diff --git "a/trainer_state.json" "b/trainer_state.json" --- "a/trainer_state.json" +++ "b/trainer_state.json" @@ -1,7141 +1,3661 @@ { - "best_metric": 0.6008606553077698, - "best_model_checkpoint": "ai-light-dance_singing3_ft_wav2vec2-large-xlsr-53-v1/checkpoint-11560", - "epoch": 20.0, - "global_step": 11560, + "best_metric": 0.5690802335739136, + "best_model_checkpoint": "ai-light-dance_singing3_ft_wav2vec2-large-xlsr-53-v1/checkpoint-3168", + "epoch": 19.997402597402598, + "global_step": 5760, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { - "epoch": 0.02, - "learning_rate": 1.0000000000000002e-06, - "loss": 0.6655, + "epoch": 0.03, + "learning_rate": 1.0000000000000001e-07, + "loss": 0.5283, "step": 10 }, { - "epoch": 0.03, - "learning_rate": 2.0000000000000003e-06, - "loss": 0.6386, + "epoch": 0.07, + "learning_rate": 2.0000000000000002e-07, + "loss": 0.528, "step": 20 }, { - "epoch": 0.05, - "learning_rate": 3e-06, - "loss": 0.6362, + "epoch": 0.1, + "learning_rate": 2.9000000000000003e-07, + "loss": 0.5047, "step": 30 }, { - "epoch": 0.07, - "learning_rate": 3.900000000000001e-06, - "loss": 0.6472, + "epoch": 0.14, + "learning_rate": 3.9e-07, + "loss": 0.5226, "step": 40 }, { - "epoch": 0.09, - "learning_rate": 4.9000000000000005e-06, - "loss": 0.6737, + "epoch": 0.17, + "learning_rate": 4.900000000000001e-07, + "loss": 0.5815, "step": 50 }, { - "epoch": 0.1, - "learning_rate": 5.9e-06, - "loss": 0.6551, + "epoch": 0.21, + "learning_rate": 5.900000000000001e-07, + "loss": 0.4874, "step": 60 }, { - "epoch": 0.12, - "learning_rate": 6.9e-06, - "loss": 0.6678, + "epoch": 0.24, + "learning_rate": 6.900000000000001e-07, + "loss": 0.5212, "step": 70 }, { - "epoch": 0.14, - "learning_rate": 7.9e-06, - "loss": 0.6285, + "epoch": 0.28, + "learning_rate": 7.900000000000001e-07, + "loss": 0.5171, "step": 80 }, { - "epoch": 0.16, - "learning_rate": 8.900000000000001e-06, - "loss": 0.6751, + "epoch": 0.31, + "learning_rate": 8.900000000000001e-07, + "loss": 0.5332, "step": 90 }, { - "epoch": 0.17, - "learning_rate": 9.9e-06, - "loss": 0.7244, + "epoch": 0.35, + "learning_rate": 9.9e-07, + "loss": 0.563, "step": 100 }, { - "epoch": 0.19, - "learning_rate": 9.99214659685864e-06, - "loss": 0.6706, + "epoch": 0.38, + "learning_rate": 1.0900000000000002e-06, + "loss": 0.5127, "step": 110 }, { - "epoch": 0.21, - "learning_rate": 9.983420593368238e-06, - "loss": 0.6412, + "epoch": 0.42, + "learning_rate": 1.19e-06, + "loss": 0.5071, "step": 120 }, { - "epoch": 0.22, - "learning_rate": 9.974694589877837e-06, - "loss": 0.6213, + "epoch": 0.45, + "learning_rate": 1.2900000000000001e-06, + "loss": 0.5518, "step": 130 }, { - "epoch": 0.24, - "learning_rate": 9.965968586387436e-06, - "loss": 0.6782, + "epoch": 0.48, + "learning_rate": 1.3900000000000002e-06, + "loss": 0.5234, "step": 140 }, { - "epoch": 0.26, - "learning_rate": 9.958115183246074e-06, - "loss": 0.7375, + "epoch": 0.52, + "learning_rate": 1.4900000000000001e-06, + "loss": 0.5274, "step": 150 }, { - "epoch": 0.28, - "learning_rate": 9.949389179755673e-06, - "loss": 0.731, + "epoch": 0.55, + "learning_rate": 1.5900000000000002e-06, + "loss": 0.529, "step": 160 }, { - "epoch": 0.29, - "learning_rate": 9.940663176265271e-06, - "loss": 0.7039, + "epoch": 0.59, + "learning_rate": 1.6900000000000003e-06, + "loss": 0.5659, "step": 170 }, { - "epoch": 0.31, - "learning_rate": 9.93193717277487e-06, - "loss": 0.6615, + "epoch": 0.62, + "learning_rate": 1.79e-06, + "loss": 0.4647, "step": 180 }, { - "epoch": 0.33, - "learning_rate": 9.92321116928447e-06, - "loss": 0.6966, + "epoch": 0.66, + "learning_rate": 1.8900000000000001e-06, + "loss": 0.4986, "step": 190 }, { - "epoch": 0.35, - "learning_rate": 9.914485165794067e-06, - "loss": 0.7347, + "epoch": 0.69, + "learning_rate": 1.9900000000000004e-06, + "loss": 0.5279, "step": 200 }, { - "epoch": 0.36, - "learning_rate": 9.905759162303665e-06, - "loss": 0.6558, + "epoch": 0.73, + "learning_rate": 2.09e-06, + "loss": 0.5265, "step": 210 }, { - "epoch": 0.38, - "learning_rate": 9.897033158813264e-06, - "loss": 0.6703, + "epoch": 0.76, + "learning_rate": 2.19e-06, + "loss": 0.5067, "step": 220 }, { - "epoch": 0.4, - "learning_rate": 9.888307155322863e-06, - "loss": 0.7173, + "epoch": 0.8, + "learning_rate": 2.29e-06, + "loss": 0.4832, "step": 230 }, { - "epoch": 0.42, - "learning_rate": 9.87958115183246e-06, - "loss": 0.6881, + "epoch": 0.83, + "learning_rate": 2.39e-06, + "loss": 0.5363, "step": 240 }, { - "epoch": 0.43, - "learning_rate": 9.87085514834206e-06, - "loss": 0.6411, + "epoch": 0.87, + "learning_rate": 2.4900000000000003e-06, + "loss": 0.5589, "step": 250 }, { - "epoch": 0.45, - "learning_rate": 9.86212914485166e-06, - "loss": 0.7406, + "epoch": 0.9, + "learning_rate": 2.59e-06, + "loss": 0.5353, "step": 260 }, { - "epoch": 0.47, - "learning_rate": 9.853403141361257e-06, - "loss": 0.7658, + "epoch": 0.94, + "learning_rate": 2.6900000000000005e-06, + "loss": 0.5202, "step": 270 }, { - "epoch": 0.48, - "learning_rate": 9.844677137870856e-06, - "loss": 0.7319, + "epoch": 0.97, + "learning_rate": 2.7900000000000004e-06, + "loss": 0.5352, "step": 280 }, { - "epoch": 0.5, - "learning_rate": 9.835951134380455e-06, - "loss": 0.6201, + "epoch": 1.0, + "eval_loss": 0.602483868598938, + "eval_runtime": 129.9404, + "eval_samples_per_second": 4.094, + "eval_steps_per_second": 1.024, + "eval_wer": 0.3327119624606287, + "step": 288 + }, + { + "epoch": 1.01, + "learning_rate": 2.89e-06, + "loss": 0.5971, "step": 290 }, { - "epoch": 0.52, - "learning_rate": 9.827225130890053e-06, - "loss": 0.7429, + "epoch": 1.04, + "learning_rate": 2.99e-06, + "loss": 0.5174, "step": 300 }, { - "epoch": 0.54, - "learning_rate": 9.818499127399652e-06, - "loss": 0.6145, + "epoch": 1.08, + "learning_rate": 3.09e-06, + "loss": 0.5195, "step": 310 }, { - "epoch": 0.55, - "learning_rate": 9.809773123909251e-06, - "loss": 0.7034, + "epoch": 1.11, + "learning_rate": 3.1900000000000004e-06, + "loss": 0.5103, "step": 320 }, { - "epoch": 0.57, - "learning_rate": 9.801047120418849e-06, - "loss": 0.5801, + "epoch": 1.15, + "learning_rate": 3.2900000000000003e-06, + "loss": 0.546, "step": 330 }, { - "epoch": 0.59, - "learning_rate": 9.792321116928447e-06, - "loss": 0.6591, + "epoch": 1.18, + "learning_rate": 3.3900000000000006e-06, + "loss": 0.5034, "step": 340 }, { - "epoch": 0.61, - "learning_rate": 9.783595113438046e-06, - "loss": 0.7283, + "epoch": 1.21, + "learning_rate": 3.49e-06, + "loss": 0.4854, "step": 350 }, { - "epoch": 0.62, - "learning_rate": 9.774869109947645e-06, - "loss": 0.6782, + "epoch": 1.25, + "learning_rate": 3.5900000000000004e-06, + "loss": 0.4965, "step": 360 }, { - "epoch": 0.64, - "learning_rate": 9.766143106457243e-06, - "loss": 0.6953, + "epoch": 1.28, + "learning_rate": 3.6900000000000002e-06, + "loss": 0.4936, "step": 370 }, { - "epoch": 0.66, - "learning_rate": 9.757417102966842e-06, - "loss": 0.6408, + "epoch": 1.32, + "learning_rate": 3.79e-06, + "loss": 0.5177, "step": 380 }, { - "epoch": 0.67, - "learning_rate": 9.748691099476441e-06, - "loss": 0.6515, + "epoch": 1.35, + "learning_rate": 3.89e-06, + "loss": 0.5723, "step": 390 }, { - "epoch": 0.69, - "learning_rate": 9.739965095986039e-06, - "loss": 0.6411, + "epoch": 1.39, + "learning_rate": 3.990000000000001e-06, + "loss": 0.5412, "step": 400 }, { - "epoch": 0.71, - "learning_rate": 9.731239092495638e-06, - "loss": 0.7034, + "epoch": 1.42, + "learning_rate": 4.09e-06, + "loss": 0.5021, "step": 410 }, { - "epoch": 0.73, - "learning_rate": 9.722513089005237e-06, - "loss": 0.649, + "epoch": 1.46, + "learning_rate": 4.1900000000000005e-06, + "loss": 0.5059, "step": 420 }, { - "epoch": 0.74, - "learning_rate": 9.713787085514835e-06, - "loss": 0.6486, + "epoch": 1.49, + "learning_rate": 4.2900000000000004e-06, + "loss": 0.5326, "step": 430 }, { - "epoch": 0.76, - "learning_rate": 9.705061082024434e-06, - "loss": 0.7005, + "epoch": 1.53, + "learning_rate": 4.39e-06, + "loss": 0.5409, "step": 440 }, { - "epoch": 0.78, - "learning_rate": 9.696335078534033e-06, - "loss": 0.7006, + "epoch": 1.56, + "learning_rate": 4.49e-06, + "loss": 0.5114, "step": 450 }, { - "epoch": 0.8, - "learning_rate": 9.687609075043631e-06, - "loss": 0.6438, + "epoch": 1.6, + "learning_rate": 4.590000000000001e-06, + "loss": 0.5381, "step": 460 }, { - "epoch": 0.81, - "learning_rate": 9.678883071553228e-06, - "loss": 0.6592, + "epoch": 1.63, + "learning_rate": 4.69e-06, + "loss": 0.5171, "step": 470 }, { - "epoch": 0.83, - "learning_rate": 9.670157068062828e-06, - "loss": 0.6678, + "epoch": 1.66, + "learning_rate": 4.79e-06, + "loss": 0.5357, "step": 480 }, { - "epoch": 0.85, - "learning_rate": 9.661431064572427e-06, - "loss": 0.683, + "epoch": 1.7, + "learning_rate": 4.890000000000001e-06, + "loss": 0.567, "step": 490 }, { - "epoch": 0.87, - "learning_rate": 9.652705061082025e-06, - "loss": 0.7025, + "epoch": 1.73, + "learning_rate": 4.9900000000000005e-06, + "loss": 0.5421, "step": 500 }, { - "epoch": 0.88, - "learning_rate": 9.644851657940664e-06, - "loss": 0.7112, + "epoch": 1.77, + "learning_rate": 5.09e-06, + "loss": 0.5293, "step": 510 }, { - "epoch": 0.9, - "learning_rate": 9.636125654450262e-06, - "loss": 0.6762, + "epoch": 1.8, + "learning_rate": 5.19e-06, + "loss": 0.523, "step": 520 }, { - "epoch": 0.92, - "learning_rate": 9.62739965095986e-06, - "loss": 0.6436, + "epoch": 1.84, + "learning_rate": 5.290000000000001e-06, + "loss": 0.5464, "step": 530 }, { - "epoch": 0.93, - "learning_rate": 9.61867364746946e-06, - "loss": 0.6524, + "epoch": 1.87, + "learning_rate": 5.390000000000001e-06, + "loss": 0.5473, "step": 540 }, { - "epoch": 0.95, - "learning_rate": 9.609947643979058e-06, - "loss": 0.7047, + "epoch": 1.91, + "learning_rate": 5.490000000000001e-06, + "loss": 0.5302, "step": 550 }, { - "epoch": 0.97, - "learning_rate": 9.601221640488657e-06, - "loss": 0.6413, + "epoch": 1.94, + "learning_rate": 5.590000000000001e-06, + "loss": 0.506, "step": 560 }, { - "epoch": 0.99, - "learning_rate": 9.592495636998256e-06, - "loss": 0.6364, + "epoch": 1.98, + "learning_rate": 5.69e-06, + "loss": 0.5365, "step": 570 }, { - "epoch": 1.0, - "eval_loss": 0.6806110143661499, - "eval_runtime": 130.1347, - "eval_samples_per_second": 4.088, - "eval_steps_per_second": 0.515, - "eval_wer": 0.4167255897666645, - "step": 578 + "epoch": 2.0, + "eval_loss": 0.6085500717163086, + "eval_runtime": 133.1913, + "eval_samples_per_second": 3.994, + "eval_steps_per_second": 0.999, + "eval_wer": 0.34801054187825414, + "step": 576 }, { - "epoch": 1.0, - "learning_rate": 9.583769633507854e-06, - "loss": 0.7132, + "epoch": 2.01, + "learning_rate": 5.7900000000000005e-06, + "loss": 0.5768, "step": 580 }, { - "epoch": 1.02, - "learning_rate": 9.575043630017453e-06, - "loss": 0.661, + "epoch": 2.05, + "learning_rate": 5.89e-06, + "loss": 0.541, "step": 590 }, { - "epoch": 1.04, - "learning_rate": 9.566317626527052e-06, - "loss": 0.7, + "epoch": 2.08, + "learning_rate": 5.99e-06, + "loss": 0.5115, "step": 600 }, { - "epoch": 1.06, - "learning_rate": 9.55759162303665e-06, - "loss": 0.6402, + "epoch": 2.12, + "learning_rate": 6.09e-06, + "loss": 0.5329, "step": 610 }, { - "epoch": 1.07, - "learning_rate": 9.548865619546247e-06, - "loss": 0.6832, + "epoch": 2.15, + "learning_rate": 6.190000000000001e-06, + "loss": 0.5549, "step": 620 }, { - "epoch": 1.09, - "learning_rate": 9.540139616055847e-06, - "loss": 0.653, + "epoch": 2.19, + "learning_rate": 6.290000000000001e-06, + "loss": 0.5476, "step": 630 }, { - "epoch": 1.11, - "learning_rate": 9.531413612565446e-06, - "loss": 0.6902, + "epoch": 2.22, + "learning_rate": 6.390000000000001e-06, + "loss": 0.5519, "step": 640 }, { - "epoch": 1.12, - "learning_rate": 9.522687609075043e-06, - "loss": 0.6262, + "epoch": 2.26, + "learning_rate": 6.4900000000000005e-06, + "loss": 0.5358, "step": 650 }, { - "epoch": 1.14, - "learning_rate": 9.513961605584643e-06, - "loss": 0.688, + "epoch": 2.29, + "learning_rate": 6.5900000000000004e-06, + "loss": 0.5354, "step": 660 }, { - "epoch": 1.16, - "learning_rate": 9.505235602094242e-06, - "loss": 0.6777, + "epoch": 2.33, + "learning_rate": 6.690000000000001e-06, + "loss": 0.5671, "step": 670 }, { - "epoch": 1.18, - "learning_rate": 9.49650959860384e-06, - "loss": 0.6755, + "epoch": 2.36, + "learning_rate": 6.790000000000001e-06, + "loss": 0.6002, "step": 680 }, { - "epoch": 1.19, - "learning_rate": 9.487783595113439e-06, - "loss": 0.6325, + "epoch": 2.39, + "learning_rate": 6.89e-06, + "loss": 0.5269, "step": 690 }, { - "epoch": 1.21, - "learning_rate": 9.479057591623038e-06, - "loss": 0.6346, + "epoch": 2.43, + "learning_rate": 6.99e-06, + "loss": 0.4979, "step": 700 }, { - "epoch": 1.23, - "learning_rate": 9.470331588132636e-06, - "loss": 0.6016, + "epoch": 2.46, + "learning_rate": 7.09e-06, + "loss": 0.5041, "step": 710 }, { - "epoch": 1.25, - "learning_rate": 9.461605584642235e-06, - "loss": 0.6944, + "epoch": 2.5, + "learning_rate": 7.190000000000001e-06, + "loss": 0.5332, "step": 720 }, { - "epoch": 1.26, - "learning_rate": 9.452879581151834e-06, - "loss": 0.692, + "epoch": 2.53, + "learning_rate": 7.2900000000000005e-06, + "loss": 0.578, "step": 730 }, { - "epoch": 1.28, - "learning_rate": 9.444153577661432e-06, - "loss": 0.6802, + "epoch": 2.57, + "learning_rate": 7.39e-06, + "loss": 0.5307, "step": 740 }, { - "epoch": 1.3, - "learning_rate": 9.43542757417103e-06, - "loss": 0.642, + "epoch": 2.6, + "learning_rate": 7.48e-06, + "loss": 0.5141, "step": 750 }, { - "epoch": 1.31, - "learning_rate": 9.426701570680629e-06, - "loss": 0.6653, + "epoch": 2.64, + "learning_rate": 7.58e-06, + "loss": 0.5475, "step": 760 }, { - "epoch": 1.33, - "learning_rate": 9.417975567190228e-06, - "loss": 0.6658, + "epoch": 2.67, + "learning_rate": 7.680000000000001e-06, + "loss": 0.5673, "step": 770 }, { - "epoch": 1.35, - "learning_rate": 9.409249563699825e-06, - "loss": 0.7396, + "epoch": 2.71, + "learning_rate": 7.78e-06, + "loss": 0.613, "step": 780 }, { - "epoch": 1.37, - "learning_rate": 9.400523560209425e-06, - "loss": 0.6764, + "epoch": 2.74, + "learning_rate": 7.88e-06, + "loss": 0.5118, "step": 790 }, { - "epoch": 1.38, - "learning_rate": 9.391797556719024e-06, - "loss": 0.6731, + "epoch": 2.78, + "learning_rate": 7.980000000000002e-06, + "loss": 0.5186, "step": 800 }, { - "epoch": 1.4, - "learning_rate": 9.383071553228621e-06, - "loss": 0.6267, + "epoch": 2.81, + "learning_rate": 8.08e-06, + "loss": 0.5052, "step": 810 }, { - "epoch": 1.42, - "learning_rate": 9.37434554973822e-06, - "loss": 0.6536, + "epoch": 2.85, + "learning_rate": 8.18e-06, + "loss": 0.5333, "step": 820 }, { - "epoch": 1.44, - "learning_rate": 9.36561954624782e-06, - "loss": 0.6999, + "epoch": 2.88, + "learning_rate": 8.28e-06, + "loss": 0.5374, "step": 830 }, { - "epoch": 1.45, - "learning_rate": 9.356893542757418e-06, - "loss": 0.684, + "epoch": 2.91, + "learning_rate": 8.380000000000001e-06, + "loss": 0.5103, "step": 840 }, { - "epoch": 1.47, - "learning_rate": 9.348167539267017e-06, - "loss": 0.6323, + "epoch": 2.95, + "learning_rate": 8.48e-06, + "loss": 0.5278, "step": 850 }, { - "epoch": 1.49, - "learning_rate": 9.339441535776616e-06, - "loss": 0.6375, + "epoch": 2.98, + "learning_rate": 8.580000000000001e-06, + "loss": 0.5359, "step": 860 }, { - "epoch": 1.51, - "learning_rate": 9.330715532286214e-06, - "loss": 0.6428, + "epoch": 3.0, + "eval_loss": 0.6110973954200745, + "eval_runtime": 129.4368, + "eval_samples_per_second": 4.11, + "eval_steps_per_second": 1.028, + "eval_wer": 0.3363116282059523, + "step": 864 + }, + { + "epoch": 3.02, + "learning_rate": 8.68e-06, + "loss": 0.5897, "step": 870 }, { - "epoch": 1.52, - "learning_rate": 9.321989528795811e-06, - "loss": 0.7098, + "epoch": 3.06, + "learning_rate": 8.78e-06, + "loss": 0.5366, "step": 880 }, { - "epoch": 1.54, - "learning_rate": 9.31326352530541e-06, - "loss": 0.6982, + "epoch": 3.09, + "learning_rate": 8.880000000000001e-06, + "loss": 0.5459, "step": 890 }, { - "epoch": 1.56, - "learning_rate": 9.30453752181501e-06, - "loss": 0.7052, + "epoch": 3.12, + "learning_rate": 8.98e-06, + "loss": 0.5517, "step": 900 }, { - "epoch": 1.57, - "learning_rate": 9.295811518324607e-06, - "loss": 0.6462, + "epoch": 3.16, + "learning_rate": 9.080000000000001e-06, + "loss": 0.5796, "step": 910 }, { - "epoch": 1.59, - "learning_rate": 9.287085514834206e-06, - "loss": 0.6441, + "epoch": 3.19, + "learning_rate": 9.180000000000002e-06, + "loss": 0.5524, "step": 920 }, { - "epoch": 1.61, - "learning_rate": 9.278359511343806e-06, - "loss": 0.8151, + "epoch": 3.23, + "learning_rate": 9.280000000000001e-06, + "loss": 0.5449, "step": 930 }, { - "epoch": 1.63, - "learning_rate": 9.269633507853403e-06, - "loss": 0.6238, + "epoch": 3.26, + "learning_rate": 9.38e-06, + "loss": 0.522, "step": 940 }, { - "epoch": 1.64, - "learning_rate": 9.260907504363003e-06, - "loss": 0.6523, + "epoch": 3.3, + "learning_rate": 9.48e-06, + "loss": 0.5555, "step": 950 }, { - "epoch": 1.66, - "learning_rate": 9.252181500872602e-06, - "loss": 0.6742, + "epoch": 3.33, + "learning_rate": 9.58e-06, + "loss": 0.5739, "step": 960 }, { - "epoch": 1.68, - "learning_rate": 9.2434554973822e-06, - "loss": 0.7055, + "epoch": 3.37, + "learning_rate": 9.68e-06, + "loss": 0.583, "step": 970 }, { - "epoch": 1.7, - "learning_rate": 9.234729493891799e-06, - "loss": 0.6799, + "epoch": 3.4, + "learning_rate": 9.780000000000001e-06, + "loss": 0.562, "step": 980 }, { - "epoch": 1.71, - "learning_rate": 9.226003490401398e-06, - "loss": 0.6864, + "epoch": 3.44, + "learning_rate": 9.88e-06, + "loss": 0.5047, "step": 990 }, { - "epoch": 1.73, - "learning_rate": 9.217277486910995e-06, - "loss": 0.681, + "epoch": 3.47, + "learning_rate": 9.980000000000001e-06, + "loss": 0.5506, "step": 1000 }, { - "epoch": 1.75, - "learning_rate": 9.208551483420593e-06, - "loss": 0.6355, + "epoch": 3.51, + "learning_rate": 9.983193277310925e-06, + "loss": 0.5454, "step": 1010 }, { - "epoch": 1.76, - "learning_rate": 9.199825479930192e-06, - "loss": 0.7063, + "epoch": 3.54, + "learning_rate": 9.962184873949581e-06, + "loss": 0.5662, "step": 1020 }, { - "epoch": 1.78, - "learning_rate": 9.191099476439792e-06, - "loss": 0.7285, + "epoch": 3.57, + "learning_rate": 9.941176470588236e-06, + "loss": 0.5528, "step": 1030 }, { - "epoch": 1.8, - "learning_rate": 9.182373472949389e-06, - "loss": 0.6999, + "epoch": 3.61, + "learning_rate": 9.920168067226892e-06, + "loss": 0.5295, "step": 1040 }, { - "epoch": 1.82, - "learning_rate": 9.173647469458988e-06, - "loss": 0.649, + "epoch": 3.64, + "learning_rate": 9.899159663865548e-06, + "loss": 0.533, "step": 1050 }, { - "epoch": 1.83, - "learning_rate": 9.164921465968588e-06, - "loss": 0.6578, + "epoch": 3.68, + "learning_rate": 9.878151260504203e-06, + "loss": 0.5563, "step": 1060 }, { - "epoch": 1.85, - "learning_rate": 9.156195462478185e-06, - "loss": 0.6518, + "epoch": 3.71, + "learning_rate": 9.857142857142859e-06, + "loss": 0.5019, "step": 1070 }, { - "epoch": 1.87, - "learning_rate": 9.147469458987784e-06, - "loss": 0.6811, + "epoch": 3.75, + "learning_rate": 9.836134453781513e-06, + "loss": 0.5082, "step": 1080 }, { - "epoch": 1.89, - "learning_rate": 9.138743455497384e-06, - "loss": 0.6681, + "epoch": 3.78, + "learning_rate": 9.815126050420168e-06, + "loss": 0.5142, "step": 1090 }, { - "epoch": 1.9, - "learning_rate": 9.130017452006981e-06, - "loss": 0.6644, + "epoch": 3.82, + "learning_rate": 9.794117647058824e-06, + "loss": 0.5551, "step": 1100 }, { - "epoch": 1.92, - "learning_rate": 9.12129144851658e-06, - "loss": 0.6534, + "epoch": 3.85, + "learning_rate": 9.77310924369748e-06, + "loss": 0.4937, "step": 1110 }, { - "epoch": 1.94, - "learning_rate": 9.11256544502618e-06, - "loss": 0.6434, + "epoch": 3.89, + "learning_rate": 9.752100840336135e-06, + "loss": 0.5637, "step": 1120 }, { - "epoch": 1.96, - "learning_rate": 9.103839441535777e-06, - "loss": 0.6994, + "epoch": 3.92, + "learning_rate": 9.731092436974791e-06, + "loss": 0.5431, "step": 1130 }, { - "epoch": 1.97, - "learning_rate": 9.095113438045377e-06, - "loss": 0.707, + "epoch": 3.96, + "learning_rate": 9.710084033613445e-06, + "loss": 0.5548, "step": 1140 }, { - "epoch": 1.99, - "learning_rate": 9.086387434554974e-06, - "loss": 0.6466, + "epoch": 3.99, + "learning_rate": 9.689075630252102e-06, + "loss": 0.5395, "step": 1150 }, { - "epoch": 2.0, - "eval_loss": 0.7113747000694275, - "eval_runtime": 131.0653, - "eval_samples_per_second": 4.059, - "eval_steps_per_second": 0.511, - "eval_wer": 0.4237963617664074, - "step": 1156 + "epoch": 4.0, + "eval_loss": 0.6081970930099487, + "eval_runtime": 131.6568, + "eval_samples_per_second": 4.041, + "eval_steps_per_second": 1.01, + "eval_wer": 0.3415825673330334, + "step": 1152 }, { - "epoch": 2.01, - "learning_rate": 9.078534031413612e-06, - "loss": 0.7015, + "epoch": 4.03, + "learning_rate": 9.668067226890758e-06, + "loss": 0.5705, "step": 1160 }, { - "epoch": 2.02, - "learning_rate": 9.069808027923211e-06, - "loss": 0.6715, + "epoch": 4.06, + "learning_rate": 9.647058823529412e-06, + "loss": 0.5435, "step": 1170 }, { - "epoch": 2.04, - "learning_rate": 9.06108202443281e-06, - "loss": 0.6755, + "epoch": 4.1, + "learning_rate": 9.626050420168068e-06, + "loss": 0.4879, "step": 1180 }, { - "epoch": 2.06, - "learning_rate": 9.052356020942408e-06, - "loss": 0.6481, + "epoch": 4.13, + "learning_rate": 9.605042016806723e-06, + "loss": 0.5412, "step": 1190 }, { - "epoch": 2.08, - "learning_rate": 9.043630017452007e-06, - "loss": 0.6963, + "epoch": 4.17, + "learning_rate": 9.584033613445379e-06, + "loss": 0.5287, "step": 1200 }, { - "epoch": 2.09, - "learning_rate": 9.034904013961607e-06, - "loss": 0.6539, + "epoch": 4.2, + "learning_rate": 9.563025210084035e-06, + "loss": 0.5543, "step": 1210 }, { - "epoch": 2.11, - "learning_rate": 9.026178010471204e-06, - "loss": 0.7036, + "epoch": 4.24, + "learning_rate": 9.54201680672269e-06, + "loss": 0.5252, "step": 1220 }, { - "epoch": 2.13, - "learning_rate": 9.017452006980803e-06, - "loss": 0.6114, + "epoch": 4.27, + "learning_rate": 9.521008403361344e-06, + "loss": 0.5034, "step": 1230 }, { - "epoch": 2.15, - "learning_rate": 9.008726003490403e-06, - "loss": 0.6247, + "epoch": 4.3, + "learning_rate": 9.5e-06, + "loss": 0.5514, "step": 1240 }, { - "epoch": 2.16, - "learning_rate": 9e-06, - "loss": 0.6935, + "epoch": 4.34, + "learning_rate": 9.478991596638657e-06, + "loss": 0.5893, "step": 1250 }, { - "epoch": 2.18, - "learning_rate": 8.9912739965096e-06, - "loss": 0.6991, + "epoch": 4.37, + "learning_rate": 9.457983193277311e-06, + "loss": 0.5859, "step": 1260 }, { - "epoch": 2.2, - "learning_rate": 8.982547993019199e-06, - "loss": 0.647, + "epoch": 4.41, + "learning_rate": 9.436974789915967e-06, + "loss": 0.5404, "step": 1270 }, { - "epoch": 2.21, - "learning_rate": 8.973821989528796e-06, - "loss": 0.6767, + "epoch": 4.44, + "learning_rate": 9.415966386554622e-06, + "loss": 0.5362, "step": 1280 }, { - "epoch": 2.23, - "learning_rate": 8.965095986038394e-06, - "loss": 0.6765, + "epoch": 4.48, + "learning_rate": 9.394957983193278e-06, + "loss": 0.5252, "step": 1290 }, { - "epoch": 2.25, - "learning_rate": 8.956369982547993e-06, - "loss": 0.657, + "epoch": 4.51, + "learning_rate": 9.373949579831934e-06, + "loss": 0.5532, "step": 1300 }, { - "epoch": 2.27, - "learning_rate": 8.947643979057592e-06, - "loss": 0.7333, + "epoch": 4.55, + "learning_rate": 9.352941176470589e-06, + "loss": 0.5605, "step": 1310 }, { - "epoch": 2.28, - "learning_rate": 8.93891797556719e-06, - "loss": 0.6675, + "epoch": 4.58, + "learning_rate": 9.331932773109245e-06, + "loss": 0.5585, "step": 1320 }, { - "epoch": 2.3, - "learning_rate": 8.93019197207679e-06, - "loss": 0.6853, + "epoch": 4.62, + "learning_rate": 9.3109243697479e-06, + "loss": 0.5177, "step": 1330 }, { - "epoch": 2.32, - "learning_rate": 8.921465968586388e-06, - "loss": 0.6333, + "epoch": 4.65, + "learning_rate": 9.289915966386556e-06, + "loss": 0.5592, "step": 1340 }, { - "epoch": 2.34, - "learning_rate": 8.912739965095986e-06, - "loss": 0.671, + "epoch": 4.69, + "learning_rate": 9.268907563025212e-06, + "loss": 0.5556, "step": 1350 }, { - "epoch": 2.35, - "learning_rate": 8.904013961605585e-06, - "loss": 0.7246, + "epoch": 4.72, + "learning_rate": 9.247899159663866e-06, + "loss": 0.5304, "step": 1360 }, { - "epoch": 2.37, - "learning_rate": 8.895287958115185e-06, - "loss": 0.668, + "epoch": 4.75, + "learning_rate": 9.226890756302523e-06, + "loss": 0.5362, "step": 1370 }, { - "epoch": 2.39, - "learning_rate": 8.886561954624782e-06, - "loss": 0.6453, + "epoch": 4.79, + "learning_rate": 9.205882352941177e-06, + "loss": 0.5443, "step": 1380 }, { - "epoch": 2.4, - "learning_rate": 8.877835951134381e-06, - "loss": 0.6504, + "epoch": 4.82, + "learning_rate": 9.184873949579832e-06, + "loss": 0.5354, "step": 1390 }, { - "epoch": 2.42, - "learning_rate": 8.86910994764398e-06, - "loss": 0.7074, + "epoch": 4.86, + "learning_rate": 9.163865546218488e-06, + "loss": 0.5676, "step": 1400 }, { - "epoch": 2.44, - "learning_rate": 8.860383944153578e-06, - "loss": 0.6474, + "epoch": 4.89, + "learning_rate": 9.142857142857144e-06, + "loss": 0.5385, "step": 1410 }, { - "epoch": 2.46, - "learning_rate": 8.851657940663177e-06, - "loss": 0.6634, + "epoch": 4.93, + "learning_rate": 9.121848739495798e-06, + "loss": 0.5218, "step": 1420 }, { - "epoch": 2.47, - "learning_rate": 8.842931937172775e-06, - "loss": 0.6065, + "epoch": 4.96, + "learning_rate": 9.100840336134455e-06, + "loss": 0.5123, "step": 1430 }, { - "epoch": 2.49, - "learning_rate": 8.834205933682374e-06, - "loss": 0.6308, + "epoch": 5.0, + "learning_rate": 9.07983193277311e-06, + "loss": 0.5692, + "step": 1440 + }, + { + "epoch": 5.0, + "eval_loss": 0.5948615074157715, + "eval_runtime": 131.2811, + "eval_samples_per_second": 4.052, + "eval_steps_per_second": 1.013, + "eval_wer": 0.33181204602429776, "step": 1440 }, { - "epoch": 2.51, - "learning_rate": 8.825479930191972e-06, - "loss": 0.6334, + "epoch": 5.03, + "learning_rate": 9.058823529411765e-06, + "loss": 0.5519, "step": 1450 }, { - "epoch": 2.53, - "learning_rate": 8.816753926701571e-06, - "loss": 0.7137, + "epoch": 5.07, + "learning_rate": 9.037815126050421e-06, + "loss": 0.5099, "step": 1460 }, { - "epoch": 2.54, - "learning_rate": 8.80802792321117e-06, - "loss": 0.6967, + "epoch": 5.1, + "learning_rate": 9.016806722689076e-06, + "loss": 0.4983, "step": 1470 }, { - "epoch": 2.56, - "learning_rate": 8.799301919720768e-06, - "loss": 0.6364, + "epoch": 5.14, + "learning_rate": 8.995798319327732e-06, + "loss": 0.5158, "step": 1480 }, { - "epoch": 2.58, - "learning_rate": 8.790575916230367e-06, - "loss": 0.6574, + "epoch": 5.17, + "learning_rate": 8.974789915966388e-06, + "loss": 0.5672, "step": 1490 }, { - "epoch": 2.6, - "learning_rate": 8.781849912739966e-06, - "loss": 0.7073, + "epoch": 5.21, + "learning_rate": 8.953781512605043e-06, + "loss": 0.5244, "step": 1500 }, { - "epoch": 2.61, - "learning_rate": 8.773123909249564e-06, - "loss": 0.6355, + "epoch": 5.24, + "learning_rate": 8.932773109243699e-06, + "loss": 0.4964, "step": 1510 }, { - "epoch": 2.63, - "learning_rate": 8.764397905759163e-06, - "loss": 0.6496, + "epoch": 5.28, + "learning_rate": 8.911764705882354e-06, + "loss": 0.5016, "step": 1520 }, { - "epoch": 2.65, - "learning_rate": 8.755671902268763e-06, - "loss": 0.6596, + "epoch": 5.31, + "learning_rate": 8.890756302521008e-06, + "loss": 0.532, "step": 1530 }, { - "epoch": 2.66, - "learning_rate": 8.74694589877836e-06, - "loss": 0.6953, + "epoch": 5.35, + "learning_rate": 8.869747899159664e-06, + "loss": 0.5434, "step": 1540 }, { - "epoch": 2.68, - "learning_rate": 8.73821989528796e-06, - "loss": 0.6462, + "epoch": 5.38, + "learning_rate": 8.84873949579832e-06, + "loss": 0.5334, "step": 1550 }, { - "epoch": 2.7, - "learning_rate": 8.729493891797557e-06, - "loss": 0.6362, + "epoch": 5.42, + "learning_rate": 8.827731092436975e-06, + "loss": 0.5062, "step": 1560 }, { - "epoch": 2.72, - "learning_rate": 8.720767888307156e-06, - "loss": 0.669, + "epoch": 5.45, + "learning_rate": 8.806722689075631e-06, + "loss": 0.5314, "step": 1570 }, { - "epoch": 2.73, - "learning_rate": 8.712041884816754e-06, - "loss": 0.589, + "epoch": 5.48, + "learning_rate": 8.785714285714286e-06, + "loss": 0.5543, "step": 1580 }, { - "epoch": 2.75, - "learning_rate": 8.703315881326353e-06, - "loss": 0.6468, + "epoch": 5.52, + "learning_rate": 8.764705882352942e-06, + "loss": 0.5478, "step": 1590 }, { - "epoch": 2.77, - "learning_rate": 8.694589877835952e-06, - "loss": 0.6115, + "epoch": 5.55, + "learning_rate": 8.743697478991598e-06, + "loss": 0.5225, "step": 1600 }, { - "epoch": 2.79, - "learning_rate": 8.68586387434555e-06, - "loss": 0.6202, + "epoch": 5.59, + "learning_rate": 8.722689075630252e-06, + "loss": 0.5084, "step": 1610 }, { - "epoch": 2.8, - "learning_rate": 8.677137870855149e-06, - "loss": 0.6699, + "epoch": 5.62, + "learning_rate": 8.701680672268909e-06, + "loss": 0.5226, "step": 1620 }, { - "epoch": 2.82, - "learning_rate": 8.668411867364748e-06, - "loss": 0.6347, + "epoch": 5.66, + "learning_rate": 8.680672268907563e-06, + "loss": 0.5181, "step": 1630 }, { - "epoch": 2.84, - "learning_rate": 8.659685863874346e-06, - "loss": 0.6264, + "epoch": 5.69, + "learning_rate": 8.65966386554622e-06, + "loss": 0.5648, "step": 1640 }, { - "epoch": 2.85, - "learning_rate": 8.650959860383945e-06, - "loss": 0.6338, + "epoch": 5.73, + "learning_rate": 8.638655462184876e-06, + "loss": 0.537, "step": 1650 }, { - "epoch": 2.87, - "learning_rate": 8.642233856893544e-06, - "loss": 0.7019, + "epoch": 5.76, + "learning_rate": 8.61764705882353e-06, + "loss": 0.5747, "step": 1660 }, { - "epoch": 2.89, - "learning_rate": 8.633507853403142e-06, - "loss": 0.6375, + "epoch": 5.8, + "learning_rate": 8.596638655462186e-06, + "loss": 0.5388, "step": 1670 }, { - "epoch": 2.91, - "learning_rate": 8.624781849912741e-06, - "loss": 0.6317, - "step": 1680 + "epoch": 5.83, + "learning_rate": 8.57563025210084e-06, + "loss": 0.5336, + "step": 1680 }, { - "epoch": 2.92, - "learning_rate": 8.616055846422339e-06, - "loss": 0.6738, + "epoch": 5.87, + "learning_rate": 8.554621848739497e-06, + "loss": 0.5465, "step": 1690 }, { - "epoch": 2.94, - "learning_rate": 8.607329842931938e-06, - "loss": 0.6408, + "epoch": 5.9, + "learning_rate": 8.533613445378151e-06, + "loss": 0.5231, "step": 1700 }, { - "epoch": 2.96, - "learning_rate": 8.598603839441536e-06, - "loss": 0.6505, + "epoch": 5.94, + "learning_rate": 8.512605042016808e-06, + "loss": 0.5345, "step": 1710 }, { - "epoch": 2.98, - "learning_rate": 8.589877835951135e-06, - "loss": 0.618, + "epoch": 5.97, + "learning_rate": 8.491596638655462e-06, + "loss": 0.5592, "step": 1720 }, { - "epoch": 2.99, - "learning_rate": 8.581151832460734e-06, - "loss": 0.6371, - "step": 1730 + "epoch": 6.0, + "eval_loss": 0.6045897006988525, + "eval_runtime": 132.0329, + "eval_samples_per_second": 4.029, + "eval_steps_per_second": 1.007, + "eval_wer": 0.3322620042424632, + "step": 1728 }, { - "epoch": 3.0, - "eval_loss": 0.7032670974731445, - "eval_runtime": 129.4318, - "eval_samples_per_second": 4.11, - "eval_steps_per_second": 0.518, - "eval_wer": 0.4318313299479334, - "step": 1734 + "epoch": 6.01, + "learning_rate": 8.470588235294118e-06, + "loss": 0.5959, + "step": 1730 }, { - "epoch": 3.01, - "learning_rate": 8.572425828970332e-06, - "loss": 0.6566, + "epoch": 6.04, + "learning_rate": 8.449579831932774e-06, + "loss": 0.5532, "step": 1740 }, { - "epoch": 3.03, - "learning_rate": 8.563699825479931e-06, - "loss": 0.6582, + "epoch": 6.08, + "learning_rate": 8.428571428571429e-06, + "loss": 0.5406, "step": 1750 }, { - "epoch": 3.04, - "learning_rate": 8.55497382198953e-06, - "loss": 0.6444, + "epoch": 6.11, + "learning_rate": 8.407563025210085e-06, + "loss": 0.5133, "step": 1760 }, { - "epoch": 3.06, - "learning_rate": 8.546247818499128e-06, - "loss": 0.7375, + "epoch": 6.15, + "learning_rate": 8.38655462184874e-06, + "loss": 0.5282, "step": 1770 }, { - "epoch": 3.08, - "learning_rate": 8.537521815008727e-06, - "loss": 0.688, + "epoch": 6.18, + "learning_rate": 8.365546218487396e-06, + "loss": 0.5323, "step": 1780 }, { - "epoch": 3.1, - "learning_rate": 8.528795811518326e-06, - "loss": 0.6782, + "epoch": 6.21, + "learning_rate": 8.344537815126052e-06, + "loss": 0.5416, "step": 1790 }, { - "epoch": 3.11, - "learning_rate": 8.520069808027924e-06, - "loss": 0.653, + "epoch": 6.25, + "learning_rate": 8.323529411764707e-06, + "loss": 0.469, "step": 1800 }, { - "epoch": 3.13, - "learning_rate": 8.511343804537523e-06, - "loss": 0.6175, + "epoch": 6.28, + "learning_rate": 8.302521008403363e-06, + "loss": 0.4816, "step": 1810 }, { - "epoch": 3.15, - "learning_rate": 8.502617801047122e-06, - "loss": 0.6632, + "epoch": 6.32, + "learning_rate": 8.281512605042017e-06, + "loss": 0.5499, "step": 1820 }, { - "epoch": 3.17, - "learning_rate": 8.49389179755672e-06, - "loss": 0.6808, + "epoch": 6.35, + "learning_rate": 8.260504201680672e-06, + "loss": 0.5841, "step": 1830 }, { - "epoch": 3.18, - "learning_rate": 8.485165794066317e-06, - "loss": 0.6445, + "epoch": 6.39, + "learning_rate": 8.239495798319328e-06, + "loss": 0.5092, "step": 1840 }, { - "epoch": 3.2, - "learning_rate": 8.476439790575917e-06, - "loss": 0.5956, + "epoch": 6.42, + "learning_rate": 8.218487394957984e-06, + "loss": 0.5167, "step": 1850 }, { - "epoch": 3.22, - "learning_rate": 8.467713787085516e-06, - "loss": 0.6072, + "epoch": 6.46, + "learning_rate": 8.197478991596639e-06, + "loss": 0.5315, "step": 1860 }, { - "epoch": 3.24, - "learning_rate": 8.458987783595114e-06, - "loss": 0.6568, + "epoch": 6.49, + "learning_rate": 8.176470588235295e-06, + "loss": 0.4991, "step": 1870 }, { - "epoch": 3.25, - "learning_rate": 8.450261780104713e-06, - "loss": 0.6643, + "epoch": 6.53, + "learning_rate": 8.155462184873951e-06, + "loss": 0.5728, "step": 1880 }, { - "epoch": 3.27, - "learning_rate": 8.441535776614312e-06, - "loss": 0.7481, + "epoch": 6.56, + "learning_rate": 8.134453781512605e-06, + "loss": 0.534, "step": 1890 }, { - "epoch": 3.29, - "learning_rate": 8.43280977312391e-06, - "loss": 0.669, + "epoch": 6.6, + "learning_rate": 8.113445378151262e-06, + "loss": 0.5377, "step": 1900 }, { - "epoch": 3.3, - "learning_rate": 8.424083769633509e-06, - "loss": 0.6675, + "epoch": 6.63, + "learning_rate": 8.092436974789916e-06, + "loss": 0.5285, "step": 1910 }, { - "epoch": 3.32, - "learning_rate": 8.415357766143108e-06, - "loss": 0.6338, + "epoch": 6.66, + "learning_rate": 8.071428571428572e-06, + "loss": 0.5373, "step": 1920 }, { - "epoch": 3.34, - "learning_rate": 8.406631762652706e-06, - "loss": 0.6446, + "epoch": 6.7, + "learning_rate": 8.050420168067229e-06, + "loss": 0.5336, "step": 1930 }, { - "epoch": 3.36, - "learning_rate": 8.398778359511345e-06, - "loss": 0.6956, + "epoch": 6.73, + "learning_rate": 8.029411764705883e-06, + "loss": 0.4982, "step": 1940 }, { - "epoch": 3.37, - "learning_rate": 8.390052356020943e-06, - "loss": 0.6845, + "epoch": 6.77, + "learning_rate": 8.00840336134454e-06, + "loss": 0.534, "step": 1950 }, { - "epoch": 3.39, - "learning_rate": 8.381326352530542e-06, - "loss": 0.6379, + "epoch": 6.8, + "learning_rate": 7.987394957983194e-06, + "loss": 0.4873, "step": 1960 }, { - "epoch": 3.41, - "learning_rate": 8.372600349040141e-06, - "loss": 0.6526, + "epoch": 6.84, + "learning_rate": 7.966386554621848e-06, + "loss": 0.5475, "step": 1970 }, { - "epoch": 3.43, - "learning_rate": 8.363874345549739e-06, - "loss": 0.6498, + "epoch": 6.87, + "learning_rate": 7.945378151260504e-06, + "loss": 0.5666, "step": 1980 }, { - "epoch": 3.44, - "learning_rate": 8.355148342059336e-06, - "loss": 0.6664, + "epoch": 6.91, + "learning_rate": 7.92436974789916e-06, + "loss": 0.5062, "step": 1990 }, { - "epoch": 3.46, - "learning_rate": 8.346422338568936e-06, - "loss": 0.6951, + "epoch": 6.94, + "learning_rate": 7.903361344537815e-06, + "loss": 0.4941, "step": 2000 }, { - "epoch": 3.48, - "learning_rate": 8.337696335078535e-06, - "loss": 0.6097, + "epoch": 6.98, + "learning_rate": 7.882352941176471e-06, + "loss": 0.5172, "step": 2010 }, { - "epoch": 3.49, - "learning_rate": 8.328970331588133e-06, - "loss": 0.6301, + "epoch": 7.0, + "eval_loss": 0.5837918519973755, + "eval_runtime": 131.1239, + "eval_samples_per_second": 4.057, + "eval_steps_per_second": 1.014, + "eval_wer": 0.31850613871569067, + "step": 2016 + }, + { + "epoch": 7.01, + "learning_rate": 7.861344537815126e-06, + "loss": 0.5684, "step": 2020 }, { - "epoch": 3.51, - "learning_rate": 8.320244328097732e-06, - "loss": 0.6253, + "epoch": 7.05, + "learning_rate": 7.840336134453782e-06, + "loss": 0.5345, "step": 2030 }, { - "epoch": 3.53, - "learning_rate": 8.311518324607331e-06, - "loss": 0.6924, + "epoch": 7.08, + "learning_rate": 7.819327731092438e-06, + "loss": 0.4978, "step": 2040 }, { - "epoch": 3.55, - "learning_rate": 8.302792321116929e-06, - "loss": 0.6185, + "epoch": 7.12, + "learning_rate": 7.798319327731093e-06, + "loss": 0.5151, "step": 2050 }, { - "epoch": 3.56, - "learning_rate": 8.294066317626528e-06, - "loss": 0.6114, + "epoch": 7.15, + "learning_rate": 7.777310924369749e-06, + "loss": 0.519, "step": 2060 }, { - "epoch": 3.58, - "learning_rate": 8.285340314136127e-06, - "loss": 0.6305, + "epoch": 7.19, + "learning_rate": 7.756302521008405e-06, + "loss": 0.5289, "step": 2070 }, { - "epoch": 3.6, - "learning_rate": 8.276614310645725e-06, - "loss": 0.6085, + "epoch": 7.22, + "learning_rate": 7.73529411764706e-06, + "loss": 0.5416, "step": 2080 }, { - "epoch": 3.62, - "learning_rate": 8.267888307155324e-06, - "loss": 0.6608, + "epoch": 7.26, + "learning_rate": 7.714285714285716e-06, + "loss": 0.5011, "step": 2090 }, { - "epoch": 3.63, - "learning_rate": 8.259162303664923e-06, - "loss": 0.6095, + "epoch": 7.29, + "learning_rate": 7.69327731092437e-06, + "loss": 0.5269, "step": 2100 }, { - "epoch": 3.65, - "learning_rate": 8.25043630017452e-06, - "loss": 0.6031, + "epoch": 7.33, + "learning_rate": 7.672268907563026e-06, + "loss": 0.5394, "step": 2110 }, { - "epoch": 3.67, - "learning_rate": 8.241710296684118e-06, - "loss": 0.6122, + "epoch": 7.36, + "learning_rate": 7.651260504201681e-06, + "loss": 0.5125, "step": 2120 }, { - "epoch": 3.69, - "learning_rate": 8.232984293193718e-06, - "loss": 0.6317, + "epoch": 7.39, + "learning_rate": 7.630252100840337e-06, + "loss": 0.5249, "step": 2130 }, { - "epoch": 3.7, - "learning_rate": 8.224258289703317e-06, - "loss": 0.597, + "epoch": 7.43, + "learning_rate": 7.6092436974789916e-06, + "loss": 0.5097, "step": 2140 }, { - "epoch": 3.72, - "learning_rate": 8.215532286212914e-06, - "loss": 0.6278, + "epoch": 7.46, + "learning_rate": 7.588235294117648e-06, + "loss": 0.5186, "step": 2150 }, { - "epoch": 3.74, - "learning_rate": 8.206806282722514e-06, - "loss": 0.6403, + "epoch": 7.5, + "learning_rate": 7.567226890756303e-06, + "loss": 0.5011, "step": 2160 }, { - "epoch": 3.75, - "learning_rate": 8.198080279232113e-06, - "loss": 0.6566, + "epoch": 7.53, + "learning_rate": 7.5462184873949584e-06, + "loss": 0.5464, "step": 2170 }, { - "epoch": 3.77, - "learning_rate": 8.18935427574171e-06, - "loss": 0.6089, + "epoch": 7.57, + "learning_rate": 7.525210084033614e-06, + "loss": 0.5232, "step": 2180 }, { - "epoch": 3.79, - "learning_rate": 8.18062827225131e-06, - "loss": 0.6692, + "epoch": 7.6, + "learning_rate": 7.504201680672269e-06, + "loss": 0.4948, "step": 2190 }, { - "epoch": 3.81, - "learning_rate": 8.171902268760909e-06, - "loss": 0.6096, + "epoch": 7.64, + "learning_rate": 7.483193277310925e-06, + "loss": 0.5235, "step": 2200 }, { - "epoch": 3.82, - "learning_rate": 8.163176265270507e-06, - "loss": 0.5987, + "epoch": 7.67, + "learning_rate": 7.462184873949581e-06, + "loss": 0.5354, "step": 2210 }, { - "epoch": 3.84, - "learning_rate": 8.154450261780106e-06, - "loss": 0.6237, + "epoch": 7.71, + "learning_rate": 7.441176470588236e-06, + "loss": 0.544, "step": 2220 }, { - "epoch": 3.86, - "learning_rate": 8.145724258289705e-06, - "loss": 0.6206, + "epoch": 7.74, + "learning_rate": 7.420168067226891e-06, + "loss": 0.5063, "step": 2230 }, { - "epoch": 3.88, - "learning_rate": 8.136998254799303e-06, - "loss": 0.6444, + "epoch": 7.78, + "learning_rate": 7.3991596638655475e-06, + "loss": 0.5067, "step": 2240 }, { - "epoch": 3.89, - "learning_rate": 8.1282722513089e-06, - "loss": 0.5927, + "epoch": 7.81, + "learning_rate": 7.378151260504203e-06, + "loss": 0.4885, "step": 2250 }, { - "epoch": 3.91, - "learning_rate": 8.1195462478185e-06, - "loss": 0.6132, + "epoch": 7.85, + "learning_rate": 7.357142857142858e-06, + "loss": 0.5029, "step": 2260 }, { - "epoch": 3.93, - "learning_rate": 8.110820244328099e-06, - "loss": 0.6335, + "epoch": 7.88, + "learning_rate": 7.336134453781513e-06, + "loss": 0.5642, "step": 2270 }, { - "epoch": 3.94, - "learning_rate": 8.102094240837696e-06, - "loss": 0.6204, + "epoch": 7.91, + "learning_rate": 7.315126050420168e-06, + "loss": 0.5236, "step": 2280 }, { - "epoch": 3.96, - "learning_rate": 8.093368237347296e-06, - "loss": 0.6246, + "epoch": 7.95, + "learning_rate": 7.294117647058823e-06, + "loss": 0.4973, "step": 2290 }, { - "epoch": 3.98, - "learning_rate": 8.084642233856895e-06, - "loss": 0.5686, + "epoch": 7.98, + "learning_rate": 7.27310924369748e-06, + "loss": 0.5108, "step": 2300 }, { - "epoch": 4.0, - "learning_rate": 8.075916230366492e-06, - "loss": 0.7171, - "step": 2310 + "epoch": 8.0, + "eval_loss": 0.6065585613250732, + "eval_runtime": 133.6934, + "eval_samples_per_second": 3.979, + "eval_steps_per_second": 0.995, + "eval_wer": 0.3211416082792312, + "step": 2304 }, { - "epoch": 4.0, - "eval_loss": 0.6488160490989685, - "eval_runtime": 130.2289, - "eval_samples_per_second": 4.085, - "eval_steps_per_second": 0.514, - "eval_wer": 0.39699170791283667, - "step": 2312 + "epoch": 8.02, + "learning_rate": 7.252100840336135e-06, + "loss": 0.5786, + "step": 2310 }, { - "epoch": 4.01, - "learning_rate": 8.067190226876092e-06, - "loss": 0.6293, + "epoch": 8.06, + "learning_rate": 7.23109243697479e-06, + "loss": 0.5407, "step": 2320 }, { - "epoch": 4.03, - "learning_rate": 8.058464223385691e-06, - "loss": 0.6144, + "epoch": 8.09, + "learning_rate": 7.210084033613446e-06, + "loss": 0.4974, "step": 2330 }, { - "epoch": 4.05, - "learning_rate": 8.049738219895288e-06, - "loss": 0.5733, + "epoch": 8.12, + "learning_rate": 7.189075630252102e-06, + "loss": 0.499, "step": 2340 }, { - "epoch": 4.07, - "learning_rate": 8.041012216404888e-06, - "loss": 0.6863, + "epoch": 8.16, + "learning_rate": 7.168067226890757e-06, + "loss": 0.5148, "step": 2350 }, { - "epoch": 4.08, - "learning_rate": 8.032286212914487e-06, - "loss": 0.6221, + "epoch": 8.19, + "learning_rate": 7.1470588235294125e-06, + "loss": 0.5337, "step": 2360 }, { - "epoch": 4.1, - "learning_rate": 8.023560209424085e-06, - "loss": 0.6876, + "epoch": 8.23, + "learning_rate": 7.126050420168068e-06, + "loss": 0.5337, "step": 2370 }, { - "epoch": 4.12, - "learning_rate": 8.014834205933682e-06, - "loss": 0.6303, + "epoch": 8.26, + "learning_rate": 7.105042016806723e-06, + "loss": 0.4786, "step": 2380 }, { - "epoch": 4.13, - "learning_rate": 8.006108202443281e-06, - "loss": 0.6016, + "epoch": 8.3, + "learning_rate": 7.084033613445379e-06, + "loss": 0.519, "step": 2390 }, { - "epoch": 4.15, - "learning_rate": 7.99738219895288e-06, - "loss": 0.58, + "epoch": 8.33, + "learning_rate": 7.063025210084035e-06, + "loss": 0.535, "step": 2400 }, { - "epoch": 4.17, - "learning_rate": 7.988656195462478e-06, - "loss": 0.6173, + "epoch": 8.37, + "learning_rate": 7.04201680672269e-06, + "loss": 0.5086, "step": 2410 }, { - "epoch": 4.19, - "learning_rate": 7.979930191972077e-06, - "loss": 0.6092, + "epoch": 8.4, + "learning_rate": 7.0210084033613446e-06, + "loss": 0.5093, "step": 2420 }, { - "epoch": 4.2, - "learning_rate": 7.971204188481677e-06, - "loss": 0.6311, + "epoch": 8.44, + "learning_rate": 7e-06, + "loss": 0.475, "step": 2430 }, { - "epoch": 4.22, - "learning_rate": 7.962478184991274e-06, - "loss": 0.6298, + "epoch": 8.47, + "learning_rate": 6.978991596638656e-06, + "loss": 0.5045, "step": 2440 }, { - "epoch": 4.24, - "learning_rate": 7.953752181500874e-06, - "loss": 0.6152, + "epoch": 8.51, + "learning_rate": 6.9579831932773114e-06, + "loss": 0.5231, "step": 2450 }, { - "epoch": 4.26, - "learning_rate": 7.945026178010473e-06, - "loss": 0.6223, + "epoch": 8.54, + "learning_rate": 6.936974789915967e-06, + "loss": 0.4882, "step": 2460 }, { - "epoch": 4.27, - "learning_rate": 7.93630017452007e-06, - "loss": 0.667, + "epoch": 8.57, + "learning_rate": 6.915966386554622e-06, + "loss": 0.5021, "step": 2470 }, { - "epoch": 4.29, - "learning_rate": 7.92757417102967e-06, - "loss": 0.5928, + "epoch": 8.61, + "learning_rate": 6.8949579831932775e-06, + "loss": 0.5154, "step": 2480 }, { - "epoch": 4.31, - "learning_rate": 7.918848167539269e-06, - "loss": 0.5741, + "epoch": 8.64, + "learning_rate": 6.873949579831934e-06, + "loss": 0.5142, "step": 2490 }, { - "epoch": 4.33, - "learning_rate": 7.910122164048866e-06, - "loss": 0.6001, + "epoch": 8.68, + "learning_rate": 6.852941176470589e-06, + "loss": 0.5319, "step": 2500 }, { - "epoch": 4.34, - "learning_rate": 7.901396160558464e-06, - "loss": 0.7027, + "epoch": 8.71, + "learning_rate": 6.831932773109244e-06, + "loss": 0.5445, "step": 2510 }, { - "epoch": 4.36, - "learning_rate": 7.892670157068063e-06, - "loss": 0.712, + "epoch": 8.75, + "learning_rate": 6.8109243697479e-06, + "loss": 0.5134, "step": 2520 }, { - "epoch": 4.38, - "learning_rate": 7.883944153577662e-06, - "loss": 0.6157, + "epoch": 8.78, + "learning_rate": 6.789915966386556e-06, + "loss": 0.5008, "step": 2530 }, { - "epoch": 4.39, - "learning_rate": 7.87521815008726e-06, - "loss": 0.5956, + "epoch": 8.82, + "learning_rate": 6.768907563025211e-06, + "loss": 0.4979, "step": 2540 }, { - "epoch": 4.41, - "learning_rate": 7.86649214659686e-06, - "loss": 0.623, + "epoch": 8.85, + "learning_rate": 6.7478991596638666e-06, + "loss": 0.5292, "step": 2550 }, { - "epoch": 4.43, - "learning_rate": 7.857766143106459e-06, - "loss": 0.6355, + "epoch": 8.89, + "learning_rate": 6.726890756302522e-06, + "loss": 0.5521, "step": 2560 }, { - "epoch": 4.45, - "learning_rate": 7.849040139616056e-06, - "loss": 0.6509, + "epoch": 8.92, + "learning_rate": 6.705882352941176e-06, + "loss": 0.5019, "step": 2570 }, { - "epoch": 4.46, - "learning_rate": 7.840314136125655e-06, - "loss": 0.6159, + "epoch": 8.96, + "learning_rate": 6.684873949579832e-06, + "loss": 0.5079, "step": 2580 }, { - "epoch": 4.48, - "learning_rate": 7.831588132635255e-06, - "loss": 0.6519, + "epoch": 8.99, + "learning_rate": 6.663865546218488e-06, + "loss": 0.4981, "step": 2590 }, { - "epoch": 4.5, - "learning_rate": 7.822862129144852e-06, - "loss": 0.649, + "epoch": 9.0, + "eval_loss": 0.5958260297775269, + "eval_runtime": 131.9283, + "eval_samples_per_second": 4.032, + "eval_steps_per_second": 1.008, + "eval_wer": 0.31638490711576783, + "step": 2592 + }, + { + "epoch": 9.03, + "learning_rate": 6.642857142857143e-06, + "loss": 0.523, "step": 2600 }, { - "epoch": 4.52, - "learning_rate": 7.814136125654451e-06, - "loss": 0.6529, + "epoch": 9.06, + "learning_rate": 6.621848739495799e-06, + "loss": 0.503, "step": 2610 }, { - "epoch": 4.53, - "learning_rate": 7.80541012216405e-06, - "loss": 0.6602, + "epoch": 9.1, + "learning_rate": 6.600840336134454e-06, + "loss": 0.5157, "step": 2620 }, { - "epoch": 4.55, - "learning_rate": 7.796684118673648e-06, - "loss": 0.644, + "epoch": 9.13, + "learning_rate": 6.57983193277311e-06, + "loss": 0.501, "step": 2630 }, { - "epoch": 4.57, - "learning_rate": 7.787958115183246e-06, - "loss": 0.615, + "epoch": 9.17, + "learning_rate": 6.5588235294117655e-06, + "loss": 0.5101, "step": 2640 }, { - "epoch": 4.58, - "learning_rate": 7.779232111692845e-06, - "loss": 0.64, + "epoch": 9.2, + "learning_rate": 6.537815126050421e-06, + "loss": 0.5184, "step": 2650 }, { - "epoch": 4.6, - "learning_rate": 7.770506108202444e-06, - "loss": 0.6304, + "epoch": 9.24, + "learning_rate": 6.516806722689076e-06, + "loss": 0.5178, "step": 2660 }, { - "epoch": 4.62, - "learning_rate": 7.761780104712042e-06, - "loss": 0.6394, + "epoch": 9.27, + "learning_rate": 6.4957983193277315e-06, + "loss": 0.4754, "step": 2670 }, { - "epoch": 4.64, - "learning_rate": 7.753054101221641e-06, - "loss": 0.6335, + "epoch": 9.3, + "learning_rate": 6.474789915966388e-06, + "loss": 0.5115, "step": 2680 }, { - "epoch": 4.65, - "learning_rate": 7.74432809773124e-06, - "loss": 0.6404, + "epoch": 9.34, + "learning_rate": 6.453781512605043e-06, + "loss": 0.5263, "step": 2690 }, { - "epoch": 4.67, - "learning_rate": 7.735602094240838e-06, - "loss": 0.6368, + "epoch": 9.37, + "learning_rate": 6.432773109243698e-06, + "loss": 0.5424, "step": 2700 }, { - "epoch": 4.69, - "learning_rate": 7.726876090750437e-06, - "loss": 0.6742, + "epoch": 9.41, + "learning_rate": 6.411764705882354e-06, + "loss": 0.4897, "step": 2710 }, { - "epoch": 4.71, - "learning_rate": 7.718150087260035e-06, - "loss": 0.6193, + "epoch": 9.44, + "learning_rate": 6.390756302521008e-06, + "loss": 0.4653, "step": 2720 }, { - "epoch": 4.72, - "learning_rate": 7.709424083769634e-06, - "loss": 0.6625, + "epoch": 9.48, + "learning_rate": 6.3697478991596636e-06, + "loss": 0.524, "step": 2730 }, { - "epoch": 4.74, - "learning_rate": 7.700698080279233e-06, - "loss": 0.623, + "epoch": 9.51, + "learning_rate": 6.34873949579832e-06, + "loss": 0.4786, "step": 2740 }, { - "epoch": 4.76, - "learning_rate": 7.691972076788831e-06, - "loss": 0.6042, + "epoch": 9.55, + "learning_rate": 6.327731092436975e-06, + "loss": 0.5414, "step": 2750 }, { - "epoch": 4.78, - "learning_rate": 7.68324607329843e-06, - "loss": 0.6694, + "epoch": 9.58, + "learning_rate": 6.3067226890756304e-06, + "loss": 0.4963, "step": 2760 }, { - "epoch": 4.79, - "learning_rate": 7.67452006980803e-06, - "loss": 0.6461, + "epoch": 9.62, + "learning_rate": 6.285714285714286e-06, + "loss": 0.483, "step": 2770 }, { - "epoch": 4.81, - "learning_rate": 7.665794066317627e-06, - "loss": 0.6445, + "epoch": 9.65, + "learning_rate": 6.264705882352942e-06, + "loss": 0.5189, "step": 2780 }, { - "epoch": 4.83, - "learning_rate": 7.657068062827225e-06, - "loss": 0.6295, + "epoch": 9.69, + "learning_rate": 6.243697478991597e-06, + "loss": 0.5364, "step": 2790 }, { - "epoch": 4.84, - "learning_rate": 7.648342059336824e-06, - "loss": 0.627, + "epoch": 9.72, + "learning_rate": 6.222689075630253e-06, + "loss": 0.5128, "step": 2800 }, { - "epoch": 4.86, - "learning_rate": 7.639616055846423e-06, - "loss": 0.6522, + "epoch": 9.75, + "learning_rate": 6.201680672268908e-06, + "loss": 0.512, "step": 2810 }, { - "epoch": 4.88, - "learning_rate": 7.63089005235602e-06, - "loss": 0.5977, + "epoch": 9.79, + "learning_rate": 6.180672268907563e-06, + "loss": 0.4781, "step": 2820 }, { - "epoch": 4.9, - "learning_rate": 7.62216404886562e-06, - "loss": 0.6507, + "epoch": 9.82, + "learning_rate": 6.1596638655462195e-06, + "loss": 0.5159, "step": 2830 }, { - "epoch": 4.91, - "learning_rate": 7.613438045375219e-06, - "loss": 0.6277, + "epoch": 9.86, + "learning_rate": 6.138655462184875e-06, + "loss": 0.5274, "step": 2840 }, { - "epoch": 4.93, - "learning_rate": 7.6047120418848176e-06, - "loss": 0.5772, + "epoch": 9.89, + "learning_rate": 6.11764705882353e-06, + "loss": 0.494, "step": 2850 }, { - "epoch": 4.95, - "learning_rate": 7.595986038394416e-06, - "loss": 0.6038, + "epoch": 9.93, + "learning_rate": 6.0966386554621856e-06, + "loss": 0.5349, "step": 2860 }, { - "epoch": 4.97, - "learning_rate": 7.587260034904015e-06, - "loss": 0.6321, + "epoch": 9.96, + "learning_rate": 6.07563025210084e-06, + "loss": 0.5046, "step": 2870 }, { - "epoch": 4.98, - "learning_rate": 7.578534031413614e-06, - "loss": 0.6387, + "epoch": 10.0, + "learning_rate": 6.054621848739496e-06, + "loss": 0.5193, "step": 2880 }, { - "epoch": 5.0, - "learning_rate": 7.569808027923212e-06, - "loss": 0.6228, - "step": 2890 + "epoch": 10.0, + "eval_loss": 0.5888818502426147, + "eval_runtime": 132.9604, + "eval_samples_per_second": 4.001, + "eval_steps_per_second": 1.0, + "eval_wer": 0.3143922350067494, + "step": 2880 }, { - "epoch": 5.0, - "eval_loss": 0.6369497179985046, - "eval_runtime": 131.6121, - "eval_samples_per_second": 4.042, - "eval_steps_per_second": 0.509, - "eval_wer": 0.38580703220415247, + "epoch": 10.03, + "learning_rate": 6.033613445378152e-06, + "loss": 0.4943, "step": 2890 }, { - "epoch": 5.02, - "learning_rate": 7.561082024432811e-06, - "loss": 0.6214, + "epoch": 10.07, + "learning_rate": 6.012605042016807e-06, + "loss": 0.478, "step": 2900 }, { - "epoch": 5.03, - "learning_rate": 7.552356020942409e-06, - "loss": 0.5504, + "epoch": 10.1, + "learning_rate": 5.991596638655462e-06, + "loss": 0.4735, "step": 2910 }, { - "epoch": 5.05, - "learning_rate": 7.543630017452007e-06, - "loss": 0.5698, + "epoch": 10.14, + "learning_rate": 5.970588235294118e-06, + "loss": 0.5069, "step": 2920 }, { - "epoch": 5.07, - "learning_rate": 7.534904013961606e-06, - "loss": 0.5852, + "epoch": 10.17, + "learning_rate": 5.949579831932774e-06, + "loss": 0.5395, "step": 2930 }, { - "epoch": 5.09, - "learning_rate": 7.526178010471205e-06, - "loss": 0.6374, + "epoch": 10.21, + "learning_rate": 5.928571428571429e-06, + "loss": 0.47, "step": 2940 }, { - "epoch": 5.1, - "learning_rate": 7.517452006980803e-06, - "loss": 0.6181, + "epoch": 10.24, + "learning_rate": 5.9075630252100845e-06, + "loss": 0.4725, "step": 2950 }, { - "epoch": 5.12, - "learning_rate": 7.508726003490402e-06, - "loss": 0.6154, + "epoch": 10.28, + "learning_rate": 5.88655462184874e-06, + "loss": 0.4931, "step": 2960 }, { - "epoch": 5.14, - "learning_rate": 7.500000000000001e-06, - "loss": 0.5678, + "epoch": 10.31, + "learning_rate": 5.865546218487396e-06, + "loss": 0.4759, "step": 2970 }, { - "epoch": 5.16, - "learning_rate": 7.4912739965095994e-06, - "loss": 0.6184, + "epoch": 10.35, + "learning_rate": 5.844537815126051e-06, + "loss": 0.529, "step": 2980 }, { - "epoch": 5.17, - "learning_rate": 7.482547993019198e-06, - "loss": 0.6863, + "epoch": 10.38, + "learning_rate": 5.823529411764707e-06, + "loss": 0.5001, "step": 2990 }, { - "epoch": 5.19, - "learning_rate": 7.473821989528796e-06, - "loss": 0.6103, + "epoch": 10.42, + "learning_rate": 5.802521008403362e-06, + "loss": 0.5389, "step": 3000 }, { - "epoch": 5.21, - "learning_rate": 7.4650959860383955e-06, - "loss": 0.592, + "epoch": 10.45, + "learning_rate": 5.781512605042017e-06, + "loss": 0.5405, "step": 3010 }, { - "epoch": 5.22, - "learning_rate": 7.456369982547994e-06, - "loss": 0.5662, + "epoch": 10.48, + "learning_rate": 5.760504201680672e-06, + "loss": 0.5206, "step": 3020 }, { - "epoch": 5.24, - "learning_rate": 7.447643979057592e-06, - "loss": 0.6428, + "epoch": 10.52, + "learning_rate": 5.739495798319328e-06, + "loss": 0.557, "step": 3030 }, { - "epoch": 5.26, - "learning_rate": 7.43891797556719e-06, - "loss": 0.6139, + "epoch": 10.55, + "learning_rate": 5.7184873949579834e-06, + "loss": 0.5228, "step": 3040 }, { - "epoch": 5.28, - "learning_rate": 7.430191972076789e-06, - "loss": 0.6185, + "epoch": 10.59, + "learning_rate": 5.697478991596639e-06, + "loss": 0.5261, "step": 3050 }, { - "epoch": 5.29, - "learning_rate": 7.421465968586388e-06, - "loss": 0.5885, + "epoch": 10.62, + "learning_rate": 5.676470588235294e-06, + "loss": 0.4701, "step": 3060 }, { - "epoch": 5.31, - "learning_rate": 7.412739965095986e-06, - "loss": 0.5653, + "epoch": 10.66, + "learning_rate": 5.65546218487395e-06, + "loss": 0.4977, "step": 3070 }, { - "epoch": 5.33, - "learning_rate": 7.404013961605585e-06, - "loss": 0.5897, + "epoch": 10.69, + "learning_rate": 5.634453781512606e-06, + "loss": 0.5027, "step": 3080 }, { - "epoch": 5.35, - "learning_rate": 7.395287958115184e-06, - "loss": 0.6161, + "epoch": 10.73, + "learning_rate": 5.613445378151261e-06, + "loss": 0.506, "step": 3090 }, { - "epoch": 5.36, - "learning_rate": 7.386561954624782e-06, - "loss": 0.637, + "epoch": 10.76, + "learning_rate": 5.592436974789916e-06, + "loss": 0.5185, "step": 3100 }, { - "epoch": 5.38, - "learning_rate": 7.377835951134381e-06, - "loss": 0.6338, + "epoch": 10.8, + "learning_rate": 5.571428571428572e-06, + "loss": 0.5177, "step": 3110 }, { - "epoch": 5.4, - "learning_rate": 7.36910994764398e-06, - "loss": 0.6049, + "epoch": 10.83, + "learning_rate": 5.550420168067228e-06, + "loss": 0.5073, "step": 3120 }, { - "epoch": 5.42, - "learning_rate": 7.360383944153578e-06, - "loss": 0.6478, + "epoch": 10.87, + "learning_rate": 5.529411764705883e-06, + "loss": 0.525, "step": 3130 }, { - "epoch": 5.43, - "learning_rate": 7.351657940663177e-06, - "loss": 0.6265, + "epoch": 10.9, + "learning_rate": 5.5084033613445386e-06, + "loss": 0.5193, "step": 3140 }, { - "epoch": 5.45, - "learning_rate": 7.342931937172776e-06, - "loss": 0.6207, + "epoch": 10.94, + "learning_rate": 5.487394957983194e-06, + "loss": 0.4716, "step": 3150 }, { - "epoch": 5.47, - "learning_rate": 7.334205933682374e-06, - "loss": 0.5683, + "epoch": 10.97, + "learning_rate": 5.466386554621848e-06, + "loss": 0.4988, "step": 3160 }, { - "epoch": 5.48, - "learning_rate": 7.325479930191972e-06, - "loss": 0.6437, + "epoch": 11.0, + "eval_loss": 0.5690802335739136, + "eval_runtime": 134.0737, + "eval_samples_per_second": 3.968, + "eval_steps_per_second": 0.992, + "eval_wer": 0.31066400977052133, + "step": 3168 + }, + { + "epoch": 11.01, + "learning_rate": 5.445378151260505e-06, + "loss": 0.5836, "step": 3170 }, { - "epoch": 5.5, - "learning_rate": 7.316753926701571e-06, - "loss": 0.597, + "epoch": 11.04, + "learning_rate": 5.42436974789916e-06, + "loss": 0.4776, "step": 3180 }, { - "epoch": 5.52, - "learning_rate": 7.3080279232111695e-06, - "loss": 0.625, + "epoch": 11.08, + "learning_rate": 5.403361344537815e-06, + "loss": 0.4806, "step": 3190 }, { - "epoch": 5.54, - "learning_rate": 7.299301919720768e-06, - "loss": 0.5945, + "epoch": 11.11, + "learning_rate": 5.382352941176471e-06, + "loss": 0.4893, "step": 3200 }, { - "epoch": 5.55, - "learning_rate": 7.290575916230367e-06, - "loss": 0.6067, + "epoch": 11.15, + "learning_rate": 5.361344537815126e-06, + "loss": 0.5033, "step": 3210 }, { - "epoch": 5.57, - "learning_rate": 7.2818499127399655e-06, - "loss": 0.5843, + "epoch": 11.18, + "learning_rate": 5.340336134453782e-06, + "loss": 0.5383, "step": 3220 }, { - "epoch": 5.59, - "learning_rate": 7.273123909249564e-06, - "loss": 0.6162, + "epoch": 11.21, + "learning_rate": 5.3193277310924375e-06, + "loss": 0.5112, "step": 3230 }, { - "epoch": 5.61, - "learning_rate": 7.264397905759163e-06, - "loss": 0.6684, + "epoch": 11.25, + "learning_rate": 5.298319327731093e-06, + "loss": 0.5234, "step": 3240 }, { - "epoch": 5.62, - "learning_rate": 7.255671902268762e-06, - "loss": 0.6008, + "epoch": 11.28, + "learning_rate": 5.277310924369748e-06, + "loss": 0.4606, "step": 3250 }, { - "epoch": 5.64, - "learning_rate": 7.24694589877836e-06, - "loss": 0.5918, + "epoch": 11.32, + "learning_rate": 5.256302521008404e-06, + "loss": 0.4986, "step": 3260 }, { - "epoch": 5.66, - "learning_rate": 7.238219895287959e-06, - "loss": 0.6507, + "epoch": 11.35, + "learning_rate": 5.23529411764706e-06, + "loss": 0.5325, "step": 3270 }, { - "epoch": 5.67, - "learning_rate": 7.229493891797558e-06, - "loss": 0.6306, + "epoch": 11.39, + "learning_rate": 5.214285714285715e-06, + "loss": 0.5185, "step": 3280 }, { - "epoch": 5.69, - "learning_rate": 7.220767888307156e-06, - "loss": 0.628, + "epoch": 11.42, + "learning_rate": 5.19327731092437e-06, + "loss": 0.459, "step": 3290 }, { - "epoch": 5.71, - "learning_rate": 7.212041884816755e-06, - "loss": 0.6352, + "epoch": 11.46, + "learning_rate": 5.172268907563026e-06, + "loss": 0.5038, "step": 3300 }, { - "epoch": 5.73, - "learning_rate": 7.203315881326353e-06, - "loss": 0.6274, + "epoch": 11.49, + "learning_rate": 5.15126050420168e-06, + "loss": 0.5127, "step": 3310 }, { - "epoch": 5.74, - "learning_rate": 7.194589877835951e-06, - "loss": 0.5974, + "epoch": 11.53, + "learning_rate": 5.1302521008403364e-06, + "loss": 0.5238, "step": 3320 }, { - "epoch": 5.76, - "learning_rate": 7.18586387434555e-06, - "loss": 0.5933, + "epoch": 11.56, + "learning_rate": 5.109243697478992e-06, + "loss": 0.4919, "step": 3330 }, { - "epoch": 5.78, - "learning_rate": 7.177137870855149e-06, - "loss": 0.6824, + "epoch": 11.6, + "learning_rate": 5.088235294117647e-06, + "loss": 0.4719, "step": 3340 }, { - "epoch": 5.8, - "learning_rate": 7.1684118673647474e-06, - "loss": 0.6037, + "epoch": 11.63, + "learning_rate": 5.0672268907563025e-06, + "loss": 0.4991, "step": 3350 }, { - "epoch": 5.81, - "learning_rate": 7.159685863874346e-06, - "loss": 0.6859, + "epoch": 11.66, + "learning_rate": 5.046218487394959e-06, + "loss": 0.5098, "step": 3360 }, { - "epoch": 5.83, - "learning_rate": 7.150959860383945e-06, - "loss": 0.6239, + "epoch": 11.7, + "learning_rate": 5.025210084033614e-06, + "loss": 0.5066, "step": 3370 }, { - "epoch": 5.85, - "learning_rate": 7.1422338568935435e-06, - "loss": 0.6418, + "epoch": 11.73, + "learning_rate": 5.004201680672269e-06, + "loss": 0.4925, "step": 3380 }, { - "epoch": 5.87, - "learning_rate": 7.133507853403142e-06, - "loss": 0.6558, + "epoch": 11.77, + "learning_rate": 4.983193277310925e-06, + "loss": 0.4996, "step": 3390 }, { - "epoch": 5.88, - "learning_rate": 7.124781849912741e-06, - "loss": 0.5835, + "epoch": 11.8, + "learning_rate": 4.96218487394958e-06, + "loss": 0.5129, "step": 3400 }, { - "epoch": 5.9, - "learning_rate": 7.11605584642234e-06, - "loss": 0.6402, + "epoch": 11.84, + "learning_rate": 4.941176470588236e-06, + "loss": 0.4944, "step": 3410 }, { - "epoch": 5.92, - "learning_rate": 7.107329842931938e-06, - "loss": 0.5827, + "epoch": 11.87, + "learning_rate": 4.920168067226891e-06, + "loss": 0.4929, "step": 3420 }, { - "epoch": 5.93, - "learning_rate": 7.098603839441537e-06, - "loss": 0.5966, + "epoch": 11.91, + "learning_rate": 4.899159663865546e-06, + "loss": 0.4904, "step": 3430 }, { - "epoch": 5.95, - "learning_rate": 7.089877835951135e-06, - "loss": 0.643, + "epoch": 11.94, + "learning_rate": 4.878151260504202e-06, + "loss": 0.4992, "step": 3440 }, { - "epoch": 5.97, - "learning_rate": 7.081151832460733e-06, - "loss": 0.6258, + "epoch": 11.98, + "learning_rate": 4.857142857142858e-06, + "loss": 0.4966, "step": 3450 }, { - "epoch": 5.99, - "learning_rate": 7.072425828970332e-06, - "loss": 0.6004, - "step": 3460 + "epoch": 12.0, + "eval_loss": 0.5908366441726685, + "eval_runtime": 130.8263, + "eval_samples_per_second": 4.066, + "eval_steps_per_second": 1.017, + "eval_wer": 0.31272096162499197, + "step": 3456 }, { - "epoch": 6.0, - "eval_loss": 0.6312674880027771, - "eval_runtime": 129.1633, - "eval_samples_per_second": 4.119, - "eval_steps_per_second": 0.519, - "eval_wer": 0.4087549013305907, - "step": 3468 + "epoch": 12.01, + "learning_rate": 4.836134453781513e-06, + "loss": 0.5368, + "step": 3460 }, { - "epoch": 6.0, - "learning_rate": 7.063699825479931e-06, - "loss": 0.6183, + "epoch": 12.05, + "learning_rate": 4.815126050420168e-06, + "loss": 0.509, "step": 3470 }, { - "epoch": 6.02, - "learning_rate": 7.054973821989529e-06, - "loss": 0.6564, + "epoch": 12.08, + "learning_rate": 4.7941176470588245e-06, + "loss": 0.4688, "step": 3480 }, { - "epoch": 6.04, - "learning_rate": 7.046247818499128e-06, - "loss": 0.6068, + "epoch": 12.12, + "learning_rate": 4.77310924369748e-06, + "loss": 0.4622, "step": 3490 }, { - "epoch": 6.06, - "learning_rate": 7.037521815008726e-06, - "loss": 0.5793, + "epoch": 12.15, + "learning_rate": 4.752100840336134e-06, + "loss": 0.506, "step": 3500 }, { - "epoch": 6.07, - "learning_rate": 7.028795811518325e-06, - "loss": 0.5795, + "epoch": 12.19, + "learning_rate": 4.7310924369747905e-06, + "loss": 0.4969, "step": 3510 }, { - "epoch": 6.09, - "learning_rate": 7.020942408376964e-06, - "loss": 0.7028, + "epoch": 12.22, + "learning_rate": 4.710084033613446e-06, + "loss": 0.4873, "step": 3520 }, { - "epoch": 6.11, - "learning_rate": 7.0122164048865625e-06, - "loss": 0.6506, + "epoch": 12.26, + "learning_rate": 4.689075630252101e-06, + "loss": 0.489, "step": 3530 }, { - "epoch": 6.12, - "learning_rate": 7.003490401396162e-06, - "loss": 0.6417, + "epoch": 12.29, + "learning_rate": 4.6680672268907565e-06, + "loss": 0.4942, "step": 3540 }, { - "epoch": 6.14, - "learning_rate": 6.99476439790576e-06, - "loss": 0.558, + "epoch": 12.33, + "learning_rate": 4.647058823529412e-06, + "loss": 0.4939, "step": 3550 }, { - "epoch": 6.16, - "learning_rate": 6.9860383944153585e-06, - "loss": 0.5936, + "epoch": 12.36, + "learning_rate": 4.626050420168068e-06, + "loss": 0.5565, "step": 3560 }, { - "epoch": 6.18, - "learning_rate": 6.977312390924958e-06, - "loss": 0.598, + "epoch": 12.39, + "learning_rate": 4.6050420168067226e-06, + "loss": 0.5037, "step": 3570 }, { - "epoch": 6.19, - "learning_rate": 6.968586387434556e-06, - "loss": 0.6364, + "epoch": 12.43, + "learning_rate": 4.584033613445379e-06, + "loss": 0.4923, "step": 3580 }, { - "epoch": 6.21, - "learning_rate": 6.959860383944154e-06, - "loss": 0.5339, + "epoch": 12.46, + "learning_rate": 4.563025210084034e-06, + "loss": 0.4823, "step": 3590 }, { - "epoch": 6.23, - "learning_rate": 6.951134380453752e-06, - "loss": 0.5823, + "epoch": 12.5, + "learning_rate": 4.5420168067226894e-06, + "loss": 0.5044, "step": 3600 }, { - "epoch": 6.25, - "learning_rate": 6.942408376963351e-06, - "loss": 0.5901, + "epoch": 12.53, + "learning_rate": 4.521008403361345e-06, + "loss": 0.4982, "step": 3610 }, { - "epoch": 6.26, - "learning_rate": 6.93368237347295e-06, - "loss": 0.6247, + "epoch": 12.57, + "learning_rate": 4.5e-06, + "loss": 0.4721, "step": 3620 }, { - "epoch": 6.28, - "learning_rate": 6.924956369982548e-06, - "loss": 0.5914, + "epoch": 12.6, + "learning_rate": 4.478991596638656e-06, + "loss": 0.4986, "step": 3630 }, { - "epoch": 6.3, - "learning_rate": 6.916230366492147e-06, - "loss": 0.5591, + "epoch": 12.64, + "learning_rate": 4.457983193277312e-06, + "loss": 0.4926, "step": 3640 }, { - "epoch": 6.31, - "learning_rate": 6.907504363001746e-06, - "loss": 0.5877, + "epoch": 12.67, + "learning_rate": 4.436974789915966e-06, + "loss": 0.4886, "step": 3650 }, { - "epoch": 6.33, - "learning_rate": 6.898778359511344e-06, - "loss": 0.6368, + "epoch": 12.71, + "learning_rate": 4.415966386554622e-06, + "loss": 0.492, "step": 3660 }, { - "epoch": 6.35, - "learning_rate": 6.890052356020943e-06, - "loss": 0.6138, + "epoch": 12.74, + "learning_rate": 4.394957983193278e-06, + "loss": 0.5318, "step": 3670 }, { - "epoch": 6.37, - "learning_rate": 6.881326352530542e-06, - "loss": 0.5976, + "epoch": 12.78, + "learning_rate": 4.373949579831933e-06, + "loss": 0.4788, "step": 3680 }, { - "epoch": 6.38, - "learning_rate": 6.87260034904014e-06, - "loss": 0.6226, + "epoch": 12.81, + "learning_rate": 4.352941176470588e-06, + "loss": 0.4754, "step": 3690 }, { - "epoch": 6.4, - "learning_rate": 6.863874345549739e-06, - "loss": 0.648, + "epoch": 12.85, + "learning_rate": 4.3319327731092446e-06, + "loss": 0.5176, "step": 3700 }, { - "epoch": 6.42, - "learning_rate": 6.855148342059338e-06, - "loss": 0.5767, + "epoch": 12.88, + "learning_rate": 4.3109243697479e-06, + "loss": 0.5241, "step": 3710 }, { - "epoch": 6.44, - "learning_rate": 6.846422338568936e-06, - "loss": 0.6528, + "epoch": 12.91, + "learning_rate": 4.289915966386554e-06, + "loss": 0.4888, "step": 3720 }, { - "epoch": 6.45, - "learning_rate": 6.837696335078534e-06, - "loss": 0.5804, + "epoch": 12.95, + "learning_rate": 4.268907563025211e-06, + "loss": 0.4823, "step": 3730 }, { - "epoch": 6.47, - "learning_rate": 6.8289703315881325e-06, - "loss": 0.5692, + "epoch": 12.98, + "learning_rate": 4.247899159663866e-06, + "loss": 0.4801, "step": 3740 }, { - "epoch": 6.49, - "learning_rate": 6.820244328097732e-06, - "loss": 0.5992, + "epoch": 13.0, + "eval_loss": 0.58124178647995, + "eval_runtime": 131.995, + "eval_samples_per_second": 4.03, + "eval_steps_per_second": 1.008, + "eval_wer": 0.31098540849778233, + "step": 3744 + }, + { + "epoch": 13.02, + "learning_rate": 4.226890756302521e-06, + "loss": 0.5296, "step": 3750 }, { - "epoch": 6.51, - "learning_rate": 6.81151832460733e-06, - "loss": 0.6044, + "epoch": 13.06, + "learning_rate": 4.205882352941177e-06, + "loss": 0.5093, "step": 3760 }, { - "epoch": 6.52, - "learning_rate": 6.8027923211169286e-06, - "loss": 0.6106, + "epoch": 13.09, + "learning_rate": 4.184873949579833e-06, + "loss": 0.4723, "step": 3770 }, { - "epoch": 6.54, - "learning_rate": 6.794066317626528e-06, - "loss": 0.6039, + "epoch": 13.12, + "learning_rate": 4.163865546218488e-06, + "loss": 0.457, "step": 3780 }, { - "epoch": 6.56, - "learning_rate": 6.785340314136126e-06, - "loss": 0.616, + "epoch": 13.16, + "learning_rate": 4.1428571428571435e-06, + "loss": 0.4917, "step": 3790 }, { - "epoch": 6.57, - "learning_rate": 6.776614310645725e-06, - "loss": 0.5769, + "epoch": 13.19, + "learning_rate": 4.121848739495799e-06, + "loss": 0.5177, "step": 3800 }, { - "epoch": 6.59, - "learning_rate": 6.767888307155324e-06, - "loss": 0.5896, + "epoch": 13.23, + "learning_rate": 4.100840336134454e-06, + "loss": 0.4622, "step": 3810 }, { - "epoch": 6.61, - "learning_rate": 6.759162303664922e-06, - "loss": 0.6858, + "epoch": 13.26, + "learning_rate": 4.0798319327731095e-06, + "loss": 0.5018, "step": 3820 }, { - "epoch": 6.63, - "learning_rate": 6.750436300174521e-06, - "loss": 0.6426, + "epoch": 13.3, + "learning_rate": 4.058823529411765e-06, + "loss": 0.4981, "step": 3830 }, { - "epoch": 6.64, - "learning_rate": 6.74171029668412e-06, - "loss": 0.6174, + "epoch": 13.33, + "learning_rate": 4.03781512605042e-06, + "loss": 0.4718, "step": 3840 }, { - "epoch": 6.66, - "learning_rate": 6.732984293193718e-06, - "loss": 0.6534, + "epoch": 13.37, + "learning_rate": 4.016806722689076e-06, + "loss": 0.4832, "step": 3850 }, { - "epoch": 6.68, - "learning_rate": 6.724258289703316e-06, - "loss": 0.647, + "epoch": 13.4, + "learning_rate": 3.995798319327732e-06, + "loss": 0.5347, "step": 3860 }, { - "epoch": 6.7, - "learning_rate": 6.715532286212914e-06, - "loss": 0.6479, + "epoch": 13.44, + "learning_rate": 3.974789915966386e-06, + "loss": 0.4505, "step": 3870 }, { - "epoch": 6.71, - "learning_rate": 6.706806282722514e-06, - "loss": 0.585, + "epoch": 13.47, + "learning_rate": 3.953781512605042e-06, + "loss": 0.4903, "step": 3880 }, { - "epoch": 6.73, - "learning_rate": 6.698080279232112e-06, - "loss": 0.5917, + "epoch": 13.51, + "learning_rate": 3.932773109243698e-06, + "loss": 0.5229, "step": 3890 }, { - "epoch": 6.75, - "learning_rate": 6.6893542757417104e-06, - "loss": 0.6117, + "epoch": 13.54, + "learning_rate": 3.911764705882353e-06, + "loss": 0.5046, "step": 3900 }, { - "epoch": 6.76, - "learning_rate": 6.68062827225131e-06, - "loss": 0.5873, + "epoch": 13.57, + "learning_rate": 3.8907563025210084e-06, + "loss": 0.4707, "step": 3910 }, { - "epoch": 6.78, - "learning_rate": 6.671902268760908e-06, - "loss": 0.6346, + "epoch": 13.61, + "learning_rate": 3.869747899159665e-06, + "loss": 0.4842, "step": 3920 }, { - "epoch": 6.8, - "learning_rate": 6.6631762652705065e-06, - "loss": 0.5899, + "epoch": 13.64, + "learning_rate": 3.84873949579832e-06, + "loss": 0.4994, "step": 3930 }, { - "epoch": 6.82, - "learning_rate": 6.654450261780106e-06, - "loss": 0.6398, + "epoch": 13.68, + "learning_rate": 3.8277310924369745e-06, + "loss": 0.5029, "step": 3940 }, { - "epoch": 6.83, - "learning_rate": 6.645724258289704e-06, - "loss": 0.5839, + "epoch": 13.71, + "learning_rate": 3.8067226890756302e-06, + "loss": 0.4938, "step": 3950 }, { - "epoch": 6.85, - "learning_rate": 6.636998254799303e-06, - "loss": 0.6711, + "epoch": 13.75, + "learning_rate": 3.785714285714286e-06, + "loss": 0.5022, "step": 3960 }, { - "epoch": 6.87, - "learning_rate": 6.628272251308902e-06, - "loss": 0.6474, + "epoch": 13.78, + "learning_rate": 3.7647058823529414e-06, + "loss": 0.4759, "step": 3970 }, { - "epoch": 6.89, - "learning_rate": 6.6195462478185e-06, - "loss": 0.5458, + "epoch": 13.82, + "learning_rate": 3.743697478991597e-06, + "loss": 0.5091, "step": 3980 }, { - "epoch": 6.9, - "learning_rate": 6.610820244328098e-06, - "loss": 0.5972, + "epoch": 13.85, + "learning_rate": 3.7226890756302525e-06, + "loss": 0.4843, "step": 3990 }, { - "epoch": 6.92, - "learning_rate": 6.602094240837696e-06, - "loss": 0.5973, + "epoch": 13.89, + "learning_rate": 3.7016806722689082e-06, + "loss": 0.5043, "step": 4000 }, { - "epoch": 6.94, - "learning_rate": 6.5933682373472955e-06, - "loss": 0.5747, + "epoch": 13.92, + "learning_rate": 3.6806722689075636e-06, + "loss": 0.4884, "step": 4010 }, { - "epoch": 6.96, - "learning_rate": 6.584642233856894e-06, - "loss": 0.651, + "epoch": 13.96, + "learning_rate": 3.6596638655462185e-06, + "loss": 0.4764, "step": 4020 }, { - "epoch": 6.97, - "learning_rate": 6.575916230366492e-06, - "loss": 0.5707, + "epoch": 13.99, + "learning_rate": 3.6386554621848743e-06, + "loss": 0.5025, "step": 4030 }, { - "epoch": 6.99, - "learning_rate": 6.567190226876092e-06, - "loss": 0.6074, - "step": 4040 + "epoch": 14.0, + "eval_loss": 0.5805001258850098, + "eval_runtime": 132.3722, + "eval_samples_per_second": 4.019, + "eval_steps_per_second": 1.005, + "eval_wer": 0.3046217136980138, + "step": 4032 }, { - "epoch": 7.0, - "eval_loss": 0.6294800639152527, - "eval_runtime": 128.6238, - "eval_samples_per_second": 4.136, - "eval_steps_per_second": 0.521, - "eval_wer": 0.37937905765893165, - "step": 4046 + "epoch": 14.03, + "learning_rate": 3.6176470588235296e-06, + "loss": 0.5226, + "step": 4040 }, { - "epoch": 7.01, - "learning_rate": 6.55846422338569e-06, - "loss": 0.5778, + "epoch": 14.06, + "learning_rate": 3.5966386554621854e-06, + "loss": 0.4922, "step": 4050 }, { - "epoch": 7.02, - "learning_rate": 6.549738219895288e-06, - "loss": 0.638, + "epoch": 14.1, + "learning_rate": 3.5756302521008407e-06, + "loss": 0.4673, "step": 4060 }, { - "epoch": 7.04, - "learning_rate": 6.541012216404888e-06, - "loss": 0.5909, + "epoch": 14.13, + "learning_rate": 3.554621848739496e-06, + "loss": 0.5106, "step": 4070 }, { - "epoch": 7.06, - "learning_rate": 6.532286212914486e-06, - "loss": 0.5937, + "epoch": 14.17, + "learning_rate": 3.533613445378152e-06, + "loss": 0.4949, "step": 4080 }, { - "epoch": 7.08, - "learning_rate": 6.5235602094240845e-06, - "loss": 0.5668, + "epoch": 14.2, + "learning_rate": 3.5126050420168067e-06, + "loss": 0.4841, "step": 4090 }, { - "epoch": 7.09, - "learning_rate": 6.514834205933684e-06, - "loss": 0.5907, + "epoch": 14.24, + "learning_rate": 3.491596638655462e-06, + "loss": 0.4896, "step": 4100 }, { - "epoch": 7.11, - "learning_rate": 6.506108202443282e-06, - "loss": 0.5832, + "epoch": 14.27, + "learning_rate": 3.470588235294118e-06, + "loss": 0.483, "step": 4110 }, { - "epoch": 7.13, - "learning_rate": 6.49738219895288e-06, - "loss": 0.5773, + "epoch": 14.3, + "learning_rate": 3.449579831932773e-06, + "loss": 0.4755, "step": 4120 }, { - "epoch": 7.15, - "learning_rate": 6.488656195462478e-06, - "loss": 0.592, + "epoch": 14.34, + "learning_rate": 3.428571428571429e-06, + "loss": 0.4812, "step": 4130 }, { - "epoch": 7.16, - "learning_rate": 6.4799301919720765e-06, - "loss": 0.5961, + "epoch": 14.37, + "learning_rate": 3.4075630252100843e-06, + "loss": 0.4835, "step": 4140 }, { - "epoch": 7.18, - "learning_rate": 6.471204188481676e-06, - "loss": 0.6471, + "epoch": 14.41, + "learning_rate": 3.38655462184874e-06, + "loss": 0.4613, "step": 4150 }, { - "epoch": 7.2, - "learning_rate": 6.462478184991274e-06, - "loss": 0.5656, + "epoch": 14.44, + "learning_rate": 3.3655462184873954e-06, + "loss": 0.4349, "step": 4160 }, { - "epoch": 7.21, - "learning_rate": 6.453752181500873e-06, - "loss": 0.5576, + "epoch": 14.48, + "learning_rate": 3.3445378151260503e-06, + "loss": 0.4943, "step": 4170 }, { - "epoch": 7.23, - "learning_rate": 6.445026178010472e-06, - "loss": 0.5471, + "epoch": 14.51, + "learning_rate": 3.323529411764706e-06, + "loss": 0.4983, "step": 4180 }, { - "epoch": 7.25, - "learning_rate": 6.43630017452007e-06, - "loss": 0.6564, + "epoch": 14.55, + "learning_rate": 3.3025210084033614e-06, + "loss": 0.4892, "step": 4190 }, { - "epoch": 7.27, - "learning_rate": 6.427574171029669e-06, - "loss": 0.6392, + "epoch": 14.58, + "learning_rate": 3.281512605042017e-06, + "loss": 0.5268, "step": 4200 }, { - "epoch": 7.28, - "learning_rate": 6.418848167539268e-06, - "loss": 0.6312, + "epoch": 14.62, + "learning_rate": 3.2605042016806726e-06, + "loss": 0.488, "step": 4210 }, { - "epoch": 7.3, - "learning_rate": 6.410122164048866e-06, - "loss": 0.5795, + "epoch": 14.65, + "learning_rate": 3.2394957983193283e-06, + "loss": 0.4964, "step": 4220 }, { - "epoch": 7.32, - "learning_rate": 6.401396160558465e-06, - "loss": 0.5597, + "epoch": 14.69, + "learning_rate": 3.2184873949579837e-06, + "loss": 0.5091, "step": 4230 }, { - "epoch": 7.34, - "learning_rate": 6.392670157068064e-06, - "loss": 0.6218, + "epoch": 14.72, + "learning_rate": 3.1974789915966386e-06, + "loss": 0.4816, "step": 4240 }, { - "epoch": 7.35, - "learning_rate": 6.3839441535776624e-06, - "loss": 0.5992, + "epoch": 14.75, + "learning_rate": 3.1764705882352943e-06, + "loss": 0.4648, "step": 4250 }, { - "epoch": 7.37, - "learning_rate": 6.37521815008726e-06, - "loss": 0.6167, + "epoch": 14.79, + "learning_rate": 3.1554621848739497e-06, + "loss": 0.4498, "step": 4260 }, { - "epoch": 7.39, - "learning_rate": 6.366492146596858e-06, - "loss": 0.5695, + "epoch": 14.82, + "learning_rate": 3.1344537815126055e-06, + "loss": 0.4997, "step": 4270 }, { - "epoch": 7.4, - "learning_rate": 6.357766143106458e-06, - "loss": 0.5906, + "epoch": 14.86, + "learning_rate": 3.113445378151261e-06, + "loss": 0.5136, "step": 4280 }, { - "epoch": 7.42, - "learning_rate": 6.349040139616056e-06, - "loss": 0.5929, + "epoch": 14.89, + "learning_rate": 3.092436974789916e-06, + "loss": 0.5009, "step": 4290 }, { - "epoch": 7.44, - "learning_rate": 6.3403141361256545e-06, - "loss": 0.6004, + "epoch": 14.93, + "learning_rate": 3.071428571428572e-06, + "loss": 0.4597, "step": 4300 }, { - "epoch": 7.46, - "learning_rate": 6.331588132635254e-06, - "loss": 0.6006, + "epoch": 14.96, + "learning_rate": 3.0504201680672273e-06, + "loss": 0.4724, "step": 4310 }, { - "epoch": 7.47, - "learning_rate": 6.322862129144852e-06, - "loss": 0.5843, + "epoch": 15.0, + "learning_rate": 3.0294117647058826e-06, + "loss": 0.5048, + "step": 4320 + }, + { + "epoch": 15.0, + "eval_loss": 0.5906367301940918, + "eval_runtime": 133.6961, + "eval_samples_per_second": 3.979, + "eval_steps_per_second": 0.995, + "eval_wer": 0.313363759079514, "step": 4320 }, { - "epoch": 7.49, - "learning_rate": 6.314136125654451e-06, - "loss": 0.5975, + "epoch": 15.03, + "learning_rate": 3.008403361344538e-06, + "loss": 0.5216, "step": 4330 }, { - "epoch": 7.51, - "learning_rate": 6.30541012216405e-06, - "loss": 0.577, + "epoch": 15.07, + "learning_rate": 2.9873949579831933e-06, + "loss": 0.4814, "step": 4340 }, { - "epoch": 7.53, - "learning_rate": 6.296684118673648e-06, - "loss": 0.639, + "epoch": 15.1, + "learning_rate": 2.966386554621849e-06, + "loss": 0.4715, "step": 4350 }, { - "epoch": 7.54, - "learning_rate": 6.287958115183247e-06, - "loss": 0.5891, + "epoch": 15.14, + "learning_rate": 2.9453781512605044e-06, + "loss": 0.5006, "step": 4360 }, { - "epoch": 7.56, - "learning_rate": 6.279232111692846e-06, - "loss": 0.5674, + "epoch": 15.17, + "learning_rate": 2.92436974789916e-06, + "loss": 0.5171, "step": 4370 }, { - "epoch": 7.58, - "learning_rate": 6.270506108202444e-06, - "loss": 0.5703, + "epoch": 15.21, + "learning_rate": 2.9033613445378155e-06, + "loss": 0.4885, "step": 4380 }, { - "epoch": 7.6, - "learning_rate": 6.261780104712042e-06, - "loss": 0.6225, + "epoch": 15.24, + "learning_rate": 2.8823529411764704e-06, + "loss": 0.4806, "step": 4390 }, { - "epoch": 7.61, - "learning_rate": 6.25305410122164e-06, - "loss": 0.6027, + "epoch": 15.28, + "learning_rate": 2.861344537815126e-06, + "loss": 0.4416, "step": 4400 }, { - "epoch": 7.63, - "learning_rate": 6.2443280977312396e-06, - "loss": 0.5792, + "epoch": 15.31, + "learning_rate": 2.8403361344537815e-06, + "loss": 0.4799, "step": 4410 }, { - "epoch": 7.65, - "learning_rate": 6.235602094240838e-06, - "loss": 0.5622, + "epoch": 15.35, + "learning_rate": 2.8193277310924373e-06, + "loss": 0.4737, "step": 4420 }, { - "epoch": 7.66, - "learning_rate": 6.226876090750436e-06, - "loss": 0.6008, + "epoch": 15.38, + "learning_rate": 2.7983193277310926e-06, + "loss": 0.4705, "step": 4430 }, { - "epoch": 7.68, - "learning_rate": 6.218150087260036e-06, - "loss": 0.5678, + "epoch": 15.42, + "learning_rate": 2.7773109243697484e-06, + "loss": 0.4735, "step": 4440 }, { - "epoch": 7.7, - "learning_rate": 6.209424083769634e-06, - "loss": 0.6323, + "epoch": 15.45, + "learning_rate": 2.7563025210084037e-06, + "loss": 0.4891, "step": 4450 }, { - "epoch": 7.72, - "learning_rate": 6.2006980802792325e-06, - "loss": 0.5445, + "epoch": 15.48, + "learning_rate": 2.7352941176470595e-06, + "loss": 0.4813, "step": 4460 }, { - "epoch": 7.73, - "learning_rate": 6.191972076788832e-06, - "loss": 0.6158, + "epoch": 15.52, + "learning_rate": 2.7142857142857144e-06, + "loss": 0.496, "step": 4470 }, { - "epoch": 7.75, - "learning_rate": 6.18324607329843e-06, - "loss": 0.5954, + "epoch": 15.55, + "learning_rate": 2.6932773109243698e-06, + "loss": 0.4669, "step": 4480 }, { - "epoch": 7.77, - "learning_rate": 6.1745200698080285e-06, - "loss": 0.5546, + "epoch": 15.59, + "learning_rate": 2.6722689075630255e-06, + "loss": 0.4831, "step": 4490 }, { - "epoch": 7.79, - "learning_rate": 6.165794066317628e-06, - "loss": 0.6714, + "epoch": 15.62, + "learning_rate": 2.651260504201681e-06, + "loss": 0.4524, "step": 4500 }, { - "epoch": 7.8, - "learning_rate": 6.157068062827226e-06, - "loss": 0.5493, + "epoch": 15.66, + "learning_rate": 2.6302521008403362e-06, + "loss": 0.4802, "step": 4510 }, { - "epoch": 7.82, - "learning_rate": 6.148342059336824e-06, - "loss": 0.5918, + "epoch": 15.69, + "learning_rate": 2.609243697478992e-06, + "loss": 0.4813, "step": 4520 }, { - "epoch": 7.84, - "learning_rate": 6.139616055846422e-06, - "loss": 0.5873, + "epoch": 15.73, + "learning_rate": 2.5882352941176473e-06, + "loss": 0.4836, "step": 4530 }, { - "epoch": 7.85, - "learning_rate": 6.1308900523560214e-06, - "loss": 0.613, + "epoch": 15.76, + "learning_rate": 2.5672268907563027e-06, + "loss": 0.4962, "step": 4540 }, { - "epoch": 7.87, - "learning_rate": 6.12216404886562e-06, - "loss": 0.6376, + "epoch": 15.8, + "learning_rate": 2.546218487394958e-06, + "loss": 0.4936, "step": 4550 }, { - "epoch": 7.89, - "learning_rate": 6.113438045375218e-06, - "loss": 0.5883, + "epoch": 15.83, + "learning_rate": 2.5252100840336134e-06, + "loss": 0.4919, "step": 4560 }, { - "epoch": 7.91, - "learning_rate": 6.1047120418848175e-06, - "loss": 0.5745, + "epoch": 15.87, + "learning_rate": 2.504201680672269e-06, + "loss": 0.543, "step": 4570 }, { - "epoch": 7.92, - "learning_rate": 6.095986038394416e-06, - "loss": 0.5425, + "epoch": 15.9, + "learning_rate": 2.4831932773109245e-06, + "loss": 0.514, "step": 4580 }, { - "epoch": 7.94, - "learning_rate": 6.087260034904014e-06, - "loss": 0.6141, + "epoch": 15.94, + "learning_rate": 2.4642857142857147e-06, + "loss": 0.4819, "step": 4590 }, { - "epoch": 7.96, - "learning_rate": 6.078534031413614e-06, - "loss": 0.6054, + "epoch": 15.97, + "learning_rate": 2.44327731092437e-06, + "loss": 0.4772, "step": 4600 }, { - "epoch": 7.98, - "learning_rate": 6.069808027923212e-06, - "loss": 0.606, - "step": 4610 + "epoch": 16.0, + "eval_loss": 0.5693268775939941, + "eval_runtime": 132.7597, + "eval_samples_per_second": 4.007, + "eval_steps_per_second": 1.002, + "eval_wer": 0.3009577682072379, + "step": 4608 }, { - "epoch": 7.99, - "learning_rate": 6.0610820244328104e-06, - "loss": 0.6075, - "step": 4620 + "epoch": 16.01, + "learning_rate": 2.4222689075630254e-06, + "loss": 0.5274, + "step": 4610 }, { - "epoch": 8.0, - "eval_loss": 0.6280089020729065, - "eval_runtime": 129.7845, - "eval_samples_per_second": 4.099, - "eval_steps_per_second": 0.516, - "eval_wer": 0.3692228578774828, - "step": 4624 + "epoch": 16.04, + "learning_rate": 2.4012605042016807e-06, + "loss": 0.4773, + "step": 4620 }, { - "epoch": 8.01, - "learning_rate": 6.052356020942409e-06, - "loss": 0.5259, + "epoch": 16.08, + "learning_rate": 2.380252100840336e-06, + "loss": 0.4692, "step": 4630 }, { - "epoch": 8.03, - "learning_rate": 6.043630017452008e-06, - "loss": 0.5915, + "epoch": 16.11, + "learning_rate": 2.359243697478992e-06, + "loss": 0.4599, "step": 4640 }, { - "epoch": 8.04, - "learning_rate": 6.0349040139616065e-06, - "loss": 0.5615, + "epoch": 16.15, + "learning_rate": 2.338235294117647e-06, + "loss": 0.4805, "step": 4650 }, { - "epoch": 8.06, - "learning_rate": 6.026178010471204e-06, - "loss": 0.5799, + "epoch": 16.18, + "learning_rate": 2.317226890756303e-06, + "loss": 0.4887, "step": 4660 }, { - "epoch": 8.08, - "learning_rate": 6.0174520069808025e-06, - "loss": 0.5899, + "epoch": 16.21, + "learning_rate": 2.296218487394958e-06, + "loss": 0.4909, "step": 4670 }, { - "epoch": 8.1, - "learning_rate": 6.008726003490402e-06, - "loss": 0.6507, + "epoch": 16.25, + "learning_rate": 2.2752100840336136e-06, + "loss": 0.4625, "step": 4680 }, { - "epoch": 8.11, - "learning_rate": 6e-06, - "loss": 0.6065, + "epoch": 16.28, + "learning_rate": 2.254201680672269e-06, + "loss": 0.477, "step": 4690 }, { - "epoch": 8.13, - "learning_rate": 5.9912739965095986e-06, - "loss": 0.6088, + "epoch": 16.32, + "learning_rate": 2.2331932773109248e-06, + "loss": 0.4909, "step": 4700 }, { - "epoch": 8.15, - "learning_rate": 5.982547993019198e-06, - "loss": 0.5498, + "epoch": 16.35, + "learning_rate": 2.21218487394958e-06, + "loss": 0.518, "step": 4710 }, { - "epoch": 8.17, - "learning_rate": 5.973821989528796e-06, - "loss": 0.6099, + "epoch": 16.39, + "learning_rate": 2.1911764705882354e-06, + "loss": 0.4702, "step": 4720 }, { - "epoch": 8.18, - "learning_rate": 5.965095986038395e-06, - "loss": 0.5739, + "epoch": 16.42, + "learning_rate": 2.1701680672268908e-06, + "loss": 0.49, "step": 4730 }, { - "epoch": 8.2, - "learning_rate": 5.956369982547994e-06, - "loss": 0.6152, + "epoch": 16.46, + "learning_rate": 2.1491596638655466e-06, + "loss": 0.4882, "step": 4740 }, { - "epoch": 8.22, - "learning_rate": 5.947643979057592e-06, - "loss": 0.5395, + "epoch": 16.49, + "learning_rate": 2.128151260504202e-06, + "loss": 0.4719, "step": 4750 }, { - "epoch": 8.24, - "learning_rate": 5.938917975567191e-06, - "loss": 0.6203, + "epoch": 16.53, + "learning_rate": 2.1071428571428572e-06, + "loss": 0.5152, "step": 4760 }, { - "epoch": 8.25, - "learning_rate": 5.93019197207679e-06, - "loss": 0.5697, + "epoch": 16.56, + "learning_rate": 2.086134453781513e-06, + "loss": 0.4672, "step": 4770 }, { - "epoch": 8.27, - "learning_rate": 5.921465968586388e-06, - "loss": 0.6056, + "epoch": 16.6, + "learning_rate": 2.065126050420168e-06, + "loss": 0.4704, "step": 4780 }, { - "epoch": 8.29, - "learning_rate": 5.912739965095986e-06, - "loss": 0.5887, + "epoch": 16.63, + "learning_rate": 2.0441176470588237e-06, + "loss": 0.4678, "step": 4790 }, { - "epoch": 8.3, - "learning_rate": 5.904013961605584e-06, - "loss": 0.5593, + "epoch": 16.66, + "learning_rate": 2.023109243697479e-06, + "loss": 0.4737, "step": 4800 }, { - "epoch": 8.32, - "learning_rate": 5.895287958115184e-06, - "loss": 0.5574, + "epoch": 16.7, + "learning_rate": 2.002100840336135e-06, + "loss": 0.4608, "step": 4810 }, { - "epoch": 8.34, - "learning_rate": 5.886561954624782e-06, - "loss": 0.6184, + "epoch": 16.73, + "learning_rate": 1.98109243697479e-06, + "loss": 0.5042, "step": 4820 }, { - "epoch": 8.36, - "learning_rate": 5.8778359511343804e-06, - "loss": 0.5504, + "epoch": 16.77, + "learning_rate": 1.9600840336134455e-06, + "loss": 0.5033, "step": 4830 }, { - "epoch": 8.37, - "learning_rate": 5.86910994764398e-06, - "loss": 0.5687, + "epoch": 16.8, + "learning_rate": 1.9390756302521013e-06, + "loss": 0.4717, "step": 4840 }, { - "epoch": 8.39, - "learning_rate": 5.860383944153578e-06, - "loss": 0.5746, + "epoch": 16.84, + "learning_rate": 1.9180672268907566e-06, + "loss": 0.4869, "step": 4850 }, { - "epoch": 8.41, - "learning_rate": 5.8516579406631765e-06, - "loss": 0.5833, + "epoch": 16.87, + "learning_rate": 1.897058823529412e-06, + "loss": 0.4963, "step": 4860 }, { - "epoch": 8.43, - "learning_rate": 5.842931937172776e-06, - "loss": 0.6979, + "epoch": 16.91, + "learning_rate": 1.8760504201680673e-06, + "loss": 0.4411, "step": 4870 }, { - "epoch": 8.44, - "learning_rate": 5.834205933682374e-06, - "loss": 0.5845, + "epoch": 16.94, + "learning_rate": 1.8550420168067228e-06, + "loss": 0.4456, "step": 4880 }, { - "epoch": 8.46, - "learning_rate": 5.825479930191973e-06, - "loss": 0.5722, + "epoch": 16.98, + "learning_rate": 1.8340336134453782e-06, + "loss": 0.4748, "step": 4890 }, { - "epoch": 8.48, - "learning_rate": 5.816753926701572e-06, - "loss": 0.5544, + "epoch": 17.0, + "eval_loss": 0.5707325339317322, + "eval_runtime": 133.2653, + "eval_samples_per_second": 3.992, + "eval_steps_per_second": 0.998, + "eval_wer": 0.30282188082535194, + "step": 4896 + }, + { + "epoch": 17.01, + "learning_rate": 1.8130252100840337e-06, + "loss": 0.5172, "step": 4900 }, { - "epoch": 8.49, - "learning_rate": 5.80802792321117e-06, - "loss": 0.5956, + "epoch": 17.05, + "learning_rate": 1.7920168067226893e-06, + "loss": 0.4909, "step": 4910 }, { - "epoch": 8.51, - "learning_rate": 5.799301919720768e-06, - "loss": 0.5671, + "epoch": 17.08, + "learning_rate": 1.7710084033613448e-06, + "loss": 0.4509, "step": 4920 }, { - "epoch": 8.53, - "learning_rate": 5.790575916230366e-06, - "loss": 0.5987, + "epoch": 17.12, + "learning_rate": 1.75e-06, + "loss": 0.4508, "step": 4930 }, { - "epoch": 8.55, - "learning_rate": 5.7818499127399655e-06, - "loss": 0.6094, + "epoch": 17.15, + "learning_rate": 1.7289915966386555e-06, + "loss": 0.459, "step": 4940 }, { - "epoch": 8.56, - "learning_rate": 5.773123909249564e-06, - "loss": 0.553, + "epoch": 17.19, + "learning_rate": 1.707983193277311e-06, + "loss": 0.4737, "step": 4950 }, { - "epoch": 8.58, - "learning_rate": 5.764397905759162e-06, - "loss": 0.5838, + "epoch": 17.22, + "learning_rate": 1.6869747899159666e-06, + "loss": 0.4742, "step": 4960 }, { - "epoch": 8.6, - "learning_rate": 5.755671902268762e-06, - "loss": 0.6091, + "epoch": 17.26, + "learning_rate": 1.665966386554622e-06, + "loss": 0.4753, "step": 4970 }, { - "epoch": 8.62, - "learning_rate": 5.74694589877836e-06, - "loss": 0.6311, + "epoch": 17.29, + "learning_rate": 1.6449579831932775e-06, + "loss": 0.4688, "step": 4980 }, { - "epoch": 8.63, - "learning_rate": 5.738219895287958e-06, - "loss": 0.5699, + "epoch": 17.33, + "learning_rate": 1.6239495798319329e-06, + "loss": 0.4689, "step": 4990 }, { - "epoch": 8.65, - "learning_rate": 5.729493891797558e-06, - "loss": 0.5734, + "epoch": 17.36, + "learning_rate": 1.6029411764705884e-06, + "loss": 0.4724, "step": 5000 }, { - "epoch": 8.67, - "learning_rate": 5.720767888307156e-06, - "loss": 0.6414, + "epoch": 17.39, + "learning_rate": 1.5819327731092438e-06, + "loss": 0.4744, "step": 5010 }, { - "epoch": 8.69, - "learning_rate": 5.7120418848167545e-06, - "loss": 0.5658, + "epoch": 17.43, + "learning_rate": 1.5609243697478993e-06, + "loss": 0.4824, "step": 5020 }, { - "epoch": 8.7, - "learning_rate": 5.703315881326354e-06, - "loss": 0.6379, + "epoch": 17.46, + "learning_rate": 1.5399159663865549e-06, + "loss": 0.4699, "step": 5030 }, { - "epoch": 8.72, - "learning_rate": 5.694589877835952e-06, - "loss": 0.5763, + "epoch": 17.5, + "learning_rate": 1.51890756302521e-06, + "loss": 0.5166, "step": 5040 }, { - "epoch": 8.74, - "learning_rate": 5.68586387434555e-06, - "loss": 0.5111, + "epoch": 17.53, + "learning_rate": 1.4978991596638656e-06, + "loss": 0.4653, "step": 5050 }, { - "epoch": 8.75, - "learning_rate": 5.677137870855148e-06, - "loss": 0.6402, + "epoch": 17.57, + "learning_rate": 1.4768907563025211e-06, + "loss": 0.4809, "step": 5060 }, { - "epoch": 8.77, - "learning_rate": 5.668411867364747e-06, - "loss": 0.5852, + "epoch": 17.6, + "learning_rate": 1.4558823529411767e-06, + "loss": 0.4605, "step": 5070 }, { - "epoch": 8.79, - "learning_rate": 5.659685863874346e-06, - "loss": 0.5842, + "epoch": 17.64, + "learning_rate": 1.434873949579832e-06, + "loss": 0.4774, "step": 5080 }, { - "epoch": 8.81, - "learning_rate": 5.650959860383944e-06, - "loss": 0.5961, + "epoch": 17.67, + "learning_rate": 1.4159663865546218e-06, + "loss": 0.483, "step": 5090 }, { - "epoch": 8.82, - "learning_rate": 5.6422338568935435e-06, - "loss": 0.5954, + "epoch": 17.71, + "learning_rate": 1.3949579831932774e-06, + "loss": 0.5036, "step": 5100 }, { - "epoch": 8.84, - "learning_rate": 5.633507853403142e-06, - "loss": 0.5916, + "epoch": 17.74, + "learning_rate": 1.373949579831933e-06, + "loss": 0.466, "step": 5110 }, { - "epoch": 8.86, - "learning_rate": 5.62478184991274e-06, - "loss": 0.6174, + "epoch": 17.78, + "learning_rate": 1.3529411764705883e-06, + "loss": 0.4475, "step": 5120 }, { - "epoch": 8.88, - "learning_rate": 5.6160558464223396e-06, - "loss": 0.6009, + "epoch": 17.81, + "learning_rate": 1.3319327731092438e-06, + "loss": 0.4919, "step": 5130 }, { - "epoch": 8.89, - "learning_rate": 5.607329842931938e-06, - "loss": 0.5935, + "epoch": 17.85, + "learning_rate": 1.3109243697478994e-06, + "loss": 0.4884, "step": 5140 }, { - "epoch": 8.91, - "learning_rate": 5.598603839441536e-06, - "loss": 0.5522, + "epoch": 17.88, + "learning_rate": 1.289915966386555e-06, + "loss": 0.4847, "step": 5150 }, { - "epoch": 8.93, - "learning_rate": 5.589877835951135e-06, - "loss": 0.557, + "epoch": 17.91, + "learning_rate": 1.26890756302521e-06, + "loss": 0.4636, "step": 5160 }, { - "epoch": 8.94, - "learning_rate": 5.581151832460734e-06, - "loss": 0.5974, + "epoch": 17.95, + "learning_rate": 1.2478991596638656e-06, + "loss": 0.4921, "step": 5170 }, { - "epoch": 8.96, - "learning_rate": 5.5724258289703325e-06, - "loss": 0.5857, + "epoch": 17.98, + "learning_rate": 1.226890756302521e-06, + "loss": 0.4745, "step": 5180 }, { - "epoch": 8.98, - "learning_rate": 5.56369982547993e-06, - "loss": 0.5547, - "step": 5190 + "epoch": 18.0, + "eval_loss": 0.570929229259491, + "eval_runtime": 132.8501, + "eval_samples_per_second": 4.005, + "eval_steps_per_second": 1.001, + "eval_wer": 0.301921964389021, + "step": 5184 }, { - "epoch": 9.0, - "learning_rate": 5.5549738219895284e-06, - "loss": 0.5936, - "step": 5200 + "epoch": 18.02, + "learning_rate": 1.2058823529411765e-06, + "loss": 0.5153, + "step": 5190 }, { - "epoch": 9.0, - "eval_loss": 0.6200985312461853, - "eval_runtime": 130.5461, - "eval_samples_per_second": 4.075, - "eval_steps_per_second": 0.513, - "eval_wer": 0.36478755544128044, - "step": 5202 + "epoch": 18.06, + "learning_rate": 1.184873949579832e-06, + "loss": 0.4541, + "step": 5200 }, { - "epoch": 9.01, - "learning_rate": 5.546247818499128e-06, - "loss": 0.5623, + "epoch": 18.09, + "learning_rate": 1.1638655462184874e-06, + "loss": 0.474, "step": 5210 }, { - "epoch": 9.03, - "learning_rate": 5.537521815008726e-06, - "loss": 0.5816, + "epoch": 18.12, + "learning_rate": 1.142857142857143e-06, + "loss": 0.4608, "step": 5220 }, { - "epoch": 9.05, - "learning_rate": 5.5287958115183245e-06, - "loss": 0.5459, + "epoch": 18.16, + "learning_rate": 1.1218487394957985e-06, + "loss": 0.4932, "step": 5230 }, { - "epoch": 9.07, - "learning_rate": 5.520069808027924e-06, - "loss": 0.6094, + "epoch": 18.19, + "learning_rate": 1.1008403361344539e-06, + "loss": 0.4808, "step": 5240 }, { - "epoch": 9.08, - "learning_rate": 5.511343804537522e-06, - "loss": 0.5642, + "epoch": 18.23, + "learning_rate": 1.0798319327731094e-06, + "loss": 0.4657, "step": 5250 }, { - "epoch": 9.1, - "learning_rate": 5.502617801047121e-06, - "loss": 0.5674, + "epoch": 18.26, + "learning_rate": 1.0588235294117648e-06, + "loss": 0.4375, "step": 5260 }, { - "epoch": 9.12, - "learning_rate": 5.49389179755672e-06, - "loss": 0.5631, + "epoch": 18.3, + "learning_rate": 1.0378151260504203e-06, + "loss": 0.4985, "step": 5270 }, { - "epoch": 9.13, - "learning_rate": 5.485165794066318e-06, - "loss": 0.564, + "epoch": 18.33, + "learning_rate": 1.0168067226890757e-06, + "loss": 0.505, "step": 5280 }, { - "epoch": 9.15, - "learning_rate": 5.476439790575917e-06, - "loss": 0.5648, + "epoch": 18.37, + "learning_rate": 9.95798319327731e-07, + "loss": 0.4977, "step": 5290 }, { - "epoch": 9.17, - "learning_rate": 5.467713787085516e-06, - "loss": 0.5776, + "epoch": 18.4, + "learning_rate": 9.747899159663866e-07, + "loss": 0.4706, "step": 5300 }, { - "epoch": 9.19, - "learning_rate": 5.458987783595114e-06, - "loss": 0.5484, + "epoch": 18.44, + "learning_rate": 9.537815126050421e-07, + "loss": 0.4435, "step": 5310 }, { - "epoch": 9.2, - "learning_rate": 5.450261780104712e-06, - "loss": 0.6154, + "epoch": 18.47, + "learning_rate": 9.327731092436976e-07, + "loss": 0.4855, "step": 5320 }, { - "epoch": 9.22, - "learning_rate": 5.44153577661431e-06, - "loss": 0.5423, + "epoch": 18.51, + "learning_rate": 9.117647058823529e-07, + "loss": 0.4624, "step": 5330 }, { - "epoch": 9.24, - "learning_rate": 5.4328097731239096e-06, - "loss": 0.5731, + "epoch": 18.54, + "learning_rate": 8.907563025210085e-07, + "loss": 0.5086, "step": 5340 }, { - "epoch": 9.26, - "learning_rate": 5.424083769633508e-06, - "loss": 0.6023, + "epoch": 18.57, + "learning_rate": 8.697478991596639e-07, + "loss": 0.4876, "step": 5350 }, { - "epoch": 9.27, - "learning_rate": 5.415357766143106e-06, - "loss": 0.6208, + "epoch": 18.61, + "learning_rate": 8.487394957983195e-07, + "loss": 0.4862, "step": 5360 }, { - "epoch": 9.29, - "learning_rate": 5.406631762652706e-06, - "loss": 0.6127, + "epoch": 18.64, + "learning_rate": 8.277310924369748e-07, + "loss": 0.4466, "step": 5370 }, { - "epoch": 9.31, - "learning_rate": 5.397905759162304e-06, - "loss": 0.5352, + "epoch": 18.68, + "learning_rate": 8.067226890756304e-07, + "loss": 0.4713, "step": 5380 }, { - "epoch": 9.33, - "learning_rate": 5.3891797556719025e-06, - "loss": 0.587, + "epoch": 18.71, + "learning_rate": 7.857142857142857e-07, + "loss": 0.4689, "step": 5390 }, { - "epoch": 9.34, - "learning_rate": 5.380453752181502e-06, - "loss": 0.5836, + "epoch": 18.75, + "learning_rate": 7.647058823529413e-07, + "loss": 0.4741, "step": 5400 }, { - "epoch": 9.36, - "learning_rate": 5.3717277486911e-06, - "loss": 0.5529, + "epoch": 18.78, + "learning_rate": 7.436974789915967e-07, + "loss": 0.4395, "step": 5410 }, { - "epoch": 9.38, - "learning_rate": 5.3630017452006986e-06, - "loss": 0.562, + "epoch": 18.82, + "learning_rate": 7.226890756302522e-07, + "loss": 0.4878, "step": 5420 }, { - "epoch": 9.39, - "learning_rate": 5.354275741710298e-06, - "loss": 0.5406, + "epoch": 18.85, + "learning_rate": 7.016806722689076e-07, + "loss": 0.5032, "step": 5430 }, { - "epoch": 9.41, - "learning_rate": 5.345549738219896e-06, - "loss": 0.5648, + "epoch": 18.89, + "learning_rate": 6.806722689075631e-07, + "loss": 0.5066, "step": 5440 }, { - "epoch": 9.43, - "learning_rate": 5.336823734729494e-06, - "loss": 0.588, + "epoch": 18.92, + "learning_rate": 6.596638655462185e-07, + "loss": 0.4665, "step": 5450 }, { - "epoch": 9.45, - "learning_rate": 5.328097731239092e-06, - "loss": 0.6405, + "epoch": 18.96, + "learning_rate": 6.38655462184874e-07, + "loss": 0.4561, "step": 5460 }, { - "epoch": 9.46, - "learning_rate": 5.3193717277486915e-06, - "loss": 0.5604, + "epoch": 18.99, + "learning_rate": 6.176470588235295e-07, + "loss": 0.4548, "step": 5470 }, { - "epoch": 9.48, - "learning_rate": 5.31064572425829e-06, - "loss": 0.5593, + "epoch": 19.0, + "eval_loss": 0.571989119052887, + "eval_runtime": 132.9619, + "eval_samples_per_second": 4.001, + "eval_steps_per_second": 1.0, + "eval_wer": 0.30044353024362025, + "step": 5472 + }, + { + "epoch": 19.03, + "learning_rate": 5.96638655462185e-07, + "loss": 0.4904, "step": 5480 }, { - "epoch": 9.5, - "learning_rate": 5.301919720767888e-06, - "loss": 0.584, + "epoch": 19.06, + "learning_rate": 5.756302521008403e-07, + "loss": 0.4573, "step": 5490 }, { - "epoch": 9.52, - "learning_rate": 5.2931937172774875e-06, - "loss": 0.554, + "epoch": 19.1, + "learning_rate": 5.546218487394958e-07, + "loss": 0.4248, "step": 5500 }, { - "epoch": 9.53, - "learning_rate": 5.284467713787086e-06, - "loss": 0.6116, + "epoch": 19.13, + "learning_rate": 5.336134453781513e-07, + "loss": 0.4934, "step": 5510 }, { - "epoch": 9.55, - "learning_rate": 5.275741710296684e-06, - "loss": 0.5867, + "epoch": 19.17, + "learning_rate": 5.126050420168068e-07, + "loss": 0.4961, "step": 5520 }, { - "epoch": 9.57, - "learning_rate": 5.267015706806284e-06, - "loss": 0.5499, + "epoch": 19.2, + "learning_rate": 4.915966386554622e-07, + "loss": 0.4658, "step": 5530 }, { - "epoch": 9.58, - "learning_rate": 5.258289703315882e-06, - "loss": 0.6865, + "epoch": 19.24, + "learning_rate": 4.7058823529411767e-07, + "loss": 0.5053, "step": 5540 }, { - "epoch": 9.6, - "learning_rate": 5.2495636998254804e-06, - "loss": 0.6052, + "epoch": 19.27, + "learning_rate": 4.4957983193277317e-07, + "loss": 0.4687, "step": 5550 }, { - "epoch": 9.62, - "learning_rate": 5.24083769633508e-06, - "loss": 0.5486, + "epoch": 19.3, + "learning_rate": 4.285714285714286e-07, + "loss": 0.469, "step": 5560 }, { - "epoch": 9.64, - "learning_rate": 5.232111692844678e-06, - "loss": 0.5514, + "epoch": 19.34, + "learning_rate": 4.0756302521008407e-07, + "loss": 0.4819, "step": 5570 }, { - "epoch": 9.65, - "learning_rate": 5.2233856893542765e-06, - "loss": 0.5305, + "epoch": 19.37, + "learning_rate": 3.865546218487395e-07, + "loss": 0.4649, "step": 5580 }, { - "epoch": 9.67, - "learning_rate": 5.214659685863874e-06, - "loss": 0.5845, + "epoch": 19.41, + "learning_rate": 3.65546218487395e-07, + "loss": 0.4558, "step": 5590 }, { - "epoch": 9.69, - "learning_rate": 5.205933682373473e-06, - "loss": 0.6297, + "epoch": 19.44, + "learning_rate": 3.4453781512605047e-07, + "loss": 0.4492, "step": 5600 }, { - "epoch": 9.71, - "learning_rate": 5.197207678883072e-06, - "loss": 0.5702, + "epoch": 19.48, + "learning_rate": 3.235294117647059e-07, + "loss": 0.4556, "step": 5610 }, { - "epoch": 9.72, - "learning_rate": 5.18848167539267e-06, - "loss": 0.5413, + "epoch": 19.51, + "learning_rate": 3.0252100840336137e-07, + "loss": 0.4847, "step": 5620 }, { - "epoch": 9.74, - "learning_rate": 5.179755671902269e-06, - "loss": 0.5657, + "epoch": 19.55, + "learning_rate": 2.815126050420168e-07, + "loss": 0.4762, "step": 5630 }, { - "epoch": 9.76, - "learning_rate": 5.171029668411868e-06, - "loss": 0.5988, + "epoch": 19.58, + "learning_rate": 2.605042016806723e-07, + "loss": 0.4659, "step": 5640 }, { - "epoch": 9.78, - "learning_rate": 5.162303664921466e-06, - "loss": 0.5773, + "epoch": 19.62, + "learning_rate": 2.3949579831932777e-07, + "loss": 0.4651, "step": 5650 }, { - "epoch": 9.79, - "learning_rate": 5.1535776614310655e-06, - "loss": 0.5875, + "epoch": 19.65, + "learning_rate": 2.184873949579832e-07, + "loss": 0.4878, "step": 5660 }, { - "epoch": 9.81, - "learning_rate": 5.144851657940664e-06, - "loss": 0.5854, + "epoch": 19.69, + "learning_rate": 1.9747899159663866e-07, + "loss": 0.4809, "step": 5670 }, { - "epoch": 9.83, - "learning_rate": 5.136125654450262e-06, - "loss": 0.5237, + "epoch": 19.72, + "learning_rate": 1.7647058823529414e-07, + "loss": 0.4631, "step": 5680 }, { - "epoch": 9.84, - "learning_rate": 5.127399650959861e-06, - "loss": 0.6, + "epoch": 19.75, + "learning_rate": 1.554621848739496e-07, + "loss": 0.4607, "step": 5690 }, { - "epoch": 9.86, - "learning_rate": 5.11867364746946e-06, - "loss": 0.5789, + "epoch": 19.79, + "learning_rate": 1.3445378151260506e-07, + "loss": 0.4423, "step": 5700 }, { - "epoch": 9.88, - "learning_rate": 5.109947643979058e-06, - "loss": 0.5632, + "epoch": 19.82, + "learning_rate": 1.1344537815126051e-07, + "loss": 0.502, "step": 5710 }, { - "epoch": 9.9, - "learning_rate": 5.101221640488656e-06, - "loss": 0.6019, + "epoch": 19.86, + "learning_rate": 9.243697478991597e-08, + "loss": 0.5033, "step": 5720 }, { - "epoch": 9.91, - "learning_rate": 5.092495636998254e-06, - "loss": 0.5425, + "epoch": 19.89, + "learning_rate": 7.142857142857144e-08, + "loss": 0.4968, "step": 5730 }, { - "epoch": 9.93, - "learning_rate": 5.083769633507854e-06, - "loss": 0.5459, + "epoch": 19.93, + "learning_rate": 5.04201680672269e-08, + "loss": 0.4721, "step": 5740 }, { - "epoch": 9.95, - "learning_rate": 5.075043630017452e-06, - "loss": 0.5626, + "epoch": 19.96, + "learning_rate": 2.9411764705882354e-08, + "loss": 0.4608, "step": 5750 }, { - "epoch": 9.97, - "learning_rate": 5.0663176265270505e-06, - "loss": 0.5368, + "epoch": 20.0, + "learning_rate": 8.403361344537816e-09, + "loss": 0.4619, "step": 5760 }, - { - "epoch": 9.98, - "learning_rate": 5.05759162303665e-06, - "loss": 0.5416, - "step": 5770 - }, - { - "epoch": 10.0, - "learning_rate": 5.048865619546248e-06, - "loss": 0.5737, - "step": 5780 - }, - { - "epoch": 10.0, - "eval_loss": 0.628906786441803, - "eval_runtime": 131.2825, - "eval_samples_per_second": 4.052, - "eval_steps_per_second": 0.51, - "eval_wer": 0.35122452915086455, - "step": 5780 - }, - { - "epoch": 10.02, - "learning_rate": 5.0401396160558465e-06, - "loss": 0.5558, - "step": 5790 - }, - { - "epoch": 10.03, - "learning_rate": 5.031413612565446e-06, - "loss": 0.5318, - "step": 5800 - }, - { - "epoch": 10.05, - "learning_rate": 5.022687609075044e-06, - "loss": 0.565, - "step": 5810 - }, - { - "epoch": 10.07, - "learning_rate": 5.013961605584643e-06, - "loss": 0.5785, - "step": 5820 - }, - { - "epoch": 10.09, - "learning_rate": 5.005235602094242e-06, - "loss": 0.5715, - "step": 5830 - }, - { - "epoch": 10.1, - "learning_rate": 4.9965095986038394e-06, - "loss": 0.5372, - "step": 5840 - }, - { - "epoch": 10.12, - "learning_rate": 4.987783595113439e-06, - "loss": 0.5486, - "step": 5850 - }, - { - "epoch": 10.14, - "learning_rate": 4.979057591623037e-06, - "loss": 0.5391, - "step": 5860 - }, - { - "epoch": 10.16, - "learning_rate": 4.9703315881326355e-06, - "loss": 0.5498, - "step": 5870 - }, - { - "epoch": 10.17, - "learning_rate": 4.961605584642235e-06, - "loss": 0.6428, - "step": 5880 - }, - { - "epoch": 10.19, - "learning_rate": 4.952879581151832e-06, - "loss": 0.5417, - "step": 5890 - }, - { - "epoch": 10.21, - "learning_rate": 4.944153577661432e-06, - "loss": 0.568, - "step": 5900 - }, - { - "epoch": 10.22, - "learning_rate": 4.93542757417103e-06, - "loss": 0.5522, - "step": 5910 - }, - { - "epoch": 10.24, - "learning_rate": 4.926701570680628e-06, - "loss": 0.5315, - "step": 5920 - }, - { - "epoch": 10.26, - "learning_rate": 4.917975567190228e-06, - "loss": 0.6103, - "step": 5930 - }, - { - "epoch": 10.28, - "learning_rate": 4.909249563699826e-06, - "loss": 0.5107, - "step": 5940 - }, - { - "epoch": 10.29, - "learning_rate": 4.9005235602094245e-06, - "loss": 0.5268, - "step": 5950 - }, - { - "epoch": 10.31, - "learning_rate": 4.891797556719023e-06, - "loss": 0.5426, - "step": 5960 - }, - { - "epoch": 10.33, - "learning_rate": 4.883071553228621e-06, - "loss": 0.5607, - "step": 5970 - }, - { - "epoch": 10.35, - "learning_rate": 4.8743455497382206e-06, - "loss": 0.5723, - "step": 5980 - }, - { - "epoch": 10.36, - "learning_rate": 4.865619546247819e-06, - "loss": 0.608, - "step": 5990 - }, - { - "epoch": 10.38, - "learning_rate": 4.856893542757417e-06, - "loss": 0.5573, - "step": 6000 - }, - { - "epoch": 10.4, - "learning_rate": 4.848167539267017e-06, - "loss": 0.5354, - "step": 6010 - }, - { - "epoch": 10.42, - "learning_rate": 4.839441535776614e-06, - "loss": 0.5828, - "step": 6020 - }, - { - "epoch": 10.43, - "learning_rate": 4.8307155322862135e-06, - "loss": 0.6045, - "step": 6030 - }, - { - "epoch": 10.45, - "learning_rate": 4.821989528795812e-06, - "loss": 0.6047, - "step": 6040 - }, - { - "epoch": 10.47, - "learning_rate": 4.81326352530541e-06, - "loss": 0.5948, - "step": 6050 - }, - { - "epoch": 10.48, - "learning_rate": 4.8045375218150096e-06, - "loss": 0.5492, - "step": 6060 - }, - { - "epoch": 10.5, - "learning_rate": 4.795811518324608e-06, - "loss": 0.5865, - "step": 6070 - }, - { - "epoch": 10.52, - "learning_rate": 4.787085514834206e-06, - "loss": 0.6774, - "step": 6080 - }, - { - "epoch": 10.54, - "learning_rate": 4.778359511343805e-06, - "loss": 0.5863, - "step": 6090 - }, - { - "epoch": 10.55, - "learning_rate": 4.769633507853403e-06, - "loss": 0.5393, - "step": 6100 - }, - { - "epoch": 10.57, - "learning_rate": 4.7609075043630025e-06, - "loss": 0.5126, - "step": 6110 - }, - { - "epoch": 10.59, - "learning_rate": 4.752181500872601e-06, - "loss": 0.6073, - "step": 6120 - }, - { - "epoch": 10.61, - "learning_rate": 4.743455497382199e-06, - "loss": 0.5568, - "step": 6130 - }, - { - "epoch": 10.62, - "learning_rate": 4.7347294938917985e-06, - "loss": 0.6342, - "step": 6140 - }, - { - "epoch": 10.64, - "learning_rate": 4.726003490401396e-06, - "loss": 0.599, - "step": 6150 - }, - { - "epoch": 10.66, - "learning_rate": 4.717277486910995e-06, - "loss": 0.5452, - "step": 6160 - }, - { - "epoch": 10.67, - "learning_rate": 4.708551483420594e-06, - "loss": 0.5403, - "step": 6170 - }, - { - "epoch": 10.69, - "learning_rate": 4.699825479930192e-06, - "loss": 0.6477, - "step": 6180 - }, - { - "epoch": 10.71, - "learning_rate": 4.6910994764397914e-06, - "loss": 0.5814, - "step": 6190 - }, - { - "epoch": 10.73, - "learning_rate": 4.68237347294939e-06, - "loss": 0.5788, - "step": 6200 - }, - { - "epoch": 10.74, - "learning_rate": 4.673647469458988e-06, - "loss": 0.5816, - "step": 6210 - }, - { - "epoch": 10.76, - "learning_rate": 4.664921465968587e-06, - "loss": 0.5791, - "step": 6220 - }, - { - "epoch": 10.78, - "learning_rate": 4.656195462478185e-06, - "loss": 0.5815, - "step": 6230 - }, - { - "epoch": 10.8, - "learning_rate": 4.6474694589877835e-06, - "loss": 0.5384, - "step": 6240 - }, - { - "epoch": 10.81, - "learning_rate": 4.638743455497383e-06, - "loss": 0.5579, - "step": 6250 - }, - { - "epoch": 10.83, - "learning_rate": 4.630017452006981e-06, - "loss": 0.5562, - "step": 6260 - }, - { - "epoch": 10.85, - "learning_rate": 4.6212914485165796e-06, - "loss": 0.5668, - "step": 6270 - }, - { - "epoch": 10.87, - "learning_rate": 4.612565445026179e-06, - "loss": 0.5383, - "step": 6280 - }, - { - "epoch": 10.88, - "learning_rate": 4.603839441535776e-06, - "loss": 0.5839, - "step": 6290 - }, - { - "epoch": 10.9, - "learning_rate": 4.595113438045376e-06, - "loss": 0.5591, - "step": 6300 - }, - { - "epoch": 10.92, - "learning_rate": 4.586387434554974e-06, - "loss": 0.4957, - "step": 6310 - }, - { - "epoch": 10.93, - "learning_rate": 4.5776614310645725e-06, - "loss": 0.5707, - "step": 6320 - }, - { - "epoch": 10.95, - "learning_rate": 4.568935427574172e-06, - "loss": 0.6379, - "step": 6330 - }, - { - "epoch": 10.97, - "learning_rate": 4.56020942408377e-06, - "loss": 0.5519, - "step": 6340 - }, - { - "epoch": 10.99, - "learning_rate": 4.5514834205933686e-06, - "loss": 0.5711, - "step": 6350 - }, - { - "epoch": 11.0, - "eval_loss": 0.6287660002708435, - "eval_runtime": 130.3298, - "eval_samples_per_second": 4.082, - "eval_steps_per_second": 0.514, - "eval_wer": 0.35450279616892716, - "step": 6358 - }, - { - "epoch": 11.0, - "learning_rate": 4.542757417102967e-06, - "loss": 0.6257, - "step": 6360 - }, - { - "epoch": 11.02, - "learning_rate": 4.534031413612565e-06, - "loss": 0.5251, - "step": 6370 - }, - { - "epoch": 11.04, - "learning_rate": 4.525305410122165e-06, - "loss": 0.5406, - "step": 6380 - }, - { - "epoch": 11.06, - "learning_rate": 4.516579406631763e-06, - "loss": 0.5524, - "step": 6390 - }, - { - "epoch": 11.07, - "learning_rate": 4.5078534031413615e-06, - "loss": 0.543, - "step": 6400 - }, - { - "epoch": 11.09, - "learning_rate": 4.499127399650961e-06, - "loss": 0.6248, - "step": 6410 - }, - { - "epoch": 11.11, - "learning_rate": 4.490401396160558e-06, - "loss": 0.5466, - "step": 6420 - }, - { - "epoch": 11.12, - "learning_rate": 4.4816753926701575e-06, - "loss": 0.5765, - "step": 6430 - }, - { - "epoch": 11.14, - "learning_rate": 4.472949389179756e-06, - "loss": 0.5593, - "step": 6440 - }, - { - "epoch": 11.16, - "learning_rate": 4.464223385689354e-06, - "loss": 0.5573, - "step": 6450 - }, - { - "epoch": 11.18, - "learning_rate": 4.455497382198954e-06, - "loss": 0.6085, - "step": 6460 - }, - { - "epoch": 11.19, - "learning_rate": 4.446771378708552e-06, - "loss": 0.6057, - "step": 6470 - }, - { - "epoch": 11.21, - "learning_rate": 4.4380453752181504e-06, - "loss": 0.5567, - "step": 6480 - }, - { - "epoch": 11.23, - "learning_rate": 4.429319371727749e-06, - "loss": 0.503, - "step": 6490 - }, - { - "epoch": 11.25, - "learning_rate": 4.420593368237347e-06, - "loss": 0.5559, - "step": 6500 - }, - { - "epoch": 11.26, - "learning_rate": 4.4118673647469465e-06, - "loss": 0.5608, - "step": 6510 - }, - { - "epoch": 11.28, - "learning_rate": 4.403141361256545e-06, - "loss": 0.5464, - "step": 6520 - }, - { - "epoch": 11.3, - "learning_rate": 4.394415357766143e-06, - "loss": 0.5575, - "step": 6530 - }, - { - "epoch": 11.31, - "learning_rate": 4.385689354275743e-06, - "loss": 0.5326, - "step": 6540 - }, - { - "epoch": 11.33, - "learning_rate": 4.37696335078534e-06, - "loss": 0.588, - "step": 6550 - }, - { - "epoch": 11.35, - "learning_rate": 4.368237347294939e-06, - "loss": 0.6382, - "step": 6560 - }, - { - "epoch": 11.37, - "learning_rate": 4.359511343804538e-06, - "loss": 0.5361, - "step": 6570 - }, - { - "epoch": 11.38, - "learning_rate": 4.350785340314136e-06, - "loss": 0.5631, - "step": 6580 - }, - { - "epoch": 11.4, - "learning_rate": 4.3420593368237355e-06, - "loss": 0.5738, - "step": 6590 - }, - { - "epoch": 11.42, - "learning_rate": 4.333333333333334e-06, - "loss": 0.539, - "step": 6600 - }, - { - "epoch": 11.44, - "learning_rate": 4.324607329842932e-06, - "loss": 0.5762, - "step": 6610 - }, - { - "epoch": 11.45, - "learning_rate": 4.315881326352531e-06, - "loss": 0.5492, - "step": 6620 - }, - { - "epoch": 11.47, - "learning_rate": 4.307155322862129e-06, - "loss": 0.5452, - "step": 6630 - }, - { - "epoch": 11.49, - "learning_rate": 4.298429319371728e-06, - "loss": 0.5705, - "step": 6640 - }, - { - "epoch": 11.51, - "learning_rate": 4.289703315881327e-06, - "loss": 0.5864, - "step": 6650 - }, - { - "epoch": 11.52, - "learning_rate": 4.280977312390925e-06, - "loss": 0.6123, - "step": 6660 - }, - { - "epoch": 11.54, - "learning_rate": 4.2722513089005245e-06, - "loss": 0.5317, - "step": 6670 - }, - { - "epoch": 11.56, - "learning_rate": 4.263525305410123e-06, - "loss": 0.539, - "step": 6680 - }, - { - "epoch": 11.57, - "learning_rate": 4.254799301919721e-06, - "loss": 0.5819, - "step": 6690 - }, - { - "epoch": 11.59, - "learning_rate": 4.24607329842932e-06, - "loss": 0.6156, - "step": 6700 - }, - { - "epoch": 11.61, - "learning_rate": 4.237347294938918e-06, - "loss": 0.5322, - "step": 6710 - }, - { - "epoch": 11.63, - "learning_rate": 4.228621291448517e-06, - "loss": 0.5929, - "step": 6720 - }, - { - "epoch": 11.64, - "learning_rate": 4.219895287958116e-06, - "loss": 0.5201, - "step": 6730 - }, - { - "epoch": 11.66, - "learning_rate": 4.211169284467714e-06, - "loss": 0.5173, - "step": 6740 - }, - { - "epoch": 11.68, - "learning_rate": 4.202443280977313e-06, - "loss": 0.5418, - "step": 6750 - }, - { - "epoch": 11.7, - "learning_rate": 4.193717277486911e-06, - "loss": 0.5923, - "step": 6760 - }, - { - "epoch": 11.71, - "learning_rate": 4.1849912739965094e-06, - "loss": 0.5684, - "step": 6770 - }, - { - "epoch": 11.73, - "learning_rate": 4.176265270506109e-06, - "loss": 0.5318, - "step": 6780 - }, - { - "epoch": 11.75, - "learning_rate": 4.167539267015707e-06, - "loss": 0.5565, - "step": 6790 - }, - { - "epoch": 11.76, - "learning_rate": 4.1588132635253055e-06, - "loss": 0.5476, - "step": 6800 - }, - { - "epoch": 11.78, - "learning_rate": 4.150087260034905e-06, - "loss": 0.5456, - "step": 6810 - }, - { - "epoch": 11.8, - "learning_rate": 4.141361256544502e-06, - "loss": 0.5741, - "step": 6820 - }, - { - "epoch": 11.82, - "learning_rate": 4.132635253054102e-06, - "loss": 0.5374, - "step": 6830 - }, - { - "epoch": 11.83, - "learning_rate": 4.1239092495637e-06, - "loss": 0.5784, - "step": 6840 - }, - { - "epoch": 11.85, - "learning_rate": 4.115183246073298e-06, - "loss": 0.5694, - "step": 6850 - }, - { - "epoch": 11.87, - "learning_rate": 4.106457242582898e-06, - "loss": 0.5448, - "step": 6860 - }, - { - "epoch": 11.89, - "learning_rate": 4.097731239092496e-06, - "loss": 0.5357, - "step": 6870 - }, - { - "epoch": 11.9, - "learning_rate": 4.0890052356020945e-06, - "loss": 0.5875, - "step": 6880 - }, - { - "epoch": 11.92, - "learning_rate": 4.080279232111693e-06, - "loss": 0.5448, - "step": 6890 - }, - { - "epoch": 11.94, - "learning_rate": 4.071553228621291e-06, - "loss": 0.5345, - "step": 6900 - }, - { - "epoch": 11.96, - "learning_rate": 4.062827225130891e-06, - "loss": 0.5928, - "step": 6910 - }, - { - "epoch": 11.97, - "learning_rate": 4.054101221640489e-06, - "loss": 0.5589, - "step": 6920 - }, - { - "epoch": 11.99, - "learning_rate": 4.045375218150087e-06, - "loss": 0.5617, - "step": 6930 - }, - { - "epoch": 12.0, - "eval_loss": 0.6307947635650635, - "eval_runtime": 131.8534, - "eval_samples_per_second": 4.035, - "eval_steps_per_second": 0.508, - "eval_wer": 0.366201709841229, - "step": 6936 - }, - { - "epoch": 12.01, - "learning_rate": 4.036649214659687e-06, - "loss": 0.6541, - "step": 6940 - }, - { - "epoch": 12.02, - "learning_rate": 4.027923211169284e-06, - "loss": 0.5362, - "step": 6950 - }, - { - "epoch": 12.04, - "learning_rate": 4.0191972076788835e-06, - "loss": 0.5366, - "step": 6960 - }, - { - "epoch": 12.06, - "learning_rate": 4.010471204188482e-06, - "loss": 0.5454, - "step": 6970 - }, - { - "epoch": 12.08, - "learning_rate": 4.00174520069808e-06, - "loss": 0.526, - "step": 6980 - }, - { - "epoch": 12.09, - "learning_rate": 3.9930191972076796e-06, - "loss": 0.5745, - "step": 6990 - }, - { - "epoch": 12.11, - "learning_rate": 3.984293193717278e-06, - "loss": 0.5303, - "step": 7000 - }, - { - "epoch": 12.13, - "learning_rate": 3.975567190226876e-06, - "loss": 0.5295, - "step": 7010 - }, - { - "epoch": 12.15, - "learning_rate": 3.966841186736475e-06, - "loss": 0.4992, - "step": 7020 - }, - { - "epoch": 12.16, - "learning_rate": 3.958115183246073e-06, - "loss": 0.563, - "step": 7030 - }, - { - "epoch": 12.18, - "learning_rate": 3.9493891797556725e-06, - "loss": 0.5468, - "step": 7040 - }, - { - "epoch": 12.2, - "learning_rate": 3.940663176265271e-06, - "loss": 0.5138, - "step": 7050 - }, - { - "epoch": 12.21, - "learning_rate": 3.931937172774869e-06, - "loss": 0.5332, - "step": 7060 - }, - { - "epoch": 12.23, - "learning_rate": 3.9232111692844685e-06, - "loss": 0.5498, - "step": 7070 - }, - { - "epoch": 12.25, - "learning_rate": 3.914485165794067e-06, - "loss": 0.5773, - "step": 7080 - }, - { - "epoch": 12.27, - "learning_rate": 3.905759162303665e-06, - "loss": 0.5773, - "step": 7090 - }, - { - "epoch": 12.28, - "learning_rate": 3.897033158813264e-06, - "loss": 0.6018, - "step": 7100 - }, - { - "epoch": 12.3, - "learning_rate": 3.888307155322862e-06, - "loss": 0.5239, - "step": 7110 - }, - { - "epoch": 12.32, - "learning_rate": 3.8795811518324614e-06, - "loss": 0.5497, - "step": 7120 - }, - { - "epoch": 12.34, - "learning_rate": 3.87085514834206e-06, - "loss": 0.5496, - "step": 7130 - }, - { - "epoch": 12.35, - "learning_rate": 3.862129144851658e-06, - "loss": 0.5697, - "step": 7140 - }, - { - "epoch": 12.37, - "learning_rate": 3.853403141361257e-06, - "loss": 0.5426, - "step": 7150 - }, - { - "epoch": 12.39, - "learning_rate": 3.844677137870855e-06, - "loss": 0.5383, - "step": 7160 - }, - { - "epoch": 12.4, - "learning_rate": 3.835951134380454e-06, - "loss": 0.5318, - "step": 7170 - }, - { - "epoch": 12.42, - "learning_rate": 3.827225130890053e-06, - "loss": 0.5686, - "step": 7180 - }, - { - "epoch": 12.44, - "learning_rate": 3.818499127399651e-06, - "loss": 0.6092, - "step": 7190 - }, - { - "epoch": 12.46, - "learning_rate": 3.80977312390925e-06, - "loss": 0.5413, - "step": 7200 - }, - { - "epoch": 12.47, - "learning_rate": 3.801047120418849e-06, - "loss": 0.5403, - "step": 7210 - }, - { - "epoch": 12.49, - "learning_rate": 3.792321116928447e-06, - "loss": 0.5302, - "step": 7220 - }, - { - "epoch": 12.51, - "learning_rate": 3.7835951134380457e-06, - "loss": 0.567, - "step": 7230 - }, - { - "epoch": 12.53, - "learning_rate": 3.774869109947644e-06, - "loss": 0.5477, - "step": 7240 - }, - { - "epoch": 12.54, - "learning_rate": 3.766143106457243e-06, - "loss": 0.5329, - "step": 7250 - }, - { - "epoch": 12.56, - "learning_rate": 3.7574171029668417e-06, - "loss": 0.501, - "step": 7260 - }, - { - "epoch": 12.58, - "learning_rate": 3.74869109947644e-06, - "loss": 0.5299, - "step": 7270 - }, - { - "epoch": 12.6, - "learning_rate": 3.739965095986039e-06, - "loss": 0.5564, - "step": 7280 - }, - { - "epoch": 12.61, - "learning_rate": 3.731239092495637e-06, - "loss": 0.5387, - "step": 7290 - }, - { - "epoch": 12.63, - "learning_rate": 3.722513089005236e-06, - "loss": 0.5107, - "step": 7300 - }, - { - "epoch": 12.65, - "learning_rate": 3.7137870855148346e-06, - "loss": 0.5493, - "step": 7310 - }, - { - "epoch": 12.66, - "learning_rate": 3.705061082024433e-06, - "loss": 0.5872, - "step": 7320 - }, - { - "epoch": 12.68, - "learning_rate": 3.696335078534032e-06, - "loss": 0.567, - "step": 7330 - }, - { - "epoch": 12.7, - "learning_rate": 3.6876090750436307e-06, - "loss": 0.5332, - "step": 7340 - }, - { - "epoch": 12.72, - "learning_rate": 3.6788830715532287e-06, - "loss": 0.5768, - "step": 7350 - }, - { - "epoch": 12.73, - "learning_rate": 3.6701570680628275e-06, - "loss": 0.5541, - "step": 7360 - }, - { - "epoch": 12.75, - "learning_rate": 3.661431064572426e-06, - "loss": 0.5357, - "step": 7370 - }, - { - "epoch": 12.77, - "learning_rate": 3.652705061082025e-06, - "loss": 0.5611, - "step": 7380 - }, - { - "epoch": 12.79, - "learning_rate": 3.643979057591623e-06, - "loss": 0.5678, - "step": 7390 - }, - { - "epoch": 12.8, - "learning_rate": 3.635253054101222e-06, - "loss": 0.5873, - "step": 7400 - }, - { - "epoch": 12.82, - "learning_rate": 3.626527050610821e-06, - "loss": 0.5154, - "step": 7410 - }, - { - "epoch": 12.84, - "learning_rate": 3.617801047120419e-06, - "loss": 0.5613, - "step": 7420 - }, - { - "epoch": 12.85, - "learning_rate": 3.6090750436300177e-06, - "loss": 0.5612, - "step": 7430 - }, - { - "epoch": 12.87, - "learning_rate": 3.600349040139616e-06, - "loss": 0.5748, - "step": 7440 - }, - { - "epoch": 12.89, - "learning_rate": 3.591623036649215e-06, - "loss": 0.5558, - "step": 7450 - }, - { - "epoch": 12.91, - "learning_rate": 3.5828970331588138e-06, - "loss": 0.5078, - "step": 7460 - }, - { - "epoch": 12.92, - "learning_rate": 3.574171029668412e-06, - "loss": 0.5341, - "step": 7470 - }, - { - "epoch": 12.94, - "learning_rate": 3.565445026178011e-06, - "loss": 0.5523, - "step": 7480 - }, - { - "epoch": 12.96, - "learning_rate": 3.556719022687609e-06, - "loss": 0.6565, - "step": 7490 - }, - { - "epoch": 12.98, - "learning_rate": 3.547993019197208e-06, - "loss": 0.5453, - "step": 7500 - }, - { - "epoch": 12.99, - "learning_rate": 3.5392670157068067e-06, - "loss": 0.5796, - "step": 7510 - }, - { - "epoch": 13.0, - "eval_loss": 0.6088328957557678, - "eval_runtime": 130.8634, - "eval_samples_per_second": 4.065, - "eval_steps_per_second": 0.512, - "eval_wer": 0.35456707591437936, - "step": 7514 - }, - { - "epoch": 13.01, - "learning_rate": 3.530541012216405e-06, - "loss": 0.5424, - "step": 7520 - }, - { - "epoch": 13.03, - "learning_rate": 3.521815008726004e-06, - "loss": 0.5603, - "step": 7530 - }, - { - "epoch": 13.04, - "learning_rate": 3.5130890052356028e-06, - "loss": 0.5056, - "step": 7540 - }, - { - "epoch": 13.06, - "learning_rate": 3.5043630017452007e-06, - "loss": 0.5545, - "step": 7550 - }, - { - "epoch": 13.08, - "learning_rate": 3.4956369982547996e-06, - "loss": 0.5405, - "step": 7560 - }, - { - "epoch": 13.1, - "learning_rate": 3.486910994764398e-06, - "loss": 0.5882, - "step": 7570 - }, - { - "epoch": 13.11, - "learning_rate": 3.478184991273997e-06, - "loss": 0.5724, - "step": 7580 - }, - { - "epoch": 13.13, - "learning_rate": 3.4694589877835957e-06, - "loss": 0.4917, - "step": 7590 - }, - { - "epoch": 13.15, - "learning_rate": 3.460732984293194e-06, - "loss": 0.5158, - "step": 7600 - }, - { - "epoch": 13.17, - "learning_rate": 3.452006980802793e-06, - "loss": 0.5663, - "step": 7610 - }, - { - "epoch": 13.18, - "learning_rate": 3.443280977312391e-06, - "loss": 0.548, - "step": 7620 - }, - { - "epoch": 13.2, - "learning_rate": 3.4345549738219897e-06, - "loss": 0.5283, - "step": 7630 - }, - { - "epoch": 13.22, - "learning_rate": 3.425828970331588e-06, - "loss": 0.5352, - "step": 7640 - }, - { - "epoch": 13.24, - "learning_rate": 3.417102966841187e-06, - "loss": 0.5521, - "step": 7650 - }, - { - "epoch": 13.25, - "learning_rate": 3.408376963350786e-06, - "loss": 0.5661, - "step": 7660 - }, - { - "epoch": 13.27, - "learning_rate": 3.3996509598603842e-06, - "loss": 0.558, - "step": 7670 - }, - { - "epoch": 13.29, - "learning_rate": 3.390924956369983e-06, - "loss": 0.5781, - "step": 7680 - }, - { - "epoch": 13.3, - "learning_rate": 3.382198952879581e-06, - "loss": 0.54, - "step": 7690 - }, - { - "epoch": 13.32, - "learning_rate": 3.37347294938918e-06, - "loss": 0.5581, - "step": 7700 - }, - { - "epoch": 13.34, - "learning_rate": 3.3647469458987787e-06, - "loss": 0.5102, - "step": 7710 - }, - { - "epoch": 13.36, - "learning_rate": 3.356020942408377e-06, - "loss": 0.5571, - "step": 7720 - }, - { - "epoch": 13.37, - "learning_rate": 3.347294938917976e-06, - "loss": 0.5396, - "step": 7730 - }, - { - "epoch": 13.39, - "learning_rate": 3.3385689354275748e-06, - "loss": 0.5201, - "step": 7740 - }, - { - "epoch": 13.41, - "learning_rate": 3.3298429319371728e-06, - "loss": 0.5083, - "step": 7750 - }, - { - "epoch": 13.43, - "learning_rate": 3.3211169284467716e-06, - "loss": 0.5673, - "step": 7760 - }, - { - "epoch": 13.44, - "learning_rate": 3.31239092495637e-06, - "loss": 0.5473, - "step": 7770 - }, - { - "epoch": 13.46, - "learning_rate": 3.303664921465969e-06, - "loss": 0.5995, - "step": 7780 - }, - { - "epoch": 13.48, - "learning_rate": 3.2949389179755677e-06, - "loss": 0.4521, - "step": 7790 - }, - { - "epoch": 13.49, - "learning_rate": 3.286212914485166e-06, - "loss": 0.544, - "step": 7800 - }, - { - "epoch": 13.51, - "learning_rate": 3.277486910994765e-06, - "loss": 0.652, - "step": 7810 - }, - { - "epoch": 13.53, - "learning_rate": 3.268760907504363e-06, - "loss": 0.5719, - "step": 7820 - }, - { - "epoch": 13.55, - "learning_rate": 3.2600349040139618e-06, - "loss": 0.534, - "step": 7830 - }, - { - "epoch": 13.56, - "learning_rate": 3.2513089005235606e-06, - "loss": 0.5381, - "step": 7840 - }, - { - "epoch": 13.58, - "learning_rate": 3.242582897033159e-06, - "loss": 0.5832, - "step": 7850 - }, - { - "epoch": 13.6, - "learning_rate": 3.233856893542758e-06, - "loss": 0.5826, - "step": 7860 - }, - { - "epoch": 13.62, - "learning_rate": 3.2251308900523567e-06, - "loss": 0.5127, - "step": 7870 - }, - { - "epoch": 13.63, - "learning_rate": 3.216404886561955e-06, - "loss": 0.5002, - "step": 7880 - }, - { - "epoch": 13.65, - "learning_rate": 3.2076788830715535e-06, - "loss": 0.5427, - "step": 7890 - }, - { - "epoch": 13.67, - "learning_rate": 3.198952879581152e-06, - "loss": 0.5291, - "step": 7900 - }, - { - "epoch": 13.69, - "learning_rate": 3.1902268760907507e-06, - "loss": 0.5359, - "step": 7910 - }, - { - "epoch": 13.7, - "learning_rate": 3.181500872600349e-06, - "loss": 0.5195, - "step": 7920 - }, - { - "epoch": 13.72, - "learning_rate": 3.172774869109948e-06, - "loss": 0.5302, - "step": 7930 - }, - { - "epoch": 13.74, - "learning_rate": 3.164048865619547e-06, - "loss": 0.5541, - "step": 7940 - }, - { - "epoch": 13.75, - "learning_rate": 3.155322862129145e-06, - "loss": 0.528, - "step": 7950 - }, - { - "epoch": 13.77, - "learning_rate": 3.1465968586387436e-06, - "loss": 0.5817, - "step": 7960 - }, - { - "epoch": 13.79, - "learning_rate": 3.137870855148342e-06, - "loss": 0.5765, - "step": 7970 - }, - { - "epoch": 13.81, - "learning_rate": 3.129144851657941e-06, - "loss": 0.5618, - "step": 7980 - }, - { - "epoch": 13.82, - "learning_rate": 3.1204188481675397e-06, - "loss": 0.5077, - "step": 7990 - }, - { - "epoch": 13.84, - "learning_rate": 3.111692844677138e-06, - "loss": 0.564, - "step": 8000 - }, - { - "epoch": 13.86, - "learning_rate": 3.102966841186737e-06, - "loss": 0.5054, - "step": 8010 - }, - { - "epoch": 13.88, - "learning_rate": 3.094240837696335e-06, - "loss": 0.5833, - "step": 8020 - }, - { - "epoch": 13.89, - "learning_rate": 3.0855148342059338e-06, - "loss": 0.5431, - "step": 8030 - }, - { - "epoch": 13.91, - "learning_rate": 3.0767888307155326e-06, - "loss": 0.5378, - "step": 8040 - }, - { - "epoch": 13.93, - "learning_rate": 3.068062827225131e-06, - "loss": 0.5032, - "step": 8050 - }, - { - "epoch": 13.94, - "learning_rate": 3.05933682373473e-06, - "loss": 0.5718, - "step": 8060 - }, - { - "epoch": 13.96, - "learning_rate": 3.0506108202443287e-06, - "loss": 0.5637, - "step": 8070 - }, - { - "epoch": 13.98, - "learning_rate": 3.041884816753927e-06, - "loss": 0.5446, - "step": 8080 - }, - { - "epoch": 14.0, - "learning_rate": 3.0331588132635255e-06, - "loss": 0.5401, - "step": 8090 - }, - { - "epoch": 14.0, - "eval_loss": 0.6190069913864136, - "eval_runtime": 130.6585, - "eval_samples_per_second": 4.072, - "eval_steps_per_second": 0.513, - "eval_wer": 0.34878189882368066, - "step": 8092 - }, - { - "epoch": 14.01, - "learning_rate": 3.024432809773124e-06, - "loss": 0.5353, - "step": 8100 - }, - { - "epoch": 14.03, - "learning_rate": 3.0157068062827228e-06, - "loss": 0.5543, - "step": 8110 - }, - { - "epoch": 14.05, - "learning_rate": 3.0069808027923216e-06, - "loss": 0.5276, - "step": 8120 - }, - { - "epoch": 14.07, - "learning_rate": 2.99825479930192e-06, - "loss": 0.5447, - "step": 8130 - }, - { - "epoch": 14.08, - "learning_rate": 2.989528795811519e-06, - "loss": 0.5545, - "step": 8140 - }, - { - "epoch": 14.1, - "learning_rate": 2.980802792321117e-06, - "loss": 0.535, - "step": 8150 - }, - { - "epoch": 14.12, - "learning_rate": 2.9720767888307157e-06, - "loss": 0.5551, - "step": 8160 - }, - { - "epoch": 14.13, - "learning_rate": 2.963350785340314e-06, - "loss": 0.5346, - "step": 8170 - }, - { - "epoch": 14.15, - "learning_rate": 2.954624781849913e-06, - "loss": 0.5122, - "step": 8180 - }, - { - "epoch": 14.17, - "learning_rate": 2.9458987783595117e-06, - "loss": 0.5379, - "step": 8190 - }, - { - "epoch": 14.19, - "learning_rate": 2.93717277486911e-06, - "loss": 0.5619, - "step": 8200 - }, - { - "epoch": 14.2, - "learning_rate": 2.928446771378709e-06, - "loss": 0.5564, - "step": 8210 - }, - { - "epoch": 14.22, - "learning_rate": 2.919720767888307e-06, - "loss": 0.5542, - "step": 8220 - }, - { - "epoch": 14.24, - "learning_rate": 2.910994764397906e-06, - "loss": 0.5168, - "step": 8230 - }, - { - "epoch": 14.26, - "learning_rate": 2.9022687609075047e-06, - "loss": 0.5253, - "step": 8240 - }, - { - "epoch": 14.27, - "learning_rate": 2.893542757417103e-06, - "loss": 0.5264, - "step": 8250 - }, - { - "epoch": 14.29, - "learning_rate": 2.884816753926702e-06, - "loss": 0.5465, - "step": 8260 - }, - { - "epoch": 14.31, - "learning_rate": 2.8760907504363007e-06, - "loss": 0.5212, - "step": 8270 - }, - { - "epoch": 14.33, - "learning_rate": 2.867364746945899e-06, - "loss": 0.5431, - "step": 8280 - }, - { - "epoch": 14.34, - "learning_rate": 2.8586387434554976e-06, - "loss": 0.5308, - "step": 8290 - }, - { - "epoch": 14.36, - "learning_rate": 2.849912739965096e-06, - "loss": 0.5601, - "step": 8300 - }, - { - "epoch": 14.38, - "learning_rate": 2.841186736474695e-06, - "loss": 0.565, - "step": 8310 - }, - { - "epoch": 14.39, - "learning_rate": 2.8324607329842936e-06, - "loss": 0.4971, - "step": 8320 - }, - { - "epoch": 14.41, - "learning_rate": 2.823734729493892e-06, - "loss": 0.5669, - "step": 8330 - }, - { - "epoch": 14.43, - "learning_rate": 2.815008726003491e-06, - "loss": 0.5455, - "step": 8340 - }, - { - "epoch": 14.45, - "learning_rate": 2.806282722513089e-06, - "loss": 0.5567, - "step": 8350 - }, - { - "epoch": 14.46, - "learning_rate": 2.7975567190226877e-06, - "loss": 0.4743, - "step": 8360 - }, - { - "epoch": 14.48, - "learning_rate": 2.7888307155322865e-06, - "loss": 0.4855, - "step": 8370 - }, - { - "epoch": 14.5, - "learning_rate": 2.780104712041885e-06, - "loss": 0.5383, - "step": 8380 - }, - { - "epoch": 14.52, - "learning_rate": 2.7713787085514838e-06, - "loss": 0.5498, - "step": 8390 - }, - { - "epoch": 14.53, - "learning_rate": 2.7626527050610826e-06, - "loss": 0.582, - "step": 8400 - }, - { - "epoch": 14.55, - "learning_rate": 2.753926701570681e-06, - "loss": 0.5563, - "step": 8410 - }, - { - "epoch": 14.57, - "learning_rate": 2.7452006980802794e-06, - "loss": 0.5164, - "step": 8420 - }, - { - "epoch": 14.58, - "learning_rate": 2.736474694589878e-06, - "loss": 0.5519, - "step": 8430 - }, - { - "epoch": 14.6, - "learning_rate": 2.7277486910994767e-06, - "loss": 0.589, - "step": 8440 - }, - { - "epoch": 14.62, - "learning_rate": 2.719022687609075e-06, - "loss": 0.4949, - "step": 8450 - }, - { - "epoch": 14.64, - "learning_rate": 2.710296684118674e-06, - "loss": 0.5641, - "step": 8460 - }, - { - "epoch": 14.65, - "learning_rate": 2.7015706806282728e-06, - "loss": 0.5603, - "step": 8470 - }, - { - "epoch": 14.67, - "learning_rate": 2.692844677137871e-06, - "loss": 0.5616, - "step": 8480 - }, - { - "epoch": 14.69, - "learning_rate": 2.6841186736474696e-06, - "loss": 0.5677, - "step": 8490 - }, - { - "epoch": 14.71, - "learning_rate": 2.675392670157068e-06, - "loss": 0.5598, - "step": 8500 - }, - { - "epoch": 14.72, - "learning_rate": 2.666666666666667e-06, - "loss": 0.5005, - "step": 8510 - }, - { - "epoch": 14.74, - "learning_rate": 2.6579406631762657e-06, - "loss": 0.5137, - "step": 8520 - }, - { - "epoch": 14.76, - "learning_rate": 2.649214659685864e-06, - "loss": 0.5274, - "step": 8530 - }, - { - "epoch": 14.78, - "learning_rate": 2.640488656195463e-06, - "loss": 0.557, - "step": 8540 - }, - { - "epoch": 14.79, - "learning_rate": 2.631762652705061e-06, - "loss": 0.548, - "step": 8550 - }, - { - "epoch": 14.81, - "learning_rate": 2.6230366492146597e-06, - "loss": 0.5355, - "step": 8560 - }, - { - "epoch": 14.83, - "learning_rate": 2.6143106457242586e-06, - "loss": 0.54, - "step": 8570 - }, - { - "epoch": 14.84, - "learning_rate": 2.605584642233857e-06, - "loss": 0.5716, - "step": 8580 - }, - { - "epoch": 14.86, - "learning_rate": 2.596858638743456e-06, - "loss": 0.6191, - "step": 8590 - }, - { - "epoch": 14.88, - "learning_rate": 2.5881326352530546e-06, - "loss": 0.5551, - "step": 8600 - }, - { - "epoch": 14.9, - "learning_rate": 2.579406631762653e-06, - "loss": 0.5742, - "step": 8610 - }, - { - "epoch": 14.91, - "learning_rate": 2.5706806282722515e-06, - "loss": 0.5109, - "step": 8620 - }, - { - "epoch": 14.93, - "learning_rate": 2.56195462478185e-06, - "loss": 0.5588, - "step": 8630 - }, - { - "epoch": 14.95, - "learning_rate": 2.5532286212914487e-06, - "loss": 0.5456, - "step": 8640 - }, - { - "epoch": 14.97, - "learning_rate": 2.5445026178010475e-06, - "loss": 0.5422, - "step": 8650 - }, - { - "epoch": 14.98, - "learning_rate": 2.535776614310646e-06, - "loss": 0.5138, - "step": 8660 - }, - { - "epoch": 15.0, - "learning_rate": 2.527050610820245e-06, - "loss": 0.5137, - "step": 8670 - }, - { - "epoch": 15.0, - "eval_loss": 0.6050498485565186, - "eval_runtime": 132.2666, - "eval_samples_per_second": 4.022, - "eval_steps_per_second": 0.507, - "eval_wer": 0.33971845471491935, - "step": 8670 - }, - { - "epoch": 15.02, - "learning_rate": 2.518324607329843e-06, - "loss": 0.5396, - "step": 8680 - }, - { - "epoch": 15.03, - "learning_rate": 2.5095986038394416e-06, - "loss": 0.4991, - "step": 8690 - }, - { - "epoch": 15.05, - "learning_rate": 2.50087260034904e-06, - "loss": 0.5382, - "step": 8700 - }, - { - "epoch": 15.07, - "learning_rate": 2.492146596858639e-06, - "loss": 0.5248, - "step": 8710 - }, - { - "epoch": 15.09, - "learning_rate": 2.4834205933682377e-06, - "loss": 0.5947, - "step": 8720 - }, - { - "epoch": 15.1, - "learning_rate": 2.474694589877836e-06, - "loss": 0.5028, - "step": 8730 - }, - { - "epoch": 15.12, - "learning_rate": 2.4659685863874345e-06, - "loss": 0.5295, - "step": 8740 - }, - { - "epoch": 15.14, - "learning_rate": 2.4572425828970333e-06, - "loss": 0.5336, - "step": 8750 - }, - { - "epoch": 15.16, - "learning_rate": 2.448516579406632e-06, - "loss": 0.5267, - "step": 8760 - }, - { - "epoch": 15.17, - "learning_rate": 2.4397905759162306e-06, - "loss": 0.5743, - "step": 8770 - }, - { - "epoch": 15.19, - "learning_rate": 2.431064572425829e-06, - "loss": 0.5594, - "step": 8780 - }, - { - "epoch": 15.21, - "learning_rate": 2.422338568935428e-06, - "loss": 0.4936, - "step": 8790 - }, - { - "epoch": 15.22, - "learning_rate": 2.4136125654450263e-06, - "loss": 0.4595, - "step": 8800 - }, - { - "epoch": 15.24, - "learning_rate": 2.404886561954625e-06, - "loss": 0.4994, - "step": 8810 - }, - { - "epoch": 15.26, - "learning_rate": 2.3961605584642235e-06, - "loss": 0.5461, - "step": 8820 - }, - { - "epoch": 15.28, - "learning_rate": 2.3874345549738223e-06, - "loss": 0.5587, - "step": 8830 - }, - { - "epoch": 15.29, - "learning_rate": 2.3787085514834207e-06, - "loss": 0.5478, - "step": 8840 - }, - { - "epoch": 15.31, - "learning_rate": 2.3699825479930196e-06, - "loss": 0.4762, - "step": 8850 - }, - { - "epoch": 15.33, - "learning_rate": 2.361256544502618e-06, - "loss": 0.569, - "step": 8860 - }, - { - "epoch": 15.35, - "learning_rate": 2.3525305410122164e-06, - "loss": 0.5346, - "step": 8870 - }, - { - "epoch": 15.36, - "learning_rate": 2.3438045375218152e-06, - "loss": 0.4776, - "step": 8880 - }, - { - "epoch": 15.38, - "learning_rate": 2.335078534031414e-06, - "loss": 0.5498, - "step": 8890 - }, - { - "epoch": 15.4, - "learning_rate": 2.3263525305410125e-06, - "loss": 0.5457, - "step": 8900 - }, - { - "epoch": 15.42, - "learning_rate": 2.317626527050611e-06, - "loss": 0.56, - "step": 8910 - }, - { - "epoch": 15.43, - "learning_rate": 2.3089005235602097e-06, - "loss": 0.553, - "step": 8920 - }, - { - "epoch": 15.45, - "learning_rate": 2.3001745200698086e-06, - "loss": 0.5421, - "step": 8930 - }, - { - "epoch": 15.47, - "learning_rate": 2.2914485165794065e-06, - "loss": 0.5574, - "step": 8940 - }, - { - "epoch": 15.48, - "learning_rate": 2.2827225130890054e-06, - "loss": 0.5357, - "step": 8950 - }, - { - "epoch": 15.5, - "learning_rate": 2.2739965095986042e-06, - "loss": 0.5297, - "step": 8960 - }, - { - "epoch": 15.52, - "learning_rate": 2.2652705061082026e-06, - "loss": 0.5442, - "step": 8970 - }, - { - "epoch": 15.54, - "learning_rate": 2.256544502617801e-06, - "loss": 0.5945, - "step": 8980 - }, - { - "epoch": 15.55, - "learning_rate": 2.2478184991274e-06, - "loss": 0.5285, - "step": 8990 - }, - { - "epoch": 15.57, - "learning_rate": 2.2390924956369983e-06, - "loss": 0.517, - "step": 9000 - }, - { - "epoch": 15.59, - "learning_rate": 2.230366492146597e-06, - "loss": 0.5206, - "step": 9010 - }, - { - "epoch": 15.61, - "learning_rate": 2.2216404886561955e-06, - "loss": 0.5397, - "step": 9020 - }, - { - "epoch": 15.62, - "learning_rate": 2.2129144851657944e-06, - "loss": 0.4597, - "step": 9030 - }, - { - "epoch": 15.64, - "learning_rate": 2.2041884816753928e-06, - "loss": 0.5415, - "step": 9040 - }, - { - "epoch": 15.66, - "learning_rate": 2.1954624781849916e-06, - "loss": 0.4796, - "step": 9050 - }, - { - "epoch": 15.67, - "learning_rate": 2.18673647469459e-06, - "loss": 0.5345, - "step": 9060 - }, - { - "epoch": 15.69, - "learning_rate": 2.1780104712041884e-06, - "loss": 0.5424, - "step": 9070 - }, - { - "epoch": 15.71, - "learning_rate": 2.1692844677137873e-06, - "loss": 0.5421, - "step": 9080 - }, - { - "epoch": 15.73, - "learning_rate": 2.160558464223386e-06, - "loss": 0.4838, - "step": 9090 - }, - { - "epoch": 15.74, - "learning_rate": 2.1518324607329845e-06, - "loss": 0.5087, - "step": 9100 - }, - { - "epoch": 15.76, - "learning_rate": 2.143106457242583e-06, - "loss": 0.5319, - "step": 9110 - }, - { - "epoch": 15.78, - "learning_rate": 2.1343804537521818e-06, - "loss": 0.6137, - "step": 9120 - }, - { - "epoch": 15.8, - "learning_rate": 2.1256544502617806e-06, - "loss": 0.5591, - "step": 9130 - }, - { - "epoch": 15.81, - "learning_rate": 2.116928446771379e-06, - "loss": 0.5355, - "step": 9140 - }, - { - "epoch": 15.83, - "learning_rate": 2.1082024432809774e-06, - "loss": 0.52, - "step": 9150 - }, - { - "epoch": 15.85, - "learning_rate": 2.0994764397905762e-06, - "loss": 0.5187, - "step": 9160 - }, - { - "epoch": 15.87, - "learning_rate": 2.0907504363001747e-06, - "loss": 0.6128, - "step": 9170 - }, - { - "epoch": 15.88, - "learning_rate": 2.0820244328097735e-06, - "loss": 0.5245, - "step": 9180 - }, - { - "epoch": 15.9, - "learning_rate": 2.073298429319372e-06, - "loss": 0.6331, - "step": 9190 - }, - { - "epoch": 15.92, - "learning_rate": 2.0645724258289703e-06, - "loss": 0.5127, - "step": 9200 - }, - { - "epoch": 15.93, - "learning_rate": 2.055846422338569e-06, - "loss": 0.5414, - "step": 9210 - }, - { - "epoch": 15.95, - "learning_rate": 2.0471204188481676e-06, - "loss": 0.5952, - "step": 9220 - }, - { - "epoch": 15.97, - "learning_rate": 2.0383944153577664e-06, - "loss": 0.5299, - "step": 9230 - }, - { - "epoch": 15.99, - "learning_rate": 2.029668411867365e-06, - "loss": 0.5038, - "step": 9240 - }, - { - "epoch": 16.0, - "eval_loss": 0.6019043922424316, - "eval_runtime": 130.2022, - "eval_samples_per_second": 4.086, - "eval_steps_per_second": 0.515, - "eval_wer": 0.3413897280966767, - "step": 9248 - }, - { - "epoch": 16.0, - "learning_rate": 2.0209424083769636e-06, - "loss": 0.4864, - "step": 9250 - }, - { - "epoch": 16.02, - "learning_rate": 2.012216404886562e-06, - "loss": 0.5364, - "step": 9260 - }, - { - "epoch": 16.04, - "learning_rate": 2.0034904013961605e-06, - "loss": 0.51, - "step": 9270 - }, - { - "epoch": 16.06, - "learning_rate": 1.9947643979057593e-06, - "loss": 0.4869, - "step": 9280 - }, - { - "epoch": 16.07, - "learning_rate": 1.986038394415358e-06, - "loss": 0.4882, - "step": 9290 - }, - { - "epoch": 16.09, - "learning_rate": 1.9773123909249565e-06, - "loss": 0.5641, - "step": 9300 - }, - { - "epoch": 16.11, - "learning_rate": 1.968586387434555e-06, - "loss": 0.5042, - "step": 9310 - }, - { - "epoch": 16.12, - "learning_rate": 1.9598603839441538e-06, - "loss": 0.5145, - "step": 9320 - }, - { - "epoch": 16.14, - "learning_rate": 1.9511343804537526e-06, - "loss": 0.5179, - "step": 9330 - }, - { - "epoch": 16.16, - "learning_rate": 1.942408376963351e-06, - "loss": 0.5445, - "step": 9340 - }, - { - "epoch": 16.18, - "learning_rate": 1.9336823734729494e-06, - "loss": 0.5599, - "step": 9350 - }, - { - "epoch": 16.19, - "learning_rate": 1.9249563699825483e-06, - "loss": 0.5552, - "step": 9360 - }, - { - "epoch": 16.21, - "learning_rate": 1.9162303664921467e-06, - "loss": 0.5113, - "step": 9370 - }, - { - "epoch": 16.23, - "learning_rate": 1.9075043630017455e-06, - "loss": 0.5163, - "step": 9380 - }, - { - "epoch": 16.25, - "learning_rate": 1.898778359511344e-06, - "loss": 0.5468, - "step": 9390 - }, - { - "epoch": 16.26, - "learning_rate": 1.8900523560209423e-06, - "loss": 0.5454, - "step": 9400 - }, - { - "epoch": 16.28, - "learning_rate": 1.8813263525305412e-06, - "loss": 0.5191, - "step": 9410 - }, - { - "epoch": 16.3, - "learning_rate": 1.8726003490401398e-06, - "loss": 0.5209, - "step": 9420 - }, - { - "epoch": 16.31, - "learning_rate": 1.8638743455497384e-06, - "loss": 0.5336, - "step": 9430 - }, - { - "epoch": 16.33, - "learning_rate": 1.8551483420593368e-06, - "loss": 0.5381, - "step": 9440 - }, - { - "epoch": 16.35, - "learning_rate": 1.8464223385689357e-06, - "loss": 0.5876, - "step": 9450 - }, - { - "epoch": 16.37, - "learning_rate": 1.8376963350785343e-06, - "loss": 0.5021, - "step": 9460 - }, - { - "epoch": 16.38, - "learning_rate": 1.8289703315881327e-06, - "loss": 0.5492, - "step": 9470 - }, - { - "epoch": 16.4, - "learning_rate": 1.8202443280977313e-06, - "loss": 0.515, - "step": 9480 - }, - { - "epoch": 16.42, - "learning_rate": 1.8115183246073302e-06, - "loss": 0.5622, - "step": 9490 - }, - { - "epoch": 16.44, - "learning_rate": 1.8027923211169286e-06, - "loss": 0.5772, - "step": 9500 - }, - { - "epoch": 16.45, - "learning_rate": 1.7940663176265272e-06, - "loss": 0.5155, - "step": 9510 - }, - { - "epoch": 16.47, - "learning_rate": 1.7853403141361258e-06, - "loss": 0.5359, - "step": 9520 - }, - { - "epoch": 16.49, - "learning_rate": 1.7766143106457244e-06, - "loss": 0.5634, - "step": 9530 - }, - { - "epoch": 16.51, - "learning_rate": 1.7678883071553229e-06, - "loss": 0.5628, - "step": 9540 - }, - { - "epoch": 16.52, - "learning_rate": 1.7591623036649217e-06, - "loss": 0.5789, - "step": 9550 - }, - { - "epoch": 16.54, - "learning_rate": 1.7504363001745203e-06, - "loss": 0.5261, - "step": 9560 - }, - { - "epoch": 16.56, - "learning_rate": 1.7417102966841187e-06, - "loss": 0.5353, - "step": 9570 - }, - { - "epoch": 16.57, - "learning_rate": 1.7329842931937173e-06, - "loss": 0.5638, - "step": 9580 - }, - { - "epoch": 16.59, - "learning_rate": 1.7242582897033162e-06, - "loss": 0.5454, - "step": 9590 - }, - { - "epoch": 16.61, - "learning_rate": 1.7155322862129146e-06, - "loss": 0.5318, - "step": 9600 - }, - { - "epoch": 16.63, - "learning_rate": 1.7068062827225132e-06, - "loss": 0.5227, - "step": 9610 - }, - { - "epoch": 16.64, - "learning_rate": 1.6980802792321118e-06, - "loss": 0.4739, - "step": 9620 - }, - { - "epoch": 16.66, - "learning_rate": 1.6893542757417107e-06, - "loss": 0.4917, - "step": 9630 - }, - { - "epoch": 16.68, - "learning_rate": 1.680628272251309e-06, - "loss": 0.5405, - "step": 9640 - }, - { - "epoch": 16.7, - "learning_rate": 1.6719022687609077e-06, - "loss": 0.4907, - "step": 9650 - }, - { - "epoch": 16.71, - "learning_rate": 1.6631762652705063e-06, - "loss": 0.5849, - "step": 9660 - }, - { - "epoch": 16.73, - "learning_rate": 1.6544502617801047e-06, - "loss": 0.5934, - "step": 9670 - }, - { - "epoch": 16.75, - "learning_rate": 1.6457242582897034e-06, - "loss": 0.4727, - "step": 9680 - }, - { - "epoch": 16.76, - "learning_rate": 1.6369982547993022e-06, - "loss": 0.5012, - "step": 9690 - }, - { - "epoch": 16.78, - "learning_rate": 1.6282722513089006e-06, - "loss": 0.5727, - "step": 9700 - }, - { - "epoch": 16.8, - "learning_rate": 1.6195462478184992e-06, - "loss": 0.498, - "step": 9710 - }, - { - "epoch": 16.82, - "learning_rate": 1.6108202443280978e-06, - "loss": 0.562, - "step": 9720 - }, - { - "epoch": 16.83, - "learning_rate": 1.6020942408376967e-06, - "loss": 0.5346, - "step": 9730 - }, - { - "epoch": 16.85, - "learning_rate": 1.593368237347295e-06, - "loss": 0.5542, - "step": 9740 - }, - { - "epoch": 16.87, - "learning_rate": 1.5846422338568937e-06, - "loss": 0.5613, - "step": 9750 - }, - { - "epoch": 16.89, - "learning_rate": 1.5759162303664923e-06, - "loss": 0.5093, - "step": 9760 - }, - { - "epoch": 16.9, - "learning_rate": 1.5671902268760907e-06, - "loss": 0.4826, - "step": 9770 - }, - { - "epoch": 16.92, - "learning_rate": 1.5584642233856896e-06, - "loss": 0.5193, - "step": 9780 - }, - { - "epoch": 16.94, - "learning_rate": 1.5497382198952882e-06, - "loss": 0.5452, - "step": 9790 - }, - { - "epoch": 16.96, - "learning_rate": 1.5410122164048866e-06, - "loss": 0.5558, - "step": 9800 - }, - { - "epoch": 16.97, - "learning_rate": 1.5322862129144852e-06, - "loss": 0.5058, - "step": 9810 - }, - { - "epoch": 16.99, - "learning_rate": 1.5235602094240839e-06, - "loss": 0.5303, - "step": 9820 - }, - { - "epoch": 17.0, - "eval_loss": 0.6061190366744995, - "eval_runtime": 129.5204, - "eval_samples_per_second": 4.107, - "eval_steps_per_second": 0.517, - "eval_wer": 0.34563219129652245, - "step": 9826 - }, - { - "epoch": 17.01, - "learning_rate": 1.5157068062827227e-06, - "loss": 0.6035, - "step": 9830 - }, - { - "epoch": 17.02, - "learning_rate": 1.5069808027923214e-06, - "loss": 0.5436, - "step": 9840 - }, - { - "epoch": 17.04, - "learning_rate": 1.4982547993019198e-06, - "loss": 0.5601, - "step": 9850 - }, - { - "epoch": 17.06, - "learning_rate": 1.4895287958115184e-06, - "loss": 0.5334, - "step": 9860 - }, - { - "epoch": 17.08, - "learning_rate": 1.4808027923211172e-06, - "loss": 0.5023, - "step": 9870 - }, - { - "epoch": 17.09, - "learning_rate": 1.4720767888307156e-06, - "loss": 0.5493, - "step": 9880 - }, - { - "epoch": 17.11, - "learning_rate": 1.4633507853403143e-06, - "loss": 0.5355, - "step": 9890 - }, - { - "epoch": 17.13, - "learning_rate": 1.4546247818499129e-06, - "loss": 0.4531, - "step": 9900 - }, - { - "epoch": 17.15, - "learning_rate": 1.4458987783595115e-06, - "loss": 0.5017, - "step": 9910 - }, - { - "epoch": 17.16, - "learning_rate": 1.43717277486911e-06, - "loss": 0.5344, - "step": 9920 - }, - { - "epoch": 17.18, - "learning_rate": 1.4284467713787087e-06, - "loss": 0.4956, - "step": 9930 - }, - { - "epoch": 17.2, - "learning_rate": 1.4197207678883074e-06, - "loss": 0.5686, - "step": 9940 - }, - { - "epoch": 17.21, - "learning_rate": 1.4109947643979058e-06, - "loss": 0.5124, - "step": 9950 - }, - { - "epoch": 17.23, - "learning_rate": 1.4022687609075044e-06, - "loss": 0.5322, - "step": 9960 - }, - { - "epoch": 17.25, - "learning_rate": 1.3935427574171032e-06, - "loss": 0.5349, - "step": 9970 - }, - { - "epoch": 17.27, - "learning_rate": 1.3848167539267016e-06, - "loss": 0.5571, - "step": 9980 - }, - { - "epoch": 17.28, - "learning_rate": 1.3760907504363003e-06, - "loss": 0.5222, - "step": 9990 - }, - { - "epoch": 17.3, - "learning_rate": 1.3673647469458989e-06, - "loss": 0.5174, - "step": 10000 - }, - { - "epoch": 17.32, - "learning_rate": 1.3586387434554977e-06, - "loss": 0.502, - "step": 10010 - }, - { - "epoch": 17.34, - "learning_rate": 1.349912739965096e-06, - "loss": 0.4861, - "step": 10020 - }, - { - "epoch": 17.35, - "learning_rate": 1.3411867364746948e-06, - "loss": 0.5348, - "step": 10030 - }, - { - "epoch": 17.37, - "learning_rate": 1.3324607329842934e-06, - "loss": 0.5186, - "step": 10040 - }, - { - "epoch": 17.39, - "learning_rate": 1.3237347294938918e-06, - "loss": 0.5009, - "step": 10050 - }, - { - "epoch": 17.4, - "learning_rate": 1.3150087260034904e-06, - "loss": 0.5177, - "step": 10060 - }, - { - "epoch": 17.42, - "learning_rate": 1.3062827225130893e-06, - "loss": 0.5734, - "step": 10070 - }, - { - "epoch": 17.44, - "learning_rate": 1.2975567190226877e-06, - "loss": 0.5008, - "step": 10080 - }, - { - "epoch": 17.46, - "learning_rate": 1.2888307155322863e-06, - "loss": 0.5276, - "step": 10090 - }, - { - "epoch": 17.47, - "learning_rate": 1.280104712041885e-06, - "loss": 0.5469, - "step": 10100 - }, - { - "epoch": 17.49, - "learning_rate": 1.2713787085514837e-06, - "loss": 0.5398, - "step": 10110 - }, - { - "epoch": 17.51, - "learning_rate": 1.2626527050610822e-06, - "loss": 0.5883, - "step": 10120 - }, - { - "epoch": 17.53, - "learning_rate": 1.2539267015706808e-06, - "loss": 0.5326, - "step": 10130 - }, - { - "epoch": 17.54, - "learning_rate": 1.2452006980802794e-06, - "loss": 0.5396, - "step": 10140 - }, - { - "epoch": 17.56, - "learning_rate": 1.236474694589878e-06, - "loss": 0.5167, - "step": 10150 - }, - { - "epoch": 17.58, - "learning_rate": 1.2277486910994764e-06, - "loss": 0.5202, - "step": 10160 - }, - { - "epoch": 17.6, - "learning_rate": 1.219022687609075e-06, - "loss": 0.5075, - "step": 10170 - }, - { - "epoch": 17.61, - "learning_rate": 1.2102966841186737e-06, - "loss": 0.5561, - "step": 10180 - }, - { - "epoch": 17.63, - "learning_rate": 1.2015706806282723e-06, - "loss": 0.5276, - "step": 10190 - }, - { - "epoch": 17.65, - "learning_rate": 1.192844677137871e-06, - "loss": 0.4895, - "step": 10200 - }, - { - "epoch": 17.66, - "learning_rate": 1.1841186736474695e-06, - "loss": 0.4994, - "step": 10210 - }, - { - "epoch": 17.68, - "learning_rate": 1.1753926701570682e-06, - "loss": 0.5567, - "step": 10220 - }, - { - "epoch": 17.7, - "learning_rate": 1.1666666666666668e-06, - "loss": 0.5653, - "step": 10230 - }, - { - "epoch": 17.72, - "learning_rate": 1.1579406631762654e-06, - "loss": 0.5314, - "step": 10240 - }, - { - "epoch": 17.73, - "learning_rate": 1.149214659685864e-06, - "loss": 0.5415, - "step": 10250 - }, - { - "epoch": 17.75, - "learning_rate": 1.1404886561954627e-06, - "loss": 0.5135, - "step": 10260 - }, - { - "epoch": 17.77, - "learning_rate": 1.131762652705061e-06, - "loss": 0.5054, - "step": 10270 - }, - { - "epoch": 17.79, - "learning_rate": 1.1230366492146597e-06, - "loss": 0.562, - "step": 10280 - }, - { - "epoch": 17.8, - "learning_rate": 1.1143106457242583e-06, - "loss": 0.5183, - "step": 10290 - }, - { - "epoch": 17.82, - "learning_rate": 1.105584642233857e-06, - "loss": 0.531, - "step": 10300 - }, - { - "epoch": 17.84, - "learning_rate": 1.0968586387434556e-06, - "loss": 0.5516, - "step": 10310 - }, - { - "epoch": 17.85, - "learning_rate": 1.0881326352530542e-06, - "loss": 0.5528, - "step": 10320 - }, - { - "epoch": 17.87, - "learning_rate": 1.0794066317626528e-06, - "loss": 0.5452, - "step": 10330 - }, - { - "epoch": 17.89, - "learning_rate": 1.0706806282722514e-06, - "loss": 0.4932, - "step": 10340 - }, - { - "epoch": 17.91, - "learning_rate": 1.06195462478185e-06, - "loss": 0.5541, - "step": 10350 - }, - { - "epoch": 17.92, - "learning_rate": 1.0532286212914487e-06, - "loss": 0.5646, - "step": 10360 - }, - { - "epoch": 17.94, - "learning_rate": 1.044502617801047e-06, - "loss": 0.5169, - "step": 10370 - }, - { - "epoch": 17.96, - "learning_rate": 1.035776614310646e-06, - "loss": 0.533, - "step": 10380 - }, - { - "epoch": 17.98, - "learning_rate": 1.0270506108202443e-06, - "loss": 0.4859, - "step": 10390 - }, - { - "epoch": 17.99, - "learning_rate": 1.0183246073298432e-06, - "loss": 0.5401, - "step": 10400 - }, - { - "epoch": 18.0, - "eval_loss": 0.6051313281059265, - "eval_runtime": 130.7444, - "eval_samples_per_second": 4.069, - "eval_steps_per_second": 0.512, - "eval_wer": 0.34100404962396347, - "step": 10404 - }, - { - "epoch": 18.01, - "learning_rate": 1.0095986038394416e-06, - "loss": 0.5043, - "step": 10410 - }, - { - "epoch": 18.03, - "learning_rate": 1.0008726003490402e-06, - "loss": 0.5156, - "step": 10420 - }, - { - "epoch": 18.04, - "learning_rate": 9.921465968586388e-07, - "loss": 0.5203, - "step": 10430 - }, - { - "epoch": 18.06, - "learning_rate": 9.834205933682374e-07, - "loss": 0.5351, - "step": 10440 - }, - { - "epoch": 18.08, - "learning_rate": 9.74694589877836e-07, - "loss": 0.5186, - "step": 10450 - }, - { - "epoch": 18.1, - "learning_rate": 9.659685863874347e-07, - "loss": 0.5792, - "step": 10460 - }, - { - "epoch": 18.11, - "learning_rate": 9.57242582897033e-07, - "loss": 0.5372, - "step": 10470 - }, - { - "epoch": 18.13, - "learning_rate": 9.485165794066318e-07, - "loss": 0.5196, - "step": 10480 - }, - { - "epoch": 18.15, - "learning_rate": 9.397905759162305e-07, - "loss": 0.5388, - "step": 10490 - }, - { - "epoch": 18.17, - "learning_rate": 9.310645724258291e-07, - "loss": 0.5527, - "step": 10500 - }, - { - "epoch": 18.18, - "learning_rate": 9.223385689354276e-07, - "loss": 0.5534, - "step": 10510 - }, - { - "epoch": 18.2, - "learning_rate": 9.136125654450262e-07, - "loss": 0.5178, - "step": 10520 - }, - { - "epoch": 18.22, - "learning_rate": 9.048865619546248e-07, - "loss": 0.4728, - "step": 10530 - }, - { - "epoch": 18.24, - "learning_rate": 8.961605584642235e-07, - "loss": 0.5262, - "step": 10540 - }, - { - "epoch": 18.25, - "learning_rate": 8.874345549738221e-07, - "loss": 0.5408, - "step": 10550 - }, - { - "epoch": 18.27, - "learning_rate": 8.787085514834207e-07, - "loss": 0.5592, - "step": 10560 - }, - { - "epoch": 18.29, - "learning_rate": 8.699825479930192e-07, - "loss": 0.5394, - "step": 10570 - }, - { - "epoch": 18.3, - "learning_rate": 8.612565445026178e-07, - "loss": 0.4525, - "step": 10580 - }, - { - "epoch": 18.32, - "learning_rate": 8.525305410122165e-07, - "loss": 0.5343, - "step": 10590 - }, - { - "epoch": 18.34, - "learning_rate": 8.438045375218151e-07, - "loss": 0.5522, - "step": 10600 - }, - { - "epoch": 18.36, - "learning_rate": 8.350785340314137e-07, - "loss": 0.5574, - "step": 10610 - }, - { - "epoch": 18.37, - "learning_rate": 8.263525305410122e-07, - "loss": 0.5128, - "step": 10620 - }, - { - "epoch": 18.39, - "learning_rate": 8.17626527050611e-07, - "loss": 0.5077, - "step": 10630 - }, - { - "epoch": 18.41, - "learning_rate": 8.089005235602095e-07, - "loss": 0.522, - "step": 10640 - }, - { - "epoch": 18.43, - "learning_rate": 8.001745200698081e-07, - "loss": 0.5084, - "step": 10650 - }, - { - "epoch": 18.44, - "learning_rate": 7.914485165794067e-07, - "loss": 0.5284, - "step": 10660 - }, - { - "epoch": 18.46, - "learning_rate": 7.827225130890052e-07, - "loss": 0.5423, - "step": 10670 - }, - { - "epoch": 18.48, - "learning_rate": 7.73996509598604e-07, - "loss": 0.5091, - "step": 10680 - }, - { - "epoch": 18.49, - "learning_rate": 7.652705061082025e-07, - "loss": 0.5135, - "step": 10690 - }, - { - "epoch": 18.51, - "learning_rate": 7.565445026178011e-07, - "loss": 0.5221, - "step": 10700 - }, - { - "epoch": 18.53, - "learning_rate": 7.478184991273997e-07, - "loss": 0.5039, - "step": 10710 - }, - { - "epoch": 18.55, - "learning_rate": 7.390924956369982e-07, - "loss": 0.5479, - "step": 10720 - }, - { - "epoch": 18.56, - "learning_rate": 7.30366492146597e-07, - "loss": 0.5199, - "step": 10730 - }, - { - "epoch": 18.58, - "learning_rate": 7.216404886561955e-07, - "loss": 0.5071, - "step": 10740 - }, - { - "epoch": 18.6, - "learning_rate": 7.129144851657942e-07, - "loss": 0.5522, - "step": 10750 - }, - { - "epoch": 18.62, - "learning_rate": 7.041884816753927e-07, - "loss": 0.58, - "step": 10760 - }, - { - "epoch": 18.63, - "learning_rate": 6.954624781849913e-07, - "loss": 0.4925, - "step": 10770 - }, - { - "epoch": 18.65, - "learning_rate": 6.8673647469459e-07, - "loss": 0.5188, - "step": 10780 - }, - { - "epoch": 18.67, - "learning_rate": 6.780104712041885e-07, - "loss": 0.4897, - "step": 10790 - }, - { - "epoch": 18.69, - "learning_rate": 6.692844677137872e-07, - "loss": 0.524, - "step": 10800 - }, - { - "epoch": 18.7, - "learning_rate": 6.605584642233857e-07, - "loss": 0.5318, - "step": 10810 - }, - { - "epoch": 18.72, - "learning_rate": 6.518324607329843e-07, - "loss": 0.4951, - "step": 10820 - }, - { - "epoch": 18.74, - "learning_rate": 6.43106457242583e-07, - "loss": 0.4568, - "step": 10830 - }, - { - "epoch": 18.75, - "learning_rate": 6.343804537521815e-07, - "loss": 0.5313, - "step": 10840 - }, - { - "epoch": 18.77, - "learning_rate": 6.256544502617802e-07, - "loss": 0.5422, - "step": 10850 - }, - { - "epoch": 18.79, - "learning_rate": 6.169284467713788e-07, - "loss": 0.5219, - "step": 10860 - }, - { - "epoch": 18.81, - "learning_rate": 6.082024432809774e-07, - "loss": 0.5134, - "step": 10870 - }, - { - "epoch": 18.82, - "learning_rate": 5.99476439790576e-07, - "loss": 0.4774, - "step": 10880 - }, - { - "epoch": 18.84, - "learning_rate": 5.907504363001745e-07, - "loss": 0.552, - "step": 10890 - }, - { - "epoch": 18.86, - "learning_rate": 5.820244328097731e-07, - "loss": 0.5729, - "step": 10900 - }, - { - "epoch": 18.88, - "learning_rate": 5.732984293193718e-07, - "loss": 0.5364, - "step": 10910 - }, - { - "epoch": 18.89, - "learning_rate": 5.645724258289704e-07, - "loss": 0.5542, - "step": 10920 - }, - { - "epoch": 18.91, - "learning_rate": 5.55846422338569e-07, - "loss": 0.4835, - "step": 10930 - }, - { - "epoch": 18.93, - "learning_rate": 5.471204188481675e-07, - "loss": 0.4924, - "step": 10940 - }, - { - "epoch": 18.94, - "learning_rate": 5.383944153577661e-07, - "loss": 0.4915, - "step": 10950 - }, - { - "epoch": 18.96, - "learning_rate": 5.296684118673648e-07, - "loss": 0.5168, - "step": 10960 - }, - { - "epoch": 18.98, - "learning_rate": 5.209424083769634e-07, - "loss": 0.5149, - "step": 10970 - }, - { - "epoch": 19.0, - "learning_rate": 5.12216404886562e-07, - "loss": 0.5394, - "step": 10980 - }, - { - "epoch": 19.0, - "eval_loss": 0.6060788035392761, - "eval_runtime": 130.6747, - "eval_samples_per_second": 4.071, - "eval_steps_per_second": 0.513, - "eval_wer": 0.33991129395127595, - "step": 10982 - }, - { - "epoch": 19.01, - "learning_rate": 5.034904013961605e-07, - "loss": 0.4956, - "step": 10990 - }, - { - "epoch": 19.03, - "learning_rate": 4.947643979057592e-07, - "loss": 0.5086, - "step": 11000 - }, - { - "epoch": 19.05, - "learning_rate": 4.860383944153578e-07, - "loss": 0.4949, - "step": 11010 - }, - { - "epoch": 19.07, - "learning_rate": 4.773123909249564e-07, - "loss": 0.5349, - "step": 11020 - }, - { - "epoch": 19.08, - "learning_rate": 4.68586387434555e-07, - "loss": 0.5417, - "step": 11030 - }, - { - "epoch": 19.1, - "learning_rate": 4.598603839441536e-07, - "loss": 0.5695, - "step": 11040 - }, - { - "epoch": 19.12, - "learning_rate": 4.511343804537522e-07, - "loss": 0.4968, - "step": 11050 - }, - { - "epoch": 19.13, - "learning_rate": 4.4240837696335083e-07, - "loss": 0.4633, - "step": 11060 - }, - { - "epoch": 19.15, - "learning_rate": 4.3368237347294946e-07, - "loss": 0.5527, - "step": 11070 - }, - { - "epoch": 19.17, - "learning_rate": 4.2495636998254803e-07, - "loss": 0.5272, - "step": 11080 - }, - { - "epoch": 19.19, - "learning_rate": 4.162303664921466e-07, - "loss": 0.5196, - "step": 11090 - }, - { - "epoch": 19.2, - "learning_rate": 4.075043630017452e-07, - "loss": 0.5283, - "step": 11100 - }, - { - "epoch": 19.22, - "learning_rate": 3.9877835951134384e-07, - "loss": 0.4837, - "step": 11110 - }, - { - "epoch": 19.24, - "learning_rate": 3.9005235602094246e-07, - "loss": 0.5378, - "step": 11120 - }, - { - "epoch": 19.26, - "learning_rate": 3.813263525305411e-07, - "loss": 0.5771, - "step": 11130 - }, - { - "epoch": 19.27, - "learning_rate": 3.726003490401396e-07, - "loss": 0.5156, - "step": 11140 - }, - { - "epoch": 19.29, - "learning_rate": 3.6387434554973823e-07, - "loss": 0.5284, - "step": 11150 - }, - { - "epoch": 19.31, - "learning_rate": 3.5514834205933685e-07, - "loss": 0.5293, - "step": 11160 - }, - { - "epoch": 19.33, - "learning_rate": 3.4642233856893547e-07, - "loss": 0.5141, - "step": 11170 - }, - { - "epoch": 19.34, - "learning_rate": 3.376963350785341e-07, - "loss": 0.5497, - "step": 11180 - }, - { - "epoch": 19.36, - "learning_rate": 3.289703315881326e-07, - "loss": 0.483, - "step": 11190 - }, - { - "epoch": 19.38, - "learning_rate": 3.2024432809773123e-07, - "loss": 0.4886, - "step": 11200 - }, - { - "epoch": 19.39, - "learning_rate": 3.1151832460732986e-07, - "loss": 0.5088, - "step": 11210 - }, - { - "epoch": 19.41, - "learning_rate": 3.027923211169285e-07, - "loss": 0.51, - "step": 11220 - }, - { - "epoch": 19.43, - "learning_rate": 2.9406631762652705e-07, - "loss": 0.5243, - "step": 11230 - }, - { - "epoch": 19.45, - "learning_rate": 2.8534031413612567e-07, - "loss": 0.5467, - "step": 11240 - }, - { - "epoch": 19.46, - "learning_rate": 2.766143106457243e-07, - "loss": 0.5319, - "step": 11250 - }, - { - "epoch": 19.48, - "learning_rate": 2.6788830715532287e-07, - "loss": 0.5202, - "step": 11260 - }, - { - "epoch": 19.5, - "learning_rate": 2.591623036649215e-07, - "loss": 0.5157, - "step": 11270 - }, - { - "epoch": 19.52, - "learning_rate": 2.5043630017452006e-07, - "loss": 0.5354, - "step": 11280 - }, - { - "epoch": 19.53, - "learning_rate": 2.417102966841187e-07, - "loss": 0.5245, - "step": 11290 - }, - { - "epoch": 19.55, - "learning_rate": 2.329842931937173e-07, - "loss": 0.5048, - "step": 11300 - }, - { - "epoch": 19.57, - "learning_rate": 2.242582897033159e-07, - "loss": 0.5001, - "step": 11310 - }, - { - "epoch": 19.58, - "learning_rate": 2.1553228621291452e-07, - "loss": 0.549, - "step": 11320 - }, - { - "epoch": 19.6, - "learning_rate": 2.068062827225131e-07, - "loss": 0.5802, - "step": 11330 - }, - { - "epoch": 19.62, - "learning_rate": 1.9808027923211171e-07, - "loss": 0.5279, - "step": 11340 - }, - { - "epoch": 19.64, - "learning_rate": 1.8935427574171034e-07, - "loss": 0.5119, - "step": 11350 - }, - { - "epoch": 19.65, - "learning_rate": 1.806282722513089e-07, - "loss": 0.4711, - "step": 11360 - }, - { - "epoch": 19.67, - "learning_rate": 1.7190226876090753e-07, - "loss": 0.5211, - "step": 11370 - }, - { - "epoch": 19.69, - "learning_rate": 1.631762652705061e-07, - "loss": 0.5347, - "step": 11380 - }, - { - "epoch": 19.71, - "learning_rate": 1.5445026178010472e-07, - "loss": 0.5219, - "step": 11390 - }, - { - "epoch": 19.72, - "learning_rate": 1.4572425828970332e-07, - "loss": 0.5322, - "step": 11400 - }, - { - "epoch": 19.74, - "learning_rate": 1.3699825479930191e-07, - "loss": 0.4966, - "step": 11410 - }, - { - "epoch": 19.76, - "learning_rate": 1.2827225130890054e-07, - "loss": 0.5378, - "step": 11420 - }, - { - "epoch": 19.78, - "learning_rate": 1.1954624781849913e-07, - "loss": 0.5918, - "step": 11430 - }, - { - "epoch": 19.79, - "learning_rate": 1.1082024432809774e-07, - "loss": 0.5185, - "step": 11440 - }, - { - "epoch": 19.81, - "learning_rate": 1.0209424083769634e-07, - "loss": 0.4573, - "step": 11450 - }, - { - "epoch": 19.83, - "learning_rate": 9.336823734729495e-08, - "loss": 0.4898, - "step": 11460 - }, - { - "epoch": 19.84, - "learning_rate": 8.464223385689354e-08, - "loss": 0.5509, - "step": 11470 - }, - { - "epoch": 19.86, - "learning_rate": 7.591623036649215e-08, - "loss": 0.5644, - "step": 11480 - }, - { - "epoch": 19.88, - "learning_rate": 6.719022687609076e-08, - "loss": 0.5216, - "step": 11490 - }, - { - "epoch": 19.9, - "learning_rate": 5.846422338568936e-08, - "loss": 0.5568, - "step": 11500 - }, - { - "epoch": 19.91, - "learning_rate": 4.9738219895287956e-08, - "loss": 0.5159, - "step": 11510 - }, - { - "epoch": 19.93, - "learning_rate": 4.1012216404886566e-08, - "loss": 0.5331, - "step": 11520 - }, - { - "epoch": 19.95, - "learning_rate": 3.228621291448517e-08, - "loss": 0.556, - "step": 11530 - }, - { - "epoch": 19.97, - "learning_rate": 2.356020942408377e-08, - "loss": 0.5162, - "step": 11540 - }, - { - "epoch": 19.98, - "learning_rate": 1.4834205933682374e-08, - "loss": 0.5134, - "step": 11550 - }, - { - "epoch": 20.0, - "learning_rate": 6.108202443280978e-09, - "loss": 0.4707, - "step": 11560 - }, { "epoch": 20.0, - "eval_loss": 0.6008606553077698, - "eval_runtime": 132.4975, - "eval_samples_per_second": 4.015, - "eval_steps_per_second": 0.506, - "eval_wer": 0.3386256990422318, - "step": 11560 + "eval_loss": 0.5728763937950134, + "eval_runtime": 133.6384, + "eval_samples_per_second": 3.981, + "eval_steps_per_second": 0.995, + "eval_wer": 0.3018576846435688, + "step": 5760 }, { "epoch": 20.0, - "step": 11560, - "total_flos": 5.45175658649426e+19, - "train_loss": 0.5838032910980567, - "train_runtime": 39615.1021, - "train_samples_per_second": 2.331, - "train_steps_per_second": 0.292 + "step": 5760, + "total_flos": 5.436639839854258e+19, + "train_loss": 0.5082269724872377, + "train_runtime": 42641.5901, + "train_samples_per_second": 2.165, + "train_steps_per_second": 0.135 } ], - "max_steps": 11560, + "max_steps": 5760, "num_train_epochs": 20, - "total_flos": 5.45175658649426e+19, + "total_flos": 5.436639839854258e+19, "trial_name": null, "trial_params": null }