": 6730 }, { "epoch": 46.8, "learning_rate": 2.1437762237762237e-07, "loss": 0.289, "step": 6740 }, { "epoch": 46.87, "learning_rate": 2.1409790209790208e-07, "loss": 0.2778, "step": 6750 }, { "epoch": 46.94, "learning_rate": 2.138181818181818e-07, "loss": 0.3037, "step": 6760 }, { "epoch": 47.0, "eval_loss": 0.4618070125579834, "eval_runtime": 126.4764, "eval_samples_per_second": 4.206, "eval_steps_per_second": 0.53, "eval_wer": 0.20653082213794433, "step": 6768 }, { "epoch": 47.01, "learning_rate": 2.1353846153846154e-07, "loss": 0.3039, "step": 6770 }, { "epoch": 47.08, "learning_rate": 2.1325874125874125e-07, "loss": 0.3021, "step": 6780 }, { "epoch": 47.15, "learning_rate": 2.12979020979021e-07, "loss": 0.2829, "step": 6790 }, { "epoch": 47.22, "learning_rate": 2.126993006993007e-07, "loss": 0.3079, "step": 6800 }, { "epoch": 47.29, "learning_rate": 2.124195804195804e-07, "loss": 0.286, "step": 6810 }, { "epoch": 47.36, "learning_rate": 2.121398601398601e-07, "loss": 0.2811, "step": 6820 }, { "epoch": 47.43, "learning_rate": 2.1186013986013985e-07, "loss": 0.2944, "step": 6830 }, { "epoch": 47.5, "learning_rate": 2.1158041958041956e-07, "loss": 0.2951, "step": 6840 }, { "epoch": 47.57, "learning_rate": 2.113006993006993e-07, "loss": 0.2947, "step": 6850 }, { "epoch": 47.64, "learning_rate": 2.1102097902097902e-07, "loss": 0.2743, "step": 6860 }, { "epoch": 47.71, "learning_rate": 2.1074125874125873e-07, "loss": 0.2979, "step": 6870 }, { "epoch": 47.78, "learning_rate": 2.1046153846153847e-07, "loss": 0.289, "step": 6880 }, { "epoch": 47.84, "learning_rate": 2.101818181818182e-07, "loss": 0.2882, "step": 6890 }, { "epoch": 47.91, "learning_rate": 2.0990209790209787e-07, "loss": 0.2833, "step": 6900 }, { "epoch": 47.98, "learning_rate": 2.096223776223776e-07, "loss": 0.3058, "step": 6910 }, { "epoch": 48.0, "eval_loss": 0.46171513199806213, "eval_runtime": 126.836, "eval_samples_per_second": 4.194, "eval_steps_per_second": 0.528, "eval_wer": 0.2062737031561355, "step": 6912 }, { "epoch": 48.06, "learning_rate": 2.0934265734265733e-07, "loss": 0.3079, "step": 6920 }, { "epoch": 48.12, "learning_rate": 2.0906293706293704e-07, "loss": 0.3026, "step": 6930 }, { "epoch": 48.19, "learning_rate": 2.0878321678321678e-07, "loss": 0.3062, "step": 6940 }, { "epoch": 48.26, "learning_rate": 2.085034965034965e-07, "loss": 0.3066, "step": 6950 }, { "epoch": 48.33, "learning_rate": 2.0822377622377624e-07, "loss": 0.3013, "step": 6960 }, { "epoch": 48.4, "learning_rate": 2.0794405594405595e-07, "loss": 0.2832, "step": 6970 }, { "epoch": 48.47, "learning_rate": 2.0766433566433567e-07, "loss": 0.2929, "step": 6980 }, { "epoch": 48.54, "learning_rate": 2.0738461538461535e-07, "loss": 0.3027, "step": 6990 }, { "epoch": 48.61, "learning_rate": 2.071048951048951e-07, "loss": 0.2835, "step": 7000 }, { "epoch": 48.68, "learning_rate": 2.068251748251748e-07, "loss": 0.2994, "step": 7010 }, { "epoch": 48.75, "learning_rate": 2.0654545454545452e-07, "loss": 0.292, "step": 7020 }, { "epoch": 48.82, "learning_rate": 2.0626573426573426e-07, "loss": 0.283, "step": 7030 }, { "epoch": 48.89, "learning_rate": 2.0598601398601398e-07, "loss": 0.2803, "step": 7040 }, { "epoch": 48.96, "learning_rate": 2.0570629370629372e-07, "loss": 0.3054, "step": 7050 }, { "epoch": 49.0, "eval_loss": 0.4615088999271393, "eval_runtime": 126.8007, "eval_samples_per_second": 4.196, "eval_steps_per_second": 0.528, "eval_wer": 0.20640226264703992, "step": 7056 }, { "epoch": 49.03, "learning_rate": 2.0542657342657343e-07, "loss": 0.2956, "step": 7060 }, { "epoch": 49.1, "learning_rate": 2.0514685314685312e-07, "loss": 0.295, "step": 7070 }, { "epoch": 49.17, "learning_rate": 2.0486713286713283e-07, "loss": 0.2949, "step": 7080 }, { "epoch": 49.24, "learning_rate": 2.0458741258741257e-07, "loss": 0.3004, "step": 7090 }, { "epoch": 49.3, "learning_rate": 2.043076923076923e-07, "loss": 0.2994, "step": 7100 }, { "epoch": 49.37, "learning_rate": 2.0402797202797203e-07, "loss": 0.2846, "step": 7110 }, { "epoch": 49.44, "learning_rate": 2.0374825174825174e-07, "loss": 0.2832, "step": 7120 }, { "epoch": 49.51, "learning_rate": 2.0346853146853146e-07, "loss": 0.2952, "step": 7130 }, { "epoch": 49.58, "learning_rate": 2.031888111888112e-07, "loss": 0.2998, "step": 7140 }, { "epoch": 49.65, "learning_rate": 2.029090909090909e-07, "loss": 0.2825, "step": 7150 }, { "epoch": 49.72, "learning_rate": 2.026293706293706e-07, "loss": 0.3092, "step": 7160 }, { "epoch": 49.79, "learning_rate": 2.0234965034965031e-07, "loss": 0.2957, "step": 7170 }, { "epoch": 49.86, "learning_rate": 2.0206993006993005e-07, "loss": 0.282, "step": 7180 }, { "epoch": 49.93, "learning_rate": 2.0179020979020977e-07, "loss": 0.29, "step": 7190 }, { "epoch": 50.0, "learning_rate": 2.015104895104895e-07, "loss": 0.3006, "step": 7200 }, { "epoch": 50.0, "eval_loss": 0.4634318947792053, "eval_runtime": 126.434, "eval_samples_per_second": 4.208, "eval_steps_per_second": 0.53, "eval_wer": 0.2071093398470142, "step": 7200 }, { "epoch": 50.07, "learning_rate": 2.0123076923076922e-07, "loss": 0.3064, "step": 7210 }, { "epoch": 50.14, "learning_rate": 2.0095104895104896e-07, "loss": 0.2834, "step": 7220 }, { "epoch": 50.21, "learning_rate": 2.0067132867132868e-07, "loss": 0.2917, "step": 7230 }, { "epoch": 50.28, "learning_rate": 2.003916083916084e-07, "loss": 0.2821, "step": 7240 }, { "epoch": 50.35, "learning_rate": 2.0011188811188808e-07, "loss": 0.2878, "step": 7250 }, { "epoch": 50.42, "learning_rate": 1.9983216783216782e-07, "loss": 0.2893, "step": 7260 }, { "epoch": 50.48, "learning_rate": 1.9955244755244753e-07, "loss": 0.3086, "step": 7270 }, { "epoch": 50.55, "learning_rate": 1.9927272727272725e-07, "loss": 0.2908, "step": 7280 }, { "epoch": 50.62, "learning_rate": 1.98993006993007e-07, "loss": 0.2764, "step": 7290 }, { "epoch": 50.69, "learning_rate": 1.987132867132867e-07, "loss": 0.3057, "step": 7300 }, { "epoch": 50.76, "learning_rate": 1.9843356643356644e-07, "loss": 0.2979, "step": 7310 }, { "epoch": 50.83, "learning_rate": 1.9815384615384613e-07, "loss": 0.2902, "step": 7320 }, { "epoch": 50.9, "learning_rate": 1.9787412587412587e-07, "loss": 0.2871, "step": 7330 }, { "epoch": 50.97, "learning_rate": 1.9759440559440558e-07, "loss": 0.3009, "step": 7340 }, { "epoch": 51.0, "eval_loss": 0.4634670913219452, "eval_runtime": 126.2777, "eval_samples_per_second": 4.213, "eval_steps_per_second": 0.531, "eval_wer": 0.2062094234106833, "step": 7344 }, { "epoch": 51.04, "learning_rate": 1.9731468531468533e-07, "loss": 0.3048, "step": 7350 }, { "epoch": 51.11, "learning_rate": 1.97034965034965e-07, "loss": 0.2855, "step": 7360 }, { "epoch": 51.18, "learning_rate": 1.9675524475524475e-07, "loss": 0.3004, "step": 7370 }, { "epoch": 51.25, "learning_rate": 1.9647552447552447e-07, "loss": 0.2994, "step": 7380 }, { "epoch": 51.32, "learning_rate": 1.9619580419580418e-07, "loss": 0.3044, "step": 7390 }, { "epoch": 51.39, "learning_rate": 1.959160839160839e-07, "loss": 0.2916, "step": 7400 }, { "epoch": 51.46, "learning_rate": 1.956363636363636e-07, "loss": 0.3148, "step": 7410 }, { "epoch": 51.53, "learning_rate": 1.9535664335664335e-07, "loss": 0.2915, "step": 7420 }, { "epoch": 51.6, "learning_rate": 1.9507692307692306e-07, "loss": 0.2768, "step": 7430 }, { "epoch": 51.66, "learning_rate": 1.947972027972028e-07, "loss": 0.2853, "step": 7440 }, { "epoch": 51.73, "learning_rate": 1.945174825174825e-07, "loss": 0.2862, "step": 7450 }, { "epoch": 51.8, "learning_rate": 1.9423776223776223e-07, "loss": 0.2808, "step": 7460 }, { "epoch": 51.87, "learning_rate": 1.9395804195804195e-07, "loss": 0.2773, "step": 7470 }, { "epoch": 51.94, "learning_rate": 1.936783216783217e-07, "loss": 0.2946, "step": 7480 }, { "epoch": 52.0, "eval_loss": 0.4647519290447235, "eval_runtime": 126.8045, "eval_samples_per_second": 4.195, "eval_steps_per_second": 0.528, "eval_wer": 0.20653082213794433, "step": 7488 }, { "epoch": 52.01, "learning_rate": 1.9339860139860138e-07, "loss": 0.3101, "step": 7490 }, { "epoch": 52.08, "learning_rate": 1.9311888111888112e-07, "loss": 0.2818, "step": 7500 }, { "epoch": 52.15, "learning_rate": 1.9283916083916083e-07, "loss": 0.2878, "step": 7510 }, { "epoch": 52.22, "learning_rate": 1.9255944055944054e-07, "loss": 0.3021, "step": 7520 }, { "epoch": 52.29, "learning_rate": 1.9227972027972028e-07, "loss": 0.2909, "step": 7530 }, { "epoch": 52.36, "learning_rate": 1.9199999999999997e-07, "loss": 0.2925, "step": 7540 }, { "epoch": 52.43, "learning_rate": 1.917202797202797e-07, "loss": 0.2918, "step": 7550 }, { "epoch": 52.5, "learning_rate": 1.9144055944055943e-07, "loss": 0.2936, "step": 7560 }, { "epoch": 52.57, "learning_rate": 1.9116083916083917e-07, "loss": 0.2842, "step": 7570 }, { "epoch": 52.64, "learning_rate": 1.9088111888111886e-07, "loss": 0.2674, "step": 7580 }, { "epoch": 52.71, "learning_rate": 1.906013986013986e-07, "loss": 0.3074, "step": 7590 }, { "epoch": 52.78, "learning_rate": 1.903216783216783e-07, "loss": 0.2861, "step": 7600 }, { "epoch": 52.84, "learning_rate": 1.9004195804195805e-07, "loss": 0.2975, "step": 7610 }, { "epoch": 52.91, "learning_rate": 1.8976223776223776e-07, "loss": 0.2832, "step": 7620 }, { "epoch": 52.98, "learning_rate": 1.8948251748251748e-07, "loss": 0.3093, "step": 7630 }, { "epoch": 53.0, "eval_loss": 0.46502918004989624, "eval_runtime": 127.1543, "eval_samples_per_second": 4.184, "eval_steps_per_second": 0.527, "eval_wer": 0.20608086391977887, "step": 7632 }, { "epoch": 53.06, "learning_rate": 1.892027972027972e-07, "loss": 0.3088, "step": 7640 }, { "epoch": 53.12, "learning_rate": 1.889230769230769e-07, "loss": 0.2768, "step": 7650 }, { "epoch": 53.19, "learning_rate": 1.8864335664335665e-07, "loss": 0.2945, "step": 7660 }, { "epoch": 53.26, "learning_rate": 1.8836363636363633e-07, "loss": 0.3026, "step": 7670 }, { "epoch": 53.33, "learning_rate": 1.8808391608391608e-07, "loss": 0.2862, "step": 7680 }, { "epoch": 53.4, "learning_rate": 1.878041958041958e-07, "loss": 0.2825, "step": 7690 }, { "epoch": 53.47, "learning_rate": 1.8752447552447553e-07, "loss": 0.3046, "step": 7700 }, { "epoch": 53.54, "learning_rate": 1.8724475524475522e-07, "loss": 0.2874, "step": 7710 }, { "epoch": 53.61, "learning_rate": 1.8696503496503496e-07, "loss": 0.2968, "step": 7720 }, { "epoch": 53.68, "learning_rate": 1.8668531468531467e-07, "loss": 0.2933, "step": 7730 }, { "epoch": 53.75, "learning_rate": 1.864055944055944e-07, "loss": 0.2872, "step": 7740 }, { "epoch": 53.82, "learning_rate": 1.8612587412587413e-07, "loss": 0.2825, "step": 7750 }, { "epoch": 53.89, "learning_rate": 1.8584615384615384e-07, "loss": 0.275, "step": 7760 }, { "epoch": 53.96, "learning_rate": 1.8556643356643355e-07, "loss": 0.3132, "step": 7770 }, { "epoch": 54.0, "eval_loss": 0.46536538004875183, "eval_runtime": 126.248, "eval_samples_per_second": 4.214, "eval_steps_per_second": 0.531, "eval_wer": 0.20524522722890018, "step": 7776 }, { "epoch": 54.03, "learning_rate": 1.8528671328671327e-07, "loss": 0.312, "step": 7780 }, { "epoch": 54.1, "learning_rate": 1.85006993006993e-07, "loss": 0.2733, "step": 7790 }, { "epoch": 54.17, "learning_rate": 1.847272727272727e-07, "loss": 0.292, "step": 7800 }, { "epoch": 54.24, "learning_rate": 1.8444755244755244e-07, "loss": 0.3134, "step": 7810 }, { "epoch": 54.3, "learning_rate": 1.8416783216783215e-07, "loss": 0.2816, "step": 7820 }, { "epoch": 54.37, "learning_rate": 1.838881118881119e-07, "loss": 0.2772, "step": 7830 }, { "epoch": 54.44, "learning_rate": 1.836083916083916e-07, "loss": 0.2963, "step": 7840 }, { "epoch": 54.51, "learning_rate": 1.8332867132867132e-07, "loss": 0.2995, "step": 7850 }, { "epoch": 54.58, "learning_rate": 1.8304895104895103e-07, "loss": 0.2875, "step": 7860 }, { "epoch": 54.65, "learning_rate": 1.8276923076923077e-07, "loss": 0.2691, "step": 7870 }, { "epoch": 54.72, "learning_rate": 1.824895104895105e-07, "loss": 0.3173, "step": 7880 }, { "epoch": 54.79, "learning_rate": 1.822097902097902e-07, "loss": 0.2869, "step": 7890 }, { "epoch": 54.86, "learning_rate": 1.8193006993006992e-07, "loss": 0.3037, "step": 7900 }, { "epoch": 54.93, "learning_rate": 1.8165034965034963e-07, "loss": 0.2941, "step": 7910 }, { "epoch": 55.0, "learning_rate": 1.8137062937062937e-07, "loss": 0.2926, "step": 7920 }, { "epoch": 55.0, "eval_loss": 0.4643465280532837, "eval_runtime": 126.3747, "eval_samples_per_second": 4.21, "eval_steps_per_second": 0.53, "eval_wer": 0.20640226264703992, "step": 7920 }, { "epoch": 55.07, "learning_rate": 1.8109090909090906e-07, "loss": 0.3096, "step": 7930 }, { "epoch": 55.14, "learning_rate": 1.808111888111888e-07, "loss": 0.2752, "step": 7940 }, { "epoch": 55.21, "learning_rate": 1.8053146853146851e-07, "loss": 0.2993, "step": 7950 }, { "epoch": 55.28, "learning_rate": 1.8025174825174825e-07, "loss": 0.2942, "step": 7960 }, { "epoch": 55.35, "learning_rate": 1.7997202797202797e-07, "loss": 0.2995, "step": 7970 }, { "epoch": 55.42, "learning_rate": 1.7969230769230768e-07, "loss": 0.3057, "step": 7980 }, { "epoch": 55.48, "learning_rate": 1.794125874125874e-07, "loss": 0.2952, "step": 7990 }, { "epoch": 55.55, "learning_rate": 1.7913286713286714e-07, "loss": 0.285, "step": 8000 }, { "epoch": 55.62, "learning_rate": 1.7885314685314685e-07, "loss": 0.284, "step": 8010 }, { "epoch": 55.69, "learning_rate": 1.7857342657342657e-07, "loss": 0.3031, "step": 8020 }, { "epoch": 55.76, "learning_rate": 1.7829370629370628e-07, "loss": 0.2919, "step": 8030 }, { "epoch": 55.83, "learning_rate": 1.78013986013986e-07, "loss": 0.2903, "step": 8040 }, { "epoch": 55.9, "learning_rate": 1.7773426573426573e-07, "loss": 0.288, "step": 8050 }, { "epoch": 55.97, "learning_rate": 1.7745454545454545e-07, "loss": 0.3089, "step": 8060 }, { "epoch": 56.0, "eval_loss": 0.4637354910373688, "eval_runtime": 126.0701, "eval_samples_per_second": 4.22, "eval_steps_per_second": 0.531, "eval_wer": 0.20665938162884875, "step": 8064 }, { "epoch": 56.04, "learning_rate": 1.7717482517482516e-07, "loss": 0.3037, "step": 8070 }, { "epoch": 56.11, "learning_rate": 1.7689510489510488e-07, "loss": 0.2869, "step": 8080 }, { "epoch": 56.18, "learning_rate": 1.7661538461538462e-07, "loss": 0.2928, "step": 8090 }, { "epoch": 56.25, "learning_rate": 1.7633566433566433e-07, "loss": 0.2871, "step": 8100 }, { "epoch": 56.32, "learning_rate": 1.7605594405594404e-07, "loss": 0.2912, "step": 8110 }, { "epoch": 56.39, "learning_rate": 1.7577622377622376e-07, "loss": 0.2605, "step": 8120 }, { "epoch": 56.46, "learning_rate": 1.754965034965035e-07, "loss": 0.312, "step": 8130 }, { "epoch": 56.53, "learning_rate": 1.7521678321678321e-07, "loss": 0.2967, "step": 8140 }, { "epoch": 56.6, "learning_rate": 1.7493706293706293e-07, "loss": 0.2992, "step": 8150 }, { "epoch": 56.66, "learning_rate": 1.7465734265734264e-07, "loss": 0.2884, "step": 8160 }, { "epoch": 56.73, "learning_rate": 1.7437762237762236e-07, "loss": 0.3082, "step": 8170 }, { "epoch": 56.8, "learning_rate": 1.740979020979021e-07, "loss": 0.2947, "step": 8180 }, { "epoch": 56.87, "learning_rate": 1.738181818181818e-07, "loss": 0.2821, "step": 8190 }, { "epoch": 56.94, "learning_rate": 1.7353846153846152e-07, "loss": 0.3052, "step": 8200 }, { "epoch": 57.0, "eval_loss": 0.46311089396476746, "eval_runtime": 126.1645, "eval_samples_per_second": 4.217, "eval_steps_per_second": 0.531, "eval_wer": 0.2055666259561612, "step": 8208 }, { "epoch": 57.01, "learning_rate": 1.7325874125874124e-07, "loss": 0.2938, "step": 8210 }, { "epoch": 57.08, "learning_rate": 1.7297902097902098e-07, "loss": 0.2848, "step": 8220 }, { "epoch": 57.15, "learning_rate": 1.726993006993007e-07, "loss": 0.29, "step": 8230 }, { "epoch": 57.22, "learning_rate": 1.724195804195804e-07, "loss": 0.304, "step": 8240 }, { "epoch": 57.29, "learning_rate": 1.7213986013986012e-07, "loss": 0.3001, "step": 8250 }, { "epoch": 57.36, "learning_rate": 1.7186013986013986e-07, "loss": 0.298, "step": 8260 }, { "epoch": 57.43, "learning_rate": 1.7158041958041958e-07, "loss": 0.3071, "step": 8270 }, { "epoch": 57.5, "learning_rate": 1.713006993006993e-07, "loss": 0.3026, "step": 8280 }, { "epoch": 57.57, "learning_rate": 1.71020979020979e-07, "loss": 0.2953, "step": 8290 }, { "epoch": 57.64, "learning_rate": 1.7074125874125872e-07, "loss": 0.2899, "step": 8300 }, { "epoch": 57.71, "learning_rate": 1.7046153846153846e-07, "loss": 0.2918, "step": 8310 }, { "epoch": 57.78, "learning_rate": 1.7018181818181817e-07, "loss": 0.2881, "step": 8320 }, { "epoch": 57.84, "learning_rate": 1.699020979020979e-07, "loss": 0.2829, "step": 8330 }, { "epoch": 57.91, "learning_rate": 1.696223776223776e-07, "loss": 0.2843, "step": 8340 }, { "epoch": 57.98, "learning_rate": 1.6934265734265734e-07, "loss": 0.3037, "step": 8350 }, { "epoch": 58.0, "eval_loss": 0.46301281452178955, "eval_runtime": 127.1068, "eval_samples_per_second": 4.185, "eval_steps_per_second": 0.527, "eval_wer": 0.20518094748344798, "step": 8352 }, { "epoch": 58.06, "learning_rate": 1.6906293706293706e-07, "loss": 0.301, "step": 8360 }, { "epoch": 58.12, "learning_rate": 1.687832167832168e-07, "loss": 0.2741, "step": 8370 }, { "epoch": 58.19, "learning_rate": 1.6850349650349648e-07, "loss": 0.2996, "step": 8380 }, { "epoch": 58.26, "learning_rate": 1.6822377622377622e-07, "loss": 0.2953, "step": 8390 }, { "epoch": 58.33, "learning_rate": 1.6794405594405594e-07, "loss": 0.3018, "step": 8400 }, { "epoch": 58.4, "learning_rate": 1.6766433566433565e-07, "loss": 0.2758, "step": 8410 }, { "epoch": 58.47, "learning_rate": 1.6738461538461537e-07, "loss": 0.3074, "step": 8420 }, { "epoch": 58.54, "learning_rate": 1.671048951048951e-07, "loss": 0.2811, "step": 8430 }, { "epoch": 58.61, "learning_rate": 1.6682517482517482e-07, "loss": 0.2805, "step": 8440 }, { "epoch": 58.68, "learning_rate": 1.6654545454545454e-07, "loss": 0.2993, "step": 8450 }, { "epoch": 58.75, "learning_rate": 1.6626573426573425e-07, "loss": 0.3169, "step": 8460 }, { "epoch": 58.82, "learning_rate": 1.6598601398601396e-07, "loss": 0.2954, "step": 8470 }, { "epoch": 58.89, "learning_rate": 1.657062937062937e-07, "loss": 0.2781, "step": 8480 }, { "epoch": 58.96, "learning_rate": 1.6542657342657342e-07, "loss": 0.2948, "step": 8490 }, { "epoch": 59.0, "eval_loss": 0.464093416929245, "eval_runtime": 126.1432, "eval_samples_per_second": 4.217, "eval_steps_per_second": 0.531, "eval_wer": 0.2054380664652568, "step": 8496 }, { "epoch": 59.03, "learning_rate": 1.6514685314685316e-07, "loss": 0.3063, "step": 8500 }, { "epoch": 59.1, "learning_rate": 1.6486713286713285e-07, "loss": 0.2648, "step": 8510 }, { "epoch": 59.17, "learning_rate": 1.6458741258741259e-07, "loss": 0.2932, "step": 8520 }, { "epoch": 59.24, "learning_rate": 1.643076923076923e-07, "loss": 0.2989, "step": 8530 }, { "epoch": 59.3, "learning_rate": 1.6402797202797201e-07, "loss": 0.2996, "step": 8540 }, { "epoch": 59.37, "learning_rate": 1.6377622377622378e-07, "loss": 0.2929, "step": 8550 }, { "epoch": 59.44, "learning_rate": 1.6349650349650347e-07, "loss": 0.3099, "step": 8560 }, { "epoch": 59.51, "learning_rate": 1.632167832167832e-07, "loss": 0.3002, "step": 8570 }, { "epoch": 59.58, "learning_rate": 1.6293706293706293e-07, "loss": 0.2894, "step": 8580 }, { "epoch": 59.65, "learning_rate": 1.6265734265734267e-07, "loss": 0.2945, "step": 8590 }, { "epoch": 59.72, "learning_rate": 1.6237762237762238e-07, "loss": 0.3194, "step": 8600 }, { "epoch": 59.79, "learning_rate": 1.620979020979021e-07, "loss": 0.2807, "step": 8610 }, { "epoch": 59.86, "learning_rate": 1.618181818181818e-07, "loss": 0.2727, "step": 8620 }, { "epoch": 59.93, "learning_rate": 1.6153846153846152e-07, "loss": 0.2838, "step": 8630 }, { "epoch": 60.0, "learning_rate": 1.6125874125874126e-07, "loss": 0.2857, "step": 8640 }, { "epoch": 60.0, "eval_loss": 0.4635641872882843, "eval_runtime": 126.6479, "eval_samples_per_second": 4.201, "eval_steps_per_second": 0.529, "eval_wer": 0.20595230442887447, "step": 8640 }, { "epoch": 60.07, "learning_rate": 1.6097902097902095e-07, "loss": 0.3201, "step": 8650 }, { "epoch": 60.14, "learning_rate": 1.606993006993007e-07, "loss": 0.2764, "step": 8660 }, { "epoch": 60.21, "learning_rate": 1.604195804195804e-07, "loss": 0.2876, "step": 8670 }, { "epoch": 60.28, "learning_rate": 1.6013986013986015e-07, "loss": 0.2923, "step": 8680 }, { "epoch": 60.35, "learning_rate": 1.5986013986013983e-07, "loss": 0.2943, "step": 8690 }, { "epoch": 60.42, "learning_rate": 1.5958041958041957e-07, "loss": 0.2907, "step": 8700 }, { "epoch": 60.48, "learning_rate": 1.593006993006993e-07, "loss": 0.2967, "step": 8710 }, { "epoch": 60.55, "learning_rate": 1.5902097902097903e-07, "loss": 0.2835, "step": 8720 }, { "epoch": 60.62, "learning_rate": 1.5874125874125874e-07, "loss": 0.282, "step": 8730 }, { "epoch": 60.69, "learning_rate": 1.5846153846153846e-07, "loss": 0.3076, "step": 8740 }, { "epoch": 60.76, "learning_rate": 1.5818181818181817e-07, "loss": 0.2992, "step": 8750 }, { "epoch": 60.83, "learning_rate": 1.5790209790209789e-07, "loss": 0.2807, "step": 8760 }, { "epoch": 60.9, "learning_rate": 1.5762237762237763e-07, "loss": 0.2939, "step": 8770 }, { "epoch": 60.97, "learning_rate": 1.5734265734265731e-07, "loss": 0.2955, "step": 8780 }, { "epoch": 61.0, "eval_loss": 0.4629845917224884, "eval_runtime": 126.8679, "eval_samples_per_second": 4.193, "eval_steps_per_second": 0.528, "eval_wer": 0.20601658417432667, "step": 8784 }, { "epoch": 61.04, "learning_rate": 1.5706293706293705e-07, "loss": 0.3129, "step": 8790 }, { "epoch": 61.11, "learning_rate": 1.5678321678321677e-07, "loss": 0.2793, "step": 8800 }, { "epoch": 61.18, "learning_rate": 1.565034965034965e-07, "loss": 0.3005, "step": 8810 }, { "epoch": 61.25, "learning_rate": 1.5622377622377622e-07, "loss": 0.3016, "step": 8820 }, { "epoch": 61.32, "learning_rate": 1.5594405594405594e-07, "loss": 0.3, "step": 8830 }, { "epoch": 61.39, "learning_rate": 1.5566433566433565e-07, "loss": 0.2741, "step": 8840 }, { "epoch": 61.46, "learning_rate": 1.553846153846154e-07, "loss": 0.3015, "step": 8850 }, { "epoch": 61.53, "learning_rate": 1.551048951048951e-07, "loss": 0.2763, "step": 8860 }, { "epoch": 61.6, "learning_rate": 1.5482517482517482e-07, "loss": 0.2826, "step": 8870 }, { "epoch": 61.66, "learning_rate": 1.5454545454545453e-07, "loss": 0.2765, "step": 8880 }, { "epoch": 61.73, "learning_rate": 1.5426573426573425e-07, "loss": 0.3101, "step": 8890 }, { "epoch": 61.8, "learning_rate": 1.53986013986014e-07, "loss": 0.295, "step": 8900 }, { "epoch": 61.87, "learning_rate": 1.537062937062937e-07, "loss": 0.2948, "step": 8910 }, { "epoch": 61.94, "learning_rate": 1.5342657342657342e-07, "loss": 0.3035, "step": 8920 }, { "epoch": 62.0, "eval_loss": 0.4630157947540283, "eval_runtime": 126.2293, "eval_samples_per_second": 4.215, "eval_steps_per_second": 0.531, "eval_wer": 0.20608086391977887, "step": 8928 }, { "epoch": 62.01, "learning_rate": 1.5314685314685313e-07, "loss": 0.3241, "step": 8930 }, { "epoch": 62.08, "learning_rate": 1.5286713286713287e-07, "loss": 0.292, "step": 8940 }, { "epoch": 62.15, "learning_rate": 1.5258741258741259e-07, "loss": 0.2904, "step": 8950 }, { "epoch": 62.22, "learning_rate": 1.523076923076923e-07, "loss": 0.3085, "step": 8960 }, { "epoch": 62.29, "learning_rate": 1.5202797202797201e-07, "loss": 0.298, "step": 8970 }, { "epoch": 62.36, "learning_rate": 1.5174825174825175e-07, "loss": 0.3032, "step": 8980 }, { "epoch": 62.43, "learning_rate": 1.5146853146853147e-07, "loss": 0.2985, "step": 8990 }, { "epoch": 62.5, "learning_rate": 1.5118881118881118e-07, "loss": 0.2978, "step": 9000 }, { "epoch": 62.57, "learning_rate": 1.509090909090909e-07, "loss": 0.291, "step": 9010 }, { "epoch": 62.64, "learning_rate": 1.506293706293706e-07, "loss": 0.2892, "step": 9020 }, { "epoch": 62.71, "learning_rate": 1.5034965034965035e-07, "loss": 0.3104, "step": 9030 }, { "epoch": 62.78, "learning_rate": 1.5006993006993006e-07, "loss": 0.2988, "step": 9040 }, { "epoch": 62.84, "learning_rate": 1.4979020979020978e-07, "loss": 0.2941, "step": 9050 }, { "epoch": 62.91, "learning_rate": 1.495104895104895e-07, "loss": 0.2925, "step": 9060 }, { "epoch": 62.98, "learning_rate": 1.4923076923076923e-07, "loss": 0.2955, "step": 9070 }, { "epoch": 63.0, "eval_loss": 0.4625116288661957, "eval_runtime": 126.8407, "eval_samples_per_second": 4.194, "eval_steps_per_second": 0.528, "eval_wer": 0.20575946519251784, "step": 9072 }, { "epoch": 63.06, "learning_rate": 1.4895104895104895e-07, "loss": 0.3193, "step": 9080 }, { "epoch": 63.12, "learning_rate": 1.4867132867132866e-07, "loss": 0.2838, "step": 9090 }, { "epoch": 63.19, "learning_rate": 1.4839160839160838e-07, "loss": 0.3065, "step": 9100 }, { "epoch": 63.26, "learning_rate": 1.4811188811188812e-07, "loss": 0.305, "step": 9110 }, { "epoch": 63.33, "learning_rate": 1.4783216783216783e-07, "loss": 0.2876, "step": 9120 }, { "epoch": 63.4, "learning_rate": 1.4755244755244754e-07, "loss": 0.2886, "step": 9130 }, { "epoch": 63.47, "learning_rate": 1.4727272727272726e-07, "loss": 0.3028, "step": 9140 }, { "epoch": 63.54, "learning_rate": 1.4699300699300697e-07, "loss": 0.2901, "step": 9150 }, { "epoch": 63.61, "learning_rate": 1.467132867132867e-07, "loss": 0.3, "step": 9160 }, { "epoch": 63.68, "learning_rate": 1.4643356643356643e-07, "loss": 0.2952, "step": 9170 }, { "epoch": 63.75, "learning_rate": 1.4615384615384614e-07, "loss": 0.3018, "step": 9180 }, { "epoch": 63.82, "learning_rate": 1.4587412587412586e-07, "loss": 0.3022, "step": 9190 }, { "epoch": 63.89, "learning_rate": 1.455944055944056e-07, "loss": 0.2837, "step": 9200 }, { "epoch": 63.96, "learning_rate": 1.453146853146853e-07, "loss": 0.3138, "step": 9210 }, { "epoch": 64.0, "eval_loss": 0.46125251054763794, "eval_runtime": 127.2433, "eval_samples_per_second": 4.181, "eval_steps_per_second": 0.527, "eval_wer": 0.20479526901073472, "step": 9216 }, { "epoch": 64.03, "learning_rate": 1.4503496503496502e-07, "loss": 0.3051, "step": 9220 }, { "epoch": 64.1, "learning_rate": 1.4475524475524474e-07, "loss": 0.2925, "step": 9230 }, { "epoch": 64.17, "learning_rate": 1.4447552447552448e-07, "loss": 0.2946, "step": 9240 }, { "epoch": 64.24, "learning_rate": 1.441958041958042e-07, "loss": 0.3049, "step": 9250 }, { "epoch": 64.3, "learning_rate": 1.439160839160839e-07, "loss": 0.307, "step": 9260 }, { "epoch": 64.37, "learning_rate": 1.4363636363636362e-07, "loss": 0.2821, "step": 9270 }, { "epoch": 64.44, "learning_rate": 1.4335664335664333e-07, "loss": 0.2944, "step": 9280 }, { "epoch": 64.51, "learning_rate": 1.4307692307692308e-07, "loss": 0.2993, "step": 9290 }, { "epoch": 64.58, "learning_rate": 1.427972027972028e-07, "loss": 0.311, "step": 9300 }, { "epoch": 64.65, "learning_rate": 1.425174825174825e-07, "loss": 0.2965, "step": 9310 }, { "epoch": 64.72, "learning_rate": 1.4223776223776222e-07, "loss": 0.2972, "step": 9320 }, { "epoch": 64.79, "learning_rate": 1.4195804195804196e-07, "loss": 0.3007, "step": 9330 }, { "epoch": 64.86, "learning_rate": 1.4167832167832167e-07, "loss": 0.3029, "step": 9340 }, { "epoch": 64.93, "learning_rate": 1.413986013986014e-07, "loss": 0.303, "step": 9350 }, { "epoch": 65.0, "learning_rate": 1.411188811188811e-07, "loss": 0.2911, "step": 9360 }, { "epoch": 65.0, "eval_loss": 0.46050775051116943, "eval_runtime": 126.825, "eval_samples_per_second": 4.195, "eval_steps_per_second": 0.528, "eval_wer": 0.20479526901073472, "step": 9360 }, { "epoch": 65.07, "learning_rate": 1.4083916083916084e-07, "loss": 0.3125, "step": 9370 }, { "epoch": 65.14, "learning_rate": 1.4055944055944055e-07, "loss": 0.2933, "step": 9380 }, { "epoch": 65.21, "learning_rate": 1.4027972027972027e-07, "loss": 0.3121, "step": 9390 }, { "epoch": 65.28, "learning_rate": 1.3999999999999998e-07, "loss": 0.3041, "step": 9400 }, { "epoch": 65.35, "learning_rate": 1.397202797202797e-07, "loss": 0.2828, "step": 9410 }, { "epoch": 65.42, "learning_rate": 1.3944055944055944e-07, "loss": 0.3102, "step": 9420 }, { "epoch": 65.48, "learning_rate": 1.3916083916083915e-07, "loss": 0.3019, "step": 9430 }, { "epoch": 65.55, "learning_rate": 1.3888111888111887e-07, "loss": 0.2831, "step": 9440 }, { "epoch": 65.62, "learning_rate": 1.3860139860139858e-07, "loss": 0.2743, "step": 9450 }, { "epoch": 65.69, "learning_rate": 1.3832167832167832e-07, "loss": 0.3067, "step": 9460 }, { "epoch": 65.76, "learning_rate": 1.3804195804195803e-07, "loss": 0.2902, "step": 9470 }, { "epoch": 65.83, "learning_rate": 1.3776223776223778e-07, "loss": 0.3047, "step": 9480 }, { "epoch": 65.9, "learning_rate": 1.3748251748251746e-07, "loss": 0.2738, "step": 9490 }, { "epoch": 65.97, "learning_rate": 1.372027972027972e-07, "loss": 0.2884, "step": 9500 }, { "epoch": 66.0, "eval_loss": 0.46074166893959045, "eval_runtime": 127.3201, "eval_samples_per_second": 4.178, "eval_steps_per_second": 0.526, "eval_wer": 0.2046667095198303, "step": 9504 }, { "epoch": 66.04, "learning_rate": 1.3692307692307692e-07, "loss": 0.3166, "step": 9510 }, { "epoch": 66.11, "learning_rate": 1.3664335664335663e-07, "loss": 0.3005, "step": 9520 }, { "epoch": 66.18, "learning_rate": 1.3636363636363635e-07, "loss": 0.289, "step": 9530 }, { "epoch": 66.25, "learning_rate": 1.3608391608391606e-07, "loss": 0.3103, "step": 9540 }, { "epoch": 66.32, "learning_rate": 1.358041958041958e-07, "loss": 0.2915, "step": 9550 }, { "epoch": 66.39, "learning_rate": 1.3552447552447551e-07, "loss": 0.2933, "step": 9560 }, { "epoch": 66.46, "learning_rate": 1.3524475524475525e-07, "loss": 0.3019, "step": 9570 }, { "epoch": 66.53, "learning_rate": 1.3496503496503494e-07, "loss": 0.3058, "step": 9580 }, { "epoch": 66.6, "learning_rate": 1.3468531468531468e-07, "loss": 0.2966, "step": 9590 }, { "epoch": 66.66, "learning_rate": 1.344055944055944e-07, "loss": 0.2948, "step": 9600 }, { "epoch": 66.73, "learning_rate": 1.3412587412587414e-07, "loss": 0.298, "step": 9610 }, { "epoch": 66.8, "learning_rate": 1.3384615384615383e-07, "loss": 0.2929, "step": 9620 }, { "epoch": 66.87, "learning_rate": 1.3356643356643357e-07, "loss": 0.2844, "step": 9630 }, { "epoch": 66.94, "learning_rate": 1.3328671328671328e-07, "loss": 0.3129, "step": 9640 }, { "epoch": 67.0, "eval_loss": 0.46037429571151733, "eval_runtime": 126.4219, "eval_samples_per_second": 4.208, "eval_steps_per_second": 0.53, "eval_wer": 0.20479526901073472, "step": 9648 }, { "epoch": 67.01, "learning_rate": 1.33006993006993e-07, "loss": 0.3072, "step": 9650 }, { "epoch": 67.08, "learning_rate": 1.3272727272727273e-07, "loss": 0.2826, "step": 9660 }, { "epoch": 67.15, "learning_rate": 1.3244755244755245e-07, "loss": 0.2805, "step": 9670 }, { "epoch": 67.22, "learning_rate": 1.3216783216783216e-07, "loss": 0.3232, "step": 9680 }, { "epoch": 67.29, "learning_rate": 1.3188811188811188e-07, "loss": 0.2871, "step": 9690 }, { "epoch": 67.36, "learning_rate": 1.3160839160839162e-07, "loss": 0.2745, "step": 9700 }, { "epoch": 67.43, "learning_rate": 1.313286713286713e-07, "loss": 0.2912, "step": 9710 }, { "epoch": 67.5, "learning_rate": 1.3104895104895105e-07, "loss": 0.2991, "step": 9720 }, { "epoch": 67.57, "learning_rate": 1.3076923076923076e-07, "loss": 0.2966, "step": 9730 }, { "epoch": 67.64, "learning_rate": 1.304895104895105e-07, "loss": 0.2835, "step": 9740 }, { "epoch": 67.71, "learning_rate": 1.302097902097902e-07, "loss": 0.3056, "step": 9750 }, { "epoch": 67.78, "learning_rate": 1.2993006993006993e-07, "loss": 0.3037, "step": 9760 }, { "epoch": 67.84, "learning_rate": 1.2965034965034964e-07, "loss": 0.3004, "step": 9770 }, { "epoch": 67.91, "learning_rate": 1.2937062937062938e-07, "loss": 0.2897, "step": 9780 }, { "epoch": 67.98, "learning_rate": 1.290909090909091e-07, "loss": 0.308, "step": 9790 }, { "epoch": 68.0, "eval_loss": 0.4608405530452728, "eval_runtime": 126.1051, "eval_samples_per_second": 4.219, "eval_steps_per_second": 0.531, "eval_wer": 0.20530950697435238, "step": 9792 }, { "epoch": 68.06, "learning_rate": 1.288111888111888e-07, "loss": 0.2987, "step": 9800 }, { "epoch": 68.12, "learning_rate": 1.2853146853146852e-07, "loss": 0.2846, "step": 9810 }, { "epoch": 68.19, "learning_rate": 1.2825174825174824e-07, "loss": 0.309, "step": 9820 }, { "epoch": 68.26, "learning_rate": 1.2797202797202798e-07, "loss": 0.3028, "step": 9830 }, { "epoch": 68.33, "learning_rate": 1.2769230769230767e-07, "loss": 0.3039, "step": 9840 }, { "epoch": 68.4, "learning_rate": 1.274125874125874e-07, "loss": 0.2934, "step": 9850 }, { "epoch": 68.47, "learning_rate": 1.2713286713286712e-07, "loss": 0.3144, "step": 9860 }, { "epoch": 68.54, "learning_rate": 1.2685314685314686e-07, "loss": 0.3038, "step": 9870 }, { "epoch": 68.61, "learning_rate": 1.2657342657342658e-07, "loss": 0.2782, "step": 9880 }, { "epoch": 68.68, "learning_rate": 1.262937062937063e-07, "loss": 0.2904, "step": 9890 }, { "epoch": 68.75, "learning_rate": 1.26013986013986e-07, "loss": 0.2864, "step": 9900 }, { "epoch": 68.82, "learning_rate": 1.2573426573426574e-07, "loss": 0.2973, "step": 9910 }, { "epoch": 68.89, "learning_rate": 1.2545454545454546e-07, "loss": 0.2839, "step": 9920 }, { "epoch": 68.96, "learning_rate": 1.2517482517482517e-07, "loss": 0.2981, "step": 9930 }, { "epoch": 69.0, "eval_loss": 0.45970267057418823, "eval_runtime": 126.3105, "eval_samples_per_second": 4.212, "eval_steps_per_second": 0.53, "eval_wer": 0.20498810824709135, "step": 9936 }, { "epoch": 69.03, "learning_rate": 1.248951048951049e-07, "loss": 0.3117, "step": 9940 }, { "epoch": 69.1, "learning_rate": 1.246153846153846e-07, "loss": 0.3046, "step": 9950 }, { "epoch": 69.17, "learning_rate": 1.2433566433566434e-07, "loss": 0.2836, "step": 9960 }, { "epoch": 69.24, "learning_rate": 1.2405594405594403e-07, "loss": 0.3082, "step": 9970 }, { "epoch": 69.3, "learning_rate": 1.2377622377622377e-07, "loss": 0.2899, "step": 9980 }, { "epoch": 69.37, "learning_rate": 1.2349650349650348e-07, "loss": 0.2862, "step": 9990 }, { "epoch": 69.44, "learning_rate": 1.2321678321678322e-07, "loss": 0.3033, "step": 10000 }, { "epoch": 69.51, "learning_rate": 1.2293706293706294e-07, "loss": 0.304, "step": 10010 }, { "epoch": 69.58, "learning_rate": 1.2265734265734265e-07, "loss": 0.2993, "step": 10020 }, { "epoch": 69.65, "learning_rate": 1.2237762237762237e-07, "loss": 0.2872, "step": 10030 }, { "epoch": 69.72, "learning_rate": 1.220979020979021e-07, "loss": 0.3057, "step": 10040 }, { "epoch": 69.79, "learning_rate": 1.2181818181818182e-07, "loss": 0.2966, "step": 10050 }, { "epoch": 69.86, "learning_rate": 1.2153846153846154e-07, "loss": 0.2993, "step": 10060 }, { "epoch": 69.93, "learning_rate": 1.2125874125874125e-07, "loss": 0.2898, "step": 10070 }, { "epoch": 70.0, "learning_rate": 1.2097902097902096e-07, "loss": 0.308, "step": 10080 }, { "epoch": 70.0, "eval_loss": 0.4608989357948303, "eval_runtime": 126.2907, "eval_samples_per_second": 4.213, "eval_steps_per_second": 0.531, "eval_wer": 0.20530950697435238, "step": 10080 }, { "epoch": 70.07, "learning_rate": 1.206993006993007e-07, "loss": 0.3083, "step": 10090 }, { "epoch": 70.14, "learning_rate": 1.2041958041958042e-07, "loss": 0.2873, "step": 10100 }, { "epoch": 70.21, "learning_rate": 1.2013986013986013e-07, "loss": 0.2928, "step": 10110 }, { "epoch": 70.28, "learning_rate": 1.1986013986013985e-07, "loss": 0.3115, "step": 10120 }, { "epoch": 70.35, "learning_rate": 1.195804195804196e-07, "loss": 0.2979, "step": 10130 }, { "epoch": 70.42, "learning_rate": 1.193006993006993e-07, "loss": 0.277, "step": 10140 }, { "epoch": 70.48, "learning_rate": 1.19020979020979e-07, "loss": 0.3051, "step": 10150 }, { "epoch": 70.55, "learning_rate": 1.1874125874125873e-07, "loss": 0.2843, "step": 10160 }, { "epoch": 70.62, "learning_rate": 1.1846153846153846e-07, "loss": 0.2806, "step": 10170 }, { "epoch": 70.69, "learning_rate": 1.1818181818181818e-07, "loss": 0.3119, "step": 10180 }, { "epoch": 70.76, "learning_rate": 1.1790209790209791e-07, "loss": 0.3052, "step": 10190 }, { "epoch": 70.83, "learning_rate": 1.1762237762237761e-07, "loss": 0.305, "step": 10200 }, { "epoch": 70.9, "learning_rate": 1.1734265734265734e-07, "loss": 0.2918, "step": 10210 }, { "epoch": 70.97, "learning_rate": 1.1706293706293707e-07, "loss": 0.3059, "step": 10220 }, { "epoch": 71.0, "eval_loss": 0.45938730239868164, "eval_runtime": 126.3187, "eval_samples_per_second": 4.212, "eval_steps_per_second": 0.53, "eval_wer": 0.2046024297743781, "step": 10224 }, { "epoch": 71.04, "learning_rate": 1.1678321678321678e-07, "loss": 0.3218, "step": 10230 }, { "epoch": 71.11, "learning_rate": 1.165034965034965e-07, "loss": 0.2947, "step": 10240 }, { "epoch": 71.18, "learning_rate": 1.1622377622377622e-07, "loss": 0.2877, "step": 10250 }, { "epoch": 71.25, "learning_rate": 1.1594405594405594e-07, "loss": 0.2973, "step": 10260 }, { "epoch": 71.32, "learning_rate": 1.1566433566433566e-07, "loss": 0.2975, "step": 10270 }, { "epoch": 71.39, "learning_rate": 1.1538461538461536e-07, "loss": 0.2766, "step": 10280 }, { "epoch": 71.46, "learning_rate": 1.1510489510489509e-07, "loss": 0.3067, "step": 10290 }, { "epoch": 71.53, "learning_rate": 1.1482517482517482e-07, "loss": 0.2974, "step": 10300 }, { "epoch": 71.6, "learning_rate": 1.1454545454545455e-07, "loss": 0.3062, "step": 10310 }, { "epoch": 71.66, "learning_rate": 1.1426573426573427e-07, "loss": 0.2891, "step": 10320 }, { "epoch": 71.73, "learning_rate": 1.1398601398601397e-07, "loss": 0.3118, "step": 10330 }, { "epoch": 71.8, "learning_rate": 1.137062937062937e-07, "loss": 0.2863, "step": 10340 }, { "epoch": 71.87, "learning_rate": 1.1342657342657343e-07, "loss": 0.282, "step": 10350 }, { "epoch": 71.94, "learning_rate": 1.1314685314685316e-07, "loss": 0.3101, "step": 10360 }, { "epoch": 72.0, "eval_loss": 0.45849326252937317, "eval_runtime": 126.6189, "eval_samples_per_second": 4.202, "eval_steps_per_second": 0.529, "eval_wer": 0.20518094748344798, "step": 10368 }, { "epoch": 72.01, "learning_rate": 1.1286713286713286e-07, "loss": 0.3201, "step": 10370 }, { "epoch": 72.08, "learning_rate": 1.1258741258741258e-07, "loss": 0.3028, "step": 10380 }, { "epoch": 72.15, "learning_rate": 1.123076923076923e-07, "loss": 0.2903, "step": 10390 }, { "epoch": 72.22, "learning_rate": 1.1202797202797203e-07, "loss": 0.3197, "step": 10400 }, { "epoch": 72.29, "learning_rate": 1.1174825174825175e-07, "loss": 0.2872, "step": 10410 }, { "epoch": 72.36, "learning_rate": 1.1146853146853145e-07, "loss": 0.289, "step": 10420 }, { "epoch": 72.43, "learning_rate": 1.1118881118881118e-07, "loss": 0.3041, "step": 10430 }, { "epoch": 72.5, "learning_rate": 1.1090909090909091e-07, "loss": 0.3005, "step": 10440 }, { "epoch": 72.57, "learning_rate": 1.1062937062937064e-07, "loss": 0.2889, "step": 10450 }, { "epoch": 72.64, "learning_rate": 1.1034965034965034e-07, "loss": 0.2919, "step": 10460 }, { "epoch": 72.71, "learning_rate": 1.1006993006993006e-07, "loss": 0.3066, "step": 10470 }, { "epoch": 72.78, "learning_rate": 1.0979020979020979e-07, "loss": 0.2861, "step": 10480 }, { "epoch": 72.84, "learning_rate": 1.0951048951048952e-07, "loss": 0.2911, "step": 10490 }, { "epoch": 72.91, "learning_rate": 1.0923076923076922e-07, "loss": 0.2846, "step": 10500 }, { "epoch": 72.98, "learning_rate": 1.0895104895104895e-07, "loss": 0.2985, "step": 10510 }, { "epoch": 73.0, "eval_loss": 0.4591454565525055, "eval_runtime": 126.5058, "eval_samples_per_second": 4.205, "eval_steps_per_second": 0.53, "eval_wer": 0.20518094748344798, "step": 10512 }, { "epoch": 73.06, "learning_rate": 1.0867132867132866e-07, "loss": 0.3081, "step": 10520 }, { "epoch": 73.12, "learning_rate": 1.0839160839160839e-07, "loss": 0.2751, "step": 10530 }, { "epoch": 73.19, "learning_rate": 1.0811188811188812e-07, "loss": 0.3097, "step": 10540 }, { "epoch": 73.26, "learning_rate": 1.0783216783216782e-07, "loss": 0.291, "step": 10550 }, { "epoch": 73.33, "learning_rate": 1.0755244755244754e-07, "loss": 0.2948, "step": 10560 }, { "epoch": 73.4, "learning_rate": 1.073006993006993e-07, "loss": 0.315, "step": 10570 }, { "epoch": 73.47, "learning_rate": 1.0702097902097901e-07, "loss": 0.293, "step": 10580 }, { "epoch": 73.54, "learning_rate": 1.0674125874125874e-07, "loss": 0.2856, "step": 10590 }, { "epoch": 73.61, "learning_rate": 1.0646153846153844e-07, "loss": 0.2914, "step": 10600 }, { "epoch": 73.68, "learning_rate": 1.0618181818181817e-07, "loss": 0.2884, "step": 10610 }, { "epoch": 73.75, "learning_rate": 1.059020979020979e-07, "loss": 0.2983, "step": 10620 }, { "epoch": 73.82, "learning_rate": 1.0562237762237762e-07, "loss": 0.2777, "step": 10630 }, { "epoch": 73.89, "learning_rate": 1.0534265734265735e-07, "loss": 0.281, "step": 10640 }, { "epoch": 73.96, "learning_rate": 1.0506293706293705e-07, "loss": 0.3158, "step": 10650 }, { "epoch": 74.0, "eval_loss": 0.4581505060195923, "eval_runtime": 126.7988, "eval_samples_per_second": 4.196, "eval_steps_per_second": 0.528, "eval_wer": 0.20511666773799575, "step": 10656 }, { "epoch": 74.03, "learning_rate": 1.0478321678321678e-07, "loss": 0.3016, "step": 10660 }, { "epoch": 74.1, "learning_rate": 1.045034965034965e-07, "loss": 0.273, "step": 10670 }, { "epoch": 74.17, "learning_rate": 1.0422377622377623e-07, "loss": 0.2748, "step": 10680 }, { "epoch": 74.24, "learning_rate": 1.0394405594405593e-07, "loss": 0.3031, "step": 10690 }, { "epoch": 74.3, "learning_rate": 1.0366433566433566e-07, "loss": 0.2866, "step": 10700 }, { "epoch": 74.37, "learning_rate": 1.0338461538461538e-07, "loss": 0.2838, "step": 10710 }, { "epoch": 74.44, "learning_rate": 1.031048951048951e-07, "loss": 0.306, "step": 10720 }, { "epoch": 74.51, "learning_rate": 1.028251748251748e-07, "loss": 0.2987, "step": 10730 }, { "epoch": 74.58, "learning_rate": 1.0254545454545453e-07, "loss": 0.2983, "step": 10740 }, { "epoch": 74.65, "learning_rate": 1.0226573426573426e-07, "loss": 0.29, "step": 10750 }, { "epoch": 74.72, "learning_rate": 1.0198601398601399e-07, "loss": 0.3083, "step": 10760 }, { "epoch": 74.79, "learning_rate": 1.0170629370629371e-07, "loss": 0.3061, "step": 10770 }, { "epoch": 74.86, "learning_rate": 1.0142657342657341e-07, "loss": 0.2994, "step": 10780 }, { "epoch": 74.93, "learning_rate": 1.0114685314685314e-07, "loss": 0.3044, "step": 10790 }, { "epoch": 75.0, "learning_rate": 1.0086713286713287e-07, "loss": 0.2931, "step": 10800 }, { "epoch": 75.0, "eval_loss": 0.4587400555610657, "eval_runtime": 126.492, "eval_samples_per_second": 4.206, "eval_steps_per_second": 0.53, "eval_wer": 0.20537378671980458, "step": 10800 }, { "epoch": 75.07, "learning_rate": 1.005874125874126e-07, "loss": 0.3234, "step": 10810 }, { "epoch": 75.14, "learning_rate": 1.003076923076923e-07, "loss": 0.3, "step": 10820 }, { "epoch": 75.21, "learning_rate": 1.0002797202797202e-07, "loss": 0.3215, "step": 10830 }, { "epoch": 75.28, "learning_rate": 9.974825174825174e-08, "loss": 0.2905, "step": 10840 }, { "epoch": 75.35, "learning_rate": 9.946853146853147e-08, "loss": 0.2931, "step": 10850 }, { "epoch": 75.42, "learning_rate": 9.918881118881118e-08, "loss": 0.2915, "step": 10860 }, { "epoch": 75.48, "learning_rate": 9.890909090909091e-08, "loss": 0.3137, "step": 10870 }, { "epoch": 75.55, "learning_rate": 9.862937062937062e-08, "loss": 0.2814, "step": 10880 }, { "epoch": 75.62, "learning_rate": 9.834965034965035e-08, "loss": 0.2893, "step": 10890 }, { "epoch": 75.69, "learning_rate": 9.806993006993006e-08, "loss": 0.2893, "step": 10900 }, { "epoch": 75.76, "learning_rate": 9.779020979020979e-08, "loss": 0.3141, "step": 10910 }, { "epoch": 75.83, "learning_rate": 9.75104895104895e-08, "loss": 0.2749, "step": 10920 }, { "epoch": 75.9, "learning_rate": 9.723076923076923e-08, "loss": 0.2942, "step": 10930 }, { "epoch": 75.97, "learning_rate": 9.695104895104895e-08, "loss": 0.3084, "step": 10940 }, { "epoch": 76.0, "eval_loss": 0.4586826264858246, "eval_runtime": 126.2294, "eval_samples_per_second": 4.215, "eval_steps_per_second": 0.531, "eval_wer": 0.20370251333804718, "step": 10944 }, { "epoch": 76.04, "learning_rate": 9.667132867132867e-08, "loss": 0.3, "step": 10950 }, { "epoch": 76.11, "learning_rate": 9.639160839160839e-08, "loss": 0.2955, "step": 10960 }, { "epoch": 76.18, "learning_rate": 9.61118881118881e-08, "loss": 0.2884, "step": 10970 }, { "epoch": 76.25, "learning_rate": 9.583216783216783e-08, "loss": 0.3223, "step": 10980 }, { "epoch": 76.32, "learning_rate": 9.555244755244754e-08, "loss": 0.3112, "step": 10990 }, { "epoch": 76.39, "learning_rate": 9.527272727272727e-08, "loss": 0.2778, "step": 11000 }, { "epoch": 76.46, "learning_rate": 9.499300699300698e-08, "loss": 0.2967, "step": 11010 }, { "epoch": 76.53, "learning_rate": 9.471328671328671e-08, "loss": 0.2914, "step": 11020 }, { "epoch": 76.6, "learning_rate": 9.443356643356642e-08, "loss": 0.3008, "step": 11030 }, { "epoch": 76.66, "learning_rate": 9.415384615384615e-08, "loss": 0.2846, "step": 11040 }, { "epoch": 76.73, "learning_rate": 9.387412587412587e-08, "loss": 0.289, "step": 11050 }, { "epoch": 76.8, "learning_rate": 9.359440559440559e-08, "loss": 0.2969, "step": 11060 }, { "epoch": 76.87, "learning_rate": 9.331468531468532e-08, "loss": 0.3062, "step": 11070 }, { "epoch": 76.94, "learning_rate": 9.303496503496503e-08, "loss": 0.2957, "step": 11080 }, { "epoch": 77.0, "eval_loss": 0.4589062035083771, "eval_runtime": 126.3352, "eval_samples_per_second": 4.211, "eval_steps_per_second": 0.53, "eval_wer": 0.20479526901073472, "step": 11088 }, { "epoch": 77.01, "learning_rate": 9.275524475524475e-08, "loss": 0.3079, "step": 11090 }, { "epoch": 77.08, "learning_rate": 9.247552447552446e-08, "loss": 0.2769, "step": 11100 }, { "epoch": 77.15, "learning_rate": 9.219580419580419e-08, "loss": 0.2936, "step": 11110 }, { "epoch": 77.22, "learning_rate": 9.19160839160839e-08, "loss": 0.3019, "step": 11120 }, { "epoch": 77.29, "learning_rate": 9.163636363636363e-08, "loss": 0.2989, "step": 11130 }, { "epoch": 77.36, "learning_rate": 9.135664335664335e-08, "loss": 0.3008, "step": 11140 }, { "epoch": 77.43, "learning_rate": 9.107692307692307e-08, "loss": 0.2978, "step": 11150 }, { "epoch": 77.5, "learning_rate": 9.079720279720279e-08, "loss": 0.3022, "step": 11160 }, { "epoch": 77.57, "learning_rate": 9.051748251748251e-08, "loss": 0.3025, "step": 11170 }, { "epoch": 77.64, "learning_rate": 9.023776223776224e-08, "loss": 0.2946, "step": 11180 }, { "epoch": 77.71, "learning_rate": 8.995804195804196e-08, "loss": 0.308, "step": 11190 }, { "epoch": 77.78, "learning_rate": 8.967832167832168e-08, "loss": 0.3092, "step": 11200 }, { "epoch": 77.84, "learning_rate": 8.93986013986014e-08, "loss": 0.2922, "step": 11210 }, { "epoch": 77.91, "learning_rate": 8.911888111888111e-08, "loss": 0.291, "step": 11220 }, { "epoch": 77.98, "learning_rate": 8.883916083916083e-08, "loss": 0.2859, "step": 11230 }, { "epoch": 78.0, "eval_loss": 0.4582020044326782, "eval_runtime": 127.183, "eval_samples_per_second": 4.183, "eval_steps_per_second": 0.527, "eval_wer": 0.20511666773799575, "step": 11232 }, { "epoch": 78.06, "learning_rate": 8.855944055944055e-08, "loss": 0.3064, "step": 11240 }, { "epoch": 78.12, "learning_rate": 8.827972027972027e-08, "loss": 0.2905, "step": 11250 }, { "epoch": 78.19, "learning_rate": 8.8e-08, "loss": 0.2983, "step": 11260 }, { "epoch": 78.26, "learning_rate": 8.772027972027972e-08, "loss": 0.3015, "step": 11270 }, { "epoch": 78.33, "learning_rate": 8.744055944055944e-08, "loss": 0.3039, "step": 11280 }, { "epoch": 78.4, "learning_rate": 8.716083916083916e-08, "loss": 0.2764, "step": 11290 }, { "epoch": 78.47, "learning_rate": 8.688111888111888e-08, "loss": 0.3145, "step": 11300 }, { "epoch": 78.54, "learning_rate": 8.66013986013986e-08, "loss": 0.2833, "step": 11310 }, { "epoch": 78.61, "learning_rate": 8.632167832167832e-08, "loss": 0.2961, "step": 11320 }, { "epoch": 78.68, "learning_rate": 8.604195804195805e-08, "loss": 0.2984, "step": 11330 }, { "epoch": 78.75, "learning_rate": 8.576223776223776e-08, "loss": 0.2833, "step": 11340 }, { "epoch": 78.82, "learning_rate": 8.55104895104895e-08, "loss": 0.3026, "step": 11350 }, { "epoch": 78.89, "learning_rate": 8.523076923076923e-08, "loss": 0.2973, "step": 11360 }, { "epoch": 78.96, "learning_rate": 8.495104895104894e-08, "loss": 0.3045, "step": 11370 }, { "epoch": 79.0, "eval_loss": 0.4591695964336395, "eval_runtime": 127.0668, "eval_samples_per_second": 4.187, "eval_steps_per_second": 0.527, "eval_wer": 0.2044738702834737, "step": 11376 }, { "epoch": 79.03, "learning_rate": 8.467132867132867e-08, "loss": 0.3105, "step": 11380 }, { "epoch": 79.1, "learning_rate": 8.43916083916084e-08, "loss": 0.2964, "step": 11390 }, { "epoch": 79.17, "learning_rate": 8.411188811188811e-08, "loss": 0.2837, "step": 11400 }, { "epoch": 79.24, "learning_rate": 8.383216783216783e-08, "loss": 0.3098, "step": 11410 }, { "epoch": 79.3, "learning_rate": 8.355244755244755e-08, "loss": 0.2948, "step": 11420 }, { "epoch": 79.37, "learning_rate": 8.327272727272727e-08, "loss": 0.2814, "step": 11430 }, { "epoch": 79.44, "learning_rate": 8.299300699300698e-08, "loss": 0.3028, "step": 11440 }, { "epoch": 79.51, "learning_rate": 8.271328671328671e-08, "loss": 0.3016, "step": 11450 }, { "epoch": 79.58, "learning_rate": 8.243356643356642e-08, "loss": 0.2938, "step": 11460 }, { "epoch": 79.65, "learning_rate": 8.215384615384615e-08, "loss": 0.2807, "step": 11470 }, { "epoch": 79.72, "learning_rate": 8.187412587412586e-08, "loss": 0.3153, "step": 11480 }, { "epoch": 79.79, "learning_rate": 8.159440559440559e-08, "loss": 0.2872, "step": 11490 }, { "epoch": 79.86, "learning_rate": 8.131468531468532e-08, "loss": 0.2997, "step": 11500 }, { "epoch": 79.93, "learning_rate": 8.103496503496503e-08, "loss": 0.2976, "step": 11510 }, { "epoch": 80.0, "learning_rate": 8.075524475524476e-08, "loss": 0.3131, "step": 11520 }, { "epoch": 80.0, "eval_loss": 0.4580213129520416, "eval_runtime": 126.5789, "eval_samples_per_second": 4.203, "eval_steps_per_second": 0.529, "eval_wer": 0.20511666773799575, "step": 11520 }, { "epoch": 80.07, "learning_rate": 8.047552447552447e-08, "loss": 0.3032, "step": 11530 }, { "epoch": 80.14, "learning_rate": 8.01958041958042e-08, "loss": 0.2866, "step": 11540 }, { "epoch": 80.21, "learning_rate": 7.991608391608392e-08, "loss": 0.2993, "step": 11550 }, { "epoch": 80.28, "learning_rate": 7.963636363636363e-08, "loss": 0.2936, "step": 11560 }, { "epoch": 80.35, "learning_rate": 7.935664335664334e-08, "loss": 0.2909, "step": 11570 }, { "epoch": 80.42, "learning_rate": 7.907692307692307e-08, "loss": 0.3043, "step": 11580 }, { "epoch": 80.48, "learning_rate": 7.879720279720279e-08, "loss": 0.2954, "step": 11590 }, { "epoch": 80.55, "learning_rate": 7.851748251748251e-08, "loss": 0.2741, "step": 11600 }, { "epoch": 80.62, "learning_rate": 7.823776223776224e-08, "loss": 0.298, "step": 11610 }, { "epoch": 80.69, "learning_rate": 7.795804195804195e-08, "loss": 0.2988, "step": 11620 }, { "epoch": 80.76, "learning_rate": 7.767832167832168e-08, "loss": 0.2969, "step": 11630 }, { "epoch": 80.83, "learning_rate": 7.73986013986014e-08, "loss": 0.3072, "step": 11640 }, { "epoch": 80.9, "learning_rate": 7.711888111888112e-08, "loss": 0.2853, "step": 11650 }, { "epoch": 80.97, "learning_rate": 7.683916083916084e-08, "loss": 0.313, "step": 11660 }, { "epoch": 81.0, "eval_loss": 0.4582548439502716, "eval_runtime": 126.5277, "eval_samples_per_second": 4.205, "eval_steps_per_second": 0.53, "eval_wer": 0.20479526901073472, "step": 11664 }, { "epoch": 81.04, "learning_rate": 7.655944055944056e-08, "loss": 0.2996, "step": 11670 }, { "epoch": 81.11, "learning_rate": 7.627972027972028e-08, "loss": 0.2934, "step": 11680 }, { "epoch": 81.18, "learning_rate": 7.599999999999999e-08, "loss": 0.2945, "step": 11690 }, { "epoch": 81.25, "learning_rate": 7.57202797202797e-08, "loss": 0.3071, "step": 11700 }, { "epoch": 81.32, "learning_rate": 7.544055944055943e-08, "loss": 0.2838, "step": 11710 }, { "epoch": 81.39, "learning_rate": 7.516083916083916e-08, "loss": 0.2683, "step": 11720 }, { "epoch": 81.46, "learning_rate": 7.488111888111888e-08, "loss": 0.288, "step": 11730 }, { "epoch": 81.53, "learning_rate": 7.46013986013986e-08, "loss": 0.2999, "step": 11740 }, { "epoch": 81.6, "learning_rate": 7.432167832167832e-08, "loss": 0.2984, "step": 11750 }, { "epoch": 81.66, "learning_rate": 7.404195804195804e-08, "loss": 0.297, "step": 11760 }, { "epoch": 81.73, "learning_rate": 7.376223776223776e-08, "loss": 0.3132, "step": 11770 }, { "epoch": 81.8, "learning_rate": 7.348251748251749e-08, "loss": 0.2935, "step": 11780 }, { "epoch": 81.87, "learning_rate": 7.32027972027972e-08, "loss": 0.2914, "step": 11790 }, { "epoch": 81.94, "learning_rate": 7.292307692307693e-08, "loss": 0.2902, "step": 11800 }, { "epoch": 82.0, "eval_loss": 0.45861685276031494, "eval_runtime": 127.3264, "eval_samples_per_second": 4.178, "eval_steps_per_second": 0.526, "eval_wer": 0.20498810824709135, "step": 11808 }, { "epoch": 82.01, "learning_rate": 7.264335664335664e-08, "loss": 0.3112, "step": 11810 }, { "epoch": 82.08, "learning_rate": 7.236363636363635e-08, "loss": 0.2958, "step": 11820 }, { "epoch": 82.15, "learning_rate": 7.208391608391608e-08, "loss": 0.2857, "step": 11830 }, { "epoch": 82.22, "learning_rate": 7.18041958041958e-08, "loss": 0.3152, "step": 11840 }, { "epoch": 82.29, "learning_rate": 7.152447552447552e-08, "loss": 0.2764, "step": 11850 }, { "epoch": 82.36, "learning_rate": 7.124475524475524e-08, "loss": 0.2806, "step": 11860 }, { "epoch": 82.43, "learning_rate": 7.096503496503496e-08, "loss": 0.3162, "step": 11870 }, { "epoch": 82.5, "learning_rate": 7.068531468531468e-08, "loss": 0.303, "step": 11880 }, { "epoch": 82.57, "learning_rate": 7.04055944055944e-08, "loss": 0.3094, "step": 11890 }, { "epoch": 82.64, "learning_rate": 7.012587412587412e-08, "loss": 0.2803, "step": 11900 }, { "epoch": 82.71, "learning_rate": 6.984615384615385e-08, "loss": 0.2915, "step": 11910 }, { "epoch": 82.78, "learning_rate": 6.956643356643356e-08, "loss": 0.3082, "step": 11920 }, { "epoch": 82.84, "learning_rate": 6.928671328671329e-08, "loss": 0.3073, "step": 11930 }, { "epoch": 82.91, "learning_rate": 6.9006993006993e-08, "loss": 0.2799, "step": 11940 }, { "epoch": 82.98, "learning_rate": 6.872727272727272e-08, "loss": 0.3097, "step": 11950 }, { "epoch": 83.0, "eval_loss": 0.4590517282485962, "eval_runtime": 126.6347, "eval_samples_per_second": 4.201, "eval_steps_per_second": 0.529, "eval_wer": 0.20511666773799575, "step": 11952 }, { "epoch": 83.06, "learning_rate": 6.844755244755244e-08, "loss": 0.3045, "step": 11960 }, { "epoch": 83.12, "learning_rate": 6.816783216783216e-08, "loss": 0.297, "step": 11970 }, { "epoch": 83.19, "learning_rate": 6.788811188811189e-08, "loss": 0.2961, "step": 11980 }, { "epoch": 83.26, "learning_rate": 6.76083916083916e-08, "loss": 0.2974, "step": 11990 }, { "epoch": 83.33, "learning_rate": 6.732867132867133e-08, "loss": 0.289, "step": 12000 }, { "epoch": 83.4, "learning_rate": 6.704895104895104e-08, "loss": 0.2823, "step": 12010 }, { "epoch": 83.47, "learning_rate": 6.676923076923077e-08, "loss": 0.309, "step": 12020 }, { "epoch": 83.54, "learning_rate": 6.64895104895105e-08, "loss": 0.2801, "step": 12030 }, { "epoch": 83.61, "learning_rate": 6.620979020979021e-08, "loss": 0.2827, "step": 12040 }, { "epoch": 83.68, "learning_rate": 6.593006993006994e-08, "loss": 0.3007, "step": 12050 }, { "epoch": 83.75, "learning_rate": 6.565034965034965e-08, "loss": 0.2869, "step": 12060 }, { "epoch": 83.82, "learning_rate": 6.537062937062937e-08, "loss": 0.2979, "step": 12070 }, { "epoch": 83.89, "learning_rate": 6.509090909090908e-08, "loss": 0.2914, "step": 12080 }, { "epoch": 83.96, "learning_rate": 6.481118881118881e-08, "loss": 0.3045, "step": 12090 }, { "epoch": 84.0, "eval_loss": 0.45935899019241333, "eval_runtime": 127.2419, "eval_samples_per_second": 4.181, "eval_steps_per_second": 0.527, "eval_wer": 0.20505238799254355, "step": 12096 }, { "epoch": 84.03, "learning_rate": 6.453146853146852e-08, "loss": 0.321, "step": 12100 }, { "epoch": 84.1, "learning_rate": 6.425174825174825e-08, "loss": 0.284, "step": 12110 }, { "epoch": 84.17, "learning_rate": 6.397202797202796e-08, "loss": 0.2901, "step": 12120 }, { "epoch": 84.24, "learning_rate": 6.369230769230769e-08, "loss": 0.3056, "step": 12130 }, { "epoch": 84.3, "learning_rate": 6.341258741258742e-08, "loss": 0.2908, "step": 12140 }, { "epoch": 84.37, "learning_rate": 6.313286713286713e-08, "loss": 0.2987, "step": 12150 }, { "epoch": 84.44, "learning_rate": 6.285314685314686e-08, "loss": 0.3057, "step": 12160 }, { "epoch": 84.51, "learning_rate": 6.257342657342657e-08, "loss": 0.3013, "step": 12170 }, { "epoch": 84.58, "learning_rate": 6.22937062937063e-08, "loss": 0.287, "step": 12180 }, { "epoch": 84.65, "learning_rate": 6.201398601398601e-08, "loss": 0.2654, "step": 12190 }, { "epoch": 84.72, "learning_rate": 6.173426573426573e-08, "loss": 0.3031, "step": 12200 }, { "epoch": 84.79, "learning_rate": 6.145454545454544e-08, "loss": 0.3018, "step": 12210 }, { "epoch": 84.86, "learning_rate": 6.117482517482517e-08, "loss": 0.2994, "step": 12220 }, { "epoch": 84.93, "learning_rate": 6.089510489510488e-08, "loss": 0.299, "step": 12230 }, { "epoch": 85.0, "learning_rate": 6.061538461538461e-08, "loss": 0.3046, "step": 12240 }, { "epoch": 85.0, "eval_loss": 0.45976194739341736, "eval_runtime": 126.3075, "eval_samples_per_second": 4.212, "eval_steps_per_second": 0.53, "eval_wer": 0.20479526901073472, "step": 12240 }, { "epoch": 85.07, "learning_rate": 6.033566433566434e-08, "loss": 0.2888, "step": 12250 }, { "epoch": 85.14, "learning_rate": 6.005594405594405e-08, "loss": 0.2905, "step": 12260 }, { "epoch": 85.21, "learning_rate": 5.977622377622378e-08, "loss": 0.3019, "step": 12270 }, { "epoch": 85.28, "learning_rate": 5.949650349650349e-08, "loss": 0.2855, "step": 12280 }, { "epoch": 85.35, "learning_rate": 5.9216783216783214e-08, "loss": 0.3031, "step": 12290 }, { "epoch": 85.42, "learning_rate": 5.8937062937062935e-08, "loss": 0.3091, "step": 12300 }, { "epoch": 85.48, "learning_rate": 5.8657342657342655e-08, "loss": 0.3089, "step": 12310 }, { "epoch": 85.55, "learning_rate": 5.837762237762237e-08, "loss": 0.2853, "step": 12320 }, { "epoch": 85.62, "learning_rate": 5.8097902097902097e-08, "loss": 0.2881, "step": 12330 }, { "epoch": 85.69, "learning_rate": 5.781818181818181e-08, "loss": 0.2943, "step": 12340 }, { "epoch": 85.76, "learning_rate": 5.753846153846154e-08, "loss": 0.3018, "step": 12350 }, { "epoch": 85.83, "learning_rate": 5.725874125874126e-08, "loss": 0.3036, "step": 12360 }, { "epoch": 85.9, "learning_rate": 5.697902097902097e-08, "loss": 0.2783, "step": 12370 }, { "epoch": 85.97, "learning_rate": 5.66993006993007e-08, "loss": 0.2991, "step": 12380 }, { "epoch": 86.0, "eval_loss": 0.46038004755973816, "eval_runtime": 125.9558, "eval_samples_per_second": 4.224, "eval_steps_per_second": 0.532, "eval_wer": 0.2045381500289259, "step": 12384 }, { "epoch": 86.04, "learning_rate": 5.6419580419580414e-08, "loss": 0.2993, "step": 12390 }, { "epoch": 86.11, "learning_rate": 5.613986013986014e-08, "loss": 0.2887, "step": 12400 }, { "epoch": 86.18, "learning_rate": 5.5860139860139856e-08, "loss": 0.3004, "step": 12410 }, { "epoch": 86.25, "learning_rate": 5.558041958041958e-08, "loss": 0.3115, "step": 12420 }, { "epoch": 86.32, "learning_rate": 5.53006993006993e-08, "loss": 0.2978, "step": 12430 }, { "epoch": 86.39, "learning_rate": 5.502097902097902e-08, "loss": 0.2807, "step": 12440 }, { "epoch": 86.46, "learning_rate": 5.474125874125873e-08, "loss": 0.3022, "step": 12450 }, { "epoch": 86.53, "learning_rate": 5.446153846153846e-08, "loss": 0.2917, "step": 12460 }, { "epoch": 86.6, "learning_rate": 5.4181818181818186e-08, "loss": 0.2789, "step": 12470 }, { "epoch": 86.66, "learning_rate": 5.39020979020979e-08, "loss": 0.2849, "step": 12480 }, { "epoch": 86.73, "learning_rate": 5.362237762237762e-08, "loss": 0.3053, "step": 12490 }, { "epoch": 86.8, "learning_rate": 5.3342657342657335e-08, "loss": 0.3044, "step": 12500 }, { "epoch": 86.87, "learning_rate": 5.306293706293706e-08, "loss": 0.2877, "step": 12510 }, { "epoch": 86.94, "learning_rate": 5.2783216783216777e-08, "loss": 0.3083, "step": 12520 }, { "epoch": 87.0, "eval_loss": 0.4601919949054718, "eval_runtime": 126.7865, "eval_samples_per_second": 4.196, "eval_steps_per_second": 0.528, "eval_wer": 0.20524522722890018, "step": 12528 }, { "epoch": 87.01, "learning_rate": 5.2503496503496504e-08, "loss": 0.3103, "step": 12530 }, { "epoch": 87.08, "learning_rate": 5.222377622377622e-08, "loss": 0.3036, "step": 12540 }, { "epoch": 87.15, "learning_rate": 5.1944055944055945e-08, "loss": 0.2977, "step": 12550 }, { "epoch": 87.22, "learning_rate": 5.166433566433566e-08, "loss": 0.3073, "step": 12560 }, { "epoch": 87.29, "learning_rate": 5.138461538461538e-08, "loss": 0.2875, "step": 12570 }, { "epoch": 87.36, "learning_rate": 5.110489510489511e-08, "loss": 0.292, "step": 12580 }, { "epoch": 87.43, "learning_rate": 5.082517482517482e-08, "loss": 0.2978, "step": 12590 }, { "epoch": 87.5, "learning_rate": 5.054545454545455e-08, "loss": 0.2978, "step": 12600 }, { "epoch": 87.57, "learning_rate": 5.026573426573426e-08, "loss": 0.2922, "step": 12610 }, { "epoch": 87.64, "learning_rate": 4.9986013986013983e-08, "loss": 0.27, "step": 12620 }, { "epoch": 87.71, "learning_rate": 4.9706293706293704e-08, "loss": 0.2898, "step": 12630 }, { "epoch": 87.78, "learning_rate": 4.9426573426573425e-08, "loss": 0.3014, "step": 12640 }, { "epoch": 87.84, "learning_rate": 4.9146853146853146e-08, "loss": 0.3106, "step": 12650 }, { "epoch": 87.91, "learning_rate": 4.8867132867132866e-08, "loss": 0.2901, "step": 12660 }, { "epoch": 87.98, "learning_rate": 4.858741258741259e-08, "loss": 0.3141, "step": 12670 }, { "epoch": 88.0, "eval_loss": 0.4599737226963043, "eval_runtime": 127.428, "eval_samples_per_second": 4.175, "eval_steps_per_second": 0.526, "eval_wer": 0.20479526901073472, "step": 12672 }, { "epoch": 88.06, "learning_rate": 4.830769230769231e-08, "loss": 0.3111, "step": 12680 }, { "epoch": 88.12, "learning_rate": 4.802797202797202e-08, "loss": 0.2808, "step": 12690 }, { "epoch": 88.19, "learning_rate": 4.774825174825174e-08, "loss": 0.2962, "step": 12700 }, { "epoch": 88.26, "learning_rate": 4.746853146853146e-08, "loss": 0.292, "step": 12710 }, { "epoch": 88.33, "learning_rate": 4.7188811188811184e-08, "loss": 0.3097, "step": 12720 }, { "epoch": 88.4, "learning_rate": 4.690909090909091e-08, "loss": 0.2983, "step": 12730 }, { "epoch": 88.47, "learning_rate": 4.662937062937063e-08, "loss": 0.2983, "step": 12740 }, { "epoch": 88.54, "learning_rate": 4.6349650349650346e-08, "loss": 0.2853, "step": 12750 }, { "epoch": 88.61, "learning_rate": 4.6069930069930066e-08, "loss": 0.2886, "step": 12760 }, { "epoch": 88.68, "learning_rate": 4.579020979020979e-08, "loss": 0.2985, "step": 12770 }, { "epoch": 88.75, "learning_rate": 4.551048951048951e-08, "loss": 0.3029, "step": 12780 }, { "epoch": 88.82, "learning_rate": 4.523076923076923e-08, "loss": 0.2865, "step": 12790 }, { "epoch": 88.89, "learning_rate": 4.495104895104895e-08, "loss": 0.2925, "step": 12800 }, { "epoch": 88.96, "learning_rate": 4.467132867132867e-08, "loss": 0.2919, "step": 12810 }, { "epoch": 89.0, "eval_loss": 0.4596988260746002, "eval_runtime": 127.1935, "eval_samples_per_second": 4.183, "eval_steps_per_second": 0.527, "eval_wer": 0.20518094748344798, "step": 12816 }, { "epoch": 89.03, "learning_rate": 4.4391608391608384e-08, "loss": 0.3105, "step": 12820 }, { "epoch": 89.1, "learning_rate": 4.411188811188811e-08, "loss": 0.3109, "step": 12830 }, { "epoch": 89.17, "learning_rate": 4.383216783216783e-08, "loss": 0.2972, "step": 12840 }, { "epoch": 89.24, "learning_rate": 4.355244755244755e-08, "loss": 0.2944, "step": 12850 }, { "epoch": 89.3, "learning_rate": 4.3272727272727273e-08, "loss": 0.2852, "step": 12860 }, { "epoch": 89.37, "learning_rate": 4.2993006993006994e-08, "loss": 0.2874, "step": 12870 }, { "epoch": 89.44, "learning_rate": 4.271328671328671e-08, "loss": 0.3014, "step": 12880 }, { "epoch": 89.51, "learning_rate": 4.243356643356643e-08, "loss": 0.3053, "step": 12890 }, { "epoch": 89.58, "learning_rate": 4.215384615384615e-08, "loss": 0.2821, "step": 12900 }, { "epoch": 89.65, "learning_rate": 4.187412587412587e-08, "loss": 0.2835, "step": 12910 }, { "epoch": 89.72, "learning_rate": 4.159440559440559e-08, "loss": 0.308, "step": 12920 }, { "epoch": 89.79, "learning_rate": 4.131468531468531e-08, "loss": 0.2973, "step": 12930 }, { "epoch": 89.86, "learning_rate": 4.103496503496503e-08, "loss": 0.2898, "step": 12940 }, { "epoch": 89.93, "learning_rate": 4.075524475524475e-08, "loss": 0.3069, "step": 12950 }, { "epoch": 90.0, "learning_rate": 4.0475524475524474e-08, "loss": 0.2992, "step": 12960 }, { "epoch": 90.0, "eval_loss": 0.46025562286376953, "eval_runtime": 127.2379, "eval_samples_per_second": 4.181, "eval_steps_per_second": 0.527, "eval_wer": 0.20524522722890018, "step": 12960 }, { "epoch": 90.07, "learning_rate": 4.0195804195804194e-08, "loss": 0.2991, "step": 12970 }, { "epoch": 90.14, "learning_rate": 3.9916083916083915e-08, "loss": 0.2884, "step": 12980 }, { "epoch": 90.21, "learning_rate": 3.9636363636363636e-08, "loss": 0.3093, "step": 12990 }, { "epoch": 90.28, "learning_rate": 3.9356643356643356e-08, "loss": 0.2933, "step": 13000 }, { "epoch": 90.35, "learning_rate": 3.907692307692308e-08, "loss": 0.3, "step": 13010 }, { "epoch": 90.42, "learning_rate": 3.879720279720279e-08, "loss": 0.2949, "step": 13020 }, { "epoch": 90.48, "learning_rate": 3.851748251748251e-08, "loss": 0.3077, "step": 13030 }, { "epoch": 90.55, "learning_rate": 3.823776223776223e-08, "loss": 0.2897, "step": 13040 }, { "epoch": 90.62, "learning_rate": 3.795804195804196e-08, "loss": 0.2907, "step": 13050 }, { "epoch": 90.69, "learning_rate": 3.767832167832168e-08, "loss": 0.2963, "step": 13060 }, { "epoch": 90.76, "learning_rate": 3.73986013986014e-08, "loss": 0.2958, "step": 13070 }, { "epoch": 90.83, "learning_rate": 3.7118881118881115e-08, "loss": 0.2953, "step": 13080 }, { "epoch": 90.9, "learning_rate": 3.6839160839160836e-08, "loss": 0.2884, "step": 13090 }, { "epoch": 90.97, "learning_rate": 3.655944055944056e-08, "loss": 0.3039, "step": 13100 }, { "epoch": 91.0, "eval_loss": 0.4601438641548157, "eval_runtime": 127.1189, "eval_samples_per_second": 4.185, "eval_steps_per_second": 0.527, "eval_wer": 0.20537378671980458, "step": 13104 }, { "epoch": 91.04, "learning_rate": 3.627972027972028e-08, "loss": 0.3137, "step": 13110 }, { "epoch": 91.11, "learning_rate": 3.6e-08, "loss": 0.2943, "step": 13120 }, { "epoch": 91.18, "learning_rate": 3.572027972027972e-08, "loss": 0.3063, "step": 13130 }, { "epoch": 91.25, "learning_rate": 3.544055944055944e-08, "loss": 0.2981, "step": 13140 }, { "epoch": 91.32, "learning_rate": 3.516083916083916e-08, "loss": 0.2964, "step": 13150 }, { "epoch": 91.39, "learning_rate": 3.488111888111888e-08, "loss": 0.2825, "step": 13160 }, { "epoch": 91.46, "learning_rate": 3.46013986013986e-08, "loss": 0.2986, "step": 13170 }, { "epoch": 91.53, "learning_rate": 3.432167832167832e-08, "loss": 0.2913, "step": 13180 }, { "epoch": 91.6, "learning_rate": 3.404195804195804e-08, "loss": 0.2865, "step": 13190 }, { "epoch": 91.66, "learning_rate": 3.3762237762237764e-08, "loss": 0.2923, "step": 13200 }, { "epoch": 91.73, "learning_rate": 3.348251748251748e-08, "loss": 0.2836, "step": 13210 }, { "epoch": 91.8, "learning_rate": 3.32027972027972e-08, "loss": 0.2949, "step": 13220 }, { "epoch": 91.87, "learning_rate": 3.292307692307692e-08, "loss": 0.2809, "step": 13230 }, { "epoch": 91.94, "learning_rate": 3.264335664335664e-08, "loss": 0.2992, "step": 13240 }, { "epoch": 92.0, "eval_loss": 0.45991969108581543, "eval_runtime": 126.7835, "eval_samples_per_second": 4.196, "eval_steps_per_second": 0.528, "eval_wer": 0.2054380664652568, "step": 13248 }, { "epoch": 92.01, "learning_rate": 3.236363636363636e-08, "loss": 0.3321, "step": 13250 }, { "epoch": 92.08, "learning_rate": 3.208391608391609e-08, "loss": 0.2888, "step": 13260 }, { "epoch": 92.15, "learning_rate": 3.18041958041958e-08, "loss": 0.2927, "step": 13270 }, { "epoch": 92.22, "learning_rate": 3.152447552447552e-08, "loss": 0.3003, "step": 13280 }, { "epoch": 92.29, "learning_rate": 3.124475524475524e-08, "loss": 0.2845, "step": 13290 }, { "epoch": 92.36, "learning_rate": 3.0965034965034964e-08, "loss": 0.2936, "step": 13300 }, { "epoch": 92.43, "learning_rate": 3.0685314685314685e-08, "loss": 0.3031, "step": 13310 }, { "epoch": 92.5, "learning_rate": 3.0405594405594405e-08, "loss": 0.2987, "step": 13320 }, { "epoch": 92.57, "learning_rate": 3.0125874125874126e-08, "loss": 0.286, "step": 13330 }, { "epoch": 92.64, "learning_rate": 2.984615384615384e-08, "loss": 0.2881, "step": 13340 }, { "epoch": 92.71, "learning_rate": 2.9566433566433564e-08, "loss": 0.3053, "step": 13350 }, { "epoch": 92.78, "learning_rate": 2.928671328671328e-08, "loss": 0.3059, "step": 13360 }, { "epoch": 92.84, "learning_rate": 2.900699300699301e-08, "loss": 0.2942, "step": 13370 }, { "epoch": 92.91, "learning_rate": 2.8727272727272726e-08, "loss": 0.29, "step": 13380 }, { "epoch": 92.98, "learning_rate": 2.8447552447552447e-08, "loss": 0.305, "step": 13390 }, { "epoch": 93.0, "eval_loss": 0.4603525698184967, "eval_runtime": 126.6336, "eval_samples_per_second": 4.201, "eval_steps_per_second": 0.529, "eval_wer": 0.20511666773799575, "step": 13392 }, { "epoch": 93.06, "learning_rate": 2.8167832167832168e-08, "loss": 0.3209, "step": 13400 }, { "epoch": 93.12, "learning_rate": 2.7888111888111888e-08, "loss": 0.2795, "step": 13410 }, { "epoch": 93.19, "learning_rate": 2.7608391608391606e-08, "loss": 0.296, "step": 13420 }, { "epoch": 93.26, "learning_rate": 2.7328671328671326e-08, "loss": 0.2833, "step": 13430 }, { "epoch": 93.33, "learning_rate": 2.7048951048951047e-08, "loss": 0.3023, "step": 13440 }, { "epoch": 93.4, "learning_rate": 2.6769230769230768e-08, "loss": 0.283, "step": 13450 }, { "epoch": 93.47, "learning_rate": 2.6489510489510485e-08, "loss": 0.3023, "step": 13460 }, { "epoch": 93.54, "learning_rate": 2.6209790209790212e-08, "loss": 0.2979, "step": 13470 }, { "epoch": 93.61, "learning_rate": 2.593006993006993e-08, "loss": 0.2946, "step": 13480 }, { "epoch": 93.68, "learning_rate": 2.565034965034965e-08, "loss": 0.2985, "step": 13490 }, { "epoch": 93.75, "learning_rate": 2.537062937062937e-08, "loss": 0.3178, "step": 13500 }, { "epoch": 93.82, "learning_rate": 2.5090909090909092e-08, "loss": 0.2894, "step": 13510 }, { "epoch": 93.89, "learning_rate": 2.481118881118881e-08, "loss": 0.2838, "step": 13520 }, { "epoch": 93.96, "learning_rate": 2.453146853146853e-08, "loss": 0.2952, "step": 13530 }, { "epoch": 94.0, "eval_loss": 0.4603303372859955, "eval_runtime": 126.7335, "eval_samples_per_second": 4.198, "eval_steps_per_second": 0.529, "eval_wer": 0.2062737031561355, "step": 13536 }, { "epoch": 94.03, "learning_rate": 2.425174825174825e-08, "loss": 0.3216, "step": 13540 }, { "epoch": 94.1, "learning_rate": 2.397202797202797e-08, "loss": 0.2931, "step": 13550 }, { "epoch": 94.17, "learning_rate": 2.3692307692307692e-08, "loss": 0.2902, "step": 13560 }, { "epoch": 94.24, "learning_rate": 2.3412587412587413e-08, "loss": 0.3066, "step": 13570 }, { "epoch": 94.3, "learning_rate": 2.313286713286713e-08, "loss": 0.2903, "step": 13580 }, { "epoch": 94.37, "learning_rate": 2.285314685314685e-08, "loss": 0.2791, "step": 13590 }, { "epoch": 94.44, "learning_rate": 2.26013986013986e-08, "loss": 0.3004, "step": 13600 }, { "epoch": 94.51, "learning_rate": 2.232167832167832e-08, "loss": 0.2829, "step": 13610 }, { "epoch": 94.58, "learning_rate": 2.204195804195804e-08, "loss": 0.2881, "step": 13620 }, { "epoch": 94.65, "learning_rate": 2.1762237762237762e-08, "loss": 0.2844, "step": 13630 }, { "epoch": 94.72, "learning_rate": 2.1482517482517483e-08, "loss": 0.3061, "step": 13640 }, { "epoch": 94.79, "learning_rate": 2.12027972027972e-08, "loss": 0.2945, "step": 13650 }, { "epoch": 94.86, "learning_rate": 2.092307692307692e-08, "loss": 0.3003, "step": 13660 }, { "epoch": 94.93, "learning_rate": 2.0643356643356645e-08, "loss": 0.3008, "step": 13670 }, { "epoch": 95.0, "learning_rate": 2.0363636363636362e-08, "loss": 0.2989, "step": 13680 }, { "epoch": 95.0, "eval_loss": 0.46059897541999817, "eval_runtime": 126.9134, "eval_samples_per_second": 4.192, "eval_steps_per_second": 0.528, "eval_wer": 0.20492382850163912, "step": 13680 }, { "epoch": 95.07, "learning_rate": 2.0083916083916083e-08, "loss": 0.2972, "step": 13690 }, { "epoch": 95.14, "learning_rate": 1.9804195804195804e-08, "loss": 0.2765, "step": 13700 }, { "epoch": 95.21, "learning_rate": 1.9524475524475524e-08, "loss": 0.3002, "step": 13710 }, { "epoch": 95.28, "learning_rate": 1.9244755244755245e-08, "loss": 0.2924, "step": 13720 }, { "epoch": 95.35, "learning_rate": 1.8965034965034966e-08, "loss": 0.2927, "step": 13730 }, { "epoch": 95.42, "learning_rate": 1.8685314685314686e-08, "loss": 0.2955, "step": 13740 }, { "epoch": 95.48, "learning_rate": 1.8405594405594404e-08, "loss": 0.303, "step": 13750 }, { "epoch": 95.55, "learning_rate": 1.8125874125874124e-08, "loss": 0.3026, "step": 13760 }, { "epoch": 95.62, "learning_rate": 1.7846153846153845e-08, "loss": 0.2744, "step": 13770 }, { "epoch": 95.69, "learning_rate": 1.7566433566433566e-08, "loss": 0.307, "step": 13780 }, { "epoch": 95.76, "learning_rate": 1.7286713286713286e-08, "loss": 0.2972, "step": 13790 }, { "epoch": 95.83, "learning_rate": 1.7006993006993007e-08, "loss": 0.3019, "step": 13800 }, { "epoch": 95.9, "learning_rate": 1.6727272727272725e-08, "loss": 0.3035, "step": 13810 }, { "epoch": 95.97, "learning_rate": 1.6447552447552445e-08, "loss": 0.3033, "step": 13820 }, { "epoch": 96.0, "eval_loss": 0.460376113653183, "eval_runtime": 126.7115, "eval_samples_per_second": 4.199, "eval_steps_per_second": 0.529, "eval_wer": 0.20518094748344798, "step": 13824 }, { "epoch": 96.04, "learning_rate": 1.616783216783217e-08, "loss": 0.3057, "step": 13830 }, { "epoch": 96.11, "learning_rate": 1.5888111888111887e-08, "loss": 0.2911, "step": 13840 }, { "epoch": 96.18, "learning_rate": 1.5608391608391607e-08, "loss": 0.2967, "step": 13850 }, { "epoch": 96.25, "learning_rate": 1.5328671328671328e-08, "loss": 0.3096, "step": 13860 }, { "epoch": 96.32, "learning_rate": 1.504895104895105e-08, "loss": 0.2862, "step": 13870 }, { "epoch": 96.39, "learning_rate": 1.476923076923077e-08, "loss": 0.274, "step": 13880 }, { "epoch": 96.46, "learning_rate": 1.448951048951049e-08, "loss": 0.3141, "step": 13890 }, { "epoch": 96.53, "learning_rate": 1.4209790209790209e-08, "loss": 0.2927, "step": 13900 }, { "epoch": 96.6, "learning_rate": 1.393006993006993e-08, "loss": 0.2961, "step": 13910 }, { "epoch": 96.66, "learning_rate": 1.3650349650349649e-08, "loss": 0.2855, "step": 13920 }, { "epoch": 96.73, "learning_rate": 1.337062937062937e-08, "loss": 0.3024, "step": 13930 }, { "epoch": 96.8, "learning_rate": 1.3090909090909092e-08, "loss": 0.2958, "step": 13940 }, { "epoch": 96.87, "learning_rate": 1.2811188811188811e-08, "loss": 0.299, "step": 13950 }, { "epoch": 96.94, "learning_rate": 1.2531468531468532e-08, "loss": 0.2861, "step": 13960 }, { "epoch": 97.0, "eval_loss": 0.4600101113319397, "eval_runtime": 127.256, "eval_samples_per_second": 4.181, "eval_steps_per_second": 0.526, "eval_wer": 0.2055666259561612, "step": 13968 }, { "epoch": 97.01, "learning_rate": 1.225174825174825e-08, "loss": 0.2845, "step": 13970 }, { "epoch": 97.08, "learning_rate": 1.1972027972027971e-08, "loss": 0.3046, "step": 13980 }, { "epoch": 97.15, "learning_rate": 1.1692307692307692e-08, "loss": 0.2969, "step": 13990 }, { "epoch": 97.22, "learning_rate": 1.1412587412587413e-08, "loss": 0.3083, "step": 14000 }, { "epoch": 97.29, "learning_rate": 1.1132867132867132e-08, "loss": 0.3, "step": 14010 }, { "epoch": 97.36, "learning_rate": 1.0853146853146852e-08, "loss": 0.275, "step": 14020 }, { "epoch": 97.43, "learning_rate": 1.0573426573426573e-08, "loss": 0.294, "step": 14030 }, { "epoch": 97.5, "learning_rate": 1.0293706293706294e-08, "loss": 0.2889, "step": 14040 }, { "epoch": 97.57, "learning_rate": 1.0013986013986013e-08, "loss": 0.2804, "step": 14050 }, { "epoch": 97.64, "learning_rate": 9.734265734265734e-09, "loss": 0.2737, "step": 14060 }, { "epoch": 97.71, "learning_rate": 9.454545454545454e-09, "loss": 0.3003, "step": 14070 }, { "epoch": 97.78, "learning_rate": 9.174825174825175e-09, "loss": 0.2957, "step": 14080 }, { "epoch": 97.84, "learning_rate": 8.895104895104896e-09, "loss": 0.3133, "step": 14090 }, { "epoch": 97.91, "learning_rate": 8.615384615384615e-09, "loss": 0.2949, "step": 14100 }, { "epoch": 97.98, "learning_rate": 8.335664335664335e-09, "loss": 0.3054, "step": 14110 }, { "epoch": 98.0, "eval_loss": 0.46060049533843994, "eval_runtime": 126.6554, "eval_samples_per_second": 4.2, "eval_steps_per_second": 0.529, "eval_wer": 0.20537378671980458, "step": 14112 }, { "epoch": 98.06, "learning_rate": 8.055944055944056e-09, "loss": 0.2983, "step": 14120 }, { "epoch": 98.12, "learning_rate": 7.776223776223775e-09, "loss": 0.2832, "step": 14130 }, { "epoch": 98.19, "learning_rate": 7.496503496503496e-09, "loss": 0.3105, "step": 14140 }, { "epoch": 98.26, "learning_rate": 7.2167832167832164e-09, "loss": 0.2903, "step": 14150 }, { "epoch": 98.33, "learning_rate": 6.937062937062937e-09, "loss": 0.2966, "step": 14160 }, { "epoch": 98.4, "learning_rate": 6.657342657342657e-09, "loss": 0.2958, "step": 14170 }, { "epoch": 98.47, "learning_rate": 6.377622377622377e-09, "loss": 0.3112, "step": 14180 }, { "epoch": 98.54, "learning_rate": 6.0979020979020976e-09, "loss": 0.286, "step": 14190 }, { "epoch": 98.61, "learning_rate": 5.818181818181818e-09, "loss": 0.2921, "step": 14200 }, { "epoch": 98.68, "learning_rate": 5.538461538461538e-09, "loss": 0.2949, "step": 14210 }, { "epoch": 98.75, "learning_rate": 5.258741258741259e-09, "loss": 0.3012, "step": 14220 }, { "epoch": 98.82, "learning_rate": 4.979020979020979e-09, "loss": 0.2902, "step": 14230 }, { "epoch": 98.89, "learning_rate": 4.699300699300699e-09, "loss": 0.2787, "step": 14240 }, { "epoch": 98.96, "learning_rate": 4.419580419580419e-09, "loss": 0.2957, "step": 14250 }, { "epoch": 99.0, "eval_loss": 0.46088045835494995, "eval_runtime": 126.9714, "eval_samples_per_second": 4.19, "eval_steps_per_second": 0.528, "eval_wer": 0.2054380664652568, "step": 14256 }, { "epoch": 99.03, "learning_rate": 4.13986013986014e-09, "loss": 0.3354, "step": 14260 }, { "epoch": 99.1, "learning_rate": 3.86013986013986e-09, "loss": 0.3068, "step": 14270 }, { "epoch": 99.17, "learning_rate": 3.5804195804195805e-09, "loss": 0.2808, "step": 14280 }, { "epoch": 99.24, "learning_rate": 3.3006993006993007e-09, "loss": 0.297, "step": 14290 }, { "epoch": 99.3, "learning_rate": 3.020979020979021e-09, "loss": 0.2914, "step": 14300 }, { "epoch": 99.37, "learning_rate": 2.7412587412587413e-09, "loss": 0.3023, "step": 14310 }, { "epoch": 99.44, "learning_rate": 2.4615384615384616e-09, "loss": 0.3025, "step": 14320 }, { "epoch": 99.51, "learning_rate": 2.181818181818182e-09, "loss": 0.303, "step": 14330 }, { "epoch": 99.58, "learning_rate": 1.902097902097902e-09, "loss": 0.2888, "step": 14340 }, { "epoch": 99.65, "learning_rate": 1.6223776223776224e-09, "loss": 0.285, "step": 14350 }, { "epoch": 99.72, "learning_rate": 1.3426573426573427e-09, "loss": 0.2976, "step": 14360 }, { "epoch": 99.79, "learning_rate": 1.062937062937063e-09, "loss": 0.2858, "step": 14370 }, { "epoch": 99.86, "learning_rate": 7.832167832167832e-10, "loss": 0.288, "step": 14380 }, { "epoch": 99.93, "learning_rate": 5.034965034965035e-10, "loss": 0.3065, "step": 14390 }, { "epoch": 100.0, "learning_rate": 2.2377622377622378e-10, "loss": 0.3004, "step": 14400 }, { "epoch": 100.0, "eval_loss": 0.4606250822544098, "eval_runtime": 127.4765, "eval_samples_per_second": 4.173, "eval_steps_per_second": 0.526, "eval_wer": 0.2062737031561355, "step": 14400 }, { "epoch": 100.0, "step": 14400, "total_flos": 2.720292531062204e+20, "train_loss": 0.29647377184695667, "train_runtime": 190063.067, "train_samples_per_second": 2.429, "train_steps_per_second": 0.076 } ], "max_steps": 14400, "num_train_epochs": 100, "total_flos": 2.720292531062204e+20, "trial_name": null, "trial_params": null }