diff --git "a/exp/log/log-train-2022-06-18-10-29-38-2" "b/exp/log/log-train-2022-06-18-10-29-38-2"
new file mode 100644--- /dev/null
+++ "b/exp/log/log-train-2022-06-18-10-29-38-2"
@@ -0,0 +1,2737 @@
+2022-06-18 10:29:38,930 INFO [train.py:963] (2/4) Training started
+2022-06-18 10:29:38,930 INFO [train.py:973] (2/4) Device: cuda:2
+2022-06-18 10:29:39,185 INFO [lexicon.py:176] (2/4) Loading pre-compiled data/lang_char/Linv.pt
+2022-06-18 10:29:39,213 INFO [train.py:985] (2/4) {'best_train_loss': inf, 'best_valid_loss': inf, 'best_train_epoch': -1, 'best_valid_epoch': -1, 'batch_idx_train': 0, 'log_interval': 50, 'reset_interval': 200, 'valid_interval': 1000, 'feature_dim': 80, 'subsampling_factor': 4, 'model_warm_step': 3000, 'env_info': {'k2-version': '1.15.1', 'k2-build-type': 'Release', 'k2-with-cuda': True, 'k2-git-sha1': 'f8d2dba06c000ffee36aab5b66f24e7c9809f116', 'k2-git-date': 'Thu Apr 21 12:20:34 2022', 'lhotse-version': '1.3.0.dev+missing.version.file', 'torch-version': '1.10.0+cu102', 'torch-cuda-available': True, 'torch-cuda-version': '10.2', 'python-version': '3.8', 'icefall-git-branch': 'pruned-rnnt-aishell', 'icefall-git-sha1': 'd0a5f1d-dirty', 'icefall-git-date': 'Mon Jun 13 20:40:46 2022', 'icefall-path': '/k2-dev/fangjun/open-source/icefall-aishell', 'k2-path': '/ceph-fj/fangjun/open-source-2/k2-multi-22/k2/python/k2/__init__.py', 'lhotse-path': '/ceph-fj/fangjun/open-source-2/lhotse-jsonl/lhotse/__init__.py', 'hostname': 'de-74279-k2-train-7-0616225511-78bf4545d8-tv52r', 'IP address': '10.177.77.9'}, 'world_size': 4, 'master_port': 12356, 'tensorboard': True, 'num_epochs': 30, 'start_epoch': 1, 'start_batch': 0, 'exp_dir': PosixPath('pruned_transducer_stateless3/exp-context-size-1'), 'lang_dir': PosixPath('data/lang_char'), 'initial_lr': 0.003, 'lr_batches': 5000, 'lr_epochs': 6, 'context_size': 1, 'prune_range': 5, 'lm_scale': 0.25, 'am_scale': 0.0, 'simple_loss_scale': 0.5, 'seed': 42, 'print_diagnostics': False, 'save_every_n': 4000, 'keep_last_k': 30, 'average_period': 100, 'use_fp16': True, 'datatang_prob': 0.5, 'num_encoder_layers': 12, 'dim_feedforward': 2048, 'nhead': 8, 'encoder_dim': 512, 'decoder_dim': 512, 'joiner_dim': 512, 'max_duration': 200, 'bucketing_sampler': True, 'num_buckets': 30, 'shuffle': True, 'return_cuts': True, 'num_workers': 2, 'enable_spec_aug': True, 'spec_aug_time_warp_factor': 80, 'enable_musan': True, 'manifest_dir': PosixPath('data/fbank'), 'on_the_fly_feats': False, 'blank_id': 0, 'vocab_size': 4336}
+2022-06-18 10:29:39,214 INFO [train.py:987] (2/4) About to create model
+2022-06-18 10:29:39,965 INFO [train.py:991] (2/4) Number of model parameters: 96983734
+2022-06-18 10:29:44,914 INFO [train.py:1006] (2/4) Using DDP
+2022-06-18 10:29:46,048 INFO [aishell.py:39] (2/4) About to get train cuts from data/fbank/aishell_cuts_train.jsonl.gz
+2022-06-18 10:29:46,051 INFO [aidatatang_200zh.py:39] (2/4) About to get train cuts from data/fbank/aidatatang_cuts_train.jsonl.gz
+2022-06-18 10:29:47,936 INFO [asr_datamodule.py:163] (2/4) Enable MUSAN
+2022-06-18 10:29:47,936 INFO [asr_datamodule.py:175] (2/4) Enable SpecAugment
+2022-06-18 10:29:47,937 INFO [asr_datamodule.py:176] (2/4) Time warp factor: 80
+2022-06-18 10:29:47,937 INFO [asr_datamodule.py:188] (2/4) Num frame mask: 10
+2022-06-18 10:29:47,937 INFO [asr_datamodule.py:201] (2/4) About to create train dataset
+2022-06-18 10:29:47,937 INFO [asr_datamodule.py:229] (2/4) Using DynamicBucketingSampler.
+2022-06-18 10:29:50,229 INFO [asr_datamodule.py:238] (2/4) About to create train dataloader
+2022-06-18 10:29:50,229 INFO [asr_datamodule.py:163] (2/4) Enable MUSAN
+2022-06-18 10:29:50,230 INFO [asr_datamodule.py:175] (2/4) Enable SpecAugment
+2022-06-18 10:29:50,230 INFO [asr_datamodule.py:176] (2/4) Time warp factor: 80
+2022-06-18 10:29:50,230 INFO [asr_datamodule.py:188] (2/4) Num frame mask: 10
+2022-06-18 10:29:50,230 INFO [asr_datamodule.py:201] (2/4) About to create train dataset
+2022-06-18 10:29:50,230 INFO [asr_datamodule.py:229] (2/4) Using DynamicBucketingSampler.
+2022-06-18 10:29:53,177 INFO [asr_datamodule.py:238] (2/4) About to create train dataloader
+2022-06-18 10:29:53,178 INFO [aishell.py:45] (2/4) About to get valid cuts from data/fbank/aishell_cuts_dev.jsonl.gz
+2022-06-18 10:29:53,179 INFO [asr_datamodule.py:251] (2/4) About to create dev dataset
+2022-06-18 10:29:53,639 INFO [asr_datamodule.py:270] (2/4) About to create dev dataloader
+2022-06-18 10:29:53,640 INFO [train.py:1171] (2/4) Sanity check -- see if any of the batches in epoch 1 would cause OOM.
+2022-06-18 10:31:15,568 INFO [train.py:1081] (2/4) start training from epoch 1
+2022-06-18 10:32:07,199 INFO [train.py:874] (2/4) Epoch 1, batch 50, aishell_loss[loss=0.5064, simple_loss=1.013, pruned_loss=9.16, over 4934.00 frames.], tot_loss[loss=1.383, simple_loss=2.766, pruned_loss=8.738, over 218322.10 frames.], batch size: 64, aishell_tot_loss[loss=0.6821, simple_loss=1.364, pruned_loss=8.851, over 115903.19 frames.], datatang_tot_loss[loss=2.14, simple_loss=4.28, pruned_loss=8.638, over 116076.76 frames.], batch size: 64, lr: 3.00e-03
+2022-06-18 10:32:38,259 INFO [train.py:874] (2/4) Epoch 1, batch 100, datatang_loss[loss=0.3697, simple_loss=0.7395, pruned_loss=8.464, over 4938.00 frames.], tot_loss[loss=0.839, simple_loss=1.678, pruned_loss=8.72, over 388353.06 frames.], batch size: 50, aishell_tot_loss[loss=0.5423, simple_loss=1.085, pruned_loss=8.99, over 218418.04 frames.], datatang_tot_loss[loss=1.224, simple_loss=2.447, pruned_loss=8.461, over 218361.74 frames.], batch size: 50, lr: 3.00e-03
+2022-06-18 10:33:05,801 INFO [train.py:874] (2/4) Epoch 1, batch 150, aishell_loss[loss=0.2821, simple_loss=0.5642, pruned_loss=8.873, over 4781.00 frames.], tot_loss[loss=0.6427, simple_loss=1.285, pruned_loss=8.75, over 520694.17 frames.], batch size: 21, aishell_tot_loss[loss=0.4856, simple_loss=0.9713, pruned_loss=9.028, over 305299.97 frames.], datatang_tot_loss[loss=0.8888, simple_loss=1.778, pruned_loss=8.478, over 312133.16 frames.], batch size: 21, lr: 3.00e-03
+2022-06-18 10:33:36,984 INFO [train.py:874] (2/4) Epoch 1, batch 200, datatang_loss[loss=0.3563, simple_loss=0.7125, pruned_loss=8.358, over 4954.00 frames.], tot_loss[loss=0.5429, simple_loss=1.086, pruned_loss=8.766, over 623468.93 frames.], batch size: 91, aishell_tot_loss[loss=0.4452, simple_loss=0.8903, pruned_loss=9.08, over 396975.20 frames.], datatang_tot_loss[loss=0.7483, simple_loss=1.497, pruned_loss=8.423, over 379481.99 frames.], batch size: 91, lr: 3.00e-03
+2022-06-18 10:34:08,979 INFO [train.py:874] (2/4) Epoch 1, batch 250, aishell_loss[loss=0.3452, simple_loss=0.6904, pruned_loss=9.088, over 4971.00 frames.], tot_loss[loss=0.4793, simple_loss=0.9585, pruned_loss=8.723, over 704016.73 frames.], batch size: 48, aishell_tot_loss[loss=0.4231, simple_loss=0.8463, pruned_loss=9.062, over 461096.72 frames.], datatang_tot_loss[loss=0.6331, simple_loss=1.266, pruned_loss=8.391, over 456455.07 frames.], batch size: 48, lr: 3.00e-03
+2022-06-18 10:34:36,713 INFO [train.py:874] (2/4) Epoch 1, batch 300, datatang_loss[loss=0.3081, simple_loss=0.6161, pruned_loss=9.024, over 4930.00 frames.], tot_loss[loss=0.4378, simple_loss=0.8757, pruned_loss=8.769, over 766704.48 frames.], batch size: 83, aishell_tot_loss[loss=0.4072, simple_loss=0.8145, pruned_loss=9.051, over 523099.48 frames.], datatang_tot_loss[loss=0.5623, simple_loss=1.125, pruned_loss=8.472, over 518825.24 frames.], batch size: 83, lr: 3.00e-03
+2022-06-18 10:35:07,032 INFO [train.py:874] (2/4) Epoch 1, batch 350, aishell_loss[loss=0.3398, simple_loss=0.6796, pruned_loss=8.765, over 4924.00 frames.], tot_loss[loss=0.4046, simple_loss=0.8091, pruned_loss=8.815, over 815213.43 frames.], batch size: 41, aishell_tot_loss[loss=0.3935, simple_loss=0.7871, pruned_loss=9.025, over 569403.02 frames.], datatang_tot_loss[loss=0.503, simple_loss=1.006, pruned_loss=8.586, over 581859.90 frames.], batch size: 41, lr: 3.00e-03
+2022-06-18 10:35:37,474 INFO [train.py:874] (2/4) Epoch 1, batch 400, aishell_loss[loss=0.3395, simple_loss=0.679, pruned_loss=9.069, over 4946.00 frames.], tot_loss[loss=0.3823, simple_loss=0.7646, pruned_loss=8.83, over 853064.48 frames.], batch size: 56, aishell_tot_loss[loss=0.3822, simple_loss=0.7643, pruned_loss=9.003, over 618555.61 frames.], datatang_tot_loss[loss=0.466, simple_loss=0.932, pruned_loss=8.633, over 629369.88 frames.], batch size: 56, lr: 3.00e-03
+2022-06-18 10:36:05,194 INFO [train.py:874] (2/4) Epoch 1, batch 450, datatang_loss[loss=0.2664, simple_loss=0.5328, pruned_loss=8.813, over 4907.00 frames.], tot_loss[loss=0.363, simple_loss=0.726, pruned_loss=8.834, over 882313.54 frames.], batch size: 34, aishell_tot_loss[loss=0.3713, simple_loss=0.7425, pruned_loss=8.977, over 658339.06 frames.], datatang_tot_loss[loss=0.4346, simple_loss=0.8691, pruned_loss=8.673, over 674493.37 frames.], batch size: 34, lr: 2.99e-03
+2022-06-18 10:36:36,098 INFO [train.py:874] (2/4) Epoch 1, batch 500, aishell_loss[loss=0.3261, simple_loss=0.6523, pruned_loss=8.639, over 4948.00 frames.], tot_loss[loss=0.3503, simple_loss=0.7007, pruned_loss=8.859, over 905482.90 frames.], batch size: 64, aishell_tot_loss[loss=0.3631, simple_loss=0.7263, pruned_loss=8.964, over 701491.54 frames.], datatang_tot_loss[loss=0.4133, simple_loss=0.8267, pruned_loss=8.718, over 706956.28 frames.], batch size: 64, lr: 2.99e-03
+2022-06-18 10:37:05,651 INFO [train.py:874] (2/4) Epoch 1, batch 550, aishell_loss[loss=0.3346, simple_loss=0.6693, pruned_loss=8.924, over 4930.00 frames.], tot_loss[loss=0.3406, simple_loss=0.6812, pruned_loss=8.869, over 923001.06 frames.], batch size: 49, aishell_tot_loss[loss=0.3572, simple_loss=0.7145, pruned_loss=8.964, over 735939.20 frames.], datatang_tot_loss[loss=0.3947, simple_loss=0.7894, pruned_loss=8.735, over 738540.03 frames.], batch size: 49, lr: 2.99e-03
+2022-06-18 10:37:34,687 INFO [train.py:874] (2/4) Epoch 1, batch 600, datatang_loss[loss=0.2781, simple_loss=0.5562, pruned_loss=9.015, over 4921.00 frames.], tot_loss[loss=0.3322, simple_loss=0.6644, pruned_loss=8.893, over 936936.37 frames.], batch size: 83, aishell_tot_loss[loss=0.3521, simple_loss=0.7042, pruned_loss=8.954, over 761984.69 frames.], datatang_tot_loss[loss=0.3776, simple_loss=0.7551, pruned_loss=8.786, over 770971.02 frames.], batch size: 83, lr: 2.99e-03
+2022-06-18 10:38:06,367 INFO [train.py:874] (2/4) Epoch 1, batch 650, aishell_loss[loss=0.3239, simple_loss=0.6477, pruned_loss=8.95, over 4970.00 frames.], tot_loss[loss=0.3256, simple_loss=0.6511, pruned_loss=8.896, over 947735.03 frames.], batch size: 51, aishell_tot_loss[loss=0.3466, simple_loss=0.6933, pruned_loss=8.946, over 791341.69 frames.], datatang_tot_loss[loss=0.3654, simple_loss=0.7309, pruned_loss=8.802, over 793330.70 frames.], batch size: 51, lr: 2.99e-03
+2022-06-18 10:38:34,692 INFO [train.py:874] (2/4) Epoch 1, batch 700, datatang_loss[loss=0.2988, simple_loss=0.5977, pruned_loss=9.056, over 4934.00 frames.], tot_loss[loss=0.3198, simple_loss=0.6396, pruned_loss=8.902, over 956421.23 frames.], batch size: 79, aishell_tot_loss[loss=0.343, simple_loss=0.6861, pruned_loss=8.944, over 813651.27 frames.], datatang_tot_loss[loss=0.3532, simple_loss=0.7064, pruned_loss=8.818, over 816845.33 frames.], batch size: 79, lr: 2.99e-03
+2022-06-18 10:39:03,376 INFO [train.py:874] (2/4) Epoch 1, batch 750, aishell_loss[loss=0.3272, simple_loss=0.6544, pruned_loss=9.128, over 4950.00 frames.], tot_loss[loss=0.3138, simple_loss=0.6275, pruned_loss=8.918, over 962784.69 frames.], batch size: 45, aishell_tot_loss[loss=0.3387, simple_loss=0.6775, pruned_loss=8.947, over 832931.02 frames.], datatang_tot_loss[loss=0.3422, simple_loss=0.6844, pruned_loss=8.841, over 837562.05 frames.], batch size: 45, lr: 2.98e-03
+2022-06-18 10:39:33,598 INFO [train.py:874] (2/4) Epoch 1, batch 800, datatang_loss[loss=0.2814, simple_loss=0.5629, pruned_loss=9.14, over 4924.00 frames.], tot_loss[loss=0.3093, simple_loss=0.6186, pruned_loss=8.927, over 967698.60 frames.], batch size: 57, aishell_tot_loss[loss=0.3358, simple_loss=0.6715, pruned_loss=8.94, over 846689.07 frames.], datatang_tot_loss[loss=0.3327, simple_loss=0.6653, pruned_loss=8.868, over 858823.08 frames.], batch size: 57, lr: 2.98e-03
+2022-06-18 10:40:05,971 INFO [train.py:874] (2/4) Epoch 1, batch 850, datatang_loss[loss=0.2824, simple_loss=0.5648, pruned_loss=8.897, over 4898.00 frames.], tot_loss[loss=0.3048, simple_loss=0.6095, pruned_loss=8.926, over 971829.02 frames.], batch size: 42, aishell_tot_loss[loss=0.3328, simple_loss=0.6656, pruned_loss=8.932, over 860607.51 frames.], datatang_tot_loss[loss=0.3238, simple_loss=0.6476, pruned_loss=8.882, over 876112.24 frames.], batch size: 42, lr: 2.98e-03
+2022-06-18 10:40:33,803 INFO [train.py:874] (2/4) Epoch 1, batch 900, datatang_loss[loss=0.2606, simple_loss=0.5212, pruned_loss=8.956, over 4976.00 frames.], tot_loss[loss=0.3016, simple_loss=0.6033, pruned_loss=8.937, over 974906.58 frames.], batch size: 60, aishell_tot_loss[loss=0.3293, simple_loss=0.6585, pruned_loss=8.928, over 875856.27 frames.], datatang_tot_loss[loss=0.3173, simple_loss=0.6347, pruned_loss=8.904, over 888564.82 frames.], batch size: 60, lr: 2.98e-03
+2022-06-18 10:41:04,975 INFO [train.py:874] (2/4) Epoch 1, batch 950, datatang_loss[loss=0.287, simple_loss=0.574, pruned_loss=9.168, over 4945.00 frames.], tot_loss[loss=0.2967, simple_loss=0.5934, pruned_loss=8.952, over 977037.62 frames.], batch size: 86, aishell_tot_loss[loss=0.3265, simple_loss=0.6531, pruned_loss=8.932, over 884555.14 frames.], datatang_tot_loss[loss=0.3093, simple_loss=0.6187, pruned_loss=8.923, over 903365.34 frames.], batch size: 86, lr: 2.97e-03
+2022-06-18 10:41:36,980 INFO [train.py:874] (2/4) Epoch 1, batch 1000, datatang_loss[loss=0.2982, simple_loss=0.5963, pruned_loss=9.147, over 4959.00 frames.], tot_loss[loss=0.2936, simple_loss=0.5871, pruned_loss=8.967, over 978557.79 frames.], batch size: 45, aishell_tot_loss[loss=0.3228, simple_loss=0.6456, pruned_loss=8.935, over 896021.25 frames.], datatang_tot_loss[loss=0.304, simple_loss=0.6079, pruned_loss=8.943, over 913086.97 frames.], batch size: 45, lr: 2.97e-03
+2022-06-18 10:41:36,980 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 10:41:53,016 INFO [train.py:914] (2/4) Epoch 1, validation: loss=9.395, simple_loss=0.6242, pruned_loss=9.083, over 1622729.00 frames.
+2022-06-18 10:42:24,212 INFO [train.py:874] (2/4) Epoch 1, batch 1050, aishell_loss[loss=0.3032, simple_loss=0.6065, pruned_loss=9.208, over 4925.00 frames.], tot_loss[loss=0.2896, simple_loss=0.5791, pruned_loss=8.987, over 980562.87 frames.], batch size: 41, aishell_tot_loss[loss=0.3176, simple_loss=0.6352, pruned_loss=8.943, over 908036.93 frames.], datatang_tot_loss[loss=0.2994, simple_loss=0.5987, pruned_loss=8.966, over 920849.92 frames.], batch size: 41, lr: 2.97e-03
+2022-06-18 10:42:52,050 INFO [train.py:874] (2/4) Epoch 1, batch 1100, datatang_loss[loss=0.2724, simple_loss=0.5449, pruned_loss=9.366, over 4955.00 frames.], tot_loss[loss=0.2853, simple_loss=0.5707, pruned_loss=9.021, over 981552.53 frames.], batch size: 99, aishell_tot_loss[loss=0.313, simple_loss=0.626, pruned_loss=8.955, over 917061.60 frames.], datatang_tot_loss[loss=0.2944, simple_loss=0.5889, pruned_loss=8.999, over 928445.05 frames.], batch size: 99, lr: 2.96e-03
+2022-06-18 10:43:24,144 INFO [train.py:874] (2/4) Epoch 1, batch 1150, aishell_loss[loss=0.2984, simple_loss=0.5968, pruned_loss=9.012, over 4933.00 frames.], tot_loss[loss=0.2813, simple_loss=0.5625, pruned_loss=9.037, over 982436.12 frames.], batch size: 68, aishell_tot_loss[loss=0.3092, simple_loss=0.6185, pruned_loss=8.959, over 924291.37 frames.], datatang_tot_loss[loss=0.2892, simple_loss=0.5783, pruned_loss=9.024, over 935892.24 frames.], batch size: 68, lr: 2.96e-03
+2022-06-18 10:43:56,381 INFO [train.py:874] (2/4) Epoch 1, batch 1200, aishell_loss[loss=0.2747, simple_loss=0.5493, pruned_loss=9.165, over 4987.00 frames.], tot_loss[loss=0.2767, simple_loss=0.5533, pruned_loss=9.047, over 982870.12 frames.], batch size: 38, aishell_tot_loss[loss=0.3034, simple_loss=0.6069, pruned_loss=8.962, over 932191.84 frames.], datatang_tot_loss[loss=0.2851, simple_loss=0.5703, pruned_loss=9.044, over 940922.71 frames.], batch size: 38, lr: 2.96e-03
+2022-06-18 10:44:24,359 INFO [train.py:874] (2/4) Epoch 1, batch 1250, datatang_loss[loss=0.2426, simple_loss=0.4853, pruned_loss=9.233, over 4973.00 frames.], tot_loss[loss=0.2711, simple_loss=0.5421, pruned_loss=9.064, over 983720.16 frames.], batch size: 60, aishell_tot_loss[loss=0.2982, simple_loss=0.5963, pruned_loss=8.968, over 938481.24 frames.], datatang_tot_loss[loss=0.28, simple_loss=0.5599, pruned_loss=9.067, over 946407.49 frames.], batch size: 60, lr: 2.95e-03
+2022-06-18 10:44:55,542 INFO [train.py:874] (2/4) Epoch 1, batch 1300, datatang_loss[loss=0.2354, simple_loss=0.4707, pruned_loss=9.211, over 4960.00 frames.], tot_loss[loss=0.2652, simple_loss=0.5305, pruned_loss=9.075, over 984000.30 frames.], batch size: 67, aishell_tot_loss[loss=0.2924, simple_loss=0.5847, pruned_loss=8.969, over 944282.82 frames.], datatang_tot_loss[loss=0.2749, simple_loss=0.5497, pruned_loss=9.089, over 950668.36 frames.], batch size: 67, lr: 2.95e-03
+2022-06-18 10:45:26,699 INFO [train.py:874] (2/4) Epoch 1, batch 1350, aishell_loss[loss=0.2177, simple_loss=0.4354, pruned_loss=8.876, over 4877.00 frames.], tot_loss[loss=0.2591, simple_loss=0.5182, pruned_loss=9.08, over 984486.93 frames.], batch size: 28, aishell_tot_loss[loss=0.286, simple_loss=0.572, pruned_loss=8.971, over 949506.69 frames.], datatang_tot_loss[loss=0.2699, simple_loss=0.5397, pruned_loss=9.105, over 954627.10 frames.], batch size: 28, lr: 2.95e-03
+2022-06-18 10:45:54,352 INFO [train.py:874] (2/4) Epoch 1, batch 1400, aishell_loss[loss=0.2461, simple_loss=0.4922, pruned_loss=9.062, over 4960.00 frames.], tot_loss[loss=0.2536, simple_loss=0.5073, pruned_loss=9.089, over 985083.79 frames.], batch size: 40, aishell_tot_loss[loss=0.2806, simple_loss=0.5611, pruned_loss=8.974, over 954045.09 frames.], datatang_tot_loss[loss=0.2647, simple_loss=0.5294, pruned_loss=9.124, over 958344.33 frames.], batch size: 40, lr: 2.94e-03
+2022-06-18 10:46:26,250 INFO [train.py:874] (2/4) Epoch 1, batch 1450, aishell_loss[loss=0.2215, simple_loss=0.443, pruned_loss=8.982, over 4867.00 frames.], tot_loss[loss=0.2479, simple_loss=0.4958, pruned_loss=9.097, over 985036.41 frames.], batch size: 36, aishell_tot_loss[loss=0.2751, simple_loss=0.5502, pruned_loss=8.975, over 957728.00 frames.], datatang_tot_loss[loss=0.2594, simple_loss=0.5188, pruned_loss=9.14, over 961430.62 frames.], batch size: 36, lr: 2.94e-03
+2022-06-18 10:46:57,999 INFO [train.py:874] (2/4) Epoch 1, batch 1500, datatang_loss[loss=0.2298, simple_loss=0.4595, pruned_loss=9.431, over 4959.00 frames.], tot_loss[loss=0.2428, simple_loss=0.4855, pruned_loss=9.102, over 985249.81 frames.], batch size: 91, aishell_tot_loss[loss=0.27, simple_loss=0.5401, pruned_loss=8.972, over 960937.27 frames.], datatang_tot_loss[loss=0.2543, simple_loss=0.5086, pruned_loss=9.155, over 964407.06 frames.], batch size: 91, lr: 2.94e-03
+2022-06-18 10:47:26,503 INFO [train.py:874] (2/4) Epoch 1, batch 1550, datatang_loss[loss=0.2233, simple_loss=0.4465, pruned_loss=9.337, over 4915.00 frames.], tot_loss[loss=0.2384, simple_loss=0.4769, pruned_loss=9.103, over 985459.28 frames.], batch size: 77, aishell_tot_loss[loss=0.2659, simple_loss=0.5319, pruned_loss=8.97, over 963655.91 frames.], datatang_tot_loss[loss=0.2492, simple_loss=0.4985, pruned_loss=9.166, over 967193.39 frames.], batch size: 77, lr: 2.93e-03
+2022-06-18 10:47:58,275 INFO [train.py:874] (2/4) Epoch 1, batch 1600, aishell_loss[loss=0.2265, simple_loss=0.4531, pruned_loss=8.831, over 4962.00 frames.], tot_loss[loss=0.234, simple_loss=0.468, pruned_loss=9.11, over 985745.33 frames.], batch size: 31, aishell_tot_loss[loss=0.2617, simple_loss=0.5234, pruned_loss=8.97, over 966015.02 frames.], datatang_tot_loss[loss=0.2444, simple_loss=0.4888, pruned_loss=9.179, over 969797.31 frames.], batch size: 31, lr: 2.93e-03
+2022-06-18 10:48:31,037 INFO [train.py:874] (2/4) Epoch 1, batch 1650, aishell_loss[loss=0.219, simple_loss=0.4379, pruned_loss=8.968, over 4962.00 frames.], tot_loss[loss=0.2306, simple_loss=0.4611, pruned_loss=9.107, over 986089.78 frames.], batch size: 40, aishell_tot_loss[loss=0.2572, simple_loss=0.5143, pruned_loss=8.969, over 968658.19 frames.], datatang_tot_loss[loss=0.2405, simple_loss=0.481, pruned_loss=9.187, over 971742.66 frames.], batch size: 40, lr: 2.92e-03
+2022-06-18 10:48:59,345 INFO [train.py:874] (2/4) Epoch 1, batch 1700, aishell_loss[loss=0.1905, simple_loss=0.381, pruned_loss=8.719, over 4878.00 frames.], tot_loss[loss=0.2251, simple_loss=0.4503, pruned_loss=9.099, over 986019.18 frames.], batch size: 28, aishell_tot_loss[loss=0.2513, simple_loss=0.5025, pruned_loss=8.962, over 970718.84 frames.], datatang_tot_loss[loss=0.2362, simple_loss=0.4724, pruned_loss=9.194, over 973357.61 frames.], batch size: 28, lr: 2.92e-03
+2022-06-18 10:49:32,460 INFO [train.py:874] (2/4) Epoch 1, batch 1750, aishell_loss[loss=0.2253, simple_loss=0.4506, pruned_loss=8.864, over 4945.00 frames.], tot_loss[loss=0.2218, simple_loss=0.4436, pruned_loss=9.106, over 985578.24 frames.], batch size: 54, aishell_tot_loss[loss=0.2477, simple_loss=0.4953, pruned_loss=8.955, over 971777.72 frames.], datatang_tot_loss[loss=0.2322, simple_loss=0.4644, pruned_loss=9.206, over 975021.22 frames.], batch size: 54, lr: 2.91e-03
+2022-06-18 10:50:05,236 INFO [train.py:874] (2/4) Epoch 1, batch 1800, datatang_loss[loss=0.197, simple_loss=0.3941, pruned_loss=9.188, over 4934.00 frames.], tot_loss[loss=0.2179, simple_loss=0.4358, pruned_loss=9.103, over 985295.08 frames.], batch size: 79, aishell_tot_loss[loss=0.2435, simple_loss=0.487, pruned_loss=8.954, over 972965.58 frames.], datatang_tot_loss[loss=0.2279, simple_loss=0.4558, pruned_loss=9.21, over 976392.14 frames.], batch size: 79, lr: 2.91e-03
+2022-06-18 10:50:33,703 INFO [train.py:874] (2/4) Epoch 1, batch 1850, aishell_loss[loss=0.2169, simple_loss=0.4338, pruned_loss=8.914, over 4918.00 frames.], tot_loss[loss=0.215, simple_loss=0.4301, pruned_loss=9.101, over 985405.35 frames.], batch size: 68, aishell_tot_loss[loss=0.2405, simple_loss=0.481, pruned_loss=8.951, over 974135.58 frames.], datatang_tot_loss[loss=0.2239, simple_loss=0.4478, pruned_loss=9.211, over 977776.15 frames.], batch size: 68, lr: 2.91e-03
+2022-06-18 10:51:04,557 INFO [train.py:874] (2/4) Epoch 1, batch 1900, aishell_loss[loss=0.2233, simple_loss=0.4466, pruned_loss=8.9, over 4963.00 frames.], tot_loss[loss=0.2121, simple_loss=0.4241, pruned_loss=9.084, over 984854.53 frames.], batch size: 40, aishell_tot_loss[loss=0.2362, simple_loss=0.4724, pruned_loss=8.942, over 974922.70 frames.], datatang_tot_loss[loss=0.2209, simple_loss=0.4418, pruned_loss=9.208, over 978644.19 frames.], batch size: 40, lr: 2.90e-03
+2022-06-18 10:51:36,220 INFO [train.py:874] (2/4) Epoch 1, batch 1950, datatang_loss[loss=0.1769, simple_loss=0.3538, pruned_loss=9.18, over 4935.00 frames.], tot_loss[loss=0.2091, simple_loss=0.4181, pruned_loss=9.085, over 984814.71 frames.], batch size: 69, aishell_tot_loss[loss=0.2326, simple_loss=0.4652, pruned_loss=8.941, over 975864.62 frames.], datatang_tot_loss[loss=0.2176, simple_loss=0.4352, pruned_loss=9.21, over 979525.47 frames.], batch size: 69, lr: 2.90e-03
+2022-06-18 10:52:03,736 INFO [train.py:874] (2/4) Epoch 1, batch 2000, aishell_loss[loss=0.2002, simple_loss=0.4005, pruned_loss=8.924, over 4942.00 frames.], tot_loss[loss=0.2058, simple_loss=0.4116, pruned_loss=9.083, over 984924.08 frames.], batch size: 45, aishell_tot_loss[loss=0.2284, simple_loss=0.4569, pruned_loss=8.939, over 976865.33 frames.], datatang_tot_loss[loss=0.2144, simple_loss=0.4289, pruned_loss=9.212, over 980310.01 frames.], batch size: 45, lr: 2.89e-03
+2022-06-18 10:52:03,737 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 10:52:20,039 INFO [train.py:914] (2/4) Epoch 1, validation: loss=8.987, simple_loss=0.3601, pruned_loss=8.807, over 1622729.00 frames.
+2022-06-18 10:52:48,758 INFO [train.py:874] (2/4) Epoch 1, batch 2050, datatang_loss[loss=0.1945, simple_loss=0.389, pruned_loss=9.355, over 4924.00 frames.], tot_loss[loss=0.2031, simple_loss=0.4061, pruned_loss=9.074, over 984913.99 frames.], batch size: 81, aishell_tot_loss[loss=0.225, simple_loss=0.4501, pruned_loss=8.93, over 977566.14 frames.], datatang_tot_loss[loss=0.2114, simple_loss=0.4229, pruned_loss=9.208, over 981028.51 frames.], batch size: 81, lr: 2.89e-03
+2022-06-18 10:53:19,255 INFO [train.py:874] (2/4) Epoch 1, batch 2100, datatang_loss[loss=0.1689, simple_loss=0.3379, pruned_loss=9.112, over 4909.00 frames.], tot_loss[loss=0.2, simple_loss=0.4, pruned_loss=9.063, over 984993.97 frames.], batch size: 47, aishell_tot_loss[loss=0.2207, simple_loss=0.4415, pruned_loss=8.923, over 978360.68 frames.], datatang_tot_loss[loss=0.2088, simple_loss=0.4176, pruned_loss=9.205, over 981673.63 frames.], batch size: 47, lr: 2.88e-03
+2022-06-18 10:53:51,102 INFO [train.py:874] (2/4) Epoch 1, batch 2150, aishell_loss[loss=0.2084, simple_loss=0.4168, pruned_loss=8.895, over 4895.00 frames.], tot_loss[loss=0.1991, simple_loss=0.3982, pruned_loss=9.06, over 985289.68 frames.], batch size: 34, aishell_tot_loss[loss=0.2186, simple_loss=0.4371, pruned_loss=8.918, over 979170.10 frames.], datatang_tot_loss[loss=0.2066, simple_loss=0.4132, pruned_loss=9.203, over 982300.99 frames.], batch size: 34, lr: 2.88e-03
+2022-06-18 10:54:18,517 INFO [train.py:874] (2/4) Epoch 1, batch 2200, aishell_loss[loss=0.1849, simple_loss=0.3698, pruned_loss=8.975, over 4912.00 frames.], tot_loss[loss=0.1977, simple_loss=0.3954, pruned_loss=9.052, over 985403.82 frames.], batch size: 52, aishell_tot_loss[loss=0.2161, simple_loss=0.4322, pruned_loss=8.913, over 979910.25 frames.], datatang_tot_loss[loss=0.2045, simple_loss=0.4089, pruned_loss=9.199, over 982760.73 frames.], batch size: 52, lr: 2.87e-03
+2022-06-18 10:54:50,317 INFO [train.py:874] (2/4) Epoch 1, batch 2250, aishell_loss[loss=0.2024, simple_loss=0.4048, pruned_loss=8.906, over 4930.00 frames.], tot_loss[loss=0.1966, simple_loss=0.3932, pruned_loss=9.033, over 985678.68 frames.], batch size: 33, aishell_tot_loss[loss=0.2126, simple_loss=0.4251, pruned_loss=8.905, over 980873.08 frames.], datatang_tot_loss[loss=0.2035, simple_loss=0.4069, pruned_loss=9.197, over 983146.57 frames.], batch size: 33, lr: 2.86e-03
+2022-06-18 10:55:21,551 INFO [train.py:874] (2/4) Epoch 1, batch 2300, aishell_loss[loss=0.2048, simple_loss=0.4097, pruned_loss=8.895, over 4970.00 frames.], tot_loss[loss=0.195, simple_loss=0.3901, pruned_loss=9.03, over 986029.77 frames.], batch size: 69, aishell_tot_loss[loss=0.21, simple_loss=0.4201, pruned_loss=8.898, over 981605.74 frames.], datatang_tot_loss[loss=0.2016, simple_loss=0.4032, pruned_loss=9.197, over 983657.06 frames.], batch size: 69, lr: 2.86e-03
+2022-06-18 10:55:50,103 INFO [train.py:874] (2/4) Epoch 1, batch 2350, aishell_loss[loss=0.2021, simple_loss=0.4043, pruned_loss=8.874, over 4964.00 frames.], tot_loss[loss=0.1924, simple_loss=0.3849, pruned_loss=9.031, over 986303.66 frames.], batch size: 48, aishell_tot_loss[loss=0.2081, simple_loss=0.4163, pruned_loss=8.894, over 982129.13 frames.], datatang_tot_loss[loss=0.1983, simple_loss=0.3966, pruned_loss=9.193, over 984175.66 frames.], batch size: 48, lr: 2.85e-03
+2022-06-18 10:56:21,660 INFO [train.py:874] (2/4) Epoch 1, batch 2400, datatang_loss[loss=0.2245, simple_loss=0.4489, pruned_loss=9.277, over 4925.00 frames.], tot_loss[loss=0.1907, simple_loss=0.3813, pruned_loss=9.025, over 986066.50 frames.], batch size: 108, aishell_tot_loss[loss=0.2053, simple_loss=0.4105, pruned_loss=8.887, over 982613.93 frames.], datatang_tot_loss[loss=0.1966, simple_loss=0.3933, pruned_loss=9.194, over 984199.11 frames.], batch size: 108, lr: 2.85e-03
+2022-06-18 10:56:52,217 INFO [train.py:874] (2/4) Epoch 1, batch 2450, datatang_loss[loss=0.1703, simple_loss=0.3406, pruned_loss=9.083, over 4955.00 frames.], tot_loss[loss=0.1895, simple_loss=0.3791, pruned_loss=9.019, over 986245.63 frames.], batch size: 25, aishell_tot_loss[loss=0.2025, simple_loss=0.405, pruned_loss=8.881, over 983298.32 frames.], datatang_tot_loss[loss=0.1956, simple_loss=0.3913, pruned_loss=9.195, over 984353.73 frames.], batch size: 25, lr: 2.84e-03
+2022-06-18 10:57:21,457 INFO [train.py:874] (2/4) Epoch 1, batch 2500, datatang_loss[loss=0.1768, simple_loss=0.3536, pruned_loss=9.213, over 4915.00 frames.], tot_loss[loss=0.1877, simple_loss=0.3755, pruned_loss=9.022, over 985860.45 frames.], batch size: 75, aishell_tot_loss[loss=0.2004, simple_loss=0.4008, pruned_loss=8.875, over 983340.75 frames.], datatang_tot_loss[loss=0.1937, simple_loss=0.3874, pruned_loss=9.194, over 984466.54 frames.], batch size: 75, lr: 2.84e-03
+2022-06-18 10:57:53,490 INFO [train.py:874] (2/4) Epoch 1, batch 2550, aishell_loss[loss=0.1722, simple_loss=0.3444, pruned_loss=8.827, over 4913.00 frames.], tot_loss[loss=0.186, simple_loss=0.3721, pruned_loss=9.018, over 985828.81 frames.], batch size: 41, aishell_tot_loss[loss=0.1978, simple_loss=0.3957, pruned_loss=8.873, over 983680.86 frames.], datatang_tot_loss[loss=0.1922, simple_loss=0.3843, pruned_loss=9.195, over 984562.46 frames.], batch size: 41, lr: 2.83e-03
+2022-06-18 10:58:22,414 INFO [train.py:874] (2/4) Epoch 1, batch 2600, datatang_loss[loss=0.1591, simple_loss=0.3182, pruned_loss=9.165, over 4884.00 frames.], tot_loss[loss=0.1844, simple_loss=0.3688, pruned_loss=9.011, over 985272.03 frames.], batch size: 52, aishell_tot_loss[loss=0.1957, simple_loss=0.3914, pruned_loss=8.866, over 983450.67 frames.], datatang_tot_loss[loss=0.1903, simple_loss=0.3807, pruned_loss=9.193, over 984616.45 frames.], batch size: 52, lr: 2.83e-03
+2022-06-18 10:58:52,575 INFO [train.py:874] (2/4) Epoch 1, batch 2650, datatang_loss[loss=0.1634, simple_loss=0.3269, pruned_loss=9.149, over 4943.00 frames.], tot_loss[loss=0.183, simple_loss=0.366, pruned_loss=9.014, over 985321.66 frames.], batch size: 34, aishell_tot_loss[loss=0.1948, simple_loss=0.3896, pruned_loss=8.861, over 983627.21 frames.], datatang_tot_loss[loss=0.1879, simple_loss=0.3757, pruned_loss=9.189, over 984761.02 frames.], batch size: 34, lr: 2.82e-03
+2022-06-18 10:59:23,853 INFO [train.py:874] (2/4) Epoch 1, batch 2700, aishell_loss[loss=0.1756, simple_loss=0.3512, pruned_loss=8.83, over 4977.00 frames.], tot_loss[loss=0.1811, simple_loss=0.3623, pruned_loss=9.01, over 985536.69 frames.], batch size: 30, aishell_tot_loss[loss=0.1929, simple_loss=0.3859, pruned_loss=8.852, over 984123.41 frames.], datatang_tot_loss[loss=0.1858, simple_loss=0.3716, pruned_loss=9.19, over 984752.45 frames.], batch size: 30, lr: 2.81e-03
+2022-06-18 10:59:52,354 INFO [train.py:874] (2/4) Epoch 1, batch 2750, aishell_loss[loss=0.1755, simple_loss=0.3509, pruned_loss=8.784, over 4973.00 frames.], tot_loss[loss=0.1797, simple_loss=0.3594, pruned_loss=8.997, over 985631.11 frames.], batch size: 39, aishell_tot_loss[loss=0.1913, simple_loss=0.3826, pruned_loss=8.846, over 984219.98 frames.], datatang_tot_loss[loss=0.1839, simple_loss=0.3679, pruned_loss=9.183, over 985023.29 frames.], batch size: 39, lr: 2.81e-03
+2022-06-18 11:00:23,029 INFO [train.py:874] (2/4) Epoch 1, batch 2800, aishell_loss[loss=0.1583, simple_loss=0.3165, pruned_loss=8.777, over 4866.00 frames.], tot_loss[loss=0.1785, simple_loss=0.3569, pruned_loss=8.99, over 985423.47 frames.], batch size: 35, aishell_tot_loss[loss=0.1889, simple_loss=0.3779, pruned_loss=8.839, over 984064.41 frames.], datatang_tot_loss[loss=0.1831, simple_loss=0.3662, pruned_loss=9.181, over 985209.22 frames.], batch size: 35, lr: 2.80e-03
+2022-06-18 11:00:54,637 INFO [train.py:874] (2/4) Epoch 1, batch 2850, aishell_loss[loss=0.1856, simple_loss=0.3711, pruned_loss=8.83, over 4858.00 frames.], tot_loss[loss=0.1775, simple_loss=0.355, pruned_loss=8.982, over 985361.17 frames.], batch size: 36, aishell_tot_loss[loss=0.1871, simple_loss=0.3742, pruned_loss=8.833, over 984090.94 frames.], datatang_tot_loss[loss=0.182, simple_loss=0.3641, pruned_loss=9.179, over 985340.36 frames.], batch size: 36, lr: 2.80e-03
+2022-06-18 11:01:22,288 INFO [train.py:874] (2/4) Epoch 1, batch 2900, datatang_loss[loss=0.2125, simple_loss=0.425, pruned_loss=9.298, over 4954.00 frames.], tot_loss[loss=0.1782, simple_loss=0.3564, pruned_loss=8.989, over 985442.64 frames.], batch size: 99, aishell_tot_loss[loss=0.1864, simple_loss=0.3729, pruned_loss=8.827, over 984312.59 frames.], datatang_tot_loss[loss=0.1817, simple_loss=0.3635, pruned_loss=9.185, over 985345.85 frames.], batch size: 99, lr: 2.79e-03
+2022-06-18 11:01:53,457 INFO [train.py:874] (2/4) Epoch 1, batch 2950, datatang_loss[loss=0.1979, simple_loss=0.3959, pruned_loss=9.022, over 4968.00 frames.], tot_loss[loss=0.1775, simple_loss=0.3549, pruned_loss=8.993, over 985674.08 frames.], batch size: 45, aishell_tot_loss[loss=0.185, simple_loss=0.37, pruned_loss=8.821, over 984458.84 frames.], datatang_tot_loss[loss=0.1811, simple_loss=0.3622, pruned_loss=9.184, over 985572.71 frames.], batch size: 45, lr: 2.78e-03
+2022-06-18 11:02:24,451 INFO [train.py:874] (2/4) Epoch 1, batch 3000, aishell_loss[loss=9.028, simple_loss=0.3571, pruned_loss=8.85, over 4967.00 frames.], tot_loss[loss=0.22, simple_loss=0.3508, pruned_loss=8.988, over 985116.18 frames.], batch size: 64, aishell_tot_loss[loss=0.228, simple_loss=0.3667, pruned_loss=8.814, over 984365.61 frames.], datatang_tot_loss[loss=0.1792, simple_loss=0.3585, pruned_loss=9.184, over 985228.42 frames.], batch size: 64, lr: 2.78e-03
+2022-06-18 11:02:24,452 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 11:02:40,210 INFO [train.py:914] (2/4) Epoch 1, validation: loss=4.481, simple_loss=0.3323, pruned_loss=4.315, over 1622729.00 frames.
+2022-06-18 11:03:11,375 INFO [train.py:874] (2/4) Epoch 1, batch 3050, aishell_loss[loss=0.2528, simple_loss=0.3392, pruned_loss=0.8317, over 4953.00 frames.], tot_loss[loss=0.2423, simple_loss=0.3619, pruned_loss=7.261, over 985258.69 frames.], batch size: 31, aishell_tot_loss[loss=0.2377, simple_loss=0.3724, pruned_loss=7.713, over 984643.10 frames.], datatang_tot_loss[loss=0.1966, simple_loss=0.3619, pruned_loss=8.465, over 985196.23 frames.], batch size: 31, lr: 2.77e-03
+2022-06-18 11:03:41,565 INFO [train.py:874] (2/4) Epoch 1, batch 3100, datatang_loss[loss=0.2501, simple_loss=0.3664, pruned_loss=0.6694, over 4905.00 frames.], tot_loss[loss=0.243, simple_loss=0.3581, pruned_loss=5.813, over 985416.72 frames.], batch size: 42, aishell_tot_loss[loss=0.239, simple_loss=0.3699, pruned_loss=7.018, over 984753.41 frames.], datatang_tot_loss[loss=0.2028, simple_loss=0.3593, pruned_loss=7.414, over 985330.94 frames.], batch size: 42, lr: 2.77e-03
+2022-06-18 11:04:10,466 INFO [train.py:874] (2/4) Epoch 1, batch 3150, aishell_loss[loss=0.2626, simple_loss=0.4076, pruned_loss=0.5883, over 4903.00 frames.], tot_loss[loss=0.2392, simple_loss=0.3543, pruned_loss=4.646, over 985514.15 frames.], batch size: 34, aishell_tot_loss[loss=0.2376, simple_loss=0.3662, pruned_loss=6.174, over 984740.43 frames.], datatang_tot_loss[loss=0.2049, simple_loss=0.3572, pruned_loss=6.694, over 985547.61 frames.], batch size: 34, lr: 2.76e-03
+2022-06-18 11:04:41,703 INFO [train.py:874] (2/4) Epoch 1, batch 3200, datatang_loss[loss=0.1929, simple_loss=0.3153, pruned_loss=0.3526, over 4946.00 frames.], tot_loss[loss=0.2327, simple_loss=0.35, pruned_loss=3.709, over 985762.97 frames.], batch size: 45, aishell_tot_loss[loss=0.2351, simple_loss=0.3633, pruned_loss=5.575, over 984889.35 frames.], datatang_tot_loss[loss=0.2052, simple_loss=0.354, pruned_loss=5.871, over 985738.06 frames.], batch size: 45, lr: 2.75e-03
+2022-06-18 11:05:11,784 INFO [train.py:874] (2/4) Epoch 1, batch 3250, datatang_loss[loss=0.2068, simple_loss=0.3443, pruned_loss=0.3466, over 4872.00 frames.], tot_loss[loss=0.2273, simple_loss=0.3488, pruned_loss=2.968, over 985602.98 frames.], batch size: 25, aishell_tot_loss[loss=0.2325, simple_loss=0.3613, pruned_loss=5.056, over 984959.17 frames.], datatang_tot_loss[loss=0.2056, simple_loss=0.3528, pruned_loss=5.124, over 985596.95 frames.], batch size: 25, lr: 2.75e-03
+2022-06-18 11:05:40,179 INFO [train.py:874] (2/4) Epoch 1, batch 3300, datatang_loss[loss=0.175, simple_loss=0.2988, pruned_loss=0.2562, over 4941.00 frames.], tot_loss[loss=0.2224, simple_loss=0.3478, pruned_loss=2.383, over 985639.99 frames.], batch size: 69, aishell_tot_loss[loss=0.2305, simple_loss=0.3604, pruned_loss=4.546, over 984847.85 frames.], datatang_tot_loss[loss=0.2046, simple_loss=0.3507, pruned_loss=4.512, over 985813.63 frames.], batch size: 69, lr: 2.74e-03
+2022-06-18 11:06:11,124 INFO [train.py:874] (2/4) Epoch 1, batch 3350, datatang_loss[loss=0.2027, simple_loss=0.353, pruned_loss=0.262, over 4952.00 frames.], tot_loss[loss=0.2178, simple_loss=0.347, pruned_loss=1.92, over 985476.80 frames.], batch size: 86, aishell_tot_loss[loss=0.2268, simple_loss=0.3584, pruned_loss=3.99, over 984784.97 frames.], datatang_tot_loss[loss=0.2041, simple_loss=0.3499, pruned_loss=4.071, over 985785.17 frames.], batch size: 86, lr: 2.73e-03
+2022-06-18 11:06:40,526 INFO [train.py:874] (2/4) Epoch 1, batch 3400, datatang_loss[loss=0.2068, simple_loss=0.3598, pruned_loss=0.2694, over 4958.00 frames.], tot_loss[loss=0.2132, simple_loss=0.3456, pruned_loss=1.554, over 985257.40 frames.], batch size: 91, aishell_tot_loss[loss=0.2236, simple_loss=0.357, pruned_loss=3.521, over 984757.93 frames.], datatang_tot_loss[loss=0.2028, simple_loss=0.348, pruned_loss=3.657, over 985620.62 frames.], batch size: 91, lr: 2.73e-03
+2022-06-18 11:07:10,275 INFO [train.py:874] (2/4) Epoch 1, batch 3450, aishell_loss[loss=0.2121, simple_loss=0.3673, pruned_loss=0.2846, over 4946.00 frames.], tot_loss[loss=0.2094, simple_loss=0.344, pruned_loss=1.27, over 985013.43 frames.], batch size: 56, aishell_tot_loss[loss=0.2199, simple_loss=0.3549, pruned_loss=3.078, over 984827.95 frames.], datatang_tot_loss[loss=0.2021, simple_loss=0.3465, pruned_loss=3.321, over 985330.62 frames.], batch size: 56, lr: 2.72e-03
+2022-06-18 11:07:41,568 INFO [train.py:874] (2/4) Epoch 1, batch 3500, aishell_loss[loss=0.2032, simple_loss=0.3578, pruned_loss=0.2428, over 4955.00 frames.], tot_loss[loss=0.2065, simple_loss=0.3438, pruned_loss=1.043, over 985249.79 frames.], batch size: 56, aishell_tot_loss[loss=0.2171, simple_loss=0.3533, pruned_loss=2.769, over 985101.88 frames.], datatang_tot_loss[loss=0.2016, simple_loss=0.3464, pruned_loss=2.933, over 985288.77 frames.], batch size: 56, lr: 2.72e-03
+2022-06-18 11:08:10,086 INFO [train.py:874] (2/4) Epoch 1, batch 3550, aishell_loss[loss=0.2172, simple_loss=0.3785, pruned_loss=0.2797, over 4933.00 frames.], tot_loss[loss=0.2032, simple_loss=0.3422, pruned_loss=0.8633, over 985487.13 frames.], batch size: 64, aishell_tot_loss[loss=0.214, simple_loss=0.3516, pruned_loss=2.436, over 985279.87 frames.], datatang_tot_loss[loss=0.2003, simple_loss=0.3449, pruned_loss=2.651, over 985372.40 frames.], batch size: 64, lr: 2.71e-03
+2022-06-18 11:08:41,113 INFO [train.py:874] (2/4) Epoch 1, batch 3600, datatang_loss[loss=0.1709, simple_loss=0.3022, pruned_loss=0.1982, over 4922.00 frames.], tot_loss[loss=0.2005, simple_loss=0.3408, pruned_loss=0.7228, over 985637.01 frames.], batch size: 71, aishell_tot_loss[loss=0.2119, simple_loss=0.3503, pruned_loss=2.235, over 985264.29 frames.], datatang_tot_loss[loss=0.1989, simple_loss=0.3435, pruned_loss=2.303, over 985570.06 frames.], batch size: 71, lr: 2.70e-03
+2022-06-18 11:09:13,092 INFO [train.py:874] (2/4) Epoch 1, batch 3650, datatang_loss[loss=0.1749, simple_loss=0.3108, pruned_loss=0.1945, over 4927.00 frames.], tot_loss[loss=0.1978, simple_loss=0.3392, pruned_loss=0.6106, over 985683.68 frames.], batch size: 77, aishell_tot_loss[loss=0.2094, simple_loss=0.3483, pruned_loss=2.042, over 985200.00 frames.], datatang_tot_loss[loss=0.1978, simple_loss=0.3424, pruned_loss=2.012, over 985724.72 frames.], batch size: 77, lr: 2.70e-03
+2022-06-18 11:09:42,002 INFO [train.py:874] (2/4) Epoch 1, batch 3700, aishell_loss[loss=0.2024, simple_loss=0.3603, pruned_loss=0.2231, over 4966.00 frames.], tot_loss[loss=0.196, simple_loss=0.3384, pruned_loss=0.5238, over 985821.49 frames.], batch size: 61, aishell_tot_loss[loss=0.2068, simple_loss=0.3465, pruned_loss=1.834, over 985234.05 frames.], datatang_tot_loss[loss=0.1972, simple_loss=0.342, pruned_loss=1.794, over 985894.10 frames.], batch size: 61, lr: 2.69e-03
+2022-06-18 11:10:11,812 INFO [train.py:874] (2/4) Epoch 1, batch 3750, datatang_loss[loss=0.1792, simple_loss=0.3181, pruned_loss=0.2018, over 4918.00 frames.], tot_loss[loss=0.1939, simple_loss=0.3367, pruned_loss=0.4541, over 985762.76 frames.], batch size: 77, aishell_tot_loss[loss=0.2048, simple_loss=0.3451, pruned_loss=1.678, over 985177.93 frames.], datatang_tot_loss[loss=0.1957, simple_loss=0.3403, pruned_loss=1.574, over 985923.28 frames.], batch size: 77, lr: 2.68e-03
+2022-06-18 11:10:43,290 INFO [train.py:874] (2/4) Epoch 1, batch 3800, datatang_loss[loss=0.1848, simple_loss=0.3272, pruned_loss=0.2119, over 4917.00 frames.], tot_loss[loss=0.1926, simple_loss=0.3358, pruned_loss=0.4016, over 985817.03 frames.], batch size: 77, aishell_tot_loss[loss=0.2035, simple_loss=0.3443, pruned_loss=1.533, over 985314.23 frames.], datatang_tot_loss[loss=0.1943, simple_loss=0.3388, pruned_loss=1.389, over 985876.11 frames.], batch size: 77, lr: 2.68e-03
+2022-06-18 11:11:11,666 INFO [train.py:874] (2/4) Epoch 1, batch 3850, datatang_loss[loss=0.1875, simple_loss=0.3309, pruned_loss=0.2201, over 4929.00 frames.], tot_loss[loss=0.1911, simple_loss=0.3349, pruned_loss=0.3574, over 985539.05 frames.], batch size: 62, aishell_tot_loss[loss=0.2017, simple_loss=0.3437, pruned_loss=1.364, over 985122.84 frames.], datatang_tot_loss[loss=0.1929, simple_loss=0.3371, pruned_loss=1.26, over 985831.03 frames.], batch size: 62, lr: 2.67e-03
+2022-06-18 11:11:41,826 INFO [train.py:874] (2/4) Epoch 1, batch 3900, datatang_loss[loss=0.1774, simple_loss=0.3172, pruned_loss=0.1879, over 4925.00 frames.], tot_loss[loss=0.1893, simple_loss=0.333, pruned_loss=0.3221, over 985427.76 frames.], batch size: 83, aishell_tot_loss[loss=0.2, simple_loss=0.3424, pruned_loss=1.248, over 984788.55 frames.], datatang_tot_loss[loss=0.1914, simple_loss=0.3352, pruned_loss=1.117, over 986021.40 frames.], batch size: 83, lr: 2.66e-03
+2022-06-18 11:12:10,012 INFO [train.py:874] (2/4) Epoch 1, batch 3950, aishell_loss[loss=0.1708, simple_loss=0.3117, pruned_loss=0.1489, over 4965.00 frames.], tot_loss[loss=0.1882, simple_loss=0.3324, pruned_loss=0.2926, over 985422.10 frames.], batch size: 61, aishell_tot_loss[loss=0.1978, simple_loss=0.3409, pruned_loss=1.099, over 984731.94 frames.], datatang_tot_loss[loss=0.1907, simple_loss=0.3345, pruned_loss=1.029, over 986114.27 frames.], batch size: 61, lr: 2.66e-03
+2022-06-18 11:12:39,483 INFO [train.py:874] (2/4) Epoch 1, batch 4000, datatang_loss[loss=0.1886, simple_loss=0.3363, pruned_loss=0.2045, over 4934.00 frames.], tot_loss[loss=0.1864, simple_loss=0.3305, pruned_loss=0.2685, over 985460.17 frames.], batch size: 73, aishell_tot_loss[loss=0.1957, simple_loss=0.3392, pruned_loss=0.9824, over 984815.06 frames.], datatang_tot_loss[loss=0.1895, simple_loss=0.333, pruned_loss=0.9381, over 986081.72 frames.], batch size: 73, lr: 2.65e-03
+2022-06-18 11:12:39,484 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 11:13:00,453 INFO [train.py:914] (2/4) Epoch 1, validation: loss=0.2503, simple_loss=0.2882, pruned_loss=0.1062, over 1622729.00 frames.
+2022-06-18 11:13:28,274 INFO [train.py:874] (2/4) Epoch 1, batch 4050, aishell_loss[loss=0.1838, simple_loss=0.3315, pruned_loss=0.1805, over 4963.00 frames.], tot_loss[loss=0.1852, simple_loss=0.3289, pruned_loss=0.2512, over 985691.85 frames.], batch size: 61, aishell_tot_loss[loss=0.1942, simple_loss=0.3381, pruned_loss=0.8878, over 985051.48 frames.], datatang_tot_loss[loss=0.1881, simple_loss=0.331, pruned_loss=0.8511, over 986099.84 frames.], batch size: 61, lr: 2.64e-03
+2022-06-18 11:13:55,308 INFO [train.py:874] (2/4) Epoch 1, batch 4100, aishell_loss[loss=0.1898, simple_loss=0.3432, pruned_loss=0.182, over 4941.00 frames.], tot_loss[loss=0.1845, simple_loss=0.3286, pruned_loss=0.2369, over 985375.61 frames.], batch size: 45, aishell_tot_loss[loss=0.1926, simple_loss=0.3367, pruned_loss=0.8008, over 984713.35 frames.], datatang_tot_loss[loss=0.1877, simple_loss=0.3306, pruned_loss=0.7773, over 986146.26 frames.], batch size: 45, lr: 2.64e-03
+2022-06-18 11:14:25,454 INFO [train.py:874] (2/4) Epoch 1, batch 4150, aishell_loss[loss=0.1925, simple_loss=0.343, pruned_loss=0.2097, over 4929.00 frames.], tot_loss[loss=0.1843, simple_loss=0.3283, pruned_loss=0.2286, over 984852.09 frames.], batch size: 41, aishell_tot_loss[loss=0.192, simple_loss=0.3363, pruned_loss=0.7363, over 984142.07 frames.], datatang_tot_loss[loss=0.1868, simple_loss=0.3297, pruned_loss=0.7034, over 986120.96 frames.], batch size: 41, lr: 2.63e-03
+2022-06-18 11:14:54,758 INFO [train.py:874] (2/4) Epoch 1, batch 4200, datatang_loss[loss=0.1789, simple_loss=0.322, pruned_loss=0.1787, over 4951.00 frames.], tot_loss[loss=0.1831, simple_loss=0.3268, pruned_loss=0.2174, over 984909.92 frames.], batch size: 86, aishell_tot_loss[loss=0.1904, simple_loss=0.3346, pruned_loss=0.6773, over 983782.11 frames.], datatang_tot_loss[loss=0.1859, simple_loss=0.3287, pruned_loss=0.6353, over 986454.89 frames.], batch size: 86, lr: 2.63e-03
+2022-06-18 11:16:16,055 INFO [train.py:874] (2/4) Epoch 2, batch 50, datatang_loss[loss=0.1584, simple_loss=0.2864, pruned_loss=0.1522, over 4914.00 frames.], tot_loss[loss=0.1763, simple_loss=0.3183, pruned_loss=0.1715, over 218483.70 frames.], batch size: 57, aishell_tot_loss[loss=0.1841, simple_loss=0.3337, pruned_loss=0.1728, over 120435.86 frames.], datatang_tot_loss[loss=0.168, simple_loss=0.3019, pruned_loss=0.1702, over 111717.91 frames.], batch size: 57, lr: 2.60e-03
+2022-06-18 11:16:44,617 INFO [train.py:874] (2/4) Epoch 2, batch 100, datatang_loss[loss=0.1591, simple_loss=0.2876, pruned_loss=0.1534, over 4919.00 frames.], tot_loss[loss=0.1748, simple_loss=0.3155, pruned_loss=0.17, over 388310.41 frames.], batch size: 73, aishell_tot_loss[loss=0.1819, simple_loss=0.3298, pruned_loss=0.1697, over 237363.67 frames.], datatang_tot_loss[loss=0.1661, simple_loss=0.2982, pruned_loss=0.1703, over 198898.98 frames.], batch size: 73, lr: 2.59e-03
+2022-06-18 11:17:15,564 INFO [train.py:874] (2/4) Epoch 2, batch 150, datatang_loss[loss=0.1694, simple_loss=0.3072, pruned_loss=0.1577, over 4925.00 frames.], tot_loss[loss=0.1756, simple_loss=0.3171, pruned_loss=0.1706, over 521008.80 frames.], batch size: 79, aishell_tot_loss[loss=0.1799, simple_loss=0.3263, pruned_loss=0.1679, over 341927.32 frames.], datatang_tot_loss[loss=0.1697, simple_loss=0.3046, pruned_loss=0.1742, over 274157.87 frames.], batch size: 79, lr: 2.58e-03
+2022-06-18 11:17:45,669 INFO [train.py:874] (2/4) Epoch 2, batch 200, aishell_loss[loss=0.1981, simple_loss=0.3562, pruned_loss=0.2004, over 4973.00 frames.], tot_loss[loss=0.175, simple_loss=0.3163, pruned_loss=0.1689, over 624110.99 frames.], batch size: 61, aishell_tot_loss[loss=0.1784, simple_loss=0.3237, pruned_loss=0.1657, over 423467.70 frames.], datatang_tot_loss[loss=0.1707, simple_loss=0.3067, pruned_loss=0.1734, over 351673.28 frames.], batch size: 61, lr: 2.58e-03
+2022-06-18 11:18:14,304 INFO [train.py:874] (2/4) Epoch 2, batch 250, datatang_loss[loss=0.1899, simple_loss=0.338, pruned_loss=0.2091, over 4886.00 frames.], tot_loss[loss=0.1744, simple_loss=0.3153, pruned_loss=0.1679, over 704073.42 frames.], batch size: 44, aishell_tot_loss[loss=0.1779, simple_loss=0.3228, pruned_loss=0.1644, over 487044.04 frames.], datatang_tot_loss[loss=0.1707, simple_loss=0.3069, pruned_loss=0.1728, over 429090.64 frames.], batch size: 44, lr: 2.57e-03
+2022-06-18 11:18:45,564 INFO [train.py:874] (2/4) Epoch 2, batch 300, datatang_loss[loss=0.1863, simple_loss=0.3327, pruned_loss=0.1995, over 4918.00 frames.], tot_loss[loss=0.1749, simple_loss=0.3163, pruned_loss=0.1678, over 766725.70 frames.], batch size: 47, aishell_tot_loss[loss=0.1767, simple_loss=0.3212, pruned_loss=0.1613, over 552387.79 frames.], datatang_tot_loss[loss=0.1727, simple_loss=0.3102, pruned_loss=0.1759, over 487398.47 frames.], batch size: 47, lr: 2.57e-03
+2022-06-18 11:19:14,212 INFO [train.py:874] (2/4) Epoch 2, batch 350, aishell_loss[loss=0.1628, simple_loss=0.3007, pruned_loss=0.1247, over 4955.00 frames.], tot_loss[loss=0.1754, simple_loss=0.317, pruned_loss=0.169, over 815245.05 frames.], batch size: 54, aishell_tot_loss[loss=0.1765, simple_loss=0.3207, pruned_loss=0.1614, over 610830.76 frames.], datatang_tot_loss[loss=0.1737, simple_loss=0.3117, pruned_loss=0.1781, over 537357.25 frames.], batch size: 54, lr: 2.56e-03
+2022-06-18 11:19:44,617 INFO [train.py:874] (2/4) Epoch 2, batch 400, datatang_loss[loss=0.172, simple_loss=0.3107, pruned_loss=0.1664, over 4969.00 frames.], tot_loss[loss=0.1748, simple_loss=0.3161, pruned_loss=0.1672, over 853343.44 frames.], batch size: 31, aishell_tot_loss[loss=0.1768, simple_loss=0.3214, pruned_loss=0.1609, over 648522.23 frames.], datatang_tot_loss[loss=0.1726, simple_loss=0.3102, pruned_loss=0.1751, over 598094.04 frames.], batch size: 31, lr: 2.55e-03
+2022-06-18 11:20:15,803 INFO [train.py:874] (2/4) Epoch 2, batch 450, aishell_loss[loss=0.1659, simple_loss=0.3056, pruned_loss=0.131, over 4935.00 frames.], tot_loss[loss=0.1731, simple_loss=0.3136, pruned_loss=0.1634, over 882725.46 frames.], batch size: 32, aishell_tot_loss[loss=0.1752, simple_loss=0.3189, pruned_loss=0.1576, over 692522.18 frames.], datatang_tot_loss[loss=0.1718, simple_loss=0.3091, pruned_loss=0.173, over 638765.27 frames.], batch size: 32, lr: 2.55e-03
+2022-06-18 11:20:44,454 INFO [train.py:874] (2/4) Epoch 2, batch 500, datatang_loss[loss=0.1613, simple_loss=0.2919, pruned_loss=0.1533, over 4939.00 frames.], tot_loss[loss=0.1727, simple_loss=0.313, pruned_loss=0.1619, over 905616.82 frames.], batch size: 62, aishell_tot_loss[loss=0.1749, simple_loss=0.3184, pruned_loss=0.1568, over 730868.03 frames.], datatang_tot_loss[loss=0.1714, simple_loss=0.3085, pruned_loss=0.1714, over 675094.46 frames.], batch size: 62, lr: 2.54e-03
+2022-06-18 11:21:14,565 INFO [train.py:874] (2/4) Epoch 2, batch 550, datatang_loss[loss=0.2166, simple_loss=0.3858, pruned_loss=0.2366, over 4951.00 frames.], tot_loss[loss=0.1733, simple_loss=0.3142, pruned_loss=0.1618, over 923328.63 frames.], batch size: 110, aishell_tot_loss[loss=0.1752, simple_loss=0.3191, pruned_loss=0.1565, over 759560.81 frames.], datatang_tot_loss[loss=0.1717, simple_loss=0.3092, pruned_loss=0.1706, over 713193.94 frames.], batch size: 110, lr: 2.53e-03
+2022-06-18 11:21:45,598 INFO [train.py:874] (2/4) Epoch 2, batch 600, datatang_loss[loss=0.1591, simple_loss=0.2888, pruned_loss=0.1474, over 4931.00 frames.], tot_loss[loss=0.1742, simple_loss=0.3154, pruned_loss=0.1648, over 937109.72 frames.], batch size: 50, aishell_tot_loss[loss=0.176, simple_loss=0.32, pruned_loss=0.16, over 783321.71 frames.], datatang_tot_loss[loss=0.1721, simple_loss=0.3101, pruned_loss=0.1704, over 748670.47 frames.], batch size: 50, lr: 2.53e-03
+2022-06-18 11:22:14,819 INFO [train.py:874] (2/4) Epoch 2, batch 650, aishell_loss[loss=0.1655, simple_loss=0.304, pruned_loss=0.1349, over 4930.00 frames.], tot_loss[loss=0.1737, simple_loss=0.3148, pruned_loss=0.1631, over 948221.30 frames.], batch size: 32, aishell_tot_loss[loss=0.1752, simple_loss=0.3188, pruned_loss=0.1576, over 806437.85 frames.], datatang_tot_loss[loss=0.1724, simple_loss=0.3107, pruned_loss=0.1705, over 777817.09 frames.], batch size: 32, lr: 2.52e-03
+2022-06-18 11:22:45,353 INFO [train.py:874] (2/4) Epoch 2, batch 700, datatang_loss[loss=0.2, simple_loss=0.3578, pruned_loss=0.2108, over 4923.00 frames.], tot_loss[loss=0.173, simple_loss=0.3139, pruned_loss=0.161, over 955863.39 frames.], batch size: 94, aishell_tot_loss[loss=0.1742, simple_loss=0.3173, pruned_loss=0.1554, over 826869.25 frames.], datatang_tot_loss[loss=0.1725, simple_loss=0.3111, pruned_loss=0.1698, over 802273.79 frames.], batch size: 94, lr: 2.51e-03
+2022-06-18 11:23:15,123 INFO [train.py:874] (2/4) Epoch 2, batch 750, aishell_loss[loss=0.1822, simple_loss=0.3327, pruned_loss=0.1584, over 4950.00 frames.], tot_loss[loss=0.1728, simple_loss=0.3136, pruned_loss=0.1596, over 962314.95 frames.], batch size: 58, aishell_tot_loss[loss=0.1735, simple_loss=0.3164, pruned_loss=0.1536, over 846872.36 frames.], datatang_tot_loss[loss=0.1728, simple_loss=0.3116, pruned_loss=0.1696, over 822179.32 frames.], batch size: 58, lr: 2.51e-03
+2022-06-18 11:23:44,089 INFO [train.py:874] (2/4) Epoch 2, batch 800, aishell_loss[loss=0.2021, simple_loss=0.366, pruned_loss=0.1913, over 4918.00 frames.], tot_loss[loss=0.1742, simple_loss=0.3158, pruned_loss=0.1631, over 967799.19 frames.], batch size: 78, aishell_tot_loss[loss=0.1743, simple_loss=0.3176, pruned_loss=0.1545, over 861009.40 frames.], datatang_tot_loss[loss=0.1737, simple_loss=0.3129, pruned_loss=0.1721, over 844352.13 frames.], batch size: 78, lr: 2.50e-03
+2022-06-18 11:24:15,001 INFO [train.py:874] (2/4) Epoch 2, batch 850, datatang_loss[loss=0.1817, simple_loss=0.3274, pruned_loss=0.1802, over 4938.00 frames.], tot_loss[loss=0.1734, simple_loss=0.3146, pruned_loss=0.1613, over 971557.38 frames.], batch size: 88, aishell_tot_loss[loss=0.174, simple_loss=0.3174, pruned_loss=0.1537, over 874023.27 frames.], datatang_tot_loss[loss=0.1731, simple_loss=0.3121, pruned_loss=0.1705, over 862642.82 frames.], batch size: 88, lr: 2.50e-03
+2022-06-18 11:24:45,351 INFO [train.py:874] (2/4) Epoch 2, batch 900, aishell_loss[loss=0.1846, simple_loss=0.3369, pruned_loss=0.1613, over 4872.00 frames.], tot_loss[loss=0.1725, simple_loss=0.3133, pruned_loss=0.1581, over 974481.77 frames.], batch size: 36, aishell_tot_loss[loss=0.1729, simple_loss=0.3156, pruned_loss=0.151, over 887254.55 frames.], datatang_tot_loss[loss=0.1731, simple_loss=0.3123, pruned_loss=0.1694, over 876806.58 frames.], batch size: 36, lr: 2.49e-03
+2022-06-18 11:25:13,850 INFO [train.py:874] (2/4) Epoch 2, batch 950, aishell_loss[loss=0.1758, simple_loss=0.3195, pruned_loss=0.161, over 4926.00 frames.], tot_loss[loss=0.1721, simple_loss=0.3126, pruned_loss=0.1577, over 977035.23 frames.], batch size: 58, aishell_tot_loss[loss=0.173, simple_loss=0.3157, pruned_loss=0.1511, over 898986.79 frames.], datatang_tot_loss[loss=0.1724, simple_loss=0.3112, pruned_loss=0.1682, over 889583.31 frames.], batch size: 58, lr: 2.48e-03
+2022-06-18 11:25:45,478 INFO [train.py:874] (2/4) Epoch 2, batch 1000, datatang_loss[loss=0.1703, simple_loss=0.3083, pruned_loss=0.1617, over 4920.00 frames.], tot_loss[loss=0.1717, simple_loss=0.312, pruned_loss=0.1567, over 978749.82 frames.], batch size: 57, aishell_tot_loss[loss=0.1729, simple_loss=0.3156, pruned_loss=0.1507, over 907695.24 frames.], datatang_tot_loss[loss=0.1719, simple_loss=0.3105, pruned_loss=0.1667, over 902290.67 frames.], batch size: 57, lr: 2.48e-03
+2022-06-18 11:25:45,479 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 11:26:02,498 INFO [train.py:914] (2/4) Epoch 2, validation: loss=0.2333, simple_loss=0.2891, pruned_loss=0.08878, over 1622729.00 frames.
+2022-06-18 11:26:32,160 INFO [train.py:874] (2/4) Epoch 2, batch 1050, aishell_loss[loss=0.1711, simple_loss=0.3142, pruned_loss=0.1405, over 4881.00 frames.], tot_loss[loss=0.1711, simple_loss=0.3112, pruned_loss=0.1546, over 979993.35 frames.], batch size: 34, aishell_tot_loss[loss=0.1721, simple_loss=0.3145, pruned_loss=0.1488, over 917983.10 frames.], datatang_tot_loss[loss=0.1718, simple_loss=0.3104, pruned_loss=0.166, over 910581.78 frames.], batch size: 34, lr: 2.47e-03
+2022-06-18 11:27:03,420 INFO [train.py:874] (2/4) Epoch 2, batch 1100, datatang_loss[loss=0.1803, simple_loss=0.3267, pruned_loss=0.1691, over 4925.00 frames.], tot_loss[loss=0.1711, simple_loss=0.3114, pruned_loss=0.1537, over 981275.79 frames.], batch size: 77, aishell_tot_loss[loss=0.1719, simple_loss=0.3142, pruned_loss=0.1477, over 925550.06 frames.], datatang_tot_loss[loss=0.1718, simple_loss=0.3105, pruned_loss=0.1653, over 919884.08 frames.], batch size: 77, lr: 2.46e-03
+2022-06-18 11:27:31,690 INFO [train.py:874] (2/4) Epoch 2, batch 1150, aishell_loss[loss=0.1795, simple_loss=0.3307, pruned_loss=0.1417, over 4875.00 frames.], tot_loss[loss=0.1718, simple_loss=0.3126, pruned_loss=0.1548, over 982357.03 frames.], batch size: 36, aishell_tot_loss[loss=0.1722, simple_loss=0.315, pruned_loss=0.1476, over 933490.24 frames.], datatang_tot_loss[loss=0.1721, simple_loss=0.3109, pruned_loss=0.1663, over 926785.49 frames.], batch size: 36, lr: 2.46e-03
+2022-06-18 11:28:02,546 INFO [train.py:874] (2/4) Epoch 2, batch 1200, aishell_loss[loss=0.1591, simple_loss=0.294, pruned_loss=0.1209, over 4877.00 frames.], tot_loss[loss=0.1719, simple_loss=0.313, pruned_loss=0.1546, over 983050.42 frames.], batch size: 34, aishell_tot_loss[loss=0.1724, simple_loss=0.3153, pruned_loss=0.1472, over 939080.83 frames.], datatang_tot_loss[loss=0.172, simple_loss=0.3109, pruned_loss=0.1657, over 934281.73 frames.], batch size: 34, lr: 2.45e-03
+2022-06-18 11:28:33,999 INFO [train.py:874] (2/4) Epoch 2, batch 1250, datatang_loss[loss=0.1549, simple_loss=0.2821, pruned_loss=0.1386, over 4957.00 frames.], tot_loss[loss=0.1731, simple_loss=0.3144, pruned_loss=0.1588, over 983739.93 frames.], batch size: 50, aishell_tot_loss[loss=0.1734, simple_loss=0.3165, pruned_loss=0.1511, over 944808.71 frames.], datatang_tot_loss[loss=0.1722, simple_loss=0.3111, pruned_loss=0.1659, over 940200.07 frames.], batch size: 50, lr: 2.45e-03
+2022-06-18 11:29:03,384 INFO [train.py:874] (2/4) Epoch 2, batch 1300, aishell_loss[loss=0.1379, simple_loss=0.2553, pruned_loss=0.102, over 4984.00 frames.], tot_loss[loss=0.1729, simple_loss=0.3142, pruned_loss=0.1575, over 984227.64 frames.], batch size: 25, aishell_tot_loss[loss=0.1732, simple_loss=0.3165, pruned_loss=0.1501, over 948755.37 frames.], datatang_tot_loss[loss=0.1722, simple_loss=0.3113, pruned_loss=0.1652, over 946550.32 frames.], batch size: 25, lr: 2.44e-03
+2022-06-18 11:29:33,626 INFO [train.py:874] (2/4) Epoch 2, batch 1350, datatang_loss[loss=0.1592, simple_loss=0.2914, pruned_loss=0.1347, over 4915.00 frames.], tot_loss[loss=0.171, simple_loss=0.3113, pruned_loss=0.1533, over 984610.11 frames.], batch size: 75, aishell_tot_loss[loss=0.1722, simple_loss=0.3147, pruned_loss=0.1478, over 953387.00 frames.], datatang_tot_loss[loss=0.1712, simple_loss=0.3098, pruned_loss=0.163, over 950928.59 frames.], batch size: 75, lr: 2.43e-03
+2022-06-18 11:30:05,330 INFO [train.py:874] (2/4) Epoch 2, batch 1400, aishell_loss[loss=0.1512, simple_loss=0.2814, pruned_loss=0.1049, over 4985.00 frames.], tot_loss[loss=0.1706, simple_loss=0.3107, pruned_loss=0.1524, over 984509.13 frames.], batch size: 30, aishell_tot_loss[loss=0.1717, simple_loss=0.3141, pruned_loss=0.1469, over 956686.76 frames.], datatang_tot_loss[loss=0.171, simple_loss=0.3096, pruned_loss=0.1624, over 955167.19 frames.], batch size: 30, lr: 2.43e-03
+2022-06-18 11:30:32,952 INFO [train.py:874] (2/4) Epoch 2, batch 1450, aishell_loss[loss=0.1635, simple_loss=0.3028, pruned_loss=0.1213, over 4936.00 frames.], tot_loss[loss=0.1704, simple_loss=0.3105, pruned_loss=0.1512, over 985065.82 frames.], batch size: 45, aishell_tot_loss[loss=0.1718, simple_loss=0.3144, pruned_loss=0.1461, over 960224.08 frames.], datatang_tot_loss[loss=0.1706, simple_loss=0.3088, pruned_loss=0.1614, over 958952.30 frames.], batch size: 45, lr: 2.42e-03
+2022-06-18 11:31:04,362 INFO [train.py:874] (2/4) Epoch 2, batch 1500, aishell_loss[loss=0.1665, simple_loss=0.3072, pruned_loss=0.1292, over 4888.00 frames.], tot_loss[loss=0.17, simple_loss=0.31, pruned_loss=0.1499, over 985252.95 frames.], batch size: 34, aishell_tot_loss[loss=0.1717, simple_loss=0.3143, pruned_loss=0.1455, over 962861.56 frames.], datatang_tot_loss[loss=0.1701, simple_loss=0.3082, pruned_loss=0.1598, over 962519.00 frames.], batch size: 34, lr: 2.42e-03
+2022-06-18 11:31:35,279 INFO [train.py:874] (2/4) Epoch 2, batch 1550, datatang_loss[loss=0.1864, simple_loss=0.3394, pruned_loss=0.1668, over 4875.00 frames.], tot_loss[loss=0.1694, simple_loss=0.3091, pruned_loss=0.1487, over 985125.85 frames.], batch size: 39, aishell_tot_loss[loss=0.1707, simple_loss=0.3126, pruned_loss=0.1439, over 965250.28 frames.], datatang_tot_loss[loss=0.1702, simple_loss=0.3084, pruned_loss=0.1593, over 965315.44 frames.], batch size: 39, lr: 2.41e-03
+2022-06-18 11:32:03,490 INFO [train.py:874] (2/4) Epoch 2, batch 1600, aishell_loss[loss=0.1719, simple_loss=0.3159, pruned_loss=0.1393, over 4941.00 frames.], tot_loss[loss=0.1695, simple_loss=0.3094, pruned_loss=0.1476, over 985232.52 frames.], batch size: 58, aishell_tot_loss[loss=0.1704, simple_loss=0.3123, pruned_loss=0.1424, over 967620.90 frames.], datatang_tot_loss[loss=0.1703, simple_loss=0.3087, pruned_loss=0.159, over 967725.02 frames.], batch size: 58, lr: 2.40e-03
+2022-06-18 11:32:34,644 INFO [train.py:874] (2/4) Epoch 2, batch 1650, aishell_loss[loss=0.169, simple_loss=0.3115, pruned_loss=0.1324, over 4933.00 frames.], tot_loss[loss=0.1692, simple_loss=0.309, pruned_loss=0.1471, over 985322.18 frames.], batch size: 54, aishell_tot_loss[loss=0.17, simple_loss=0.3118, pruned_loss=0.1414, over 970179.62 frames.], datatang_tot_loss[loss=0.1702, simple_loss=0.3085, pruned_loss=0.1592, over 969377.07 frames.], batch size: 54, lr: 2.40e-03
+2022-06-18 11:33:04,802 INFO [train.py:874] (2/4) Epoch 2, batch 1700, aishell_loss[loss=0.1907, simple_loss=0.3484, pruned_loss=0.1648, over 4932.00 frames.], tot_loss[loss=0.1687, simple_loss=0.3083, pruned_loss=0.1456, over 985217.86 frames.], batch size: 68, aishell_tot_loss[loss=0.1697, simple_loss=0.3114, pruned_loss=0.1401, over 971972.97 frames.], datatang_tot_loss[loss=0.1697, simple_loss=0.3078, pruned_loss=0.1583, over 971112.41 frames.], batch size: 68, lr: 2.39e-03
+2022-06-18 11:33:33,377 INFO [train.py:874] (2/4) Epoch 2, batch 1750, datatang_loss[loss=0.1481, simple_loss=0.2711, pruned_loss=0.1253, over 4966.00 frames.], tot_loss[loss=0.1675, simple_loss=0.3063, pruned_loss=0.1438, over 985154.03 frames.], batch size: 45, aishell_tot_loss[loss=0.1689, simple_loss=0.3101, pruned_loss=0.1385, over 973244.53 frames.], datatang_tot_loss[loss=0.169, simple_loss=0.3066, pruned_loss=0.1571, over 972995.50 frames.], batch size: 45, lr: 2.39e-03
+2022-06-18 11:34:05,516 INFO [train.py:874] (2/4) Epoch 2, batch 1800, datatang_loss[loss=0.36, simple_loss=0.3358, pruned_loss=0.1921, over 4970.00 frames.], tot_loss[loss=0.1979, simple_loss=0.3089, pruned_loss=0.1562, over 985014.60 frames.], batch size: 65, aishell_tot_loss[loss=0.1854, simple_loss=0.3115, pruned_loss=0.1473, over 974373.40 frames.], datatang_tot_loss[loss=0.1838, simple_loss=0.3074, pruned_loss=0.1601, over 974554.69 frames.], batch size: 65, lr: 2.38e-03
+2022-06-18 11:34:34,592 INFO [train.py:874] (2/4) Epoch 2, batch 1850, datatang_loss[loss=0.271, simple_loss=0.2821, pruned_loss=0.13, over 4904.00 frames.], tot_loss[loss=0.2239, simple_loss=0.3093, pruned_loss=0.157, over 985095.68 frames.], batch size: 64, aishell_tot_loss[loss=0.2012, simple_loss=0.3124, pruned_loss=0.1486, over 975664.56 frames.], datatang_tot_loss[loss=0.1992, simple_loss=0.3069, pruned_loss=0.1603, over 975844.31 frames.], batch size: 64, lr: 2.38e-03
+2022-06-18 11:35:04,172 INFO [train.py:874] (2/4) Epoch 2, batch 1900, datatang_loss[loss=0.3323, simple_loss=0.3117, pruned_loss=0.1765, over 4906.00 frames.], tot_loss[loss=0.2391, simple_loss=0.3086, pruned_loss=0.1531, over 985439.86 frames.], batch size: 52, aishell_tot_loss[loss=0.2124, simple_loss=0.3115, pruned_loss=0.1461, over 977202.02 frames.], datatang_tot_loss[loss=0.21, simple_loss=0.3069, pruned_loss=0.1596, over 976871.81 frames.], batch size: 52, lr: 2.37e-03
+2022-06-18 11:35:34,928 INFO [train.py:874] (2/4) Epoch 2, batch 1950, aishell_loss[loss=0.3327, simple_loss=0.3492, pruned_loss=0.1581, over 4931.00 frames.], tot_loss[loss=0.2516, simple_loss=0.3088, pruned_loss=0.1504, over 985822.40 frames.], batch size: 79, aishell_tot_loss[loss=0.2221, simple_loss=0.3113, pruned_loss=0.1441, over 978522.73 frames.], datatang_tot_loss[loss=0.2203, simple_loss=0.3071, pruned_loss=0.1589, over 977925.57 frames.], batch size: 79, lr: 2.36e-03
+2022-06-18 11:36:03,036 INFO [train.py:874] (2/4) Epoch 2, batch 2000, datatang_loss[loss=0.3029, simple_loss=0.3052, pruned_loss=0.1503, over 4968.00 frames.], tot_loss[loss=0.2621, simple_loss=0.3094, pruned_loss=0.1487, over 986242.49 frames.], batch size: 55, aishell_tot_loss[loss=0.2292, simple_loss=0.3107, pruned_loss=0.142, over 979795.04 frames.], datatang_tot_loss[loss=0.2315, simple_loss=0.3083, pruned_loss=0.159, over 978884.57 frames.], batch size: 55, lr: 2.36e-03
+2022-06-18 11:36:03,037 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 11:36:19,298 INFO [train.py:914] (2/4) Epoch 2, validation: loss=0.2142, simple_loss=0.275, pruned_loss=0.07672, over 1622729.00 frames.
+2022-06-18 11:36:48,998 INFO [train.py:874] (2/4) Epoch 2, batch 2050, datatang_loss[loss=0.2531, simple_loss=0.2782, pruned_loss=0.114, over 4929.00 frames.], tot_loss[loss=0.2694, simple_loss=0.3083, pruned_loss=0.1474, over 985629.24 frames.], batch size: 71, aishell_tot_loss[loss=0.2371, simple_loss=0.3107, pruned_loss=0.1412, over 979926.75 frames.], datatang_tot_loss[loss=0.2389, simple_loss=0.307, pruned_loss=0.1581, over 979715.48 frames.], batch size: 71, lr: 2.35e-03
+2022-06-18 11:37:18,664 INFO [train.py:874] (2/4) Epoch 2, batch 2100, aishell_loss[loss=0.2428, simple_loss=0.2813, pruned_loss=0.1022, over 4926.00 frames.], tot_loss[loss=0.2715, simple_loss=0.3057, pruned_loss=0.1437, over 985692.35 frames.], batch size: 32, aishell_tot_loss[loss=0.2415, simple_loss=0.3092, pruned_loss=0.1391, over 980584.63 frames.], datatang_tot_loss[loss=0.2442, simple_loss=0.3055, pruned_loss=0.1558, over 980475.45 frames.], batch size: 32, lr: 2.35e-03
+2022-06-18 11:37:49,773 INFO [train.py:874] (2/4) Epoch 2, batch 2150, aishell_loss[loss=0.2918, simple_loss=0.3118, pruned_loss=0.1359, over 4942.00 frames.], tot_loss[loss=0.2782, simple_loss=0.3072, pruned_loss=0.1441, over 985246.20 frames.], batch size: 54, aishell_tot_loss[loss=0.2464, simple_loss=0.3092, pruned_loss=0.1378, over 980895.35 frames.], datatang_tot_loss[loss=0.2531, simple_loss=0.3068, pruned_loss=0.1565, over 980886.39 frames.], batch size: 54, lr: 2.34e-03
+2022-06-18 11:38:20,205 INFO [train.py:874] (2/4) Epoch 2, batch 2200, datatang_loss[loss=0.3744, simple_loss=0.3444, pruned_loss=0.2022, over 4935.00 frames.], tot_loss[loss=0.282, simple_loss=0.3077, pruned_loss=0.1433, over 985008.14 frames.], batch size: 79, aishell_tot_loss[loss=0.2504, simple_loss=0.3092, pruned_loss=0.1368, over 981273.06 frames.], datatang_tot_loss[loss=0.2598, simple_loss=0.3072, pruned_loss=0.1558, over 981283.03 frames.], batch size: 79, lr: 2.33e-03
+2022-06-18 11:38:48,992 INFO [train.py:874] (2/4) Epoch 2, batch 2250, datatang_loss[loss=0.2863, simple_loss=0.2913, pruned_loss=0.1406, over 4958.00 frames.], tot_loss[loss=0.2887, simple_loss=0.309, pruned_loss=0.1461, over 985333.07 frames.], batch size: 60, aishell_tot_loss[loss=0.2578, simple_loss=0.311, pruned_loss=0.1397, over 981682.84 frames.], datatang_tot_loss[loss=0.2657, simple_loss=0.3067, pruned_loss=0.1545, over 982094.91 frames.], batch size: 60, lr: 2.33e-03
+2022-06-18 11:39:20,421 INFO [train.py:874] (2/4) Epoch 2, batch 2300, aishell_loss[loss=0.2805, simple_loss=0.3117, pruned_loss=0.1246, over 4967.00 frames.], tot_loss[loss=0.2906, simple_loss=0.3098, pruned_loss=0.145, over 985440.51 frames.], batch size: 30, aishell_tot_loss[loss=0.2621, simple_loss=0.3117, pruned_loss=0.1391, over 982106.35 frames.], datatang_tot_loss[loss=0.2699, simple_loss=0.3068, pruned_loss=0.1539, over 982601.00 frames.], batch size: 30, lr: 2.32e-03
+2022-06-18 11:39:51,122 INFO [train.py:874] (2/4) Epoch 2, batch 2350, aishell_loss[loss=0.2826, simple_loss=0.3122, pruned_loss=0.1265, over 4979.00 frames.], tot_loss[loss=0.2914, simple_loss=0.3102, pruned_loss=0.1435, over 985417.89 frames.], batch size: 30, aishell_tot_loss[loss=0.2647, simple_loss=0.312, pruned_loss=0.1379, over 982399.92 frames.], datatang_tot_loss[loss=0.2739, simple_loss=0.3072, pruned_loss=0.1529, over 982998.93 frames.], batch size: 30, lr: 2.32e-03
+2022-06-18 11:40:20,384 INFO [train.py:874] (2/4) Epoch 2, batch 2400, datatang_loss[loss=0.281, simple_loss=0.2936, pruned_loss=0.1342, over 4917.00 frames.], tot_loss[loss=0.29, simple_loss=0.3091, pruned_loss=0.1411, over 985478.92 frames.], batch size: 77, aishell_tot_loss[loss=0.2659, simple_loss=0.3117, pruned_loss=0.1362, over 982555.66 frames.], datatang_tot_loss[loss=0.2762, simple_loss=0.3065, pruned_loss=0.1514, over 983539.11 frames.], batch size: 77, lr: 2.31e-03
+2022-06-18 11:40:51,460 INFO [train.py:874] (2/4) Epoch 2, batch 2450, aishell_loss[loss=0.2772, simple_loss=0.3078, pruned_loss=0.1233, over 4885.00 frames.], tot_loss[loss=0.2878, simple_loss=0.3075, pruned_loss=0.1384, over 985392.59 frames.], batch size: 35, aishell_tot_loss[loss=0.2661, simple_loss=0.3105, pruned_loss=0.134, over 982816.69 frames.], datatang_tot_loss[loss=0.2782, simple_loss=0.3061, pruned_loss=0.1501, over 983751.18 frames.], batch size: 35, lr: 2.31e-03
+2022-06-18 11:41:22,604 INFO [train.py:874] (2/4) Epoch 2, batch 2500, datatang_loss[loss=0.3016, simple_loss=0.3104, pruned_loss=0.1465, over 4931.00 frames.], tot_loss[loss=0.2912, simple_loss=0.3085, pruned_loss=0.1403, over 985333.51 frames.], batch size: 81, aishell_tot_loss[loss=0.2682, simple_loss=0.3104, pruned_loss=0.1329, over 983149.64 frames.], datatang_tot_loss[loss=0.2835, simple_loss=0.3069, pruned_loss=0.1526, over 983880.52 frames.], batch size: 81, lr: 2.30e-03
+2022-06-18 11:41:51,392 INFO [train.py:874] (2/4) Epoch 2, batch 2550, datatang_loss[loss=0.2422, simple_loss=0.264, pruned_loss=0.1101, over 4986.00 frames.], tot_loss[loss=0.2902, simple_loss=0.3075, pruned_loss=0.1391, over 985036.44 frames.], batch size: 40, aishell_tot_loss[loss=0.2691, simple_loss=0.3097, pruned_loss=0.1317, over 983030.21 frames.], datatang_tot_loss[loss=0.2854, simple_loss=0.3066, pruned_loss=0.1521, over 984118.55 frames.], batch size: 40, lr: 2.30e-03
+2022-06-18 11:42:23,032 INFO [train.py:874] (2/4) Epoch 2, batch 2600, datatang_loss[loss=0.2369, simple_loss=0.2556, pruned_loss=0.1091, over 4968.00 frames.], tot_loss[loss=0.2923, simple_loss=0.3093, pruned_loss=0.1396, over 985272.40 frames.], batch size: 45, aishell_tot_loss[loss=0.271, simple_loss=0.3105, pruned_loss=0.1314, over 983268.97 frames.], datatang_tot_loss[loss=0.2885, simple_loss=0.3076, pruned_loss=0.1521, over 984457.77 frames.], batch size: 45, lr: 2.29e-03
+2022-06-18 11:42:52,008 INFO [train.py:874] (2/4) Epoch 2, batch 2650, datatang_loss[loss=0.278, simple_loss=0.2914, pruned_loss=0.1323, over 4934.00 frames.], tot_loss[loss=0.2909, simple_loss=0.3086, pruned_loss=0.1382, over 985358.76 frames.], batch size: 79, aishell_tot_loss[loss=0.2728, simple_loss=0.3105, pruned_loss=0.1312, over 983407.00 frames.], datatang_tot_loss[loss=0.2885, simple_loss=0.3069, pruned_loss=0.1506, over 984765.01 frames.], batch size: 79, lr: 2.28e-03
+2022-06-18 11:43:22,520 INFO [train.py:874] (2/4) Epoch 2, batch 2700, aishell_loss[loss=0.3111, simple_loss=0.3371, pruned_loss=0.1426, over 4934.00 frames.], tot_loss[loss=0.2887, simple_loss=0.3077, pruned_loss=0.136, over 985442.09 frames.], batch size: 68, aishell_tot_loss[loss=0.2721, simple_loss=0.3095, pruned_loss=0.1291, over 983555.17 frames.], datatang_tot_loss[loss=0.2894, simple_loss=0.3068, pruned_loss=0.1499, over 985019.75 frames.], batch size: 68, lr: 2.28e-03
+2022-06-18 11:43:52,867 INFO [train.py:874] (2/4) Epoch 2, batch 2750, datatang_loss[loss=0.3127, simple_loss=0.2948, pruned_loss=0.1653, over 4943.00 frames.], tot_loss[loss=0.288, simple_loss=0.3072, pruned_loss=0.1354, over 985479.82 frames.], batch size: 50, aishell_tot_loss[loss=0.273, simple_loss=0.31, pruned_loss=0.1287, over 983715.68 frames.], datatang_tot_loss[loss=0.2892, simple_loss=0.3057, pruned_loss=0.1484, over 985123.01 frames.], batch size: 50, lr: 2.27e-03
+2022-06-18 11:44:21,079 INFO [train.py:874] (2/4) Epoch 2, batch 2800, aishell_loss[loss=0.2496, simple_loss=0.2928, pruned_loss=0.1032, over 4918.00 frames.], tot_loss[loss=0.2853, simple_loss=0.3055, pruned_loss=0.1333, over 984971.33 frames.], batch size: 33, aishell_tot_loss[loss=0.2733, simple_loss=0.3095, pruned_loss=0.128, over 983527.26 frames.], datatang_tot_loss[loss=0.2877, simple_loss=0.3042, pruned_loss=0.1462, over 985033.84 frames.], batch size: 33, lr: 2.27e-03
+2022-06-18 11:44:52,571 INFO [train.py:874] (2/4) Epoch 2, batch 2850, datatang_loss[loss=0.3348, simple_loss=0.3244, pruned_loss=0.1726, over 4932.00 frames.], tot_loss[loss=0.2871, simple_loss=0.3066, pruned_loss=0.1343, over 985169.80 frames.], batch size: 69, aishell_tot_loss[loss=0.2727, simple_loss=0.3087, pruned_loss=0.1267, over 983832.17 frames.], datatang_tot_loss[loss=0.2912, simple_loss=0.3059, pruned_loss=0.1475, over 985097.62 frames.], batch size: 69, lr: 2.26e-03
+2022-06-18 11:45:23,337 INFO [train.py:874] (2/4) Epoch 2, batch 2900, aishell_loss[loss=0.2314, simple_loss=0.2741, pruned_loss=0.09438, over 4887.00 frames.], tot_loss[loss=0.2863, simple_loss=0.3069, pruned_loss=0.1333, over 985271.74 frames.], batch size: 42, aishell_tot_loss[loss=0.2719, simple_loss=0.3079, pruned_loss=0.1253, over 983984.95 frames.], datatang_tot_loss[loss=0.2926, simple_loss=0.3069, pruned_loss=0.1474, over 985236.02 frames.], batch size: 42, lr: 2.26e-03
+2022-06-18 11:45:51,132 INFO [train.py:874] (2/4) Epoch 2, batch 2950, datatang_loss[loss=0.2548, simple_loss=0.2727, pruned_loss=0.1184, over 4867.00 frames.], tot_loss[loss=0.2849, simple_loss=0.3066, pruned_loss=0.1319, over 985568.83 frames.], batch size: 39, aishell_tot_loss[loss=0.2709, simple_loss=0.3069, pruned_loss=0.1237, over 984395.36 frames.], datatang_tot_loss[loss=0.2936, simple_loss=0.3073, pruned_loss=0.1476, over 985340.05 frames.], batch size: 39, lr: 2.25e-03
+2022-06-18 11:46:22,412 INFO [train.py:874] (2/4) Epoch 2, batch 3000, aishell_loss[loss=0.2663, simple_loss=0.305, pruned_loss=0.1138, over 4929.00 frames.], tot_loss[loss=0.2816, simple_loss=0.3039, pruned_loss=0.1299, over 985560.07 frames.], batch size: 56, aishell_tot_loss[loss=0.269, simple_loss=0.3052, pruned_loss=0.1219, over 984556.72 frames.], datatang_tot_loss[loss=0.2925, simple_loss=0.306, pruned_loss=0.1461, over 985333.94 frames.], batch size: 56, lr: 2.25e-03
+2022-06-18 11:46:22,413 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 11:46:39,815 INFO [train.py:914] (2/4) Epoch 2, validation: loss=0.2091, simple_loss=0.2742, pruned_loss=0.07205, over 1622729.00 frames.
+2022-06-18 11:47:08,832 INFO [train.py:874] (2/4) Epoch 2, batch 3050, datatang_loss[loss=0.2625, simple_loss=0.2776, pruned_loss=0.1237, over 4942.00 frames.], tot_loss[loss=0.2816, simple_loss=0.304, pruned_loss=0.1298, over 985563.50 frames.], batch size: 69, aishell_tot_loss[loss=0.2694, simple_loss=0.3052, pruned_loss=0.1216, over 984442.38 frames.], datatang_tot_loss[loss=0.2924, simple_loss=0.3057, pruned_loss=0.1455, over 985596.55 frames.], batch size: 69, lr: 2.24e-03
+2022-06-18 11:47:40,865 INFO [train.py:874] (2/4) Epoch 2, batch 3100, aishell_loss[loss=0.2356, simple_loss=0.2491, pruned_loss=0.111, over 4964.00 frames.], tot_loss[loss=0.2797, simple_loss=0.3032, pruned_loss=0.1283, over 985227.66 frames.], batch size: 21, aishell_tot_loss[loss=0.2682, simple_loss=0.3043, pruned_loss=0.1202, over 984122.08 frames.], datatang_tot_loss[loss=0.2924, simple_loss=0.3054, pruned_loss=0.145, over 985716.84 frames.], batch size: 21, lr: 2.24e-03
+2022-06-18 11:48:08,412 INFO [train.py:874] (2/4) Epoch 2, batch 3150, aishell_loss[loss=0.2979, simple_loss=0.3286, pruned_loss=0.1335, over 4921.00 frames.], tot_loss[loss=0.2814, simple_loss=0.3031, pruned_loss=0.13, over 985289.55 frames.], batch size: 33, aishell_tot_loss[loss=0.2694, simple_loss=0.3047, pruned_loss=0.1207, over 984318.99 frames.], datatang_tot_loss[loss=0.2929, simple_loss=0.3045, pruned_loss=0.1453, over 985641.89 frames.], batch size: 33, lr: 2.23e-03
+2022-06-18 11:48:40,358 INFO [train.py:874] (2/4) Epoch 2, batch 3200, datatang_loss[loss=0.2305, simple_loss=0.2634, pruned_loss=0.09883, over 4980.00 frames.], tot_loss[loss=0.2795, simple_loss=0.303, pruned_loss=0.1281, over 985428.07 frames.], batch size: 40, aishell_tot_loss[loss=0.2688, simple_loss=0.3047, pruned_loss=0.1197, over 984341.16 frames.], datatang_tot_loss[loss=0.2918, simple_loss=0.3042, pruned_loss=0.1438, over 985844.04 frames.], batch size: 40, lr: 2.23e-03
+2022-06-18 11:49:11,900 INFO [train.py:874] (2/4) Epoch 2, batch 3250, datatang_loss[loss=0.3135, simple_loss=0.3284, pruned_loss=0.1493, over 4924.00 frames.], tot_loss[loss=0.2793, simple_loss=0.3036, pruned_loss=0.1275, over 985587.62 frames.], batch size: 94, aishell_tot_loss[loss=0.2689, simple_loss=0.3048, pruned_loss=0.1193, over 984705.51 frames.], datatang_tot_loss[loss=0.2914, simple_loss=0.3043, pruned_loss=0.1429, over 985746.49 frames.], batch size: 94, lr: 2.22e-03
+2022-06-18 11:49:40,186 INFO [train.py:874] (2/4) Epoch 2, batch 3300, aishell_loss[loss=0.3405, simple_loss=0.3544, pruned_loss=0.1633, over 4858.00 frames.], tot_loss[loss=0.2792, simple_loss=0.3038, pruned_loss=0.1274, over 985707.03 frames.], batch size: 37, aishell_tot_loss[loss=0.2693, simple_loss=0.3049, pruned_loss=0.1193, over 984970.81 frames.], datatang_tot_loss[loss=0.2909, simple_loss=0.3041, pruned_loss=0.1422, over 985705.48 frames.], batch size: 37, lr: 2.22e-03
+2022-06-18 11:50:11,305 INFO [train.py:874] (2/4) Epoch 2, batch 3350, datatang_loss[loss=0.2617, simple_loss=0.2737, pruned_loss=0.1249, over 4912.00 frames.], tot_loss[loss=0.276, simple_loss=0.3011, pruned_loss=0.1254, over 985649.27 frames.], batch size: 42, aishell_tot_loss[loss=0.2677, simple_loss=0.3035, pruned_loss=0.1181, over 984910.05 frames.], datatang_tot_loss[loss=0.2887, simple_loss=0.3026, pruned_loss=0.1403, over 985777.19 frames.], batch size: 42, lr: 2.21e-03
+2022-06-18 11:50:42,252 INFO [train.py:874] (2/4) Epoch 2, batch 3400, aishell_loss[loss=0.2696, simple_loss=0.3124, pruned_loss=0.1134, over 4883.00 frames.], tot_loss[loss=0.2741, simple_loss=0.3005, pruned_loss=0.1239, over 985245.03 frames.], batch size: 42, aishell_tot_loss[loss=0.2661, simple_loss=0.3024, pruned_loss=0.1168, over 984593.33 frames.], datatang_tot_loss[loss=0.2879, simple_loss=0.3025, pruned_loss=0.1392, over 985737.31 frames.], batch size: 42, lr: 2.21e-03
+2022-06-18 11:51:10,163 INFO [train.py:874] (2/4) Epoch 2, batch 3450, datatang_loss[loss=0.2861, simple_loss=0.3006, pruned_loss=0.1358, over 4947.00 frames.], tot_loss[loss=0.2743, simple_loss=0.3002, pruned_loss=0.1243, over 985264.02 frames.], batch size: 50, aishell_tot_loss[loss=0.2657, simple_loss=0.3019, pruned_loss=0.1164, over 984424.38 frames.], datatang_tot_loss[loss=0.2879, simple_loss=0.3023, pruned_loss=0.139, over 985921.38 frames.], batch size: 50, lr: 2.20e-03
+2022-06-18 11:51:42,272 INFO [train.py:874] (2/4) Epoch 2, batch 3500, aishell_loss[loss=0.2662, simple_loss=0.306, pruned_loss=0.1132, over 4988.00 frames.], tot_loss[loss=0.274, simple_loss=0.3003, pruned_loss=0.1239, over 985223.47 frames.], batch size: 30, aishell_tot_loss[loss=0.2654, simple_loss=0.3018, pruned_loss=0.1159, over 984259.60 frames.], datatang_tot_loss[loss=0.2873, simple_loss=0.302, pruned_loss=0.1383, over 986073.30 frames.], batch size: 30, lr: 2.20e-03
+2022-06-18 11:52:10,567 INFO [train.py:874] (2/4) Epoch 2, batch 3550, aishell_loss[loss=0.2407, simple_loss=0.2953, pruned_loss=0.09308, over 4977.00 frames.], tot_loss[loss=0.2723, simple_loss=0.2997, pruned_loss=0.1224, over 984486.08 frames.], batch size: 51, aishell_tot_loss[loss=0.264, simple_loss=0.3011, pruned_loss=0.1147, over 983744.37 frames.], datatang_tot_loss[loss=0.2868, simple_loss=0.3017, pruned_loss=0.1377, over 985862.47 frames.], batch size: 51, lr: 2.19e-03
+2022-06-18 11:52:42,014 INFO [train.py:874] (2/4) Epoch 2, batch 3600, aishell_loss[loss=0.2822, simple_loss=0.3168, pruned_loss=0.1238, over 4945.00 frames.], tot_loss[loss=0.2842, simple_loss=0.3026, pruned_loss=0.1329, over 984891.61 frames.], batch size: 54, aishell_tot_loss[loss=0.2658, simple_loss=0.3019, pruned_loss=0.116, over 983913.21 frames.], datatang_tot_loss[loss=0.2967, simple_loss=0.3036, pruned_loss=0.1465, over 986019.57 frames.], batch size: 54, lr: 2.19e-03
+2022-06-18 11:53:13,284 INFO [train.py:874] (2/4) Epoch 2, batch 3650, aishell_loss[loss=0.285, simple_loss=0.3193, pruned_loss=0.1253, over 4894.00 frames.], tot_loss[loss=0.2808, simple_loss=0.3023, pruned_loss=0.1297, over 985401.94 frames.], batch size: 60, aishell_tot_loss[loss=0.2655, simple_loss=0.3023, pruned_loss=0.1154, over 984249.05 frames.], datatang_tot_loss[loss=0.2944, simple_loss=0.3028, pruned_loss=0.1444, over 986227.76 frames.], batch size: 60, lr: 2.18e-03
+2022-06-18 11:53:42,106 INFO [train.py:874] (2/4) Epoch 2, batch 3700, aishell_loss[loss=0.2096, simple_loss=0.2319, pruned_loss=0.09363, over 4958.00 frames.], tot_loss[loss=0.2772, simple_loss=0.3012, pruned_loss=0.1266, over 985512.73 frames.], batch size: 21, aishell_tot_loss[loss=0.2632, simple_loss=0.3008, pruned_loss=0.1137, over 984282.83 frames.], datatang_tot_loss[loss=0.2933, simple_loss=0.3031, pruned_loss=0.143, over 986375.60 frames.], batch size: 21, lr: 2.18e-03
+2022-06-18 11:54:13,036 INFO [train.py:874] (2/4) Epoch 2, batch 3750, datatang_loss[loss=0.2939, simple_loss=0.3135, pruned_loss=0.1371, over 4914.00 frames.], tot_loss[loss=0.276, simple_loss=0.3008, pruned_loss=0.1256, over 985688.68 frames.], batch size: 64, aishell_tot_loss[loss=0.263, simple_loss=0.3009, pruned_loss=0.1133, over 984433.70 frames.], datatang_tot_loss[loss=0.2912, simple_loss=0.3023, pruned_loss=0.1411, over 986399.72 frames.], batch size: 64, lr: 2.17e-03
+2022-06-18 11:54:47,898 INFO [train.py:874] (2/4) Epoch 2, batch 3800, datatang_loss[loss=0.3143, simple_loss=0.3321, pruned_loss=0.1482, over 4949.00 frames.], tot_loss[loss=0.2758, simple_loss=0.3013, pruned_loss=0.1251, over 986027.26 frames.], batch size: 99, aishell_tot_loss[loss=0.2641, simple_loss=0.3016, pruned_loss=0.114, over 984954.40 frames.], datatang_tot_loss[loss=0.2895, simple_loss=0.302, pruned_loss=0.1395, over 986302.05 frames.], batch size: 99, lr: 2.17e-03
+2022-06-18 11:55:15,861 INFO [train.py:874] (2/4) Epoch 2, batch 3850, datatang_loss[loss=0.2864, simple_loss=0.3007, pruned_loss=0.1361, over 4924.00 frames.], tot_loss[loss=0.2748, simple_loss=0.3013, pruned_loss=0.1241, over 985269.28 frames.], batch size: 75, aishell_tot_loss[loss=0.2647, simple_loss=0.3023, pruned_loss=0.1142, over 984379.65 frames.], datatang_tot_loss[loss=0.2874, simple_loss=0.3011, pruned_loss=0.1377, over 986162.10 frames.], batch size: 75, lr: 2.16e-03
+2022-06-18 11:55:45,602 INFO [train.py:874] (2/4) Epoch 2, batch 3900, datatang_loss[loss=0.2905, simple_loss=0.2955, pruned_loss=0.1427, over 4980.00 frames.], tot_loss[loss=0.2743, simple_loss=0.3008, pruned_loss=0.1239, over 985153.15 frames.], batch size: 45, aishell_tot_loss[loss=0.2652, simple_loss=0.3022, pruned_loss=0.1146, over 984119.08 frames.], datatang_tot_loss[loss=0.2858, simple_loss=0.3006, pruned_loss=0.1362, over 986251.92 frames.], batch size: 45, lr: 2.16e-03
+2022-06-18 11:56:14,171 INFO [train.py:874] (2/4) Epoch 2, batch 3950, datatang_loss[loss=0.3008, simple_loss=0.3115, pruned_loss=0.1451, over 4929.00 frames.], tot_loss[loss=0.2725, simple_loss=0.2998, pruned_loss=0.1226, over 985165.54 frames.], batch size: 71, aishell_tot_loss[loss=0.2634, simple_loss=0.301, pruned_loss=0.1134, over 984052.19 frames.], datatang_tot_loss[loss=0.2849, simple_loss=0.3006, pruned_loss=0.1353, over 986287.15 frames.], batch size: 71, lr: 2.15e-03
+2022-06-18 11:56:45,133 INFO [train.py:874] (2/4) Epoch 2, batch 4000, aishell_loss[loss=0.2849, simple_loss=0.3321, pruned_loss=0.1189, over 4949.00 frames.], tot_loss[loss=0.2712, simple_loss=0.2991, pruned_loss=0.1216, over 985270.38 frames.], batch size: 45, aishell_tot_loss[loss=0.2629, simple_loss=0.3011, pruned_loss=0.1128, over 984288.99 frames.], datatang_tot_loss[loss=0.2833, simple_loss=0.2996, pruned_loss=0.134, over 986111.33 frames.], batch size: 45, lr: 2.15e-03
+2022-06-18 11:56:45,134 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 11:57:01,101 INFO [train.py:914] (2/4) Epoch 2, validation: loss=0.2024, simple_loss=0.2723, pruned_loss=0.06625, over 1622729.00 frames.
+2022-06-18 11:57:30,549 INFO [train.py:874] (2/4) Epoch 2, batch 4050, datatang_loss[loss=0.2521, simple_loss=0.2829, pruned_loss=0.1106, over 4966.00 frames.], tot_loss[loss=0.2711, simple_loss=0.2993, pruned_loss=0.1215, over 985395.07 frames.], batch size: 45, aishell_tot_loss[loss=0.2631, simple_loss=0.3012, pruned_loss=0.1129, over 984333.52 frames.], datatang_tot_loss[loss=0.2822, simple_loss=0.2994, pruned_loss=0.133, over 986180.70 frames.], batch size: 45, lr: 2.14e-03
+2022-06-18 11:58:00,708 INFO [train.py:874] (2/4) Epoch 2, batch 4100, aishell_loss[loss=0.3193, simple_loss=0.3452, pruned_loss=0.1467, over 4922.00 frames.], tot_loss[loss=0.2706, simple_loss=0.2993, pruned_loss=0.121, over 985161.80 frames.], batch size: 34, aishell_tot_loss[loss=0.2641, simple_loss=0.3022, pruned_loss=0.1134, over 984038.87 frames.], datatang_tot_loss[loss=0.28, simple_loss=0.2982, pruned_loss=0.1313, over 986227.34 frames.], batch size: 34, lr: 2.14e-03
+2022-06-18 11:59:09,052 INFO [train.py:874] (2/4) Epoch 3, batch 50, aishell_loss[loss=0.2213, simple_loss=0.2528, pruned_loss=0.0949, over 4882.00 frames.], tot_loss[loss=0.2454, simple_loss=0.2807, pruned_loss=0.105, over 218237.70 frames.], batch size: 21, aishell_tot_loss[loss=0.2459, simple_loss=0.2885, pruned_loss=0.1017, over 120192.11 frames.], datatang_tot_loss[loss=0.2451, simple_loss=0.2726, pruned_loss=0.1088, over 111684.23 frames.], batch size: 21, lr: 2.09e-03
+2022-06-18 11:59:39,422 INFO [train.py:874] (2/4) Epoch 3, batch 100, aishell_loss[loss=0.3479, simple_loss=0.3627, pruned_loss=0.1666, over 4919.00 frames.], tot_loss[loss=0.256, simple_loss=0.2895, pruned_loss=0.1112, over 387806.63 frames.], batch size: 78, aishell_tot_loss[loss=0.2559, simple_loss=0.297, pruned_loss=0.1074, over 210195.60 frames.], datatang_tot_loss[loss=0.2548, simple_loss=0.2817, pruned_loss=0.1139, over 225895.26 frames.], batch size: 78, lr: 2.09e-03
+2022-06-18 12:00:11,062 INFO [train.py:874] (2/4) Epoch 3, batch 150, datatang_loss[loss=0.3698, simple_loss=0.3769, pruned_loss=0.1814, over 4949.00 frames.], tot_loss[loss=0.2635, simple_loss=0.2921, pruned_loss=0.1175, over 520273.62 frames.], batch size: 109, aishell_tot_loss[loss=0.2654, simple_loss=0.3007, pruned_loss=0.115, over 273405.95 frames.], datatang_tot_loss[loss=0.2598, simple_loss=0.2845, pruned_loss=0.1175, over 341786.75 frames.], batch size: 109, lr: 2.08e-03
+2022-06-18 12:00:39,815 INFO [train.py:874] (2/4) Epoch 3, batch 200, datatang_loss[loss=0.2539, simple_loss=0.2775, pruned_loss=0.1152, over 4926.00 frames.], tot_loss[loss=0.2627, simple_loss=0.2933, pruned_loss=0.116, over 623127.17 frames.], batch size: 62, aishell_tot_loss[loss=0.2616, simple_loss=0.2993, pruned_loss=0.1119, over 387622.53 frames.], datatang_tot_loss[loss=0.2621, simple_loss=0.2858, pruned_loss=0.1192, over 388481.56 frames.], batch size: 62, lr: 2.08e-03
+2022-06-18 12:01:10,605 INFO [train.py:874] (2/4) Epoch 3, batch 250, aishell_loss[loss=0.2546, simple_loss=0.3119, pruned_loss=0.09868, over 4860.00 frames.], tot_loss[loss=0.2601, simple_loss=0.2923, pruned_loss=0.1139, over 703469.52 frames.], batch size: 37, aishell_tot_loss[loss=0.2576, simple_loss=0.297, pruned_loss=0.1091, over 457981.70 frames.], datatang_tot_loss[loss=0.2624, simple_loss=0.287, pruned_loss=0.1189, over 458867.34 frames.], batch size: 37, lr: 2.07e-03
+2022-06-18 12:01:42,549 INFO [train.py:874] (2/4) Epoch 3, batch 300, datatang_loss[loss=0.2885, simple_loss=0.3066, pruned_loss=0.1352, over 4968.00 frames.], tot_loss[loss=0.2589, simple_loss=0.2924, pruned_loss=0.1127, over 765632.44 frames.], batch size: 91, aishell_tot_loss[loss=0.2548, simple_loss=0.2962, pruned_loss=0.1067, over 528871.84 frames.], datatang_tot_loss[loss=0.2634, simple_loss=0.2877, pruned_loss=0.1196, over 511582.17 frames.], batch size: 91, lr: 2.07e-03
+2022-06-18 12:02:11,410 INFO [train.py:874] (2/4) Epoch 3, batch 350, datatang_loss[loss=0.303, simple_loss=0.3167, pruned_loss=0.1446, over 4921.00 frames.], tot_loss[loss=0.262, simple_loss=0.2944, pruned_loss=0.1148, over 814170.85 frames.], batch size: 81, aishell_tot_loss[loss=0.2554, simple_loss=0.2969, pruned_loss=0.107, over 574344.11 frames.], datatang_tot_loss[loss=0.2671, simple_loss=0.2905, pruned_loss=0.1219, over 575630.75 frames.], batch size: 81, lr: 2.06e-03
+2022-06-18 12:02:42,549 INFO [train.py:874] (2/4) Epoch 3, batch 400, aishell_loss[loss=0.2307, simple_loss=0.2818, pruned_loss=0.08975, over 4932.00 frames.], tot_loss[loss=0.2602, simple_loss=0.2931, pruned_loss=0.1137, over 852169.43 frames.], batch size: 32, aishell_tot_loss[loss=0.2533, simple_loss=0.2956, pruned_loss=0.1055, over 617394.07 frames.], datatang_tot_loss[loss=0.2666, simple_loss=0.2903, pruned_loss=0.1215, over 629200.67 frames.], batch size: 32, lr: 2.06e-03
+2022-06-18 12:03:13,290 INFO [train.py:874] (2/4) Epoch 3, batch 450, aishell_loss[loss=0.2346, simple_loss=0.284, pruned_loss=0.09261, over 4929.00 frames.], tot_loss[loss=0.2618, simple_loss=0.2952, pruned_loss=0.1142, over 881456.27 frames.], batch size: 33, aishell_tot_loss[loss=0.2551, simple_loss=0.2977, pruned_loss=0.1062, over 667053.15 frames.], datatang_tot_loss[loss=0.2675, simple_loss=0.291, pruned_loss=0.122, over 664641.49 frames.], batch size: 33, lr: 2.05e-03
+2022-06-18 12:03:41,568 INFO [train.py:874] (2/4) Epoch 3, batch 500, datatang_loss[loss=0.3268, simple_loss=0.3266, pruned_loss=0.1635, over 4888.00 frames.], tot_loss[loss=0.2635, simple_loss=0.2963, pruned_loss=0.1154, over 904641.14 frames.], batch size: 25, aishell_tot_loss[loss=0.256, simple_loss=0.2983, pruned_loss=0.1069, over 701772.90 frames.], datatang_tot_loss[loss=0.2688, simple_loss=0.2922, pruned_loss=0.1227, over 705329.29 frames.], batch size: 25, lr: 2.05e-03
+2022-06-18 12:04:12,853 INFO [train.py:874] (2/4) Epoch 3, batch 550, aishell_loss[loss=0.2318, simple_loss=0.2635, pruned_loss=0.1, over 4798.00 frames.], tot_loss[loss=0.2635, simple_loss=0.2965, pruned_loss=0.1152, over 922915.19 frames.], batch size: 24, aishell_tot_loss[loss=0.2549, simple_loss=0.2975, pruned_loss=0.1061, over 735421.58 frames.], datatang_tot_loss[loss=0.2701, simple_loss=0.2936, pruned_loss=0.1233, over 738487.75 frames.], batch size: 24, lr: 2.05e-03
+2022-06-18 12:04:44,434 INFO [train.py:874] (2/4) Epoch 3, batch 600, aishell_loss[loss=0.2636, simple_loss=0.2945, pruned_loss=0.1164, over 4867.00 frames.], tot_loss[loss=0.2663, simple_loss=0.2968, pruned_loss=0.1179, over 936821.18 frames.], batch size: 28, aishell_tot_loss[loss=0.2559, simple_loss=0.2975, pruned_loss=0.1071, over 760195.28 frames.], datatang_tot_loss[loss=0.2727, simple_loss=0.2943, pruned_loss=0.1255, over 772147.38 frames.], batch size: 28, lr: 2.04e-03
+2022-06-18 12:05:11,863 INFO [train.py:874] (2/4) Epoch 3, batch 650, datatang_loss[loss=0.2745, simple_loss=0.2949, pruned_loss=0.1271, over 4917.00 frames.], tot_loss[loss=0.2654, simple_loss=0.2967, pruned_loss=0.117, over 947767.56 frames.], batch size: 75, aishell_tot_loss[loss=0.2569, simple_loss=0.2987, pruned_loss=0.1075, over 785899.34 frames.], datatang_tot_loss[loss=0.271, simple_loss=0.2933, pruned_loss=0.1244, over 798222.54 frames.], batch size: 75, lr: 2.04e-03
+2022-06-18 12:05:43,567 INFO [train.py:874] (2/4) Epoch 3, batch 700, aishell_loss[loss=0.2739, simple_loss=0.3209, pruned_loss=0.1134, over 4876.00 frames.], tot_loss[loss=0.2642, simple_loss=0.2965, pruned_loss=0.1159, over 956139.79 frames.], batch size: 36, aishell_tot_loss[loss=0.2572, simple_loss=0.2993, pruned_loss=0.1075, over 806051.50 frames.], datatang_tot_loss[loss=0.2693, simple_loss=0.2928, pruned_loss=0.1229, over 823366.27 frames.], batch size: 36, lr: 2.03e-03
+2022-06-18 12:06:14,589 INFO [train.py:874] (2/4) Epoch 3, batch 750, aishell_loss[loss=0.2715, simple_loss=0.3033, pruned_loss=0.1198, over 4886.00 frames.], tot_loss[loss=0.2629, simple_loss=0.2966, pruned_loss=0.1146, over 962904.04 frames.], batch size: 34, aishell_tot_loss[loss=0.256, simple_loss=0.2989, pruned_loss=0.1065, over 830499.61 frames.], datatang_tot_loss[loss=0.2695, simple_loss=0.2933, pruned_loss=0.1229, over 839661.05 frames.], batch size: 34, lr: 2.03e-03
+2022-06-18 12:06:42,402 INFO [train.py:874] (2/4) Epoch 3, batch 800, datatang_loss[loss=0.2435, simple_loss=0.2845, pruned_loss=0.1012, over 4930.00 frames.], tot_loss[loss=0.2613, simple_loss=0.296, pruned_loss=0.1133, over 968018.57 frames.], batch size: 50, aishell_tot_loss[loss=0.255, simple_loss=0.2983, pruned_loss=0.1058, over 851457.87 frames.], datatang_tot_loss[loss=0.269, simple_loss=0.2933, pruned_loss=0.1224, over 854352.18 frames.], batch size: 50, lr: 2.02e-03
+2022-06-18 12:07:14,028 INFO [train.py:874] (2/4) Epoch 3, batch 850, datatang_loss[loss=0.2644, simple_loss=0.2925, pruned_loss=0.1182, over 4914.00 frames.], tot_loss[loss=0.2619, simple_loss=0.2967, pruned_loss=0.1136, over 972150.66 frames.], batch size: 57, aishell_tot_loss[loss=0.2539, simple_loss=0.2979, pruned_loss=0.1049, over 866812.46 frames.], datatang_tot_loss[loss=0.2705, simple_loss=0.2945, pruned_loss=0.1232, over 870491.43 frames.], batch size: 57, lr: 2.02e-03
+2022-06-18 12:07:44,970 INFO [train.py:874] (2/4) Epoch 3, batch 900, aishell_loss[loss=0.2653, simple_loss=0.2961, pruned_loss=0.1173, over 4958.00 frames.], tot_loss[loss=0.2629, simple_loss=0.2972, pruned_loss=0.1142, over 975457.19 frames.], batch size: 27, aishell_tot_loss[loss=0.2548, simple_loss=0.2984, pruned_loss=0.1056, over 882557.13 frames.], datatang_tot_loss[loss=0.2709, simple_loss=0.2948, pruned_loss=0.1235, over 882661.85 frames.], batch size: 27, lr: 2.02e-03
+2022-06-18 12:08:13,030 INFO [train.py:874] (2/4) Epoch 3, batch 950, datatang_loss[loss=0.2941, simple_loss=0.3089, pruned_loss=0.1396, over 4933.00 frames.], tot_loss[loss=0.2619, simple_loss=0.2965, pruned_loss=0.1136, over 977722.34 frames.], batch size: 50, aishell_tot_loss[loss=0.2543, simple_loss=0.2981, pruned_loss=0.1053, over 892404.38 frames.], datatang_tot_loss[loss=0.2698, simple_loss=0.2944, pruned_loss=0.1226, over 897026.59 frames.], batch size: 50, lr: 2.01e-03
+2022-06-18 12:08:44,383 INFO [train.py:874] (2/4) Epoch 3, batch 1000, datatang_loss[loss=0.2964, simple_loss=0.3029, pruned_loss=0.1449, over 4931.00 frames.], tot_loss[loss=0.2604, simple_loss=0.2953, pruned_loss=0.1127, over 979404.14 frames.], batch size: 57, aishell_tot_loss[loss=0.2531, simple_loss=0.2971, pruned_loss=0.1046, over 904055.63 frames.], datatang_tot_loss[loss=0.2695, simple_loss=0.2942, pruned_loss=0.1224, over 906759.58 frames.], batch size: 57, lr: 2.01e-03
+2022-06-18 12:08:44,383 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 12:09:01,536 INFO [train.py:914] (2/4) Epoch 3, validation: loss=0.1966, simple_loss=0.2677, pruned_loss=0.06274, over 1622729.00 frames.
+2022-06-18 12:09:30,024 INFO [train.py:874] (2/4) Epoch 3, batch 1050, datatang_loss[loss=0.2142, simple_loss=0.259, pruned_loss=0.08467, over 4910.00 frames.], tot_loss[loss=0.2604, simple_loss=0.295, pruned_loss=0.1129, over 980373.10 frames.], batch size: 42, aishell_tot_loss[loss=0.253, simple_loss=0.2971, pruned_loss=0.1045, over 913268.27 frames.], datatang_tot_loss[loss=0.2693, simple_loss=0.2938, pruned_loss=0.1224, over 915994.72 frames.], batch size: 42, lr: 2.00e-03
+2022-06-18 12:10:01,716 INFO [train.py:874] (2/4) Epoch 3, batch 1100, aishell_loss[loss=0.2663, simple_loss=0.3204, pruned_loss=0.1061, over 4927.00 frames.], tot_loss[loss=0.2588, simple_loss=0.2942, pruned_loss=0.1116, over 981124.18 frames.], batch size: 68, aishell_tot_loss[loss=0.2526, simple_loss=0.2969, pruned_loss=0.1041, over 922917.10 frames.], datatang_tot_loss[loss=0.2682, simple_loss=0.2929, pruned_loss=0.1217, over 922662.27 frames.], batch size: 68, lr: 2.00e-03
+2022-06-18 12:10:29,521 INFO [train.py:874] (2/4) Epoch 3, batch 1150, datatang_loss[loss=0.2733, simple_loss=0.2979, pruned_loss=0.1244, over 4943.00 frames.], tot_loss[loss=0.2605, simple_loss=0.2954, pruned_loss=0.1128, over 982301.16 frames.], batch size: 37, aishell_tot_loss[loss=0.2529, simple_loss=0.2975, pruned_loss=0.1041, over 930089.54 frames.], datatang_tot_loss[loss=0.2693, simple_loss=0.2935, pruned_loss=0.1226, over 930472.64 frames.], batch size: 37, lr: 2.00e-03
+2022-06-18 12:11:00,715 INFO [train.py:874] (2/4) Epoch 3, batch 1200, aishell_loss[loss=0.2496, simple_loss=0.2927, pruned_loss=0.1032, over 4934.00 frames.], tot_loss[loss=0.2597, simple_loss=0.2945, pruned_loss=0.1125, over 983015.55 frames.], batch size: 49, aishell_tot_loss[loss=0.2522, simple_loss=0.2964, pruned_loss=0.104, over 937035.77 frames.], datatang_tot_loss[loss=0.2692, simple_loss=0.2935, pruned_loss=0.1224, over 936548.79 frames.], batch size: 49, lr: 1.99e-03
+2022-06-18 12:11:31,967 INFO [train.py:874] (2/4) Epoch 3, batch 1250, datatang_loss[loss=0.2948, simple_loss=0.2972, pruned_loss=0.1462, over 4919.00 frames.], tot_loss[loss=0.2606, simple_loss=0.2949, pruned_loss=0.1131, over 983278.61 frames.], batch size: 64, aishell_tot_loss[loss=0.2512, simple_loss=0.2956, pruned_loss=0.1034, over 943280.32 frames.], datatang_tot_loss[loss=0.2713, simple_loss=0.2946, pruned_loss=0.124, over 941460.83 frames.], batch size: 64, lr: 1.99e-03
+2022-06-18 12:11:59,373 INFO [train.py:874] (2/4) Epoch 3, batch 1300, datatang_loss[loss=0.2849, simple_loss=0.3173, pruned_loss=0.1262, over 4917.00 frames.], tot_loss[loss=0.2594, simple_loss=0.2946, pruned_loss=0.1121, over 983983.95 frames.], batch size: 98, aishell_tot_loss[loss=0.251, simple_loss=0.2958, pruned_loss=0.1031, over 948776.28 frames.], datatang_tot_loss[loss=0.2704, simple_loss=0.294, pruned_loss=0.1234, over 946304.99 frames.], batch size: 98, lr: 1.98e-03
+2022-06-18 12:12:30,424 INFO [train.py:874] (2/4) Epoch 3, batch 1350, datatang_loss[loss=0.2771, simple_loss=0.2961, pruned_loss=0.1291, over 4925.00 frames.], tot_loss[loss=0.2593, simple_loss=0.2945, pruned_loss=0.112, over 984174.35 frames.], batch size: 71, aishell_tot_loss[loss=0.2508, simple_loss=0.2957, pruned_loss=0.1029, over 952113.27 frames.], datatang_tot_loss[loss=0.2697, simple_loss=0.294, pruned_loss=0.1227, over 951772.11 frames.], batch size: 71, lr: 1.98e-03
+2022-06-18 12:13:01,673 INFO [train.py:874] (2/4) Epoch 3, batch 1400, datatang_loss[loss=0.2459, simple_loss=0.2829, pruned_loss=0.1044, over 4924.00 frames.], tot_loss[loss=0.2583, simple_loss=0.2945, pruned_loss=0.111, over 983979.56 frames.], batch size: 81, aishell_tot_loss[loss=0.2514, simple_loss=0.2965, pruned_loss=0.1032, over 956580.95 frames.], datatang_tot_loss[loss=0.2682, simple_loss=0.2931, pruned_loss=0.1216, over 954660.40 frames.], batch size: 81, lr: 1.97e-03
+2022-06-18 12:13:29,190 INFO [train.py:874] (2/4) Epoch 3, batch 1450, aishell_loss[loss=0.2303, simple_loss=0.2816, pruned_loss=0.08944, over 4928.00 frames.], tot_loss[loss=0.2584, simple_loss=0.2947, pruned_loss=0.111, over 984158.95 frames.], batch size: 33, aishell_tot_loss[loss=0.2509, simple_loss=0.2961, pruned_loss=0.1028, over 960506.80 frames.], datatang_tot_loss[loss=0.2689, simple_loss=0.2935, pruned_loss=0.1222, over 957544.07 frames.], batch size: 33, lr: 1.97e-03
+2022-06-18 12:14:01,022 INFO [train.py:874] (2/4) Epoch 3, batch 1500, datatang_loss[loss=0.2346, simple_loss=0.2746, pruned_loss=0.09734, over 4950.00 frames.], tot_loss[loss=0.2608, simple_loss=0.2958, pruned_loss=0.1129, over 984374.01 frames.], batch size: 55, aishell_tot_loss[loss=0.2517, simple_loss=0.2969, pruned_loss=0.1032, over 963429.88 frames.], datatang_tot_loss[loss=0.2704, simple_loss=0.2939, pruned_loss=0.1235, over 960781.54 frames.], batch size: 55, lr: 1.97e-03
+2022-06-18 12:14:30,079 INFO [train.py:874] (2/4) Epoch 3, batch 1550, aishell_loss[loss=0.2504, simple_loss=0.299, pruned_loss=0.1009, over 4928.00 frames.], tot_loss[loss=0.2603, simple_loss=0.2958, pruned_loss=0.1124, over 984431.72 frames.], batch size: 49, aishell_tot_loss[loss=0.2516, simple_loss=0.2973, pruned_loss=0.1029, over 965911.62 frames.], datatang_tot_loss[loss=0.2699, simple_loss=0.2937, pruned_loss=0.1231, over 963585.26 frames.], batch size: 49, lr: 1.96e-03
+2022-06-18 12:14:59,980 INFO [train.py:874] (2/4) Epoch 3, batch 1600, datatang_loss[loss=0.2708, simple_loss=0.2962, pruned_loss=0.1226, over 4959.00 frames.], tot_loss[loss=0.2579, simple_loss=0.294, pruned_loss=0.1109, over 984605.11 frames.], batch size: 55, aishell_tot_loss[loss=0.2504, simple_loss=0.2963, pruned_loss=0.1023, over 967840.83 frames.], datatang_tot_loss[loss=0.2682, simple_loss=0.2928, pruned_loss=0.1218, over 966489.74 frames.], batch size: 55, lr: 1.96e-03
+2022-06-18 12:15:30,919 INFO [train.py:874] (2/4) Epoch 3, batch 1650, aishell_loss[loss=0.2356, simple_loss=0.2861, pruned_loss=0.09254, over 4978.00 frames.], tot_loss[loss=0.2559, simple_loss=0.2926, pruned_loss=0.1096, over 984756.01 frames.], batch size: 39, aishell_tot_loss[loss=0.2497, simple_loss=0.2955, pruned_loss=0.102, over 970051.55 frames.], datatang_tot_loss[loss=0.2668, simple_loss=0.292, pruned_loss=0.1208, over 968506.58 frames.], batch size: 39, lr: 1.96e-03
+2022-06-18 12:16:00,111 INFO [train.py:874] (2/4) Epoch 3, batch 1700, aishell_loss[loss=0.2512, simple_loss=0.3081, pruned_loss=0.09711, over 4890.00 frames.], tot_loss[loss=0.2568, simple_loss=0.2934, pruned_loss=0.1101, over 984858.88 frames.], batch size: 34, aishell_tot_loss[loss=0.2495, simple_loss=0.2954, pruned_loss=0.1018, over 971180.87 frames.], datatang_tot_loss[loss=0.2668, simple_loss=0.2925, pruned_loss=0.1205, over 971160.17 frames.], batch size: 34, lr: 1.95e-03
+2022-06-18 12:16:30,618 INFO [train.py:874] (2/4) Epoch 3, batch 1750, datatang_loss[loss=0.2764, simple_loss=0.3148, pruned_loss=0.119, over 4933.00 frames.], tot_loss[loss=0.2576, simple_loss=0.2943, pruned_loss=0.1105, over 984790.17 frames.], batch size: 94, aishell_tot_loss[loss=0.2499, simple_loss=0.296, pruned_loss=0.1019, over 972913.92 frames.], datatang_tot_loss[loss=0.267, simple_loss=0.2928, pruned_loss=0.1207, over 972588.61 frames.], batch size: 94, lr: 1.95e-03
+2022-06-18 12:17:02,287 INFO [train.py:874] (2/4) Epoch 3, batch 1800, datatang_loss[loss=0.2928, simple_loss=0.303, pruned_loss=0.1413, over 4940.00 frames.], tot_loss[loss=0.2565, simple_loss=0.2939, pruned_loss=0.1096, over 984770.18 frames.], batch size: 50, aishell_tot_loss[loss=0.249, simple_loss=0.2956, pruned_loss=0.1012, over 974122.73 frames.], datatang_tot_loss[loss=0.2668, simple_loss=0.2927, pruned_loss=0.1204, over 974195.14 frames.], batch size: 50, lr: 1.94e-03
+2022-06-18 12:17:29,486 INFO [train.py:874] (2/4) Epoch 3, batch 1850, datatang_loss[loss=0.2622, simple_loss=0.2934, pruned_loss=0.1154, over 4924.00 frames.], tot_loss[loss=0.2561, simple_loss=0.2941, pruned_loss=0.109, over 984958.18 frames.], batch size: 83, aishell_tot_loss[loss=0.2488, simple_loss=0.2958, pruned_loss=0.1009, over 975554.15 frames.], datatang_tot_loss[loss=0.2665, simple_loss=0.2927, pruned_loss=0.1201, over 975469.96 frames.], batch size: 83, lr: 1.94e-03
+2022-06-18 12:18:00,798 INFO [train.py:874] (2/4) Epoch 3, batch 1900, aishell_loss[loss=0.2044, simple_loss=0.245, pruned_loss=0.08189, over 4936.00 frames.], tot_loss[loss=0.2551, simple_loss=0.2935, pruned_loss=0.1084, over 985271.65 frames.], batch size: 25, aishell_tot_loss[loss=0.2484, simple_loss=0.2956, pruned_loss=0.1006, over 976792.11 frames.], datatang_tot_loss[loss=0.2655, simple_loss=0.2922, pruned_loss=0.1194, over 976791.59 frames.], batch size: 25, lr: 1.94e-03
+2022-06-18 12:18:32,155 INFO [train.py:874] (2/4) Epoch 3, batch 1950, aishell_loss[loss=0.2017, simple_loss=0.2559, pruned_loss=0.07377, over 4946.00 frames.], tot_loss[loss=0.2538, simple_loss=0.293, pruned_loss=0.1073, over 985620.85 frames.], batch size: 27, aishell_tot_loss[loss=0.2482, simple_loss=0.2958, pruned_loss=0.1003, over 977832.35 frames.], datatang_tot_loss[loss=0.264, simple_loss=0.2915, pruned_loss=0.1182, over 978119.29 frames.], batch size: 27, lr: 1.93e-03
+2022-06-18 12:18:59,665 INFO [train.py:874] (2/4) Epoch 3, batch 2000, aishell_loss[loss=0.239, simple_loss=0.3037, pruned_loss=0.08712, over 4913.00 frames.], tot_loss[loss=0.2558, simple_loss=0.2937, pruned_loss=0.1089, over 985766.23 frames.], batch size: 46, aishell_tot_loss[loss=0.2481, simple_loss=0.2955, pruned_loss=0.1003, over 978793.34 frames.], datatang_tot_loss[loss=0.2656, simple_loss=0.2924, pruned_loss=0.1194, over 979126.65 frames.], batch size: 46, lr: 1.93e-03
+2022-06-18 12:18:59,666 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 12:19:15,682 INFO [train.py:914] (2/4) Epoch 3, validation: loss=0.1934, simple_loss=0.2657, pruned_loss=0.0605, over 1622729.00 frames.
+2022-06-18 12:19:43,609 INFO [train.py:874] (2/4) Epoch 3, batch 2050, aishell_loss[loss=0.271, simple_loss=0.3064, pruned_loss=0.1178, over 4954.00 frames.], tot_loss[loss=0.2566, simple_loss=0.294, pruned_loss=0.1096, over 985542.16 frames.], batch size: 40, aishell_tot_loss[loss=0.2474, simple_loss=0.295, pruned_loss=0.09987, over 979235.05 frames.], datatang_tot_loss[loss=0.2665, simple_loss=0.2932, pruned_loss=0.12, over 980048.85 frames.], batch size: 40, lr: 1.92e-03
+2022-06-18 12:20:13,697 INFO [train.py:874] (2/4) Epoch 3, batch 2100, datatang_loss[loss=0.2305, simple_loss=0.264, pruned_loss=0.09854, over 4853.00 frames.], tot_loss[loss=0.2555, simple_loss=0.2933, pruned_loss=0.1089, over 985320.19 frames.], batch size: 30, aishell_tot_loss[loss=0.2478, simple_loss=0.2955, pruned_loss=0.1, over 979743.72 frames.], datatang_tot_loss[loss=0.2646, simple_loss=0.292, pruned_loss=0.1186, over 980672.43 frames.], batch size: 30, lr: 1.92e-03
+2022-06-18 12:20:44,266 INFO [train.py:874] (2/4) Epoch 3, batch 2150, datatang_loss[loss=0.2645, simple_loss=0.2807, pruned_loss=0.1241, over 4941.00 frames.], tot_loss[loss=0.2528, simple_loss=0.2918, pruned_loss=0.1069, over 985066.50 frames.], batch size: 37, aishell_tot_loss[loss=0.246, simple_loss=0.2941, pruned_loss=0.09894, over 980165.25 frames.], datatang_tot_loss[loss=0.2639, simple_loss=0.2916, pruned_loss=0.1181, over 981215.80 frames.], batch size: 37, lr: 1.92e-03
+2022-06-18 12:21:13,462 INFO [train.py:874] (2/4) Epoch 3, batch 2200, aishell_loss[loss=0.2456, simple_loss=0.2846, pruned_loss=0.1033, over 4930.00 frames.], tot_loss[loss=0.2537, simple_loss=0.2921, pruned_loss=0.1076, over 985286.28 frames.], batch size: 33, aishell_tot_loss[loss=0.2457, simple_loss=0.2941, pruned_loss=0.09864, over 980837.80 frames.], datatang_tot_loss[loss=0.2645, simple_loss=0.2917, pruned_loss=0.1187, over 981806.73 frames.], batch size: 33, lr: 1.91e-03
+2022-06-18 12:21:43,212 INFO [train.py:874] (2/4) Epoch 3, batch 2250, aishell_loss[loss=0.2419, simple_loss=0.2945, pruned_loss=0.09463, over 4974.00 frames.], tot_loss[loss=0.2518, simple_loss=0.2914, pruned_loss=0.1061, over 985224.75 frames.], batch size: 51, aishell_tot_loss[loss=0.2444, simple_loss=0.2932, pruned_loss=0.0978, over 981465.12 frames.], datatang_tot_loss[loss=0.2641, simple_loss=0.2916, pruned_loss=0.1183, over 982070.67 frames.], batch size: 51, lr: 1.91e-03
+2022-06-18 12:22:13,499 INFO [train.py:874] (2/4) Epoch 3, batch 2300, datatang_loss[loss=0.2935, simple_loss=0.3128, pruned_loss=0.1371, over 4911.00 frames.], tot_loss[loss=0.251, simple_loss=0.2909, pruned_loss=0.1055, over 984816.87 frames.], batch size: 57, aishell_tot_loss[loss=0.244, simple_loss=0.2929, pruned_loss=0.09751, over 981673.01 frames.], datatang_tot_loss[loss=0.2632, simple_loss=0.2911, pruned_loss=0.1176, over 982255.48 frames.], batch size: 57, lr: 1.91e-03
+2022-06-18 12:22:41,794 INFO [train.py:874] (2/4) Epoch 3, batch 2350, datatang_loss[loss=0.2534, simple_loss=0.2903, pruned_loss=0.1082, over 4953.00 frames.], tot_loss[loss=0.2503, simple_loss=0.2908, pruned_loss=0.1049, over 984874.34 frames.], batch size: 62, aishell_tot_loss[loss=0.2437, simple_loss=0.293, pruned_loss=0.09715, over 981924.52 frames.], datatang_tot_loss[loss=0.2622, simple_loss=0.2906, pruned_loss=0.1169, over 982715.74 frames.], batch size: 62, lr: 1.90e-03
+2022-06-18 12:23:13,294 INFO [train.py:874] (2/4) Epoch 3, batch 2400, datatang_loss[loss=0.2439, simple_loss=0.2776, pruned_loss=0.1052, over 4942.00 frames.], tot_loss[loss=0.2511, simple_loss=0.2914, pruned_loss=0.1054, over 985016.47 frames.], batch size: 50, aishell_tot_loss[loss=0.2436, simple_loss=0.2931, pruned_loss=0.09708, over 982169.29 frames.], datatang_tot_loss[loss=0.2625, simple_loss=0.291, pruned_loss=0.117, over 983219.07 frames.], batch size: 50, lr: 1.90e-03
+2022-06-18 12:23:44,005 INFO [train.py:874] (2/4) Epoch 3, batch 2450, datatang_loss[loss=0.2541, simple_loss=0.2902, pruned_loss=0.109, over 4923.00 frames.], tot_loss[loss=0.2514, simple_loss=0.2912, pruned_loss=0.1058, over 985130.83 frames.], batch size: 64, aishell_tot_loss[loss=0.2445, simple_loss=0.2939, pruned_loss=0.0976, over 982471.98 frames.], datatang_tot_loss[loss=0.2611, simple_loss=0.2898, pruned_loss=0.1162, over 983571.64 frames.], batch size: 64, lr: 1.89e-03
+2022-06-18 12:24:13,118 INFO [train.py:874] (2/4) Epoch 3, batch 2500, aishell_loss[loss=0.2581, simple_loss=0.3124, pruned_loss=0.1019, over 4943.00 frames.], tot_loss[loss=0.2517, simple_loss=0.2914, pruned_loss=0.106, over 985666.72 frames.], batch size: 49, aishell_tot_loss[loss=0.2444, simple_loss=0.294, pruned_loss=0.09739, over 982967.15 frames.], datatang_tot_loss[loss=0.2612, simple_loss=0.2899, pruned_loss=0.1163, over 984137.58 frames.], batch size: 49, lr: 1.89e-03
+2022-06-18 12:24:44,572 INFO [train.py:874] (2/4) Epoch 3, batch 2550, aishell_loss[loss=0.2324, simple_loss=0.2942, pruned_loss=0.08535, over 4941.00 frames.], tot_loss[loss=0.2526, simple_loss=0.2921, pruned_loss=0.1065, over 985748.69 frames.], batch size: 56, aishell_tot_loss[loss=0.2442, simple_loss=0.2937, pruned_loss=0.09735, over 983454.11 frames.], datatang_tot_loss[loss=0.2622, simple_loss=0.2907, pruned_loss=0.1168, over 984252.19 frames.], batch size: 56, lr: 1.89e-03
+2022-06-18 12:25:14,264 INFO [train.py:874] (2/4) Epoch 3, batch 2600, datatang_loss[loss=0.3824, simple_loss=0.3829, pruned_loss=0.1909, over 4941.00 frames.], tot_loss[loss=0.2517, simple_loss=0.2917, pruned_loss=0.1059, over 986076.36 frames.], batch size: 108, aishell_tot_loss[loss=0.2432, simple_loss=0.2934, pruned_loss=0.09656, over 983935.35 frames.], datatang_tot_loss[loss=0.2622, simple_loss=0.2906, pruned_loss=0.1169, over 984569.74 frames.], batch size: 108, lr: 1.88e-03
+2022-06-18 12:25:43,667 INFO [train.py:874] (2/4) Epoch 3, batch 2650, datatang_loss[loss=0.2382, simple_loss=0.2724, pruned_loss=0.102, over 4911.00 frames.], tot_loss[loss=0.2514, simple_loss=0.2911, pruned_loss=0.1059, over 985830.30 frames.], batch size: 64, aishell_tot_loss[loss=0.2431, simple_loss=0.293, pruned_loss=0.09657, over 984056.28 frames.], datatang_tot_loss[loss=0.2617, simple_loss=0.2902, pruned_loss=0.1166, over 984608.28 frames.], batch size: 64, lr: 1.88e-03
+2022-06-18 12:26:14,978 INFO [train.py:874] (2/4) Epoch 3, batch 2700, aishell_loss[loss=0.2618, simple_loss=0.2979, pruned_loss=0.1129, over 4943.00 frames.], tot_loss[loss=0.2499, simple_loss=0.2904, pruned_loss=0.1047, over 985509.10 frames.], batch size: 49, aishell_tot_loss[loss=0.2428, simple_loss=0.293, pruned_loss=0.09636, over 983880.23 frames.], datatang_tot_loss[loss=0.2604, simple_loss=0.2894, pruned_loss=0.1157, over 984807.85 frames.], batch size: 49, lr: 1.88e-03
+2022-06-18 12:26:44,657 INFO [train.py:874] (2/4) Epoch 3, batch 2750, aishell_loss[loss=0.2612, simple_loss=0.3034, pruned_loss=0.1096, over 4938.00 frames.], tot_loss[loss=0.2491, simple_loss=0.2904, pruned_loss=0.1039, over 985496.77 frames.], batch size: 41, aishell_tot_loss[loss=0.2433, simple_loss=0.2936, pruned_loss=0.09647, over 984091.65 frames.], datatang_tot_loss[loss=0.259, simple_loss=0.2884, pruned_loss=0.1148, over 984877.24 frames.], batch size: 41, lr: 1.87e-03
+2022-06-18 12:27:14,023 INFO [train.py:874] (2/4) Epoch 3, batch 2800, aishell_loss[loss=0.2473, simple_loss=0.3054, pruned_loss=0.09457, over 4909.00 frames.], tot_loss[loss=0.2495, simple_loss=0.29, pruned_loss=0.1044, over 985499.46 frames.], batch size: 41, aishell_tot_loss[loss=0.2428, simple_loss=0.2932, pruned_loss=0.0962, over 984229.25 frames.], datatang_tot_loss[loss=0.2593, simple_loss=0.2883, pruned_loss=0.1151, over 984969.78 frames.], batch size: 41, lr: 1.87e-03
+2022-06-18 12:27:45,784 INFO [train.py:874] (2/4) Epoch 3, batch 2850, aishell_loss[loss=0.2201, simple_loss=0.282, pruned_loss=0.07907, over 4949.00 frames.], tot_loss[loss=0.2471, simple_loss=0.2889, pruned_loss=0.1027, over 985265.54 frames.], batch size: 68, aishell_tot_loss[loss=0.241, simple_loss=0.2918, pruned_loss=0.09516, over 984102.73 frames.], datatang_tot_loss[loss=0.2587, simple_loss=0.2884, pruned_loss=0.1145, over 985056.04 frames.], batch size: 68, lr: 1.87e-03
+2022-06-18 12:28:14,898 INFO [train.py:874] (2/4) Epoch 3, batch 2900, datatang_loss[loss=0.2299, simple_loss=0.2678, pruned_loss=0.09601, over 4985.00 frames.], tot_loss[loss=0.2472, simple_loss=0.2891, pruned_loss=0.1027, over 985567.86 frames.], batch size: 37, aishell_tot_loss[loss=0.2408, simple_loss=0.2918, pruned_loss=0.09492, over 984418.57 frames.], datatang_tot_loss[loss=0.2583, simple_loss=0.2882, pruned_loss=0.1142, over 985223.59 frames.], batch size: 37, lr: 1.86e-03
+2022-06-18 12:28:45,258 INFO [train.py:874] (2/4) Epoch 3, batch 2950, datatang_loss[loss=0.2456, simple_loss=0.2744, pruned_loss=0.1085, over 4866.00 frames.], tot_loss[loss=0.2475, simple_loss=0.2887, pruned_loss=0.1032, over 985377.64 frames.], batch size: 30, aishell_tot_loss[loss=0.2416, simple_loss=0.2919, pruned_loss=0.09567, over 984418.05 frames.], datatang_tot_loss[loss=0.2571, simple_loss=0.2875, pruned_loss=0.1134, over 985194.69 frames.], batch size: 30, lr: 1.86e-03
+2022-06-18 12:29:16,450 INFO [train.py:874] (2/4) Epoch 3, batch 3000, aishell_loss[loss=0.2207, simple_loss=0.2829, pruned_loss=0.07927, over 4951.00 frames.], tot_loss[loss=0.2467, simple_loss=0.2885, pruned_loss=0.1025, over 985689.79 frames.], batch size: 45, aishell_tot_loss[loss=0.2408, simple_loss=0.2915, pruned_loss=0.09508, over 984722.12 frames.], datatang_tot_loss[loss=0.2569, simple_loss=0.2874, pruned_loss=0.1132, over 985363.26 frames.], batch size: 45, lr: 1.86e-03
+2022-06-18 12:29:16,451 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 12:29:31,998 INFO [train.py:914] (2/4) Epoch 3, validation: loss=0.1943, simple_loss=0.2686, pruned_loss=0.05999, over 1622729.00 frames.
+2022-06-18 12:30:03,145 INFO [train.py:874] (2/4) Epoch 3, batch 3050, datatang_loss[loss=0.2418, simple_loss=0.2726, pruned_loss=0.1055, over 4961.00 frames.], tot_loss[loss=0.2466, simple_loss=0.2884, pruned_loss=0.1023, over 985920.65 frames.], batch size: 55, aishell_tot_loss[loss=0.2397, simple_loss=0.2906, pruned_loss=0.09447, over 984984.41 frames.], datatang_tot_loss[loss=0.2575, simple_loss=0.288, pruned_loss=0.1135, over 985510.12 frames.], batch size: 55, lr: 1.85e-03
+2022-06-18 12:30:33,637 INFO [train.py:874] (2/4) Epoch 3, batch 3100, datatang_loss[loss=0.2242, simple_loss=0.2569, pruned_loss=0.09578, over 4919.00 frames.], tot_loss[loss=0.2459, simple_loss=0.2873, pruned_loss=0.1022, over 985933.82 frames.], batch size: 75, aishell_tot_loss[loss=0.239, simple_loss=0.2897, pruned_loss=0.09418, over 985055.78 frames.], datatang_tot_loss[loss=0.2569, simple_loss=0.2875, pruned_loss=0.1131, over 985598.52 frames.], batch size: 75, lr: 1.85e-03
+2022-06-18 12:31:02,147 INFO [train.py:874] (2/4) Epoch 3, batch 3150, datatang_loss[loss=0.3718, simple_loss=0.3652, pruned_loss=0.1892, over 4950.00 frames.], tot_loss[loss=0.2464, simple_loss=0.288, pruned_loss=0.1024, over 985900.75 frames.], batch size: 108, aishell_tot_loss[loss=0.239, simple_loss=0.29, pruned_loss=0.09399, over 985112.96 frames.], datatang_tot_loss[loss=0.2569, simple_loss=0.2876, pruned_loss=0.1131, over 985659.56 frames.], batch size: 108, lr: 1.85e-03
+2022-06-18 12:31:33,992 INFO [train.py:874] (2/4) Epoch 3, batch 3200, datatang_loss[loss=0.2213, simple_loss=0.2619, pruned_loss=0.09034, over 4888.00 frames.], tot_loss[loss=0.2477, simple_loss=0.2894, pruned_loss=0.103, over 985917.27 frames.], batch size: 25, aishell_tot_loss[loss=0.2393, simple_loss=0.2903, pruned_loss=0.09409, over 985106.27 frames.], datatang_tot_loss[loss=0.2578, simple_loss=0.2884, pruned_loss=0.1136, over 985823.09 frames.], batch size: 25, lr: 1.84e-03
+2022-06-18 12:32:03,695 INFO [train.py:874] (2/4) Epoch 3, batch 3250, aishell_loss[loss=0.194, simple_loss=0.2585, pruned_loss=0.06474, over 4970.00 frames.], tot_loss[loss=0.2462, simple_loss=0.2887, pruned_loss=0.1019, over 986005.01 frames.], batch size: 30, aishell_tot_loss[loss=0.2386, simple_loss=0.2899, pruned_loss=0.09362, over 985214.39 frames.], datatang_tot_loss[loss=0.2574, simple_loss=0.2882, pruned_loss=0.1133, over 985939.58 frames.], batch size: 30, lr: 1.84e-03
+2022-06-18 12:32:33,310 INFO [train.py:874] (2/4) Epoch 3, batch 3300, aishell_loss[loss=0.2285, simple_loss=0.2895, pruned_loss=0.08375, over 4940.00 frames.], tot_loss[loss=0.2481, simple_loss=0.2898, pruned_loss=0.1032, over 985712.85 frames.], batch size: 49, aishell_tot_loss[loss=0.2401, simple_loss=0.2912, pruned_loss=0.09455, over 984935.64 frames.], datatang_tot_loss[loss=0.2569, simple_loss=0.2879, pruned_loss=0.113, over 985992.95 frames.], batch size: 49, lr: 1.84e-03
+2022-06-18 12:33:03,642 INFO [train.py:874] (2/4) Epoch 3, batch 3350, datatang_loss[loss=0.2442, simple_loss=0.2869, pruned_loss=0.1008, over 4959.00 frames.], tot_loss[loss=0.2464, simple_loss=0.289, pruned_loss=0.1019, over 985875.73 frames.], batch size: 91, aishell_tot_loss[loss=0.2388, simple_loss=0.2904, pruned_loss=0.0936, over 985027.23 frames.], datatang_tot_loss[loss=0.2567, simple_loss=0.2878, pruned_loss=0.1128, over 986156.33 frames.], batch size: 91, lr: 1.83e-03
+2022-06-18 12:33:33,268 INFO [train.py:874] (2/4) Epoch 3, batch 3400, aishell_loss[loss=0.2293, simple_loss=0.2931, pruned_loss=0.08272, over 4901.00 frames.], tot_loss[loss=0.2454, simple_loss=0.2884, pruned_loss=0.1012, over 985684.88 frames.], batch size: 41, aishell_tot_loss[loss=0.2386, simple_loss=0.2903, pruned_loss=0.09348, over 984756.32 frames.], datatang_tot_loss[loss=0.2556, simple_loss=0.2872, pruned_loss=0.112, over 986302.72 frames.], batch size: 41, lr: 1.83e-03
+2022-06-18 12:34:03,640 INFO [train.py:874] (2/4) Epoch 3, batch 3450, datatang_loss[loss=0.2459, simple_loss=0.2795, pruned_loss=0.1062, over 4944.00 frames.], tot_loss[loss=0.2438, simple_loss=0.2873, pruned_loss=0.1002, over 985605.17 frames.], batch size: 69, aishell_tot_loss[loss=0.2371, simple_loss=0.2893, pruned_loss=0.09248, over 984943.71 frames.], datatang_tot_loss[loss=0.2552, simple_loss=0.287, pruned_loss=0.1118, over 986089.39 frames.], batch size: 69, lr: 1.83e-03
+2022-06-18 12:34:34,624 INFO [train.py:874] (2/4) Epoch 3, batch 3500, aishell_loss[loss=0.2004, simple_loss=0.2694, pruned_loss=0.06571, over 4914.00 frames.], tot_loss[loss=0.2439, simple_loss=0.2871, pruned_loss=0.1003, over 985653.89 frames.], batch size: 41, aishell_tot_loss[loss=0.2374, simple_loss=0.2896, pruned_loss=0.09263, over 984987.16 frames.], datatang_tot_loss[loss=0.2545, simple_loss=0.2863, pruned_loss=0.1113, over 986112.81 frames.], batch size: 41, lr: 1.82e-03
+2022-06-18 12:35:02,636 INFO [train.py:874] (2/4) Epoch 3, batch 3550, datatang_loss[loss=0.2403, simple_loss=0.2822, pruned_loss=0.09925, over 4955.00 frames.], tot_loss[loss=0.2438, simple_loss=0.2867, pruned_loss=0.1005, over 985801.74 frames.], batch size: 67, aishell_tot_loss[loss=0.237, simple_loss=0.2891, pruned_loss=0.09242, over 985206.25 frames.], datatang_tot_loss[loss=0.2542, simple_loss=0.2863, pruned_loss=0.111, over 986062.99 frames.], batch size: 67, lr: 1.82e-03
+2022-06-18 12:35:34,109 INFO [train.py:874] (2/4) Epoch 3, batch 3600, aishell_loss[loss=0.2094, simple_loss=0.2582, pruned_loss=0.08031, over 4981.00 frames.], tot_loss[loss=0.2458, simple_loss=0.2882, pruned_loss=0.1017, over 985786.97 frames.], batch size: 27, aishell_tot_loss[loss=0.2378, simple_loss=0.2898, pruned_loss=0.09294, over 985102.39 frames.], datatang_tot_loss[loss=0.2551, simple_loss=0.2869, pruned_loss=0.1116, over 986198.17 frames.], batch size: 27, lr: 1.82e-03
+2022-06-18 12:36:04,024 INFO [train.py:874] (2/4) Epoch 3, batch 3650, datatang_loss[loss=0.2238, simple_loss=0.2697, pruned_loss=0.08899, over 4921.00 frames.], tot_loss[loss=0.2457, simple_loss=0.2878, pruned_loss=0.1018, over 985984.61 frames.], batch size: 83, aishell_tot_loss[loss=0.2381, simple_loss=0.29, pruned_loss=0.09311, over 985204.72 frames.], datatang_tot_loss[loss=0.254, simple_loss=0.2862, pruned_loss=0.1109, over 986309.84 frames.], batch size: 83, lr: 1.81e-03
+2022-06-18 12:36:39,221 INFO [train.py:874] (2/4) Epoch 3, batch 3700, aishell_loss[loss=0.234, simple_loss=0.2922, pruned_loss=0.08786, over 4951.00 frames.], tot_loss[loss=0.2445, simple_loss=0.2868, pruned_loss=0.1011, over 985903.59 frames.], batch size: 45, aishell_tot_loss[loss=0.2378, simple_loss=0.2898, pruned_loss=0.09293, over 985263.83 frames.], datatang_tot_loss[loss=0.2526, simple_loss=0.2854, pruned_loss=0.1099, over 986194.68 frames.], batch size: 45, lr: 1.81e-03
+2022-06-18 12:37:08,716 INFO [train.py:874] (2/4) Epoch 3, batch 3750, aishell_loss[loss=0.1948, simple_loss=0.2487, pruned_loss=0.07047, over 4958.00 frames.], tot_loss[loss=0.244, simple_loss=0.2862, pruned_loss=0.1009, over 985735.21 frames.], batch size: 27, aishell_tot_loss[loss=0.2377, simple_loss=0.2896, pruned_loss=0.0929, over 985298.13 frames.], datatang_tot_loss[loss=0.2517, simple_loss=0.2849, pruned_loss=0.1092, over 985998.63 frames.], batch size: 27, lr: 1.81e-03
+2022-06-18 12:37:37,176 INFO [train.py:874] (2/4) Epoch 3, batch 3800, datatang_loss[loss=0.254, simple_loss=0.2786, pruned_loss=0.1147, over 4925.00 frames.], tot_loss[loss=0.2455, simple_loss=0.2872, pruned_loss=0.1018, over 985652.36 frames.], batch size: 71, aishell_tot_loss[loss=0.2379, simple_loss=0.2899, pruned_loss=0.093, over 985250.38 frames.], datatang_tot_loss[loss=0.2526, simple_loss=0.2856, pruned_loss=0.1098, over 985951.19 frames.], batch size: 71, lr: 1.80e-03
+2022-06-18 12:38:06,928 INFO [train.py:874] (2/4) Epoch 3, batch 3850, datatang_loss[loss=0.2506, simple_loss=0.2904, pruned_loss=0.1054, over 4913.00 frames.], tot_loss[loss=0.2451, simple_loss=0.2867, pruned_loss=0.1018, over 986149.45 frames.], batch size: 64, aishell_tot_loss[loss=0.238, simple_loss=0.2897, pruned_loss=0.09312, over 985601.15 frames.], datatang_tot_loss[loss=0.2519, simple_loss=0.2852, pruned_loss=0.1093, over 986125.63 frames.], batch size: 64, lr: 1.80e-03
+2022-06-18 12:38:36,202 INFO [train.py:874] (2/4) Epoch 3, batch 3900, datatang_loss[loss=0.2422, simple_loss=0.2652, pruned_loss=0.1096, over 4947.00 frames.], tot_loss[loss=0.2448, simple_loss=0.2871, pruned_loss=0.1012, over 986193.75 frames.], batch size: 50, aishell_tot_loss[loss=0.238, simple_loss=0.29, pruned_loss=0.09303, over 985485.29 frames.], datatang_tot_loss[loss=0.2514, simple_loss=0.2852, pruned_loss=0.1088, over 986348.80 frames.], batch size: 50, lr: 1.80e-03
+2022-06-18 12:39:04,526 INFO [train.py:874] (2/4) Epoch 3, batch 3950, aishell_loss[loss=0.2228, simple_loss=0.2738, pruned_loss=0.08585, over 4948.00 frames.], tot_loss[loss=0.2424, simple_loss=0.2858, pruned_loss=0.09951, over 986118.96 frames.], batch size: 32, aishell_tot_loss[loss=0.2372, simple_loss=0.2896, pruned_loss=0.09245, over 985269.13 frames.], datatang_tot_loss[loss=0.2495, simple_loss=0.2842, pruned_loss=0.1074, over 986517.60 frames.], batch size: 32, lr: 1.79e-03
+2022-06-18 12:39:34,735 INFO [train.py:874] (2/4) Epoch 3, batch 4000, datatang_loss[loss=0.3115, simple_loss=0.3315, pruned_loss=0.1458, over 4947.00 frames.], tot_loss[loss=0.241, simple_loss=0.2843, pruned_loss=0.09879, over 985449.38 frames.], batch size: 109, aishell_tot_loss[loss=0.2366, simple_loss=0.2889, pruned_loss=0.09209, over 984929.34 frames.], datatang_tot_loss[loss=0.2482, simple_loss=0.2831, pruned_loss=0.1067, over 986165.63 frames.], batch size: 109, lr: 1.79e-03
+2022-06-18 12:39:34,736 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 12:39:51,129 INFO [train.py:914] (2/4) Epoch 3, validation: loss=0.1881, simple_loss=0.2644, pruned_loss=0.05586, over 1622729.00 frames.
+2022-06-18 12:40:21,158 INFO [train.py:874] (2/4) Epoch 3, batch 4050, datatang_loss[loss=0.2116, simple_loss=0.2588, pruned_loss=0.08222, over 4947.00 frames.], tot_loss[loss=0.2419, simple_loss=0.2849, pruned_loss=0.09943, over 985398.28 frames.], batch size: 86, aishell_tot_loss[loss=0.2368, simple_loss=0.289, pruned_loss=0.09228, over 984701.31 frames.], datatang_tot_loss[loss=0.2485, simple_loss=0.2834, pruned_loss=0.1068, over 986309.05 frames.], batch size: 86, lr: 1.79e-03
+2022-06-18 12:40:49,595 INFO [train.py:874] (2/4) Epoch 3, batch 4100, aishell_loss[loss=0.2141, simple_loss=0.2781, pruned_loss=0.07504, over 4913.00 frames.], tot_loss[loss=0.2411, simple_loss=0.2849, pruned_loss=0.09863, over 984729.78 frames.], batch size: 41, aishell_tot_loss[loss=0.2369, simple_loss=0.2892, pruned_loss=0.0923, over 983963.35 frames.], datatang_tot_loss[loss=0.2475, simple_loss=0.2828, pruned_loss=0.1061, over 986342.12 frames.], batch size: 41, lr: 1.78e-03
+2022-06-18 12:42:03,117 INFO [train.py:874] (2/4) Epoch 4, batch 50, aishell_loss[loss=0.204, simple_loss=0.2518, pruned_loss=0.07813, over 4969.00 frames.], tot_loss[loss=0.2279, simple_loss=0.2768, pruned_loss=0.08947, over 218610.56 frames.], batch size: 25, aishell_tot_loss[loss=0.2318, simple_loss=0.2874, pruned_loss=0.0881, over 111730.96 frames.], datatang_tot_loss[loss=0.2243, simple_loss=0.267, pruned_loss=0.0908, over 120529.87 frames.], batch size: 25, lr: 1.73e-03
+2022-06-18 12:42:33,959 INFO [train.py:874] (2/4) Epoch 4, batch 100, aishell_loss[loss=0.2298, simple_loss=0.286, pruned_loss=0.08674, over 4971.00 frames.], tot_loss[loss=0.228, simple_loss=0.2765, pruned_loss=0.0898, over 388308.93 frames.], batch size: 30, aishell_tot_loss[loss=0.2311, simple_loss=0.2858, pruned_loss=0.08818, over 229430.59 frames.], datatang_tot_loss[loss=0.2245, simple_loss=0.2659, pruned_loss=0.09152, over 207113.94 frames.], batch size: 30, lr: 1.73e-03
+2022-06-18 12:43:05,271 INFO [train.py:874] (2/4) Epoch 4, batch 150, datatang_loss[loss=0.2043, simple_loss=0.2416, pruned_loss=0.08354, over 4941.00 frames.], tot_loss[loss=0.2295, simple_loss=0.2778, pruned_loss=0.09056, over 520281.67 frames.], batch size: 34, aishell_tot_loss[loss=0.2318, simple_loss=0.2866, pruned_loss=0.08853, over 311364.10 frames.], datatang_tot_loss[loss=0.2268, simple_loss=0.2685, pruned_loss=0.09249, over 305599.27 frames.], batch size: 34, lr: 1.72e-03
+2022-06-18 12:43:34,801 INFO [train.py:874] (2/4) Epoch 4, batch 200, datatang_loss[loss=0.1906, simple_loss=0.2267, pruned_loss=0.07727, over 4932.00 frames.], tot_loss[loss=0.2318, simple_loss=0.2781, pruned_loss=0.09271, over 623776.55 frames.], batch size: 25, aishell_tot_loss[loss=0.2334, simple_loss=0.2874, pruned_loss=0.08972, over 384964.38 frames.], datatang_tot_loss[loss=0.2294, simple_loss=0.2688, pruned_loss=0.09501, over 391840.65 frames.], batch size: 25, lr: 1.72e-03
+2022-06-18 12:44:05,297 INFO [train.py:874] (2/4) Epoch 4, batch 250, datatang_loss[loss=0.2744, simple_loss=0.3057, pruned_loss=0.1216, over 4972.00 frames.], tot_loss[loss=0.234, simple_loss=0.2794, pruned_loss=0.09428, over 704208.65 frames.], batch size: 45, aishell_tot_loss[loss=0.2348, simple_loss=0.2886, pruned_loss=0.0905, over 447715.54 frames.], datatang_tot_loss[loss=0.2319, simple_loss=0.2703, pruned_loss=0.09674, over 469772.01 frames.], batch size: 45, lr: 1.72e-03
+2022-06-18 12:44:35,728 INFO [train.py:874] (2/4) Epoch 4, batch 300, aishell_loss[loss=0.244, simple_loss=0.3041, pruned_loss=0.0919, over 4938.00 frames.], tot_loss[loss=0.2344, simple_loss=0.2801, pruned_loss=0.09433, over 766977.68 frames.], batch size: 68, aishell_tot_loss[loss=0.2347, simple_loss=0.2887, pruned_loss=0.09037, over 515769.07 frames.], datatang_tot_loss[loss=0.2328, simple_loss=0.2709, pruned_loss=0.09728, over 526378.59 frames.], batch size: 68, lr: 1.71e-03
+2022-06-18 12:45:04,607 INFO [train.py:874] (2/4) Epoch 4, batch 350, aishell_loss[loss=0.2203, simple_loss=0.2697, pruned_loss=0.08544, over 4797.00 frames.], tot_loss[loss=0.236, simple_loss=0.2819, pruned_loss=0.09509, over 815528.09 frames.], batch size: 24, aishell_tot_loss[loss=0.2345, simple_loss=0.2885, pruned_loss=0.09026, over 573350.01 frames.], datatang_tot_loss[loss=0.2358, simple_loss=0.2739, pruned_loss=0.09888, over 578359.36 frames.], batch size: 24, lr: 1.71e-03
+2022-06-18 12:45:35,292 INFO [train.py:874] (2/4) Epoch 4, batch 400, aishell_loss[loss=0.2127, simple_loss=0.2764, pruned_loss=0.07449, over 4898.00 frames.], tot_loss[loss=0.2356, simple_loss=0.2823, pruned_loss=0.0945, over 853032.14 frames.], batch size: 34, aishell_tot_loss[loss=0.2333, simple_loss=0.2881, pruned_loss=0.08923, over 637828.57 frames.], datatang_tot_loss[loss=0.237, simple_loss=0.2742, pruned_loss=0.09991, over 609655.41 frames.], batch size: 34, lr: 1.71e-03
+2022-06-18 12:46:05,198 INFO [train.py:874] (2/4) Epoch 4, batch 450, datatang_loss[loss=0.3643, simple_loss=0.3694, pruned_loss=0.1796, over 4940.00 frames.], tot_loss[loss=0.2362, simple_loss=0.2826, pruned_loss=0.09494, over 882134.74 frames.], batch size: 109, aishell_tot_loss[loss=0.2313, simple_loss=0.2869, pruned_loss=0.08786, over 675448.68 frames.], datatang_tot_loss[loss=0.24, simple_loss=0.2765, pruned_loss=0.1017, over 657193.01 frames.], batch size: 109, lr: 1.71e-03
+2022-06-18 12:46:35,409 INFO [train.py:874] (2/4) Epoch 4, batch 500, datatang_loss[loss=0.2655, simple_loss=0.2947, pruned_loss=0.1181, over 4931.00 frames.], tot_loss[loss=0.235, simple_loss=0.2819, pruned_loss=0.094, over 904922.89 frames.], batch size: 79, aishell_tot_loss[loss=0.2301, simple_loss=0.2859, pruned_loss=0.08719, over 711849.97 frames.], datatang_tot_loss[loss=0.2395, simple_loss=0.2768, pruned_loss=0.1011, over 695842.56 frames.], batch size: 79, lr: 1.70e-03
+2022-06-18 12:47:05,933 INFO [train.py:874] (2/4) Epoch 4, batch 550, aishell_loss[loss=0.199, simple_loss=0.2509, pruned_loss=0.07358, over 4975.00 frames.], tot_loss[loss=0.237, simple_loss=0.2837, pruned_loss=0.0952, over 923219.21 frames.], batch size: 25, aishell_tot_loss[loss=0.2303, simple_loss=0.2863, pruned_loss=0.08713, over 745469.35 frames.], datatang_tot_loss[loss=0.2423, simple_loss=0.2789, pruned_loss=0.1028, over 728947.15 frames.], batch size: 25, lr: 1.70e-03
+2022-06-18 12:47:36,166 INFO [train.py:874] (2/4) Epoch 4, batch 600, datatang_loss[loss=0.2306, simple_loss=0.2685, pruned_loss=0.09634, over 4916.00 frames.], tot_loss[loss=0.2377, simple_loss=0.2841, pruned_loss=0.09564, over 937089.68 frames.], batch size: 75, aishell_tot_loss[loss=0.2296, simple_loss=0.2856, pruned_loss=0.08681, over 772637.41 frames.], datatang_tot_loss[loss=0.244, simple_loss=0.2805, pruned_loss=0.1037, over 760385.96 frames.], batch size: 75, lr: 1.70e-03
+2022-06-18 12:48:07,245 INFO [train.py:874] (2/4) Epoch 4, batch 650, aishell_loss[loss=0.2226, simple_loss=0.2881, pruned_loss=0.07857, over 4929.00 frames.], tot_loss[loss=0.2372, simple_loss=0.2842, pruned_loss=0.09507, over 947842.49 frames.], batch size: 58, aishell_tot_loss[loss=0.2288, simple_loss=0.2849, pruned_loss=0.08632, over 798676.90 frames.], datatang_tot_loss[loss=0.2445, simple_loss=0.2817, pruned_loss=0.1037, over 785892.95 frames.], batch size: 58, lr: 1.69e-03
+2022-06-18 12:48:37,395 INFO [train.py:874] (2/4) Epoch 4, batch 700, datatang_loss[loss=0.2887, simple_loss=0.3144, pruned_loss=0.1315, over 4930.00 frames.], tot_loss[loss=0.2386, simple_loss=0.2846, pruned_loss=0.09636, over 956177.30 frames.], batch size: 42, aishell_tot_loss[loss=0.2292, simple_loss=0.285, pruned_loss=0.08665, over 817284.56 frames.], datatang_tot_loss[loss=0.2456, simple_loss=0.2823, pruned_loss=0.1045, over 812941.00 frames.], batch size: 42, lr: 1.69e-03
+2022-06-18 12:49:07,576 INFO [train.py:874] (2/4) Epoch 4, batch 750, aishell_loss[loss=0.3095, simple_loss=0.3322, pruned_loss=0.1434, over 4937.00 frames.], tot_loss[loss=0.2401, simple_loss=0.2854, pruned_loss=0.09737, over 962393.70 frames.], batch size: 49, aishell_tot_loss[loss=0.23, simple_loss=0.2859, pruned_loss=0.08706, over 830954.68 frames.], datatang_tot_loss[loss=0.2462, simple_loss=0.2829, pruned_loss=0.1047, over 839014.39 frames.], batch size: 49, lr: 1.69e-03
+2022-06-18 12:49:37,667 INFO [train.py:874] (2/4) Epoch 4, batch 800, datatang_loss[loss=0.2238, simple_loss=0.2734, pruned_loss=0.08708, over 4924.00 frames.], tot_loss[loss=0.2378, simple_loss=0.2847, pruned_loss=0.09549, over 967532.37 frames.], batch size: 83, aishell_tot_loss[loss=0.23, simple_loss=0.2862, pruned_loss=0.0869, over 852998.11 frames.], datatang_tot_loss[loss=0.2446, simple_loss=0.2819, pruned_loss=0.1036, over 852544.69 frames.], batch size: 83, lr: 1.69e-03
+2022-06-18 12:50:07,650 INFO [train.py:874] (2/4) Epoch 4, batch 850, aishell_loss[loss=0.2297, simple_loss=0.2911, pruned_loss=0.08414, over 4916.00 frames.], tot_loss[loss=0.2375, simple_loss=0.2845, pruned_loss=0.09528, over 971517.99 frames.], batch size: 46, aishell_tot_loss[loss=0.2301, simple_loss=0.2863, pruned_loss=0.08694, over 870176.37 frames.], datatang_tot_loss[loss=0.2444, simple_loss=0.2815, pruned_loss=0.1036, over 866587.58 frames.], batch size: 46, lr: 1.68e-03
+2022-06-18 12:50:37,410 INFO [train.py:874] (2/4) Epoch 4, batch 900, aishell_loss[loss=0.2332, simple_loss=0.2922, pruned_loss=0.08711, over 4918.00 frames.], tot_loss[loss=0.238, simple_loss=0.2846, pruned_loss=0.09564, over 974257.45 frames.], batch size: 33, aishell_tot_loss[loss=0.23, simple_loss=0.286, pruned_loss=0.08693, over 883995.90 frames.], datatang_tot_loss[loss=0.2452, simple_loss=0.2821, pruned_loss=0.1041, over 879972.18 frames.], batch size: 33, lr: 1.68e-03
+2022-06-18 12:51:08,855 INFO [train.py:874] (2/4) Epoch 4, batch 950, datatang_loss[loss=0.2018, simple_loss=0.2511, pruned_loss=0.07622, over 4934.00 frames.], tot_loss[loss=0.2387, simple_loss=0.2849, pruned_loss=0.09625, over 976595.16 frames.], batch size: 71, aishell_tot_loss[loss=0.2313, simple_loss=0.2869, pruned_loss=0.0878, over 895839.14 frames.], datatang_tot_loss[loss=0.2449, simple_loss=0.2817, pruned_loss=0.104, over 892329.63 frames.], batch size: 71, lr: 1.68e-03
+2022-06-18 12:51:39,318 INFO [train.py:874] (2/4) Epoch 4, batch 1000, datatang_loss[loss=0.2609, simple_loss=0.295, pruned_loss=0.1134, over 4923.00 frames.], tot_loss[loss=0.238, simple_loss=0.2843, pruned_loss=0.09579, over 978938.65 frames.], batch size: 73, aishell_tot_loss[loss=0.2308, simple_loss=0.2867, pruned_loss=0.08747, over 906373.80 frames.], datatang_tot_loss[loss=0.2446, simple_loss=0.2814, pruned_loss=0.1038, over 903731.28 frames.], batch size: 73, lr: 1.67e-03
+2022-06-18 12:51:39,318 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 12:51:55,326 INFO [train.py:914] (2/4) Epoch 4, validation: loss=0.1866, simple_loss=0.2641, pruned_loss=0.05457, over 1622729.00 frames.
+2022-06-18 12:52:26,585 INFO [train.py:874] (2/4) Epoch 4, batch 1050, aishell_loss[loss=0.2087, simple_loss=0.2726, pruned_loss=0.07238, over 4882.00 frames.], tot_loss[loss=0.2368, simple_loss=0.2835, pruned_loss=0.0951, over 980431.77 frames.], batch size: 28, aishell_tot_loss[loss=0.23, simple_loss=0.286, pruned_loss=0.08699, over 914472.72 frames.], datatang_tot_loss[loss=0.244, simple_loss=0.2814, pruned_loss=0.1033, over 914642.62 frames.], batch size: 28, lr: 1.67e-03
+2022-06-18 12:52:56,825 INFO [train.py:874] (2/4) Epoch 4, batch 1100, datatang_loss[loss=0.2339, simple_loss=0.2728, pruned_loss=0.0975, over 4921.00 frames.], tot_loss[loss=0.2355, simple_loss=0.2825, pruned_loss=0.09426, over 981341.61 frames.], batch size: 77, aishell_tot_loss[loss=0.2295, simple_loss=0.2857, pruned_loss=0.08668, over 923049.16 frames.], datatang_tot_loss[loss=0.2429, simple_loss=0.2805, pruned_loss=0.1027, over 922561.08 frames.], batch size: 77, lr: 1.67e-03
+2022-06-18 12:53:26,760 INFO [train.py:874] (2/4) Epoch 4, batch 1150, aishell_loss[loss=0.2195, simple_loss=0.2794, pruned_loss=0.07982, over 4946.00 frames.], tot_loss[loss=0.2354, simple_loss=0.2828, pruned_loss=0.09399, over 982013.94 frames.], batch size: 49, aishell_tot_loss[loss=0.2291, simple_loss=0.2856, pruned_loss=0.08631, over 931860.52 frames.], datatang_tot_loss[loss=0.2434, simple_loss=0.2806, pruned_loss=0.1031, over 928188.44 frames.], batch size: 49, lr: 1.67e-03
+2022-06-18 12:53:57,025 INFO [train.py:874] (2/4) Epoch 4, batch 1200, datatang_loss[loss=0.2087, simple_loss=0.2563, pruned_loss=0.08057, over 4936.00 frames.], tot_loss[loss=0.2341, simple_loss=0.2822, pruned_loss=0.09297, over 983004.96 frames.], batch size: 69, aishell_tot_loss[loss=0.2294, simple_loss=0.286, pruned_loss=0.08633, over 938498.90 frames.], datatang_tot_loss[loss=0.2415, simple_loss=0.2794, pruned_loss=0.1018, over 934852.33 frames.], batch size: 69, lr: 1.66e-03
+2022-06-18 12:54:27,045 INFO [train.py:874] (2/4) Epoch 4, batch 1250, datatang_loss[loss=0.2878, simple_loss=0.3117, pruned_loss=0.132, over 4614.00 frames.], tot_loss[loss=0.2345, simple_loss=0.2826, pruned_loss=0.09319, over 983504.11 frames.], batch size: 24, aishell_tot_loss[loss=0.2286, simple_loss=0.2854, pruned_loss=0.08592, over 944614.83 frames.], datatang_tot_loss[loss=0.2425, simple_loss=0.2803, pruned_loss=0.1023, over 940167.60 frames.], batch size: 24, lr: 1.66e-03
+2022-06-18 12:54:57,598 INFO [train.py:874] (2/4) Epoch 4, batch 1300, datatang_loss[loss=0.2506, simple_loss=0.2787, pruned_loss=0.1113, over 4927.00 frames.], tot_loss[loss=0.2327, simple_loss=0.2813, pruned_loss=0.09201, over 983801.20 frames.], batch size: 50, aishell_tot_loss[loss=0.2271, simple_loss=0.2843, pruned_loss=0.08501, over 950243.46 frames.], datatang_tot_loss[loss=0.2421, simple_loss=0.2798, pruned_loss=0.1022, over 944412.45 frames.], batch size: 50, lr: 1.66e-03
+2022-06-18 12:55:28,213 INFO [train.py:874] (2/4) Epoch 4, batch 1350, aishell_loss[loss=0.2148, simple_loss=0.2786, pruned_loss=0.0755, over 4942.00 frames.], tot_loss[loss=0.2315, simple_loss=0.2809, pruned_loss=0.09107, over 984480.22 frames.], batch size: 33, aishell_tot_loss[loss=0.2257, simple_loss=0.2834, pruned_loss=0.08398, over 955275.65 frames.], datatang_tot_loss[loss=0.2422, simple_loss=0.28, pruned_loss=0.1023, over 948546.92 frames.], batch size: 33, lr: 1.66e-03
+2022-06-18 12:55:57,855 INFO [train.py:874] (2/4) Epoch 4, batch 1400, datatang_loss[loss=0.2347, simple_loss=0.2723, pruned_loss=0.09855, over 4904.00 frames.], tot_loss[loss=0.2315, simple_loss=0.2807, pruned_loss=0.09121, over 984439.08 frames.], batch size: 52, aishell_tot_loss[loss=0.2254, simple_loss=0.2831, pruned_loss=0.0838, over 957862.47 frames.], datatang_tot_loss[loss=0.2415, simple_loss=0.28, pruned_loss=0.1015, over 953737.80 frames.], batch size: 52, lr: 1.65e-03
+2022-06-18 12:56:27,847 INFO [train.py:874] (2/4) Epoch 4, batch 1450, datatang_loss[loss=0.2251, simple_loss=0.2677, pruned_loss=0.09126, over 4943.00 frames.], tot_loss[loss=0.2303, simple_loss=0.2799, pruned_loss=0.09037, over 985022.92 frames.], batch size: 34, aishell_tot_loss[loss=0.2238, simple_loss=0.2817, pruned_loss=0.08296, over 962283.92 frames.], datatang_tot_loss[loss=0.2419, simple_loss=0.2802, pruned_loss=0.1017, over 956507.47 frames.], batch size: 34, lr: 1.65e-03
+2022-06-18 12:57:00,070 INFO [train.py:874] (2/4) Epoch 4, batch 1500, aishell_loss[loss=0.2474, simple_loss=0.3032, pruned_loss=0.09581, over 4922.00 frames.], tot_loss[loss=0.2317, simple_loss=0.2809, pruned_loss=0.0913, over 984887.33 frames.], batch size: 68, aishell_tot_loss[loss=0.2246, simple_loss=0.2825, pruned_loss=0.08341, over 964853.78 frames.], datatang_tot_loss[loss=0.2419, simple_loss=0.2803, pruned_loss=0.1017, over 959842.87 frames.], batch size: 68, lr: 1.65e-03
+2022-06-18 12:57:28,931 INFO [train.py:874] (2/4) Epoch 4, batch 1550, datatang_loss[loss=0.2177, simple_loss=0.2605, pruned_loss=0.0875, over 4943.00 frames.], tot_loss[loss=0.2334, simple_loss=0.2816, pruned_loss=0.09256, over 984996.72 frames.], batch size: 62, aishell_tot_loss[loss=0.2261, simple_loss=0.2829, pruned_loss=0.08466, over 967441.15 frames.], datatang_tot_loss[loss=0.242, simple_loss=0.2805, pruned_loss=0.1018, over 962602.48 frames.], batch size: 62, lr: 1.65e-03
+2022-06-18 12:57:59,844 INFO [train.py:874] (2/4) Epoch 4, batch 1600, datatang_loss[loss=0.2876, simple_loss=0.3255, pruned_loss=0.1249, over 4923.00 frames.], tot_loss[loss=0.2322, simple_loss=0.2803, pruned_loss=0.09201, over 985069.81 frames.], batch size: 94, aishell_tot_loss[loss=0.2252, simple_loss=0.2822, pruned_loss=0.08411, over 968917.59 frames.], datatang_tot_loss[loss=0.2409, simple_loss=0.2798, pruned_loss=0.101, over 966047.81 frames.], batch size: 94, lr: 1.64e-03
+2022-06-18 12:58:30,895 INFO [train.py:874] (2/4) Epoch 4, batch 1650, datatang_loss[loss=0.2428, simple_loss=0.2943, pruned_loss=0.09569, over 4954.00 frames.], tot_loss[loss=0.2319, simple_loss=0.2803, pruned_loss=0.09173, over 985217.13 frames.], batch size: 86, aishell_tot_loss[loss=0.2246, simple_loss=0.2816, pruned_loss=0.08384, over 970825.64 frames.], datatang_tot_loss[loss=0.2408, simple_loss=0.2801, pruned_loss=0.1007, over 968445.11 frames.], batch size: 86, lr: 1.64e-03
+2022-06-18 12:59:01,540 INFO [train.py:874] (2/4) Epoch 4, batch 1700, datatang_loss[loss=0.2261, simple_loss=0.2713, pruned_loss=0.0904, over 4921.00 frames.], tot_loss[loss=0.2322, simple_loss=0.2806, pruned_loss=0.09193, over 985418.23 frames.], batch size: 77, aishell_tot_loss[loss=0.2253, simple_loss=0.2825, pruned_loss=0.08408, over 972526.89 frames.], datatang_tot_loss[loss=0.2401, simple_loss=0.2795, pruned_loss=0.1003, over 970662.31 frames.], batch size: 77, lr: 1.64e-03
+2022-06-18 12:59:32,074 INFO [train.py:874] (2/4) Epoch 4, batch 1750, aishell_loss[loss=0.1934, simple_loss=0.2485, pruned_loss=0.06912, over 4975.00 frames.], tot_loss[loss=0.2314, simple_loss=0.2801, pruned_loss=0.09131, over 985555.80 frames.], batch size: 27, aishell_tot_loss[loss=0.2242, simple_loss=0.2816, pruned_loss=0.08342, over 974121.71 frames.], datatang_tot_loss[loss=0.2404, simple_loss=0.2797, pruned_loss=0.1006, over 972449.82 frames.], batch size: 27, lr: 1.63e-03
+2022-06-18 13:00:02,766 INFO [train.py:874] (2/4) Epoch 4, batch 1800, aishell_loss[loss=0.2126, simple_loss=0.2856, pruned_loss=0.06981, over 4947.00 frames.], tot_loss[loss=0.2334, simple_loss=0.2817, pruned_loss=0.09255, over 985448.28 frames.], batch size: 56, aishell_tot_loss[loss=0.2248, simple_loss=0.2821, pruned_loss=0.08381, over 975365.68 frames.], datatang_tot_loss[loss=0.2419, simple_loss=0.2809, pruned_loss=0.1014, over 973984.58 frames.], batch size: 56, lr: 1.63e-03
+2022-06-18 13:00:32,615 INFO [train.py:874] (2/4) Epoch 4, batch 1850, aishell_loss[loss=0.2589, simple_loss=0.3116, pruned_loss=0.1031, over 4859.00 frames.], tot_loss[loss=0.2322, simple_loss=0.2814, pruned_loss=0.0915, over 985194.79 frames.], batch size: 35, aishell_tot_loss[loss=0.2246, simple_loss=0.2823, pruned_loss=0.08345, over 976312.89 frames.], datatang_tot_loss[loss=0.2409, simple_loss=0.2804, pruned_loss=0.1007, over 975318.67 frames.], batch size: 35, lr: 1.63e-03
+2022-06-18 13:01:02,818 INFO [train.py:874] (2/4) Epoch 4, batch 1900, aishell_loss[loss=0.2344, simple_loss=0.302, pruned_loss=0.0834, over 4907.00 frames.], tot_loss[loss=0.234, simple_loss=0.2823, pruned_loss=0.09287, over 985320.42 frames.], batch size: 41, aishell_tot_loss[loss=0.225, simple_loss=0.2826, pruned_loss=0.08373, over 977356.53 frames.], datatang_tot_loss[loss=0.2423, simple_loss=0.2811, pruned_loss=0.1017, over 976612.59 frames.], batch size: 41, lr: 1.63e-03
+2022-06-18 13:01:34,318 INFO [train.py:874] (2/4) Epoch 4, batch 1950, aishell_loss[loss=0.2112, simple_loss=0.2676, pruned_loss=0.0774, over 4915.00 frames.], tot_loss[loss=0.2343, simple_loss=0.2821, pruned_loss=0.09331, over 984924.40 frames.], batch size: 33, aishell_tot_loss[loss=0.225, simple_loss=0.2828, pruned_loss=0.08365, over 977981.34 frames.], datatang_tot_loss[loss=0.2428, simple_loss=0.2808, pruned_loss=0.1024, over 977527.04 frames.], batch size: 33, lr: 1.62e-03
+2022-06-18 13:02:04,138 INFO [train.py:874] (2/4) Epoch 4, batch 2000, datatang_loss[loss=0.231, simple_loss=0.273, pruned_loss=0.09452, over 4920.00 frames.], tot_loss[loss=0.2332, simple_loss=0.2815, pruned_loss=0.09251, over 985066.84 frames.], batch size: 83, aishell_tot_loss[loss=0.2251, simple_loss=0.2831, pruned_loss=0.0835, over 978723.15 frames.], datatang_tot_loss[loss=0.2414, simple_loss=0.2798, pruned_loss=0.1014, over 978627.56 frames.], batch size: 83, lr: 1.62e-03
+2022-06-18 13:02:04,139 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 13:02:20,118 INFO [train.py:914] (2/4) Epoch 4, validation: loss=0.1875, simple_loss=0.2609, pruned_loss=0.05705, over 1622729.00 frames.
+2022-06-18 13:02:49,779 INFO [train.py:874] (2/4) Epoch 4, batch 2050, aishell_loss[loss=0.1781, simple_loss=0.2411, pruned_loss=0.05758, over 4816.00 frames.], tot_loss[loss=0.2328, simple_loss=0.2806, pruned_loss=0.09247, over 985145.02 frames.], batch size: 26, aishell_tot_loss[loss=0.2239, simple_loss=0.2819, pruned_loss=0.0829, over 979280.85 frames.], datatang_tot_loss[loss=0.2419, simple_loss=0.2801, pruned_loss=0.1018, over 979673.53 frames.], batch size: 26, lr: 1.62e-03
+2022-06-18 13:03:20,657 INFO [train.py:874] (2/4) Epoch 4, batch 2100, datatang_loss[loss=0.2076, simple_loss=0.2577, pruned_loss=0.07872, over 4936.00 frames.], tot_loss[loss=0.2323, simple_loss=0.2804, pruned_loss=0.0921, over 985412.28 frames.], batch size: 62, aishell_tot_loss[loss=0.2244, simple_loss=0.2827, pruned_loss=0.08308, over 979939.74 frames.], datatang_tot_loss[loss=0.2407, simple_loss=0.2791, pruned_loss=0.1011, over 980626.42 frames.], batch size: 62, lr: 1.62e-03
+2022-06-18 13:03:50,530 INFO [train.py:874] (2/4) Epoch 4, batch 2150, datatang_loss[loss=0.3363, simple_loss=0.3364, pruned_loss=0.1681, over 4893.00 frames.], tot_loss[loss=0.2344, simple_loss=0.2824, pruned_loss=0.0932, over 985239.94 frames.], batch size: 59, aishell_tot_loss[loss=0.2245, simple_loss=0.2829, pruned_loss=0.08308, over 980389.23 frames.], datatang_tot_loss[loss=0.2429, simple_loss=0.2809, pruned_loss=0.1024, over 981229.21 frames.], batch size: 59, lr: 1.61e-03
+2022-06-18 13:04:20,260 INFO [train.py:874] (2/4) Epoch 4, batch 2200, datatang_loss[loss=0.2409, simple_loss=0.2894, pruned_loss=0.09627, over 4961.00 frames.], tot_loss[loss=0.2339, simple_loss=0.2828, pruned_loss=0.09254, over 985185.75 frames.], batch size: 86, aishell_tot_loss[loss=0.2251, simple_loss=0.2834, pruned_loss=0.08345, over 980844.57 frames.], datatang_tot_loss[loss=0.2421, simple_loss=0.2809, pruned_loss=0.1017, over 981771.74 frames.], batch size: 86, lr: 1.61e-03
+2022-06-18 13:04:51,366 INFO [train.py:874] (2/4) Epoch 4, batch 2250, aishell_loss[loss=0.2245, simple_loss=0.2782, pruned_loss=0.08535, over 4861.00 frames.], tot_loss[loss=0.2337, simple_loss=0.2826, pruned_loss=0.09244, over 985209.86 frames.], batch size: 37, aishell_tot_loss[loss=0.2255, simple_loss=0.2837, pruned_loss=0.08365, over 981246.33 frames.], datatang_tot_loss[loss=0.2416, simple_loss=0.2805, pruned_loss=0.1013, over 982313.16 frames.], batch size: 37, lr: 1.61e-03
+2022-06-18 13:05:21,495 INFO [train.py:874] (2/4) Epoch 4, batch 2300, datatang_loss[loss=0.2432, simple_loss=0.285, pruned_loss=0.1007, over 4905.00 frames.], tot_loss[loss=0.2331, simple_loss=0.2819, pruned_loss=0.09215, over 985064.50 frames.], batch size: 47, aishell_tot_loss[loss=0.225, simple_loss=0.2831, pruned_loss=0.08347, over 981547.00 frames.], datatang_tot_loss[loss=0.2414, simple_loss=0.2805, pruned_loss=0.1012, over 982659.18 frames.], batch size: 47, lr: 1.61e-03
+2022-06-18 13:05:52,002 INFO [train.py:874] (2/4) Epoch 4, batch 2350, datatang_loss[loss=0.2898, simple_loss=0.3274, pruned_loss=0.1261, over 4925.00 frames.], tot_loss[loss=0.2328, simple_loss=0.2821, pruned_loss=0.09181, over 985059.64 frames.], batch size: 94, aishell_tot_loss[loss=0.2253, simple_loss=0.2834, pruned_loss=0.08359, over 982103.59 frames.], datatang_tot_loss[loss=0.2406, simple_loss=0.2803, pruned_loss=0.1005, over 982786.65 frames.], batch size: 94, lr: 1.60e-03
+2022-06-18 13:06:21,649 INFO [train.py:874] (2/4) Epoch 4, batch 2400, aishell_loss[loss=0.2303, simple_loss=0.2831, pruned_loss=0.08873, over 4914.00 frames.], tot_loss[loss=0.2321, simple_loss=0.2816, pruned_loss=0.09131, over 984782.83 frames.], batch size: 52, aishell_tot_loss[loss=0.2253, simple_loss=0.2834, pruned_loss=0.08361, over 982521.34 frames.], datatang_tot_loss[loss=0.2402, simple_loss=0.2799, pruned_loss=0.1003, over 982714.30 frames.], batch size: 52, lr: 1.60e-03
+2022-06-18 13:06:51,595 INFO [train.py:874] (2/4) Epoch 4, batch 2450, datatang_loss[loss=0.2381, simple_loss=0.2892, pruned_loss=0.0935, over 4919.00 frames.], tot_loss[loss=0.2308, simple_loss=0.2805, pruned_loss=0.0905, over 985060.61 frames.], batch size: 75, aishell_tot_loss[loss=0.2247, simple_loss=0.2829, pruned_loss=0.08325, over 983116.58 frames.], datatang_tot_loss[loss=0.2394, simple_loss=0.2792, pruned_loss=0.09986, over 982913.35 frames.], batch size: 75, lr: 1.60e-03
+2022-06-18 13:07:23,700 INFO [train.py:874] (2/4) Epoch 4, batch 2500, datatang_loss[loss=0.2126, simple_loss=0.2597, pruned_loss=0.08277, over 4929.00 frames.], tot_loss[loss=0.2334, simple_loss=0.2822, pruned_loss=0.09229, over 985122.16 frames.], batch size: 77, aishell_tot_loss[loss=0.2252, simple_loss=0.2834, pruned_loss=0.08352, over 983319.27 frames.], datatang_tot_loss[loss=0.2409, simple_loss=0.2805, pruned_loss=0.1006, over 983248.81 frames.], batch size: 77, lr: 1.60e-03
+2022-06-18 13:07:53,523 INFO [train.py:874] (2/4) Epoch 4, batch 2550, datatang_loss[loss=0.2082, simple_loss=0.2489, pruned_loss=0.08377, over 4815.00 frames.], tot_loss[loss=0.2312, simple_loss=0.2799, pruned_loss=0.09122, over 985262.62 frames.], batch size: 25, aishell_tot_loss[loss=0.2255, simple_loss=0.2834, pruned_loss=0.08374, over 983559.20 frames.], datatang_tot_loss[loss=0.2378, simple_loss=0.2782, pruned_loss=0.09866, over 983598.19 frames.], batch size: 25, lr: 1.60e-03
+2022-06-18 13:08:25,262 INFO [train.py:874] (2/4) Epoch 4, batch 2600, aishell_loss[loss=0.2442, simple_loss=0.3116, pruned_loss=0.08836, over 4977.00 frames.], tot_loss[loss=0.2312, simple_loss=0.2803, pruned_loss=0.09106, over 985308.40 frames.], batch size: 39, aishell_tot_loss[loss=0.2257, simple_loss=0.284, pruned_loss=0.08372, over 983697.24 frames.], datatang_tot_loss[loss=0.2371, simple_loss=0.278, pruned_loss=0.09808, over 983883.56 frames.], batch size: 39, lr: 1.59e-03
+2022-06-18 13:08:55,093 INFO [train.py:874] (2/4) Epoch 4, batch 2650, aishell_loss[loss=0.2939, simple_loss=0.3374, pruned_loss=0.1252, over 4964.00 frames.], tot_loss[loss=0.2327, simple_loss=0.2819, pruned_loss=0.0918, over 985494.65 frames.], batch size: 64, aishell_tot_loss[loss=0.2279, simple_loss=0.2857, pruned_loss=0.08502, over 984000.37 frames.], datatang_tot_loss[loss=0.2367, simple_loss=0.2776, pruned_loss=0.09792, over 984137.65 frames.], batch size: 64, lr: 1.59e-03
+2022-06-18 13:09:25,294 INFO [train.py:874] (2/4) Epoch 4, batch 2700, datatang_loss[loss=0.2526, simple_loss=0.2943, pruned_loss=0.1055, over 4957.00 frames.], tot_loss[loss=0.2328, simple_loss=0.2818, pruned_loss=0.09195, over 985149.25 frames.], batch size: 91, aishell_tot_loss[loss=0.2277, simple_loss=0.2855, pruned_loss=0.08499, over 984088.23 frames.], datatang_tot_loss[loss=0.237, simple_loss=0.2779, pruned_loss=0.09811, over 984013.55 frames.], batch size: 91, lr: 1.59e-03
+2022-06-18 13:09:56,954 INFO [train.py:874] (2/4) Epoch 4, batch 2750, datatang_loss[loss=0.2502, simple_loss=0.2917, pruned_loss=0.1043, over 4936.00 frames.], tot_loss[loss=0.2312, simple_loss=0.2805, pruned_loss=0.09098, over 985553.92 frames.], batch size: 88, aishell_tot_loss[loss=0.2269, simple_loss=0.285, pruned_loss=0.08443, over 984155.25 frames.], datatang_tot_loss[loss=0.2361, simple_loss=0.2772, pruned_loss=0.09747, over 984618.76 frames.], batch size: 88, lr: 1.59e-03
+2022-06-18 13:10:26,845 INFO [train.py:874] (2/4) Epoch 4, batch 2800, datatang_loss[loss=0.2161, simple_loss=0.2618, pruned_loss=0.08521, over 4927.00 frames.], tot_loss[loss=0.2304, simple_loss=0.28, pruned_loss=0.09037, over 985478.51 frames.], batch size: 50, aishell_tot_loss[loss=0.227, simple_loss=0.2854, pruned_loss=0.08435, over 984244.13 frames.], datatang_tot_loss[loss=0.2349, simple_loss=0.2763, pruned_loss=0.09678, over 984719.97 frames.], batch size: 50, lr: 1.58e-03
+2022-06-18 13:10:56,170 INFO [train.py:874] (2/4) Epoch 4, batch 2850, datatang_loss[loss=0.2323, simple_loss=0.2716, pruned_loss=0.09652, over 4899.00 frames.], tot_loss[loss=0.2298, simple_loss=0.2795, pruned_loss=0.08999, over 985536.42 frames.], batch size: 52, aishell_tot_loss[loss=0.2266, simple_loss=0.285, pruned_loss=0.08413, over 984378.33 frames.], datatang_tot_loss[loss=0.2345, simple_loss=0.276, pruned_loss=0.09649, over 984873.86 frames.], batch size: 52, lr: 1.58e-03
+2022-06-18 13:11:27,467 INFO [train.py:874] (2/4) Epoch 4, batch 2900, aishell_loss[loss=0.2527, simple_loss=0.3065, pruned_loss=0.09949, over 4933.00 frames.], tot_loss[loss=0.2295, simple_loss=0.2794, pruned_loss=0.08979, over 985533.27 frames.], batch size: 58, aishell_tot_loss[loss=0.2268, simple_loss=0.2852, pruned_loss=0.08415, over 984339.86 frames.], datatang_tot_loss[loss=0.2337, simple_loss=0.2756, pruned_loss=0.09588, over 985108.50 frames.], batch size: 58, lr: 1.58e-03
+2022-06-18 13:11:56,704 INFO [train.py:874] (2/4) Epoch 4, batch 2950, aishell_loss[loss=0.2574, simple_loss=0.3029, pruned_loss=0.1059, over 4866.00 frames.], tot_loss[loss=0.2298, simple_loss=0.2796, pruned_loss=0.08999, over 985755.46 frames.], batch size: 35, aishell_tot_loss[loss=0.2276, simple_loss=0.286, pruned_loss=0.08456, over 984669.68 frames.], datatang_tot_loss[loss=0.2328, simple_loss=0.275, pruned_loss=0.09529, over 985196.52 frames.], batch size: 35, lr: 1.58e-03
+2022-06-18 13:12:25,409 INFO [train.py:874] (2/4) Epoch 4, batch 3000, aishell_loss[loss=0.2358, simple_loss=0.2969, pruned_loss=0.08733, over 4931.00 frames.], tot_loss[loss=0.2282, simple_loss=0.2784, pruned_loss=0.08899, over 985293.27 frames.], batch size: 58, aishell_tot_loss[loss=0.226, simple_loss=0.2848, pruned_loss=0.0836, over 984381.49 frames.], datatang_tot_loss[loss=0.2326, simple_loss=0.2746, pruned_loss=0.09527, over 985192.10 frames.], batch size: 58, lr: 1.57e-03
+2022-06-18 13:12:25,410 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 13:12:41,437 INFO [train.py:914] (2/4) Epoch 4, validation: loss=0.1874, simple_loss=0.266, pruned_loss=0.05438, over 1622729.00 frames.
+2022-06-18 13:13:12,184 INFO [train.py:874] (2/4) Epoch 4, batch 3050, datatang_loss[loss=0.2412, simple_loss=0.2853, pruned_loss=0.09851, over 4956.00 frames.], tot_loss[loss=0.2291, simple_loss=0.279, pruned_loss=0.08959, over 985133.34 frames.], batch size: 55, aishell_tot_loss[loss=0.2252, simple_loss=0.2843, pruned_loss=0.08303, over 984290.57 frames.], datatang_tot_loss[loss=0.2341, simple_loss=0.2754, pruned_loss=0.09637, over 985225.82 frames.], batch size: 55, lr: 1.57e-03
+2022-06-18 13:13:42,310 INFO [train.py:874] (2/4) Epoch 4, batch 3100, aishell_loss[loss=0.2221, simple_loss=0.2804, pruned_loss=0.08192, over 4967.00 frames.], tot_loss[loss=0.2283, simple_loss=0.2785, pruned_loss=0.08902, over 985307.26 frames.], batch size: 39, aishell_tot_loss[loss=0.2254, simple_loss=0.2844, pruned_loss=0.08325, over 984260.33 frames.], datatang_tot_loss[loss=0.2329, simple_loss=0.2745, pruned_loss=0.09564, over 985547.54 frames.], batch size: 39, lr: 1.57e-03
+2022-06-18 13:14:11,457 INFO [train.py:874] (2/4) Epoch 4, batch 3150, datatang_loss[loss=0.2291, simple_loss=0.267, pruned_loss=0.09559, over 4928.00 frames.], tot_loss[loss=0.227, simple_loss=0.2782, pruned_loss=0.0879, over 985037.87 frames.], batch size: 79, aishell_tot_loss[loss=0.2237, simple_loss=0.2833, pruned_loss=0.08207, over 984208.93 frames.], datatang_tot_loss[loss=0.2331, simple_loss=0.2749, pruned_loss=0.09567, over 985420.19 frames.], batch size: 79, lr: 1.57e-03
+2022-06-18 13:14:41,507 INFO [train.py:874] (2/4) Epoch 4, batch 3200, datatang_loss[loss=0.2498, simple_loss=0.2905, pruned_loss=0.1046, over 4964.00 frames.], tot_loss[loss=0.228, simple_loss=0.2787, pruned_loss=0.0887, over 985212.50 frames.], batch size: 34, aishell_tot_loss[loss=0.2236, simple_loss=0.283, pruned_loss=0.08212, over 984291.24 frames.], datatang_tot_loss[loss=0.2337, simple_loss=0.2758, pruned_loss=0.09583, over 985537.96 frames.], batch size: 34, lr: 1.57e-03
+2022-06-18 13:15:11,998 INFO [train.py:874] (2/4) Epoch 4, batch 3250, aishell_loss[loss=0.248, simple_loss=0.3068, pruned_loss=0.09459, over 4957.00 frames.], tot_loss[loss=0.227, simple_loss=0.2784, pruned_loss=0.08784, over 985219.72 frames.], batch size: 64, aishell_tot_loss[loss=0.2226, simple_loss=0.2825, pruned_loss=0.08132, over 984272.74 frames.], datatang_tot_loss[loss=0.2334, simple_loss=0.2758, pruned_loss=0.09545, over 985629.99 frames.], batch size: 64, lr: 1.56e-03
+2022-06-18 13:15:41,369 INFO [train.py:874] (2/4) Epoch 4, batch 3300, datatang_loss[loss=0.1947, simple_loss=0.2465, pruned_loss=0.0715, over 4928.00 frames.], tot_loss[loss=0.2275, simple_loss=0.2786, pruned_loss=0.08824, over 985291.26 frames.], batch size: 79, aishell_tot_loss[loss=0.2231, simple_loss=0.2827, pruned_loss=0.08172, over 984461.75 frames.], datatang_tot_loss[loss=0.2332, simple_loss=0.2758, pruned_loss=0.09532, over 985581.84 frames.], batch size: 79, lr: 1.56e-03
+2022-06-18 13:16:12,232 INFO [train.py:874] (2/4) Epoch 4, batch 3350, aishell_loss[loss=0.1676, simple_loss=0.2329, pruned_loss=0.05114, over 4872.00 frames.], tot_loss[loss=0.2275, simple_loss=0.2787, pruned_loss=0.08816, over 985113.71 frames.], batch size: 28, aishell_tot_loss[loss=0.2233, simple_loss=0.2828, pruned_loss=0.08186, over 984265.20 frames.], datatang_tot_loss[loss=0.2329, simple_loss=0.2755, pruned_loss=0.09515, over 985654.22 frames.], batch size: 28, lr: 1.56e-03
+2022-06-18 13:16:42,775 INFO [train.py:874] (2/4) Epoch 4, batch 3400, aishell_loss[loss=0.1901, simple_loss=0.258, pruned_loss=0.06111, over 4957.00 frames.], tot_loss[loss=0.2262, simple_loss=0.2783, pruned_loss=0.08707, over 985286.11 frames.], batch size: 31, aishell_tot_loss[loss=0.2224, simple_loss=0.2824, pruned_loss=0.08118, over 984166.83 frames.], datatang_tot_loss[loss=0.2324, simple_loss=0.2752, pruned_loss=0.09479, over 986008.12 frames.], batch size: 31, lr: 1.56e-03
+2022-06-18 13:17:12,116 INFO [train.py:874] (2/4) Epoch 4, batch 3450, aishell_loss[loss=0.2484, simple_loss=0.3038, pruned_loss=0.09652, over 4961.00 frames.], tot_loss[loss=0.2272, simple_loss=0.2791, pruned_loss=0.08762, over 985073.47 frames.], batch size: 77, aishell_tot_loss[loss=0.2234, simple_loss=0.2831, pruned_loss=0.08183, over 984149.12 frames.], datatang_tot_loss[loss=0.2323, simple_loss=0.2752, pruned_loss=0.09469, over 985863.36 frames.], batch size: 77, lr: 1.55e-03
+2022-06-18 13:17:41,243 INFO [train.py:874] (2/4) Epoch 4, batch 3500, datatang_loss[loss=0.2516, simple_loss=0.2704, pruned_loss=0.1164, over 4923.00 frames.], tot_loss[loss=0.2271, simple_loss=0.2793, pruned_loss=0.0875, over 984996.82 frames.], batch size: 73, aishell_tot_loss[loss=0.2236, simple_loss=0.2832, pruned_loss=0.08197, over 984263.43 frames.], datatang_tot_loss[loss=0.2322, simple_loss=0.2751, pruned_loss=0.0946, over 985703.00 frames.], batch size: 73, lr: 1.55e-03
+2022-06-18 13:18:17,392 INFO [train.py:874] (2/4) Epoch 4, batch 3550, datatang_loss[loss=0.2368, simple_loss=0.2854, pruned_loss=0.09414, over 4918.00 frames.], tot_loss[loss=0.2281, simple_loss=0.2797, pruned_loss=0.08827, over 984884.57 frames.], batch size: 83, aishell_tot_loss[loss=0.2229, simple_loss=0.2828, pruned_loss=0.08149, over 984196.36 frames.], datatang_tot_loss[loss=0.2336, simple_loss=0.2761, pruned_loss=0.09559, over 985636.61 frames.], batch size: 83, lr: 1.55e-03
+2022-06-18 13:18:45,982 INFO [train.py:874] (2/4) Epoch 4, batch 3600, aishell_loss[loss=0.22, simple_loss=0.2838, pruned_loss=0.07808, over 4939.00 frames.], tot_loss[loss=0.2286, simple_loss=0.2803, pruned_loss=0.08839, over 984975.33 frames.], batch size: 58, aishell_tot_loss[loss=0.2233, simple_loss=0.2833, pruned_loss=0.08167, over 984550.98 frames.], datatang_tot_loss[loss=0.2336, simple_loss=0.2763, pruned_loss=0.09548, over 985396.78 frames.], batch size: 58, lr: 1.55e-03
+2022-06-18 13:19:16,707 INFO [train.py:874] (2/4) Epoch 4, batch 3650, aishell_loss[loss=0.1913, simple_loss=0.2599, pruned_loss=0.06137, over 4928.00 frames.], tot_loss[loss=0.228, simple_loss=0.2796, pruned_loss=0.08821, over 984733.90 frames.], batch size: 33, aishell_tot_loss[loss=0.2228, simple_loss=0.2829, pruned_loss=0.08136, over 984318.02 frames.], datatang_tot_loss[loss=0.2332, simple_loss=0.2764, pruned_loss=0.09506, over 985356.54 frames.], batch size: 33, lr: 1.54e-03
+2022-06-18 13:19:48,113 INFO [train.py:874] (2/4) Epoch 4, batch 3700, datatang_loss[loss=0.2434, simple_loss=0.2835, pruned_loss=0.1016, over 4898.00 frames.], tot_loss[loss=0.2258, simple_loss=0.2779, pruned_loss=0.08689, over 984473.18 frames.], batch size: 52, aishell_tot_loss[loss=0.2222, simple_loss=0.2825, pruned_loss=0.08095, over 984390.40 frames.], datatang_tot_loss[loss=0.2315, simple_loss=0.275, pruned_loss=0.09399, over 985001.65 frames.], batch size: 52, lr: 1.54e-03
+2022-06-18 13:20:16,793 INFO [train.py:874] (2/4) Epoch 4, batch 3750, datatang_loss[loss=0.2016, simple_loss=0.2534, pruned_loss=0.07489, over 4954.00 frames.], tot_loss[loss=0.2267, simple_loss=0.2787, pruned_loss=0.08736, over 984814.30 frames.], batch size: 67, aishell_tot_loss[loss=0.2232, simple_loss=0.2834, pruned_loss=0.08147, over 984720.97 frames.], datatang_tot_loss[loss=0.2312, simple_loss=0.2748, pruned_loss=0.09381, over 984969.76 frames.], batch size: 67, lr: 1.54e-03
+2022-06-18 13:20:47,145 INFO [train.py:874] (2/4) Epoch 4, batch 3800, datatang_loss[loss=0.2185, simple_loss=0.2678, pruned_loss=0.08459, over 4932.00 frames.], tot_loss[loss=0.2266, simple_loss=0.2786, pruned_loss=0.08732, over 985009.58 frames.], batch size: 62, aishell_tot_loss[loss=0.223, simple_loss=0.2833, pruned_loss=0.08134, over 984846.25 frames.], datatang_tot_loss[loss=0.2312, simple_loss=0.2747, pruned_loss=0.09383, over 985046.14 frames.], batch size: 62, lr: 1.54e-03
+2022-06-18 13:21:17,332 INFO [train.py:874] (2/4) Epoch 4, batch 3850, datatang_loss[loss=0.232, simple_loss=0.2756, pruned_loss=0.09418, over 4916.00 frames.], tot_loss[loss=0.2258, simple_loss=0.2781, pruned_loss=0.08673, over 985577.65 frames.], batch size: 57, aishell_tot_loss[loss=0.2234, simple_loss=0.2836, pruned_loss=0.0816, over 985037.13 frames.], datatang_tot_loss[loss=0.2296, simple_loss=0.2738, pruned_loss=0.09273, over 985463.97 frames.], batch size: 57, lr: 1.54e-03
+2022-06-18 13:21:46,275 INFO [train.py:874] (2/4) Epoch 4, batch 3900, aishell_loss[loss=0.2112, simple_loss=0.2817, pruned_loss=0.07038, over 4972.00 frames.], tot_loss[loss=0.2265, simple_loss=0.2784, pruned_loss=0.08727, over 985523.75 frames.], batch size: 44, aishell_tot_loss[loss=0.224, simple_loss=0.2838, pruned_loss=0.08209, over 985005.71 frames.], datatang_tot_loss[loss=0.2298, simple_loss=0.2737, pruned_loss=0.09291, over 985513.64 frames.], batch size: 44, lr: 1.53e-03
+2022-06-18 13:22:14,508 INFO [train.py:874] (2/4) Epoch 4, batch 3950, datatang_loss[loss=0.214, simple_loss=0.2683, pruned_loss=0.07988, over 4905.00 frames.], tot_loss[loss=0.2254, simple_loss=0.2778, pruned_loss=0.08648, over 985427.29 frames.], batch size: 42, aishell_tot_loss[loss=0.2226, simple_loss=0.2827, pruned_loss=0.08127, over 984926.88 frames.], datatang_tot_loss[loss=0.2299, simple_loss=0.2741, pruned_loss=0.09286, over 985566.95 frames.], batch size: 42, lr: 1.53e-03
+2022-06-18 13:22:44,204 INFO [train.py:874] (2/4) Epoch 4, batch 4000, aishell_loss[loss=0.2296, simple_loss=0.2933, pruned_loss=0.08298, over 4883.00 frames.], tot_loss[loss=0.2254, simple_loss=0.2779, pruned_loss=0.08644, over 985643.12 frames.], batch size: 47, aishell_tot_loss[loss=0.2224, simple_loss=0.2828, pruned_loss=0.08103, over 984921.46 frames.], datatang_tot_loss[loss=0.2298, simple_loss=0.2738, pruned_loss=0.09289, over 985840.96 frames.], batch size: 47, lr: 1.53e-03
+2022-06-18 13:22:44,205 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 13:23:00,877 INFO [train.py:914] (2/4) Epoch 4, validation: loss=0.1818, simple_loss=0.2607, pruned_loss=0.0515, over 1622729.00 frames.
+2022-06-18 13:23:30,358 INFO [train.py:874] (2/4) Epoch 4, batch 4050, aishell_loss[loss=0.1568, simple_loss=0.2108, pruned_loss=0.05139, over 4780.00 frames.], tot_loss[loss=0.2252, simple_loss=0.2776, pruned_loss=0.08635, over 985464.59 frames.], batch size: 21, aishell_tot_loss[loss=0.2222, simple_loss=0.2826, pruned_loss=0.08087, over 984625.13 frames.], datatang_tot_loss[loss=0.2296, simple_loss=0.2737, pruned_loss=0.09278, over 986015.36 frames.], batch size: 21, lr: 1.53e-03
+2022-06-18 13:24:00,225 INFO [train.py:874] (2/4) Epoch 4, batch 4100, datatang_loss[loss=0.2176, simple_loss=0.2685, pruned_loss=0.08333, over 4923.00 frames.], tot_loss[loss=0.2243, simple_loss=0.2764, pruned_loss=0.08612, over 985488.89 frames.], batch size: 73, aishell_tot_loss[loss=0.2214, simple_loss=0.282, pruned_loss=0.08039, over 984906.23 frames.], datatang_tot_loss[loss=0.2291, simple_loss=0.2731, pruned_loss=0.09255, over 985782.88 frames.], batch size: 73, lr: 1.53e-03
+2022-06-18 13:24:29,183 INFO [train.py:874] (2/4) Epoch 4, batch 4150, aishell_loss[loss=0.2335, simple_loss=0.2958, pruned_loss=0.08564, over 4951.00 frames.], tot_loss[loss=0.2245, simple_loss=0.2764, pruned_loss=0.08633, over 985678.99 frames.], batch size: 58, aishell_tot_loss[loss=0.2214, simple_loss=0.2816, pruned_loss=0.08059, over 985205.97 frames.], datatang_tot_loss[loss=0.2291, simple_loss=0.2732, pruned_loss=0.09244, over 985722.31 frames.], batch size: 58, lr: 1.52e-03
+2022-06-18 13:25:55,784 INFO [train.py:874] (2/4) Epoch 5, batch 50, datatang_loss[loss=0.1945, simple_loss=0.2446, pruned_loss=0.07219, over 4927.00 frames.], tot_loss[loss=0.2166, simple_loss=0.2699, pruned_loss=0.08164, over 218839.78 frames.], batch size: 71, aishell_tot_loss[loss=0.219, simple_loss=0.2807, pruned_loss=0.0786, over 98657.95 frames.], datatang_tot_loss[loss=0.2148, simple_loss=0.2621, pruned_loss=0.08373, over 133534.44 frames.], batch size: 71, lr: 1.47e-03
+2022-06-18 13:26:26,147 INFO [train.py:874] (2/4) Epoch 5, batch 100, aishell_loss[loss=0.2465, simple_loss=0.2976, pruned_loss=0.09767, over 4946.00 frames.], tot_loss[loss=0.2152, simple_loss=0.2707, pruned_loss=0.07987, over 388804.32 frames.], batch size: 56, aishell_tot_loss[loss=0.2214, simple_loss=0.2834, pruned_loss=0.07966, over 195304.29 frames.], datatang_tot_loss[loss=0.2103, simple_loss=0.2603, pruned_loss=0.08017, over 241293.06 frames.], batch size: 56, lr: 1.46e-03
+2022-06-18 13:26:56,294 INFO [train.py:874] (2/4) Epoch 5, batch 150, aishell_loss[loss=0.2142, simple_loss=0.2711, pruned_loss=0.07861, over 4913.00 frames.], tot_loss[loss=0.2132, simple_loss=0.2697, pruned_loss=0.07828, over 521108.07 frames.], batch size: 33, aishell_tot_loss[loss=0.2198, simple_loss=0.2827, pruned_loss=0.07843, over 291916.63 frames.], datatang_tot_loss[loss=0.2078, simple_loss=0.2582, pruned_loss=0.07866, over 325612.15 frames.], batch size: 33, lr: 1.46e-03
+2022-06-18 13:27:26,068 INFO [train.py:874] (2/4) Epoch 5, batch 200, aishell_loss[loss=0.2182, simple_loss=0.2816, pruned_loss=0.07743, over 4981.00 frames.], tot_loss[loss=0.2126, simple_loss=0.2687, pruned_loss=0.07824, over 624328.01 frames.], batch size: 43, aishell_tot_loss[loss=0.2175, simple_loss=0.2799, pruned_loss=0.07757, over 373910.62 frames.], datatang_tot_loss[loss=0.2087, simple_loss=0.2587, pruned_loss=0.07933, over 403316.56 frames.], batch size: 43, lr: 1.46e-03
+2022-06-18 13:27:56,608 INFO [train.py:874] (2/4) Epoch 5, batch 250, datatang_loss[loss=0.2983, simple_loss=0.3291, pruned_loss=0.1338, over 4926.00 frames.], tot_loss[loss=0.2168, simple_loss=0.2707, pruned_loss=0.08146, over 704514.99 frames.], batch size: 108, aishell_tot_loss[loss=0.2158, simple_loss=0.2775, pruned_loss=0.07707, over 437575.33 frames.], datatang_tot_loss[loss=0.2165, simple_loss=0.2641, pruned_loss=0.08447, over 479884.53 frames.], batch size: 108, lr: 1.46e-03
+2022-06-18 13:28:27,267 INFO [train.py:874] (2/4) Epoch 5, batch 300, datatang_loss[loss=0.2251, simple_loss=0.264, pruned_loss=0.09308, over 4966.00 frames.], tot_loss[loss=0.2165, simple_loss=0.2702, pruned_loss=0.08135, over 766735.22 frames.], batch size: 37, aishell_tot_loss[loss=0.2166, simple_loss=0.2778, pruned_loss=0.07771, over 504291.54 frames.], datatang_tot_loss[loss=0.2154, simple_loss=0.2628, pruned_loss=0.08398, over 537344.08 frames.], batch size: 37, lr: 1.46e-03
+2022-06-18 13:28:56,549 INFO [train.py:874] (2/4) Epoch 5, batch 350, datatang_loss[loss=0.1939, simple_loss=0.2432, pruned_loss=0.07227, over 4901.00 frames.], tot_loss[loss=0.2174, simple_loss=0.2716, pruned_loss=0.08163, over 815052.66 frames.], batch size: 47, aishell_tot_loss[loss=0.2173, simple_loss=0.2789, pruned_loss=0.07786, over 565070.86 frames.], datatang_tot_loss[loss=0.2163, simple_loss=0.2635, pruned_loss=0.0845, over 586083.92 frames.], batch size: 47, lr: 1.45e-03
+2022-06-18 13:29:27,461 INFO [train.py:874] (2/4) Epoch 5, batch 400, aishell_loss[loss=0.1401, simple_loss=0.2059, pruned_loss=0.0371, over 4840.00 frames.], tot_loss[loss=0.2158, simple_loss=0.2707, pruned_loss=0.08049, over 852194.61 frames.], batch size: 24, aishell_tot_loss[loss=0.2159, simple_loss=0.2776, pruned_loss=0.07712, over 604497.19 frames.], datatang_tot_loss[loss=0.2157, simple_loss=0.2643, pruned_loss=0.08356, over 641846.84 frames.], batch size: 24, lr: 1.45e-03
+2022-06-18 13:29:57,469 INFO [train.py:874] (2/4) Epoch 5, batch 450, aishell_loss[loss=0.2121, simple_loss=0.283, pruned_loss=0.07059, over 4966.00 frames.], tot_loss[loss=0.2166, simple_loss=0.2723, pruned_loss=0.08048, over 881894.30 frames.], batch size: 61, aishell_tot_loss[loss=0.2159, simple_loss=0.2784, pruned_loss=0.07667, over 655957.36 frames.], datatang_tot_loss[loss=0.2169, simple_loss=0.2652, pruned_loss=0.08425, over 676424.09 frames.], batch size: 61, lr: 1.45e-03
+2022-06-18 13:30:27,316 INFO [train.py:874] (2/4) Epoch 5, batch 500, datatang_loss[loss=0.2114, simple_loss=0.2637, pruned_loss=0.07956, over 4931.00 frames.], tot_loss[loss=0.2166, simple_loss=0.2719, pruned_loss=0.0806, over 905160.31 frames.], batch size: 69, aishell_tot_loss[loss=0.2162, simple_loss=0.2786, pruned_loss=0.07693, over 690501.08 frames.], datatang_tot_loss[loss=0.2165, simple_loss=0.265, pruned_loss=0.08402, over 717084.49 frames.], batch size: 69, lr: 1.45e-03
+2022-06-18 13:30:57,314 INFO [train.py:874] (2/4) Epoch 5, batch 550, aishell_loss[loss=0.2383, simple_loss=0.2993, pruned_loss=0.08862, over 4972.00 frames.], tot_loss[loss=0.217, simple_loss=0.2726, pruned_loss=0.08068, over 923280.44 frames.], batch size: 39, aishell_tot_loss[loss=0.2167, simple_loss=0.2792, pruned_loss=0.07707, over 726791.66 frames.], datatang_tot_loss[loss=0.2167, simple_loss=0.2652, pruned_loss=0.08406, over 747587.50 frames.], batch size: 39, lr: 1.45e-03
+2022-06-18 13:31:27,951 INFO [train.py:874] (2/4) Epoch 5, batch 600, aishell_loss[loss=0.2134, simple_loss=0.2772, pruned_loss=0.07481, over 4953.00 frames.], tot_loss[loss=0.2175, simple_loss=0.2734, pruned_loss=0.08079, over 937131.13 frames.], batch size: 40, aishell_tot_loss[loss=0.2171, simple_loss=0.2796, pruned_loss=0.07727, over 759601.63 frames.], datatang_tot_loss[loss=0.217, simple_loss=0.2657, pruned_loss=0.08412, over 773486.27 frames.], batch size: 40, lr: 1.44e-03
+2022-06-18 13:31:56,193 INFO [train.py:874] (2/4) Epoch 5, batch 650, aishell_loss[loss=0.1968, simple_loss=0.2673, pruned_loss=0.06315, over 4870.00 frames.], tot_loss[loss=0.2169, simple_loss=0.2728, pruned_loss=0.08044, over 947770.95 frames.], batch size: 36, aishell_tot_loss[loss=0.2165, simple_loss=0.2791, pruned_loss=0.07699, over 788993.51 frames.], datatang_tot_loss[loss=0.2169, simple_loss=0.2656, pruned_loss=0.08413, over 795739.25 frames.], batch size: 36, lr: 1.44e-03
+2022-06-18 13:32:27,640 INFO [train.py:874] (2/4) Epoch 5, batch 700, aishell_loss[loss=0.2384, simple_loss=0.2954, pruned_loss=0.09067, over 4872.00 frames.], tot_loss[loss=0.2163, simple_loss=0.2724, pruned_loss=0.08013, over 956190.85 frames.], batch size: 35, aishell_tot_loss[loss=0.2155, simple_loss=0.2781, pruned_loss=0.07644, over 814751.21 frames.], datatang_tot_loss[loss=0.2174, simple_loss=0.266, pruned_loss=0.08442, over 815572.91 frames.], batch size: 35, lr: 1.44e-03
+2022-06-18 13:32:56,845 INFO [train.py:874] (2/4) Epoch 5, batch 750, datatang_loss[loss=0.2115, simple_loss=0.262, pruned_loss=0.08056, over 4969.00 frames.], tot_loss[loss=0.2171, simple_loss=0.2728, pruned_loss=0.08075, over 962566.78 frames.], batch size: 37, aishell_tot_loss[loss=0.2151, simple_loss=0.2779, pruned_loss=0.0762, over 835372.18 frames.], datatang_tot_loss[loss=0.2188, simple_loss=0.2668, pruned_loss=0.08541, over 834956.12 frames.], batch size: 37, lr: 1.44e-03
+2022-06-18 13:33:26,627 INFO [train.py:874] (2/4) Epoch 5, batch 800, datatang_loss[loss=0.2192, simple_loss=0.2702, pruned_loss=0.08406, over 4911.00 frames.], tot_loss[loss=0.2183, simple_loss=0.2736, pruned_loss=0.08149, over 968035.46 frames.], batch size: 81, aishell_tot_loss[loss=0.2156, simple_loss=0.2781, pruned_loss=0.07653, over 852021.91 frames.], datatang_tot_loss[loss=0.2197, simple_loss=0.2677, pruned_loss=0.08584, over 854109.24 frames.], batch size: 81, lr: 1.44e-03
+2022-06-18 13:33:57,569 INFO [train.py:874] (2/4) Epoch 5, batch 850, aishell_loss[loss=0.2213, simple_loss=0.2897, pruned_loss=0.0764, over 4971.00 frames.], tot_loss[loss=0.2173, simple_loss=0.2728, pruned_loss=0.08086, over 971931.41 frames.], batch size: 39, aishell_tot_loss[loss=0.2149, simple_loss=0.2774, pruned_loss=0.07616, over 867255.62 frames.], datatang_tot_loss[loss=0.2194, simple_loss=0.2677, pruned_loss=0.08551, over 870085.39 frames.], batch size: 39, lr: 1.43e-03
+2022-06-18 13:34:26,193 INFO [train.py:874] (2/4) Epoch 5, batch 900, aishell_loss[loss=0.2281, simple_loss=0.2792, pruned_loss=0.08855, over 4887.00 frames.], tot_loss[loss=0.2176, simple_loss=0.2731, pruned_loss=0.08101, over 974834.85 frames.], batch size: 42, aishell_tot_loss[loss=0.2152, simple_loss=0.2778, pruned_loss=0.0763, over 880537.36 frames.], datatang_tot_loss[loss=0.2194, simple_loss=0.2678, pruned_loss=0.08551, over 884200.65 frames.], batch size: 42, lr: 1.43e-03
+2022-06-18 13:34:56,204 INFO [train.py:874] (2/4) Epoch 5, batch 950, datatang_loss[loss=0.1897, simple_loss=0.2548, pruned_loss=0.06226, over 4925.00 frames.], tot_loss[loss=0.2172, simple_loss=0.273, pruned_loss=0.08071, over 977520.10 frames.], batch size: 81, aishell_tot_loss[loss=0.2148, simple_loss=0.2778, pruned_loss=0.07595, over 892160.70 frames.], datatang_tot_loss[loss=0.2194, simple_loss=0.2679, pruned_loss=0.08545, over 897172.14 frames.], batch size: 81, lr: 1.43e-03
+2022-06-18 13:35:26,997 INFO [train.py:874] (2/4) Epoch 5, batch 1000, datatang_loss[loss=0.2021, simple_loss=0.2559, pruned_loss=0.07418, over 4911.00 frames.], tot_loss[loss=0.2184, simple_loss=0.2738, pruned_loss=0.08147, over 979167.01 frames.], batch size: 75, aishell_tot_loss[loss=0.2145, simple_loss=0.2774, pruned_loss=0.07578, over 902818.72 frames.], datatang_tot_loss[loss=0.2211, simple_loss=0.2692, pruned_loss=0.0865, over 907779.12 frames.], batch size: 75, lr: 1.43e-03
+2022-06-18 13:35:26,998 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 13:35:43,411 INFO [train.py:914] (2/4) Epoch 5, validation: loss=0.1786, simple_loss=0.258, pruned_loss=0.04955, over 1622729.00 frames.
+2022-06-18 13:36:13,975 INFO [train.py:874] (2/4) Epoch 5, batch 1050, aishell_loss[loss=0.2078, simple_loss=0.2657, pruned_loss=0.07495, over 4865.00 frames.], tot_loss[loss=0.2182, simple_loss=0.2733, pruned_loss=0.08155, over 980377.83 frames.], batch size: 28, aishell_tot_loss[loss=0.2143, simple_loss=0.2772, pruned_loss=0.0757, over 910499.45 frames.], datatang_tot_loss[loss=0.2211, simple_loss=0.2693, pruned_loss=0.08649, over 918635.94 frames.], batch size: 28, lr: 1.43e-03
+2022-06-18 13:36:44,160 INFO [train.py:874] (2/4) Epoch 5, batch 1100, datatang_loss[loss=0.2463, simple_loss=0.2743, pruned_loss=0.1091, over 4832.00 frames.], tot_loss[loss=0.2184, simple_loss=0.274, pruned_loss=0.08142, over 981517.10 frames.], batch size: 23, aishell_tot_loss[loss=0.2146, simple_loss=0.2775, pruned_loss=0.07582, over 920709.44 frames.], datatang_tot_loss[loss=0.2214, simple_loss=0.2696, pruned_loss=0.08657, over 925264.57 frames.], batch size: 23, lr: 1.43e-03
+2022-06-18 13:37:13,244 INFO [train.py:874] (2/4) Epoch 5, batch 1150, datatang_loss[loss=0.185, simple_loss=0.2352, pruned_loss=0.06744, over 4957.00 frames.], tot_loss[loss=0.217, simple_loss=0.2724, pruned_loss=0.08077, over 982476.38 frames.], batch size: 55, aishell_tot_loss[loss=0.2135, simple_loss=0.2764, pruned_loss=0.07534, over 927324.26 frames.], datatang_tot_loss[loss=0.2209, simple_loss=0.2692, pruned_loss=0.08627, over 933372.58 frames.], batch size: 55, lr: 1.42e-03
+2022-06-18 13:37:45,219 INFO [train.py:874] (2/4) Epoch 5, batch 1200, aishell_loss[loss=0.2369, simple_loss=0.3011, pruned_loss=0.08633, over 4913.00 frames.], tot_loss[loss=0.2176, simple_loss=0.2729, pruned_loss=0.08116, over 982760.02 frames.], batch size: 52, aishell_tot_loss[loss=0.2133, simple_loss=0.2763, pruned_loss=0.07518, over 933663.00 frames.], datatang_tot_loss[loss=0.2217, simple_loss=0.2698, pruned_loss=0.08683, over 939609.87 frames.], batch size: 52, lr: 1.42e-03
+2022-06-18 13:38:15,656 INFO [train.py:874] (2/4) Epoch 5, batch 1250, aishell_loss[loss=0.2374, simple_loss=0.3002, pruned_loss=0.08729, over 4966.00 frames.], tot_loss[loss=0.219, simple_loss=0.2743, pruned_loss=0.0819, over 983588.63 frames.], batch size: 61, aishell_tot_loss[loss=0.2143, simple_loss=0.2774, pruned_loss=0.07559, over 940056.72 frames.], datatang_tot_loss[loss=0.2223, simple_loss=0.27, pruned_loss=0.0873, over 944994.61 frames.], batch size: 61, lr: 1.42e-03
+2022-06-18 13:38:44,448 INFO [train.py:874] (2/4) Epoch 5, batch 1300, datatang_loss[loss=0.2535, simple_loss=0.2976, pruned_loss=0.1047, over 4932.00 frames.], tot_loss[loss=0.2191, simple_loss=0.2745, pruned_loss=0.08182, over 984153.11 frames.], batch size: 94, aishell_tot_loss[loss=0.2138, simple_loss=0.2772, pruned_loss=0.07517, over 945514.19 frames.], datatang_tot_loss[loss=0.223, simple_loss=0.2706, pruned_loss=0.08773, over 949807.64 frames.], batch size: 94, lr: 1.42e-03
+2022-06-18 13:39:15,003 INFO [train.py:874] (2/4) Epoch 5, batch 1350, datatang_loss[loss=0.2369, simple_loss=0.2949, pruned_loss=0.08951, over 4980.00 frames.], tot_loss[loss=0.2189, simple_loss=0.2745, pruned_loss=0.08163, over 984553.42 frames.], batch size: 37, aishell_tot_loss[loss=0.213, simple_loss=0.2765, pruned_loss=0.07469, over 949989.46 frames.], datatang_tot_loss[loss=0.2237, simple_loss=0.2714, pruned_loss=0.08803, over 954322.98 frames.], batch size: 37, lr: 1.42e-03
+2022-06-18 13:39:44,957 INFO [train.py:874] (2/4) Epoch 5, batch 1400, datatang_loss[loss=0.2633, simple_loss=0.2909, pruned_loss=0.1179, over 4954.00 frames.], tot_loss[loss=0.2187, simple_loss=0.2739, pruned_loss=0.0818, over 985006.57 frames.], batch size: 86, aishell_tot_loss[loss=0.2122, simple_loss=0.2758, pruned_loss=0.07432, over 954277.81 frames.], datatang_tot_loss[loss=0.2244, simple_loss=0.2716, pruned_loss=0.08863, over 958143.30 frames.], batch size: 86, lr: 1.41e-03
+2022-06-18 13:40:14,813 INFO [train.py:874] (2/4) Epoch 5, batch 1450, aishell_loss[loss=0.2015, simple_loss=0.2673, pruned_loss=0.06788, over 4911.00 frames.], tot_loss[loss=0.2192, simple_loss=0.2748, pruned_loss=0.08185, over 985337.15 frames.], batch size: 52, aishell_tot_loss[loss=0.2132, simple_loss=0.2771, pruned_loss=0.07469, over 958252.54 frames.], datatang_tot_loss[loss=0.2241, simple_loss=0.2714, pruned_loss=0.08846, over 961305.98 frames.], batch size: 52, lr: 1.41e-03
+2022-06-18 13:40:45,878 INFO [train.py:874] (2/4) Epoch 5, batch 1500, datatang_loss[loss=0.1962, simple_loss=0.2485, pruned_loss=0.07192, over 4933.00 frames.], tot_loss[loss=0.2194, simple_loss=0.2749, pruned_loss=0.08199, over 985403.74 frames.], batch size: 79, aishell_tot_loss[loss=0.2136, simple_loss=0.2774, pruned_loss=0.07492, over 960782.21 frames.], datatang_tot_loss[loss=0.2238, simple_loss=0.2714, pruned_loss=0.08812, over 964804.19 frames.], batch size: 79, lr: 1.41e-03
+2022-06-18 13:41:16,035 INFO [train.py:874] (2/4) Epoch 5, batch 1550, datatang_loss[loss=0.2083, simple_loss=0.2694, pruned_loss=0.07358, over 4982.00 frames.], tot_loss[loss=0.2192, simple_loss=0.2748, pruned_loss=0.08182, over 985414.13 frames.], batch size: 34, aishell_tot_loss[loss=0.2145, simple_loss=0.2782, pruned_loss=0.07546, over 963293.13 frames.], datatang_tot_loss[loss=0.2227, simple_loss=0.2708, pruned_loss=0.08733, over 967562.29 frames.], batch size: 34, lr: 1.41e-03
+2022-06-18 13:41:45,623 INFO [train.py:874] (2/4) Epoch 5, batch 1600, aishell_loss[loss=0.2417, simple_loss=0.3014, pruned_loss=0.091, over 4928.00 frames.], tot_loss[loss=0.2206, simple_loss=0.2754, pruned_loss=0.08291, over 985332.88 frames.], batch size: 46, aishell_tot_loss[loss=0.2165, simple_loss=0.2791, pruned_loss=0.07695, over 965889.56 frames.], datatang_tot_loss[loss=0.2225, simple_loss=0.2705, pruned_loss=0.08722, over 969622.24 frames.], batch size: 46, lr: 1.41e-03
+2022-06-18 13:42:15,550 INFO [train.py:874] (2/4) Epoch 5, batch 1650, datatang_loss[loss=0.2153, simple_loss=0.2719, pruned_loss=0.07935, over 4948.00 frames.], tot_loss[loss=0.2215, simple_loss=0.2762, pruned_loss=0.08339, over 985754.67 frames.], batch size: 91, aishell_tot_loss[loss=0.2164, simple_loss=0.2793, pruned_loss=0.07678, over 968111.48 frames.], datatang_tot_loss[loss=0.2236, simple_loss=0.2713, pruned_loss=0.08789, over 971932.95 frames.], batch size: 91, lr: 1.40e-03
+2022-06-18 13:42:46,145 INFO [train.py:874] (2/4) Epoch 5, batch 1700, aishell_loss[loss=0.1999, simple_loss=0.2705, pruned_loss=0.06471, over 4867.00 frames.], tot_loss[loss=0.2218, simple_loss=0.2768, pruned_loss=0.08342, over 985392.52 frames.], batch size: 36, aishell_tot_loss[loss=0.2166, simple_loss=0.2797, pruned_loss=0.07677, over 969841.38 frames.], datatang_tot_loss[loss=0.2241, simple_loss=0.2719, pruned_loss=0.08815, over 973530.44 frames.], batch size: 36, lr: 1.40e-03
+2022-06-18 13:43:15,673 INFO [train.py:874] (2/4) Epoch 5, batch 1750, datatang_loss[loss=0.1941, simple_loss=0.2488, pruned_loss=0.06973, over 4903.00 frames.], tot_loss[loss=0.2212, simple_loss=0.2761, pruned_loss=0.08311, over 985406.37 frames.], batch size: 52, aishell_tot_loss[loss=0.2163, simple_loss=0.2792, pruned_loss=0.07676, over 971449.95 frames.], datatang_tot_loss[loss=0.2241, simple_loss=0.2719, pruned_loss=0.08809, over 975161.15 frames.], batch size: 52, lr: 1.40e-03
+2022-06-18 13:43:46,684 INFO [train.py:874] (2/4) Epoch 5, batch 1800, aishell_loss[loss=0.2378, simple_loss=0.2973, pruned_loss=0.08919, over 4965.00 frames.], tot_loss[loss=0.2204, simple_loss=0.2757, pruned_loss=0.08257, over 985367.18 frames.], batch size: 44, aishell_tot_loss[loss=0.216, simple_loss=0.2792, pruned_loss=0.07642, over 972744.00 frames.], datatang_tot_loss[loss=0.2236, simple_loss=0.2718, pruned_loss=0.08765, over 976553.48 frames.], batch size: 44, lr: 1.40e-03
+2022-06-18 13:44:17,128 INFO [train.py:874] (2/4) Epoch 5, batch 1850, datatang_loss[loss=0.1934, simple_loss=0.2481, pruned_loss=0.06934, over 4941.00 frames.], tot_loss[loss=0.2202, simple_loss=0.2753, pruned_loss=0.0825, over 985292.11 frames.], batch size: 50, aishell_tot_loss[loss=0.2159, simple_loss=0.2792, pruned_loss=0.07633, over 974011.06 frames.], datatang_tot_loss[loss=0.2235, simple_loss=0.2715, pruned_loss=0.08771, over 977711.01 frames.], batch size: 50, lr: 1.40e-03
+2022-06-18 13:44:47,045 INFO [train.py:874] (2/4) Epoch 5, batch 1900, datatang_loss[loss=0.2456, simple_loss=0.2935, pruned_loss=0.0989, over 4831.00 frames.], tot_loss[loss=0.2201, simple_loss=0.2756, pruned_loss=0.08226, over 985346.68 frames.], batch size: 30, aishell_tot_loss[loss=0.2162, simple_loss=0.2797, pruned_loss=0.07635, over 975696.87 frames.], datatang_tot_loss[loss=0.2233, simple_loss=0.2713, pruned_loss=0.08767, over 978363.79 frames.], batch size: 30, lr: 1.40e-03
+2022-06-18 13:45:17,707 INFO [train.py:874] (2/4) Epoch 5, batch 1950, datatang_loss[loss=0.217, simple_loss=0.2792, pruned_loss=0.07736, over 4937.00 frames.], tot_loss[loss=0.2198, simple_loss=0.2748, pruned_loss=0.08234, over 985314.72 frames.], batch size: 88, aishell_tot_loss[loss=0.2157, simple_loss=0.2789, pruned_loss=0.07621, over 976785.89 frames.], datatang_tot_loss[loss=0.2235, simple_loss=0.2713, pruned_loss=0.08783, over 979177.51 frames.], batch size: 88, lr: 1.39e-03
+2022-06-18 13:45:46,806 INFO [train.py:874] (2/4) Epoch 5, batch 2000, aishell_loss[loss=0.2726, simple_loss=0.3243, pruned_loss=0.1105, over 4972.00 frames.], tot_loss[loss=0.2208, simple_loss=0.2759, pruned_loss=0.08289, over 985539.45 frames.], batch size: 61, aishell_tot_loss[loss=0.2166, simple_loss=0.2797, pruned_loss=0.07674, over 977899.02 frames.], datatang_tot_loss[loss=0.2238, simple_loss=0.2715, pruned_loss=0.08806, over 980055.92 frames.], batch size: 61, lr: 1.39e-03
+2022-06-18 13:45:46,807 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 13:46:03,085 INFO [train.py:914] (2/4) Epoch 5, validation: loss=0.1805, simple_loss=0.2595, pruned_loss=0.05074, over 1622729.00 frames.
+2022-06-18 13:46:32,370 INFO [train.py:874] (2/4) Epoch 5, batch 2050, datatang_loss[loss=0.2439, simple_loss=0.2909, pruned_loss=0.09851, over 4930.00 frames.], tot_loss[loss=0.2203, simple_loss=0.2758, pruned_loss=0.08239, over 985463.33 frames.], batch size: 94, aishell_tot_loss[loss=0.2164, simple_loss=0.2796, pruned_loss=0.07657, over 978688.41 frames.], datatang_tot_loss[loss=0.2236, simple_loss=0.2715, pruned_loss=0.08786, over 980714.30 frames.], batch size: 94, lr: 1.39e-03
+2022-06-18 13:47:01,877 INFO [train.py:874] (2/4) Epoch 5, batch 2100, datatang_loss[loss=0.2038, simple_loss=0.2631, pruned_loss=0.07223, over 4921.00 frames.], tot_loss[loss=0.2196, simple_loss=0.2757, pruned_loss=0.08176, over 985528.09 frames.], batch size: 64, aishell_tot_loss[loss=0.2162, simple_loss=0.2796, pruned_loss=0.07639, over 979741.32 frames.], datatang_tot_loss[loss=0.2234, simple_loss=0.2712, pruned_loss=0.08778, over 981132.80 frames.], batch size: 64, lr: 1.39e-03
+2022-06-18 13:47:31,754 INFO [train.py:874] (2/4) Epoch 5, batch 2150, aishell_loss[loss=0.216, simple_loss=0.2772, pruned_loss=0.07738, over 4942.00 frames.], tot_loss[loss=0.2177, simple_loss=0.2743, pruned_loss=0.0805, over 985141.41 frames.], batch size: 32, aishell_tot_loss[loss=0.2148, simple_loss=0.2783, pruned_loss=0.07565, over 980123.94 frames.], datatang_tot_loss[loss=0.2229, simple_loss=0.2709, pruned_loss=0.08744, over 981579.35 frames.], batch size: 32, lr: 1.39e-03
+2022-06-18 13:48:01,811 INFO [train.py:874] (2/4) Epoch 5, batch 2200, datatang_loss[loss=0.2422, simple_loss=0.2773, pruned_loss=0.1035, over 4976.00 frames.], tot_loss[loss=0.218, simple_loss=0.274, pruned_loss=0.08102, over 985204.29 frames.], batch size: 55, aishell_tot_loss[loss=0.2151, simple_loss=0.2784, pruned_loss=0.07595, over 980530.19 frames.], datatang_tot_loss[loss=0.2224, simple_loss=0.2707, pruned_loss=0.08701, over 982206.72 frames.], batch size: 55, lr: 1.39e-03
+2022-06-18 13:48:32,482 INFO [train.py:874] (2/4) Epoch 5, batch 2250, datatang_loss[loss=0.2431, simple_loss=0.2899, pruned_loss=0.09818, over 4925.00 frames.], tot_loss[loss=0.2194, simple_loss=0.2743, pruned_loss=0.08224, over 985411.47 frames.], batch size: 98, aishell_tot_loss[loss=0.216, simple_loss=0.2786, pruned_loss=0.07666, over 980858.10 frames.], datatang_tot_loss[loss=0.2226, simple_loss=0.2709, pruned_loss=0.08718, over 982953.94 frames.], batch size: 98, lr: 1.38e-03
+2022-06-18 13:49:03,295 INFO [train.py:874] (2/4) Epoch 5, batch 2300, datatang_loss[loss=0.211, simple_loss=0.2564, pruned_loss=0.08284, over 4935.00 frames.], tot_loss[loss=0.2195, simple_loss=0.2742, pruned_loss=0.08238, over 985560.51 frames.], batch size: 71, aishell_tot_loss[loss=0.2162, simple_loss=0.2789, pruned_loss=0.07673, over 981211.78 frames.], datatang_tot_loss[loss=0.2223, simple_loss=0.2706, pruned_loss=0.08703, over 983515.65 frames.], batch size: 71, lr: 1.38e-03
+2022-06-18 13:49:34,054 INFO [train.py:874] (2/4) Epoch 5, batch 2350, aishell_loss[loss=0.2009, simple_loss=0.2714, pruned_loss=0.06521, over 4884.00 frames.], tot_loss[loss=0.2174, simple_loss=0.273, pruned_loss=0.08095, over 985527.50 frames.], batch size: 28, aishell_tot_loss[loss=0.2157, simple_loss=0.2788, pruned_loss=0.07632, over 981592.10 frames.], datatang_tot_loss[loss=0.2207, simple_loss=0.2693, pruned_loss=0.08604, over 983877.30 frames.], batch size: 28, lr: 1.38e-03
+2022-06-18 13:50:03,370 INFO [train.py:874] (2/4) Epoch 5, batch 2400, datatang_loss[loss=0.2431, simple_loss=0.2939, pruned_loss=0.09614, over 4918.00 frames.], tot_loss[loss=0.2173, simple_loss=0.2732, pruned_loss=0.08071, over 985330.14 frames.], batch size: 98, aishell_tot_loss[loss=0.2148, simple_loss=0.2785, pruned_loss=0.07554, over 981907.66 frames.], datatang_tot_loss[loss=0.2212, simple_loss=0.2696, pruned_loss=0.08642, over 983994.41 frames.], batch size: 98, lr: 1.38e-03
+2022-06-18 13:50:34,568 INFO [train.py:874] (2/4) Epoch 5, batch 2450, aishell_loss[loss=0.2118, simple_loss=0.2788, pruned_loss=0.07245, over 4858.00 frames.], tot_loss[loss=0.217, simple_loss=0.2726, pruned_loss=0.08071, over 985573.66 frames.], batch size: 36, aishell_tot_loss[loss=0.2135, simple_loss=0.2771, pruned_loss=0.07494, over 982255.73 frames.], datatang_tot_loss[loss=0.2218, simple_loss=0.2703, pruned_loss=0.0867, over 984425.51 frames.], batch size: 36, lr: 1.38e-03
+2022-06-18 13:51:05,475 INFO [train.py:874] (2/4) Epoch 5, batch 2500, datatang_loss[loss=0.2207, simple_loss=0.278, pruned_loss=0.08164, over 4957.00 frames.], tot_loss[loss=0.2175, simple_loss=0.273, pruned_loss=0.08097, over 985722.35 frames.], batch size: 91, aishell_tot_loss[loss=0.2136, simple_loss=0.2772, pruned_loss=0.07504, over 982861.45 frames.], datatang_tot_loss[loss=0.222, simple_loss=0.2704, pruned_loss=0.08682, over 984498.62 frames.], batch size: 91, lr: 1.38e-03
+2022-06-18 13:51:35,069 INFO [train.py:874] (2/4) Epoch 5, batch 2550, datatang_loss[loss=0.2097, simple_loss=0.2687, pruned_loss=0.07538, over 4892.00 frames.], tot_loss[loss=0.2174, simple_loss=0.2737, pruned_loss=0.08059, over 985820.86 frames.], batch size: 59, aishell_tot_loss[loss=0.2134, simple_loss=0.277, pruned_loss=0.07485, over 983443.84 frames.], datatang_tot_loss[loss=0.2224, simple_loss=0.2709, pruned_loss=0.08694, over 984541.87 frames.], batch size: 59, lr: 1.37e-03
+2022-06-18 13:52:06,136 INFO [train.py:874] (2/4) Epoch 5, batch 2600, datatang_loss[loss=0.2042, simple_loss=0.2459, pruned_loss=0.08129, over 4965.00 frames.], tot_loss[loss=0.2165, simple_loss=0.2727, pruned_loss=0.08013, over 985798.77 frames.], batch size: 55, aishell_tot_loss[loss=0.2119, simple_loss=0.2758, pruned_loss=0.07404, over 983643.79 frames.], datatang_tot_loss[loss=0.2228, simple_loss=0.2711, pruned_loss=0.08728, over 984752.67 frames.], batch size: 55, lr: 1.37e-03
+2022-06-18 13:52:35,471 INFO [train.py:874] (2/4) Epoch 5, batch 2650, datatang_loss[loss=0.2129, simple_loss=0.277, pruned_loss=0.07442, over 4949.00 frames.], tot_loss[loss=0.217, simple_loss=0.2735, pruned_loss=0.08026, over 986045.71 frames.], batch size: 86, aishell_tot_loss[loss=0.2123, simple_loss=0.2764, pruned_loss=0.07412, over 984063.88 frames.], datatang_tot_loss[loss=0.2228, simple_loss=0.2711, pruned_loss=0.08723, over 984961.16 frames.], batch size: 86, lr: 1.37e-03
+2022-06-18 13:53:06,282 INFO [train.py:874] (2/4) Epoch 5, batch 2700, datatang_loss[loss=0.1869, simple_loss=0.2437, pruned_loss=0.06502, over 4913.00 frames.], tot_loss[loss=0.2164, simple_loss=0.2733, pruned_loss=0.07976, over 986210.44 frames.], batch size: 75, aishell_tot_loss[loss=0.2116, simple_loss=0.276, pruned_loss=0.07358, over 984397.35 frames.], datatang_tot_loss[loss=0.2229, simple_loss=0.2713, pruned_loss=0.08724, over 985164.00 frames.], batch size: 75, lr: 1.37e-03
+2022-06-18 13:53:36,276 INFO [train.py:874] (2/4) Epoch 5, batch 2750, datatang_loss[loss=0.2181, simple_loss=0.267, pruned_loss=0.08462, over 4930.00 frames.], tot_loss[loss=0.217, simple_loss=0.274, pruned_loss=0.08003, over 986286.22 frames.], batch size: 79, aishell_tot_loss[loss=0.2113, simple_loss=0.2758, pruned_loss=0.07342, over 984786.94 frames.], datatang_tot_loss[loss=0.2235, simple_loss=0.272, pruned_loss=0.0875, over 985205.37 frames.], batch size: 79, lr: 1.37e-03
+2022-06-18 13:54:05,685 INFO [train.py:874] (2/4) Epoch 5, batch 2800, datatang_loss[loss=0.2238, simple_loss=0.2588, pruned_loss=0.09436, over 4972.00 frames.], tot_loss[loss=0.2182, simple_loss=0.2748, pruned_loss=0.08087, over 986164.64 frames.], batch size: 37, aishell_tot_loss[loss=0.2116, simple_loss=0.2759, pruned_loss=0.07362, over 984761.42 frames.], datatang_tot_loss[loss=0.2245, simple_loss=0.2728, pruned_loss=0.08814, over 985415.95 frames.], batch size: 37, lr: 1.37e-03
+2022-06-18 13:54:37,280 INFO [train.py:874] (2/4) Epoch 5, batch 2850, datatang_loss[loss=0.2431, simple_loss=0.2919, pruned_loss=0.09714, over 4929.00 frames.], tot_loss[loss=0.2163, simple_loss=0.2728, pruned_loss=0.07986, over 986020.80 frames.], batch size: 94, aishell_tot_loss[loss=0.2113, simple_loss=0.2759, pruned_loss=0.07341, over 984640.38 frames.], datatang_tot_loss[loss=0.2225, simple_loss=0.271, pruned_loss=0.08701, over 985620.65 frames.], batch size: 94, lr: 1.36e-03
+2022-06-18 13:55:07,710 INFO [train.py:874] (2/4) Epoch 5, batch 2900, datatang_loss[loss=0.2111, simple_loss=0.2671, pruned_loss=0.07761, over 4928.00 frames.], tot_loss[loss=0.2161, simple_loss=0.273, pruned_loss=0.07958, over 985806.16 frames.], batch size: 71, aishell_tot_loss[loss=0.2112, simple_loss=0.2757, pruned_loss=0.07333, over 984755.82 frames.], datatang_tot_loss[loss=0.2224, simple_loss=0.2711, pruned_loss=0.08688, over 985501.07 frames.], batch size: 71, lr: 1.36e-03
+2022-06-18 13:55:36,549 INFO [train.py:874] (2/4) Epoch 5, batch 2950, aishell_loss[loss=0.2218, simple_loss=0.2877, pruned_loss=0.07798, over 4933.00 frames.], tot_loss[loss=0.2154, simple_loss=0.2726, pruned_loss=0.07912, over 985914.70 frames.], batch size: 58, aishell_tot_loss[loss=0.2113, simple_loss=0.2761, pruned_loss=0.07323, over 984913.23 frames.], datatang_tot_loss[loss=0.2215, simple_loss=0.2703, pruned_loss=0.08635, over 985620.43 frames.], batch size: 58, lr: 1.36e-03
+2022-06-18 13:56:08,960 INFO [train.py:874] (2/4) Epoch 5, batch 3000, datatang_loss[loss=0.2036, simple_loss=0.2582, pruned_loss=0.07445, over 4921.00 frames.], tot_loss[loss=0.215, simple_loss=0.2725, pruned_loss=0.07872, over 985761.05 frames.], batch size: 81, aishell_tot_loss[loss=0.2109, simple_loss=0.2761, pruned_loss=0.07289, over 984889.59 frames.], datatang_tot_loss[loss=0.2211, simple_loss=0.2699, pruned_loss=0.08617, over 985637.47 frames.], batch size: 81, lr: 1.36e-03
+2022-06-18 13:56:08,961 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 13:56:26,005 INFO [train.py:914] (2/4) Epoch 5, validation: loss=0.1806, simple_loss=0.2585, pruned_loss=0.05141, over 1622729.00 frames.
+2022-06-18 13:56:55,549 INFO [train.py:874] (2/4) Epoch 5, batch 3050, aishell_loss[loss=0.1894, simple_loss=0.2406, pruned_loss=0.06913, over 4909.00 frames.], tot_loss[loss=0.2163, simple_loss=0.2735, pruned_loss=0.07951, over 985624.77 frames.], batch size: 25, aishell_tot_loss[loss=0.2117, simple_loss=0.2769, pruned_loss=0.07327, over 985010.67 frames.], datatang_tot_loss[loss=0.2215, simple_loss=0.27, pruned_loss=0.08646, over 985482.08 frames.], batch size: 25, lr: 1.36e-03
+2022-06-18 13:57:26,861 INFO [train.py:874] (2/4) Epoch 5, batch 3100, datatang_loss[loss=0.207, simple_loss=0.2484, pruned_loss=0.08284, over 4957.00 frames.], tot_loss[loss=0.2164, simple_loss=0.2734, pruned_loss=0.0797, over 985827.45 frames.], batch size: 67, aishell_tot_loss[loss=0.2119, simple_loss=0.277, pruned_loss=0.07345, over 985160.41 frames.], datatang_tot_loss[loss=0.2211, simple_loss=0.2699, pruned_loss=0.08615, over 985627.89 frames.], batch size: 67, lr: 1.36e-03
+2022-06-18 13:57:57,617 INFO [train.py:874] (2/4) Epoch 5, batch 3150, aishell_loss[loss=0.1961, simple_loss=0.2639, pruned_loss=0.06416, over 4967.00 frames.], tot_loss[loss=0.2161, simple_loss=0.2732, pruned_loss=0.0795, over 985935.16 frames.], batch size: 51, aishell_tot_loss[loss=0.2117, simple_loss=0.2767, pruned_loss=0.0733, over 985195.15 frames.], datatang_tot_loss[loss=0.2209, simple_loss=0.27, pruned_loss=0.08589, over 985790.66 frames.], batch size: 51, lr: 1.35e-03
+2022-06-18 13:58:27,914 INFO [train.py:874] (2/4) Epoch 5, batch 3200, aishell_loss[loss=0.1858, simple_loss=0.2609, pruned_loss=0.05537, over 4894.00 frames.], tot_loss[loss=0.2156, simple_loss=0.2725, pruned_loss=0.07928, over 986038.99 frames.], batch size: 34, aishell_tot_loss[loss=0.2112, simple_loss=0.2766, pruned_loss=0.07289, over 985060.03 frames.], datatang_tot_loss[loss=0.2206, simple_loss=0.2696, pruned_loss=0.08577, over 986137.52 frames.], batch size: 34, lr: 1.35e-03
+2022-06-18 13:58:58,334 INFO [train.py:874] (2/4) Epoch 5, batch 3250, datatang_loss[loss=0.2301, simple_loss=0.2721, pruned_loss=0.09411, over 4958.00 frames.], tot_loss[loss=0.215, simple_loss=0.2726, pruned_loss=0.07873, over 986128.96 frames.], batch size: 55, aishell_tot_loss[loss=0.2114, simple_loss=0.2771, pruned_loss=0.07288, over 985191.24 frames.], datatang_tot_loss[loss=0.2198, simple_loss=0.269, pruned_loss=0.08533, over 986209.95 frames.], batch size: 55, lr: 1.35e-03
+2022-06-18 13:59:26,745 INFO [train.py:874] (2/4) Epoch 5, batch 3300, aishell_loss[loss=0.2005, simple_loss=0.2752, pruned_loss=0.06288, over 4873.00 frames.], tot_loss[loss=0.2141, simple_loss=0.2725, pruned_loss=0.0779, over 986354.50 frames.], batch size: 42, aishell_tot_loss[loss=0.2097, simple_loss=0.2759, pruned_loss=0.07177, over 985437.75 frames.], datatang_tot_loss[loss=0.2208, simple_loss=0.2697, pruned_loss=0.08594, over 986334.43 frames.], batch size: 42, lr: 1.35e-03
+2022-06-18 13:59:58,743 INFO [train.py:874] (2/4) Epoch 5, batch 3350, datatang_loss[loss=0.2164, simple_loss=0.281, pruned_loss=0.07591, over 4930.00 frames.], tot_loss[loss=0.214, simple_loss=0.2724, pruned_loss=0.07783, over 986083.44 frames.], batch size: 94, aishell_tot_loss[loss=0.2101, simple_loss=0.2764, pruned_loss=0.07194, over 985176.81 frames.], datatang_tot_loss[loss=0.22, simple_loss=0.2692, pruned_loss=0.08541, over 986412.41 frames.], batch size: 94, lr: 1.35e-03
+2022-06-18 14:00:33,632 INFO [train.py:874] (2/4) Epoch 5, batch 3400, datatang_loss[loss=0.2242, simple_loss=0.2684, pruned_loss=0.08996, over 4910.00 frames.], tot_loss[loss=0.2145, simple_loss=0.273, pruned_loss=0.07797, over 986035.22 frames.], batch size: 64, aishell_tot_loss[loss=0.2105, simple_loss=0.2764, pruned_loss=0.07224, over 985467.65 frames.], datatang_tot_loss[loss=0.2202, simple_loss=0.2693, pruned_loss=0.08549, over 986169.35 frames.], batch size: 64, lr: 1.35e-03
+2022-06-18 14:01:04,098 INFO [train.py:874] (2/4) Epoch 5, batch 3450, aishell_loss[loss=0.1822, simple_loss=0.252, pruned_loss=0.05622, over 4904.00 frames.], tot_loss[loss=0.2143, simple_loss=0.2726, pruned_loss=0.07797, over 985831.84 frames.], batch size: 34, aishell_tot_loss[loss=0.2103, simple_loss=0.2762, pruned_loss=0.07214, over 985333.09 frames.], datatang_tot_loss[loss=0.2199, simple_loss=0.2693, pruned_loss=0.08521, over 986122.30 frames.], batch size: 34, lr: 1.34e-03
+2022-06-18 14:01:33,995 INFO [train.py:874] (2/4) Epoch 5, batch 3500, aishell_loss[loss=0.2516, simple_loss=0.3117, pruned_loss=0.09575, over 4937.00 frames.], tot_loss[loss=0.2143, simple_loss=0.2725, pruned_loss=0.07806, over 985714.79 frames.], batch size: 78, aishell_tot_loss[loss=0.2098, simple_loss=0.2758, pruned_loss=0.07193, over 985306.54 frames.], datatang_tot_loss[loss=0.2201, simple_loss=0.2696, pruned_loss=0.08531, over 986047.38 frames.], batch size: 78, lr: 1.34e-03
+2022-06-18 14:02:04,009 INFO [train.py:874] (2/4) Epoch 5, batch 3550, datatang_loss[loss=0.2246, simple_loss=0.2722, pruned_loss=0.08848, over 4968.00 frames.], tot_loss[loss=0.2137, simple_loss=0.2721, pruned_loss=0.07765, over 985642.67 frames.], batch size: 60, aishell_tot_loss[loss=0.2083, simple_loss=0.2747, pruned_loss=0.07097, over 985402.88 frames.], datatang_tot_loss[loss=0.221, simple_loss=0.2701, pruned_loss=0.0859, over 985880.47 frames.], batch size: 60, lr: 1.34e-03
+2022-06-18 14:02:33,813 INFO [train.py:874] (2/4) Epoch 5, batch 3600, aishell_loss[loss=0.1888, simple_loss=0.2638, pruned_loss=0.05695, over 4870.00 frames.], tot_loss[loss=0.2144, simple_loss=0.2726, pruned_loss=0.07805, over 985666.40 frames.], batch size: 37, aishell_tot_loss[loss=0.2096, simple_loss=0.2759, pruned_loss=0.07168, over 985524.39 frames.], datatang_tot_loss[loss=0.2201, simple_loss=0.2693, pruned_loss=0.0855, over 985771.29 frames.], batch size: 37, lr: 1.34e-03
+2022-06-18 14:03:03,182 INFO [train.py:874] (2/4) Epoch 5, batch 3650, datatang_loss[loss=0.2152, simple_loss=0.2615, pruned_loss=0.08441, over 4916.00 frames.], tot_loss[loss=0.2127, simple_loss=0.2707, pruned_loss=0.07736, over 985431.22 frames.], batch size: 75, aishell_tot_loss[loss=0.2087, simple_loss=0.2748, pruned_loss=0.07135, over 985240.65 frames.], datatang_tot_loss[loss=0.219, simple_loss=0.2685, pruned_loss=0.08481, over 985808.82 frames.], batch size: 75, lr: 1.34e-03
+2022-06-18 14:03:33,725 INFO [train.py:874] (2/4) Epoch 5, batch 3700, aishell_loss[loss=0.1893, simple_loss=0.2674, pruned_loss=0.05556, over 4944.00 frames.], tot_loss[loss=0.2116, simple_loss=0.2699, pruned_loss=0.07666, over 985370.31 frames.], batch size: 40, aishell_tot_loss[loss=0.2085, simple_loss=0.2747, pruned_loss=0.07119, over 984922.39 frames.], datatang_tot_loss[loss=0.2178, simple_loss=0.2675, pruned_loss=0.08407, over 986025.53 frames.], batch size: 40, lr: 1.34e-03
+2022-06-18 14:04:03,480 INFO [train.py:874] (2/4) Epoch 5, batch 3750, aishell_loss[loss=0.2114, simple_loss=0.2828, pruned_loss=0.07004, over 4968.00 frames.], tot_loss[loss=0.211, simple_loss=0.2694, pruned_loss=0.07628, over 985407.63 frames.], batch size: 44, aishell_tot_loss[loss=0.2077, simple_loss=0.2738, pruned_loss=0.07078, over 984850.13 frames.], datatang_tot_loss[loss=0.2177, simple_loss=0.2674, pruned_loss=0.08394, over 986124.27 frames.], batch size: 44, lr: 1.34e-03
+2022-06-18 14:04:33,292 INFO [train.py:874] (2/4) Epoch 5, batch 3800, aishell_loss[loss=0.1818, simple_loss=0.2543, pruned_loss=0.05463, over 4917.00 frames.], tot_loss[loss=0.2105, simple_loss=0.2692, pruned_loss=0.07592, over 985305.29 frames.], batch size: 52, aishell_tot_loss[loss=0.2066, simple_loss=0.2729, pruned_loss=0.0701, over 984696.11 frames.], datatang_tot_loss[loss=0.2182, simple_loss=0.2677, pruned_loss=0.08436, over 986205.69 frames.], batch size: 52, lr: 1.33e-03
+2022-06-18 14:05:02,225 INFO [train.py:874] (2/4) Epoch 5, batch 3850, datatang_loss[loss=0.22, simple_loss=0.276, pruned_loss=0.08196, over 4936.00 frames.], tot_loss[loss=0.2091, simple_loss=0.2681, pruned_loss=0.07506, over 985407.35 frames.], batch size: 88, aishell_tot_loss[loss=0.2057, simple_loss=0.2723, pruned_loss=0.06956, over 984951.20 frames.], datatang_tot_loss[loss=0.2175, simple_loss=0.2668, pruned_loss=0.08409, over 986066.49 frames.], batch size: 88, lr: 1.33e-03
+2022-06-18 14:05:32,288 INFO [train.py:874] (2/4) Epoch 5, batch 3900, datatang_loss[loss=0.2298, simple_loss=0.2805, pruned_loss=0.08951, over 4941.00 frames.], tot_loss[loss=0.21, simple_loss=0.269, pruned_loss=0.07543, over 985648.69 frames.], batch size: 88, aishell_tot_loss[loss=0.2063, simple_loss=0.2729, pruned_loss=0.06986, over 985023.56 frames.], datatang_tot_loss[loss=0.2172, simple_loss=0.2667, pruned_loss=0.0838, over 986231.12 frames.], batch size: 88, lr: 1.33e-03
+2022-06-18 14:06:00,053 INFO [train.py:874] (2/4) Epoch 5, batch 3950, datatang_loss[loss=0.1788, simple_loss=0.2344, pruned_loss=0.06166, over 4975.00 frames.], tot_loss[loss=0.212, simple_loss=0.2709, pruned_loss=0.07658, over 985740.53 frames.], batch size: 45, aishell_tot_loss[loss=0.2071, simple_loss=0.2738, pruned_loss=0.07025, over 985071.32 frames.], datatang_tot_loss[loss=0.2183, simple_loss=0.2675, pruned_loss=0.0845, over 986299.72 frames.], batch size: 45, lr: 1.33e-03
+2022-06-18 14:06:30,656 INFO [train.py:874] (2/4) Epoch 5, batch 4000, aishell_loss[loss=0.2177, simple_loss=0.2845, pruned_loss=0.0754, over 4943.00 frames.], tot_loss[loss=0.2131, simple_loss=0.2719, pruned_loss=0.0772, over 985817.37 frames.], batch size: 49, aishell_tot_loss[loss=0.2079, simple_loss=0.2746, pruned_loss=0.07062, over 985290.37 frames.], datatang_tot_loss[loss=0.2185, simple_loss=0.2679, pruned_loss=0.08456, over 986160.74 frames.], batch size: 49, lr: 1.33e-03
+2022-06-18 14:06:30,657 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 14:06:46,837 INFO [train.py:914] (2/4) Epoch 5, validation: loss=0.1787, simple_loss=0.2582, pruned_loss=0.04961, over 1622729.00 frames.
+2022-06-18 14:07:16,322 INFO [train.py:874] (2/4) Epoch 5, batch 4050, aishell_loss[loss=0.2001, simple_loss=0.2706, pruned_loss=0.06477, over 4979.00 frames.], tot_loss[loss=0.215, simple_loss=0.2724, pruned_loss=0.07875, over 985569.74 frames.], batch size: 48, aishell_tot_loss[loss=0.2092, simple_loss=0.2752, pruned_loss=0.07164, over 985225.98 frames.], datatang_tot_loss[loss=0.2187, simple_loss=0.2681, pruned_loss=0.08467, over 985966.97 frames.], batch size: 48, lr: 1.33e-03
+2022-06-18 14:07:45,344 INFO [train.py:874] (2/4) Epoch 5, batch 4100, aishell_loss[loss=0.2466, simple_loss=0.3119, pruned_loss=0.09069, over 4921.00 frames.], tot_loss[loss=0.2152, simple_loss=0.2723, pruned_loss=0.07902, over 985800.16 frames.], batch size: 68, aishell_tot_loss[loss=0.2101, simple_loss=0.276, pruned_loss=0.07206, over 985259.98 frames.], datatang_tot_loss[loss=0.2182, simple_loss=0.2676, pruned_loss=0.08445, over 986168.55 frames.], batch size: 68, lr: 1.32e-03
+2022-06-18 14:08:14,088 INFO [train.py:874] (2/4) Epoch 5, batch 4150, aishell_loss[loss=0.199, simple_loss=0.2657, pruned_loss=0.06613, over 4925.00 frames.], tot_loss[loss=0.2159, simple_loss=0.2732, pruned_loss=0.07927, over 985683.25 frames.], batch size: 41, aishell_tot_loss[loss=0.2107, simple_loss=0.2767, pruned_loss=0.07236, over 984922.08 frames.], datatang_tot_loss[loss=0.2186, simple_loss=0.268, pruned_loss=0.08455, over 986406.99 frames.], batch size: 41, lr: 1.32e-03
+2022-06-18 14:09:32,219 INFO [train.py:874] (2/4) Epoch 6, batch 50, datatang_loss[loss=0.1726, simple_loss=0.2401, pruned_loss=0.05259, over 4873.00 frames.], tot_loss[loss=0.2058, simple_loss=0.2657, pruned_loss=0.07298, over 218607.45 frames.], batch size: 30, aishell_tot_loss[loss=0.2106, simple_loss=0.2768, pruned_loss=0.0722, over 120626.96 frames.], datatang_tot_loss[loss=0.2006, simple_loss=0.2535, pruned_loss=0.07385, over 111650.21 frames.], batch size: 30, lr: 1.27e-03
+2022-06-18 14:10:03,277 INFO [train.py:874] (2/4) Epoch 6, batch 100, datatang_loss[loss=0.18, simple_loss=0.2447, pruned_loss=0.05766, over 4900.00 frames.], tot_loss[loss=0.204, simple_loss=0.2641, pruned_loss=0.072, over 388678.58 frames.], batch size: 64, aishell_tot_loss[loss=0.2086, simple_loss=0.2752, pruned_loss=0.07105, over 214811.68 frames.], datatang_tot_loss[loss=0.1997, simple_loss=0.2535, pruned_loss=0.07292, over 222318.25 frames.], batch size: 64, lr: 1.27e-03
+2022-06-18 14:10:32,558 INFO [train.py:874] (2/4) Epoch 6, batch 150, datatang_loss[loss=0.1921, simple_loss=0.2541, pruned_loss=0.06504, over 4928.00 frames.], tot_loss[loss=0.2017, simple_loss=0.2616, pruned_loss=0.07089, over 520679.18 frames.], batch size: 77, aishell_tot_loss[loss=0.2063, simple_loss=0.2732, pruned_loss=0.06968, over 288007.56 frames.], datatang_tot_loss[loss=0.1982, simple_loss=0.2523, pruned_loss=0.07201, over 328855.23 frames.], batch size: 77, lr: 1.27e-03
+2022-06-18 14:11:03,732 INFO [train.py:874] (2/4) Epoch 6, batch 200, aishell_loss[loss=0.1996, simple_loss=0.2682, pruned_loss=0.06545, over 4977.00 frames.], tot_loss[loss=0.201, simple_loss=0.2612, pruned_loss=0.07045, over 623870.12 frames.], batch size: 39, aishell_tot_loss[loss=0.2048, simple_loss=0.2716, pruned_loss=0.069, over 364022.34 frames.], datatang_tot_loss[loss=0.1983, simple_loss=0.2529, pruned_loss=0.07189, over 412070.84 frames.], batch size: 39, lr: 1.26e-03
+2022-06-18 14:11:32,996 INFO [train.py:874] (2/4) Epoch 6, batch 250, datatang_loss[loss=0.2138, simple_loss=0.2686, pruned_loss=0.07951, over 4923.00 frames.], tot_loss[loss=0.2047, simple_loss=0.265, pruned_loss=0.07219, over 703646.56 frames.], batch size: 83, aishell_tot_loss[loss=0.2074, simple_loss=0.2741, pruned_loss=0.07041, over 455836.96 frames.], datatang_tot_loss[loss=0.2009, simple_loss=0.2548, pruned_loss=0.07351, over 461420.97 frames.], batch size: 83, lr: 1.26e-03
+2022-06-18 14:12:03,474 INFO [train.py:874] (2/4) Epoch 6, batch 300, datatang_loss[loss=0.1559, simple_loss=0.2159, pruned_loss=0.04799, over 4855.00 frames.], tot_loss[loss=0.2048, simple_loss=0.2647, pruned_loss=0.07244, over 766339.15 frames.], batch size: 30, aishell_tot_loss[loss=0.2073, simple_loss=0.2738, pruned_loss=0.07047, over 513584.89 frames.], datatang_tot_loss[loss=0.2016, simple_loss=0.2552, pruned_loss=0.07397, over 527916.80 frames.], batch size: 30, lr: 1.26e-03
+2022-06-18 14:12:34,548 INFO [train.py:874] (2/4) Epoch 6, batch 350, datatang_loss[loss=0.1713, simple_loss=0.2267, pruned_loss=0.05801, over 4863.00 frames.], tot_loss[loss=0.2029, simple_loss=0.2628, pruned_loss=0.07149, over 814656.63 frames.], batch size: 39, aishell_tot_loss[loss=0.2056, simple_loss=0.2724, pruned_loss=0.06938, over 566749.81 frames.], datatang_tot_loss[loss=0.2006, simple_loss=0.2539, pruned_loss=0.07363, over 583862.68 frames.], batch size: 39, lr: 1.26e-03
+2022-06-18 14:13:03,867 INFO [train.py:874] (2/4) Epoch 6, batch 400, datatang_loss[loss=0.2229, simple_loss=0.2869, pruned_loss=0.07948, over 4930.00 frames.], tot_loss[loss=0.205, simple_loss=0.2651, pruned_loss=0.07248, over 853045.50 frames.], batch size: 94, aishell_tot_loss[loss=0.2048, simple_loss=0.2718, pruned_loss=0.06889, over 621914.46 frames.], datatang_tot_loss[loss=0.2043, simple_loss=0.2574, pruned_loss=0.07561, over 626013.42 frames.], batch size: 94, lr: 1.26e-03
+2022-06-18 14:13:33,619 INFO [train.py:874] (2/4) Epoch 6, batch 450, aishell_loss[loss=0.2176, simple_loss=0.2834, pruned_loss=0.07589, over 4914.00 frames.], tot_loss[loss=0.2075, simple_loss=0.2674, pruned_loss=0.07383, over 882391.01 frames.], batch size: 68, aishell_tot_loss[loss=0.2059, simple_loss=0.2729, pruned_loss=0.06942, over 667842.05 frames.], datatang_tot_loss[loss=0.2069, simple_loss=0.2595, pruned_loss=0.07719, over 665228.25 frames.], batch size: 68, lr: 1.26e-03
+2022-06-18 14:14:05,038 INFO [train.py:874] (2/4) Epoch 6, batch 500, datatang_loss[loss=0.2122, simple_loss=0.2543, pruned_loss=0.08501, over 4882.00 frames.], tot_loss[loss=0.2077, simple_loss=0.2677, pruned_loss=0.07387, over 905411.27 frames.], batch size: 47, aishell_tot_loss[loss=0.2053, simple_loss=0.2725, pruned_loss=0.06907, over 709350.13 frames.], datatang_tot_loss[loss=0.2081, simple_loss=0.2604, pruned_loss=0.07793, over 698924.18 frames.], batch size: 47, lr: 1.26e-03
+2022-06-18 14:14:34,030 INFO [train.py:874] (2/4) Epoch 6, batch 550, aishell_loss[loss=0.206, simple_loss=0.279, pruned_loss=0.06645, over 4920.00 frames.], tot_loss[loss=0.2077, simple_loss=0.2687, pruned_loss=0.07334, over 922921.95 frames.], batch size: 52, aishell_tot_loss[loss=0.2047, simple_loss=0.2725, pruned_loss=0.06845, over 755831.30 frames.], datatang_tot_loss[loss=0.2094, simple_loss=0.2613, pruned_loss=0.07872, over 717025.18 frames.], batch size: 52, lr: 1.25e-03
+2022-06-18 14:15:04,015 INFO [train.py:874] (2/4) Epoch 6, batch 600, datatang_loss[loss=0.1917, simple_loss=0.2512, pruned_loss=0.06612, over 4928.00 frames.], tot_loss[loss=0.2083, simple_loss=0.269, pruned_loss=0.07382, over 936804.53 frames.], batch size: 71, aishell_tot_loss[loss=0.2039, simple_loss=0.2719, pruned_loss=0.06795, over 781779.39 frames.], datatang_tot_loss[loss=0.2113, simple_loss=0.263, pruned_loss=0.07978, over 749914.96 frames.], batch size: 71, lr: 1.25e-03
+2022-06-18 14:15:34,708 INFO [train.py:874] (2/4) Epoch 6, batch 650, datatang_loss[loss=0.1993, simple_loss=0.2613, pruned_loss=0.06864, over 4919.00 frames.], tot_loss[loss=0.2074, simple_loss=0.2681, pruned_loss=0.07335, over 947624.17 frames.], batch size: 83, aishell_tot_loss[loss=0.203, simple_loss=0.2709, pruned_loss=0.0675, over 804886.62 frames.], datatang_tot_loss[loss=0.2113, simple_loss=0.2634, pruned_loss=0.07957, over 778708.90 frames.], batch size: 83, lr: 1.25e-03
+2022-06-18 14:16:03,558 INFO [train.py:874] (2/4) Epoch 6, batch 700, aishell_loss[loss=0.1954, simple_loss=0.2693, pruned_loss=0.06077, over 4914.00 frames.], tot_loss[loss=0.2077, simple_loss=0.2682, pruned_loss=0.07362, over 956422.43 frames.], batch size: 46, aishell_tot_loss[loss=0.2026, simple_loss=0.2708, pruned_loss=0.06717, over 826225.53 frames.], datatang_tot_loss[loss=0.2122, simple_loss=0.264, pruned_loss=0.0802, over 803432.52 frames.], batch size: 46, lr: 1.25e-03
+2022-06-18 14:16:34,009 INFO [train.py:874] (2/4) Epoch 6, batch 750, aishell_loss[loss=0.2075, simple_loss=0.2768, pruned_loss=0.06914, over 4913.00 frames.], tot_loss[loss=0.2094, simple_loss=0.2694, pruned_loss=0.07473, over 963123.69 frames.], batch size: 41, aishell_tot_loss[loss=0.2027, simple_loss=0.2709, pruned_loss=0.06722, over 842405.96 frames.], datatang_tot_loss[loss=0.2139, simple_loss=0.2656, pruned_loss=0.08115, over 828069.21 frames.], batch size: 41, lr: 1.25e-03
+2022-06-18 14:17:06,169 INFO [train.py:874] (2/4) Epoch 6, batch 800, datatang_loss[loss=0.2164, simple_loss=0.2723, pruned_loss=0.08022, over 4895.00 frames.], tot_loss[loss=0.2092, simple_loss=0.2692, pruned_loss=0.07466, over 968275.08 frames.], batch size: 47, aishell_tot_loss[loss=0.2024, simple_loss=0.2709, pruned_loss=0.06691, over 856722.87 frames.], datatang_tot_loss[loss=0.2141, simple_loss=0.2657, pruned_loss=0.08119, over 849545.08 frames.], batch size: 47, lr: 1.25e-03
+2022-06-18 14:17:35,731 INFO [train.py:874] (2/4) Epoch 6, batch 850, datatang_loss[loss=0.1851, simple_loss=0.2452, pruned_loss=0.06246, over 4940.00 frames.], tot_loss[loss=0.2098, simple_loss=0.2692, pruned_loss=0.07521, over 972056.14 frames.], batch size: 69, aishell_tot_loss[loss=0.2034, simple_loss=0.2712, pruned_loss=0.06776, over 870647.13 frames.], datatang_tot_loss[loss=0.2139, simple_loss=0.2658, pruned_loss=0.08105, over 866833.22 frames.], batch size: 69, lr: 1.25e-03
+2022-06-18 14:18:05,650 INFO [train.py:874] (2/4) Epoch 6, batch 900, aishell_loss[loss=0.2006, simple_loss=0.2772, pruned_loss=0.06201, over 4951.00 frames.], tot_loss[loss=0.2087, simple_loss=0.269, pruned_loss=0.07418, over 975016.63 frames.], batch size: 64, aishell_tot_loss[loss=0.2029, simple_loss=0.2711, pruned_loss=0.06742, over 885460.69 frames.], datatang_tot_loss[loss=0.2135, simple_loss=0.2659, pruned_loss=0.08057, over 879431.32 frames.], batch size: 64, lr: 1.25e-03
+2022-06-18 14:18:36,686 INFO [train.py:874] (2/4) Epoch 6, batch 950, aishell_loss[loss=0.1878, simple_loss=0.2601, pruned_loss=0.05775, over 4860.00 frames.], tot_loss[loss=0.209, simple_loss=0.2691, pruned_loss=0.07442, over 977327.04 frames.], batch size: 35, aishell_tot_loss[loss=0.2033, simple_loss=0.2714, pruned_loss=0.06762, over 897601.29 frames.], datatang_tot_loss[loss=0.2136, simple_loss=0.2657, pruned_loss=0.08075, over 891525.69 frames.], batch size: 35, lr: 1.24e-03
+2022-06-18 14:19:06,991 INFO [train.py:874] (2/4) Epoch 6, batch 1000, datatang_loss[loss=0.2112, simple_loss=0.2686, pruned_loss=0.07686, over 4923.00 frames.], tot_loss[loss=0.2085, simple_loss=0.2691, pruned_loss=0.07394, over 979387.83 frames.], batch size: 57, aishell_tot_loss[loss=0.2028, simple_loss=0.2713, pruned_loss=0.06713, over 908221.43 frames.], datatang_tot_loss[loss=0.2137, simple_loss=0.2659, pruned_loss=0.08073, over 902589.22 frames.], batch size: 57, lr: 1.24e-03
+2022-06-18 14:19:06,992 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 14:19:23,421 INFO [train.py:914] (2/4) Epoch 6, validation: loss=0.175, simple_loss=0.2564, pruned_loss=0.04677, over 1622729.00 frames.
+2022-06-18 14:19:52,889 INFO [train.py:874] (2/4) Epoch 6, batch 1050, datatang_loss[loss=0.2129, simple_loss=0.2666, pruned_loss=0.0796, over 4946.00 frames.], tot_loss[loss=0.2082, simple_loss=0.2691, pruned_loss=0.07363, over 980733.86 frames.], batch size: 34, aishell_tot_loss[loss=0.2026, simple_loss=0.2711, pruned_loss=0.06705, over 917888.43 frames.], datatang_tot_loss[loss=0.2137, simple_loss=0.2662, pruned_loss=0.08059, over 911735.56 frames.], batch size: 34, lr: 1.24e-03
+2022-06-18 14:20:23,048 INFO [train.py:874] (2/4) Epoch 6, batch 1100, aishell_loss[loss=0.1886, simple_loss=0.2588, pruned_loss=0.05921, over 4831.00 frames.], tot_loss[loss=0.2096, simple_loss=0.2692, pruned_loss=0.07505, over 981834.14 frames.], batch size: 29, aishell_tot_loss[loss=0.2041, simple_loss=0.2717, pruned_loss=0.06823, over 924226.06 frames.], datatang_tot_loss[loss=0.2135, simple_loss=0.2658, pruned_loss=0.08061, over 922188.21 frames.], batch size: 29, lr: 1.24e-03
+2022-06-18 14:20:54,771 INFO [train.py:874] (2/4) Epoch 6, batch 1150, aishell_loss[loss=0.1644, simple_loss=0.2256, pruned_loss=0.05161, over 4934.00 frames.], tot_loss[loss=0.2111, simple_loss=0.2701, pruned_loss=0.07602, over 982778.40 frames.], batch size: 25, aishell_tot_loss[loss=0.2052, simple_loss=0.2724, pruned_loss=0.06905, over 930321.53 frames.], datatang_tot_loss[loss=0.2141, simple_loss=0.2665, pruned_loss=0.08081, over 930920.84 frames.], batch size: 25, lr: 1.24e-03
+2022-06-18 14:21:24,814 INFO [train.py:874] (2/4) Epoch 6, batch 1200, datatang_loss[loss=0.2079, simple_loss=0.2596, pruned_loss=0.07809, over 4969.00 frames.], tot_loss[loss=0.2111, simple_loss=0.2706, pruned_loss=0.07578, over 983481.03 frames.], batch size: 55, aishell_tot_loss[loss=0.2052, simple_loss=0.2728, pruned_loss=0.06882, over 936474.08 frames.], datatang_tot_loss[loss=0.2144, simple_loss=0.2668, pruned_loss=0.08094, over 937787.08 frames.], batch size: 55, lr: 1.24e-03
+2022-06-18 14:21:54,091 INFO [train.py:874] (2/4) Epoch 6, batch 1250, datatang_loss[loss=0.2086, simple_loss=0.2718, pruned_loss=0.07271, over 4969.00 frames.], tot_loss[loss=0.2106, simple_loss=0.2701, pruned_loss=0.07553, over 983794.74 frames.], batch size: 37, aishell_tot_loss[loss=0.2054, simple_loss=0.2729, pruned_loss=0.06894, over 941967.55 frames.], datatang_tot_loss[loss=0.2139, simple_loss=0.2663, pruned_loss=0.08069, over 943529.99 frames.], batch size: 37, lr: 1.24e-03
+2022-06-18 14:22:25,463 INFO [train.py:874] (2/4) Epoch 6, batch 1300, aishell_loss[loss=0.2077, simple_loss=0.2847, pruned_loss=0.06528, over 4904.00 frames.], tot_loss[loss=0.2081, simple_loss=0.2685, pruned_loss=0.07381, over 984194.61 frames.], batch size: 41, aishell_tot_loss[loss=0.2049, simple_loss=0.2731, pruned_loss=0.06837, over 946863.98 frames.], datatang_tot_loss[loss=0.2118, simple_loss=0.2646, pruned_loss=0.07947, over 948733.99 frames.], batch size: 41, lr: 1.23e-03
+2022-06-18 14:22:55,165 INFO [train.py:874] (2/4) Epoch 6, batch 1350, aishell_loss[loss=0.1932, simple_loss=0.2761, pruned_loss=0.05519, over 4976.00 frames.], tot_loss[loss=0.2083, simple_loss=0.2684, pruned_loss=0.07412, over 984830.49 frames.], batch size: 39, aishell_tot_loss[loss=0.2055, simple_loss=0.2737, pruned_loss=0.06867, over 950820.02 frames.], datatang_tot_loss[loss=0.2112, simple_loss=0.2639, pruned_loss=0.07927, over 953969.63 frames.], batch size: 39, lr: 1.23e-03
+2022-06-18 14:23:24,945 INFO [train.py:874] (2/4) Epoch 6, batch 1400, aishell_loss[loss=0.1977, simple_loss=0.2784, pruned_loss=0.05851, over 4910.00 frames.], tot_loss[loss=0.2072, simple_loss=0.2683, pruned_loss=0.07306, over 984881.42 frames.], batch size: 52, aishell_tot_loss[loss=0.2051, simple_loss=0.2737, pruned_loss=0.06828, over 955774.29 frames.], datatang_tot_loss[loss=0.2105, simple_loss=0.2633, pruned_loss=0.07884, over 956781.86 frames.], batch size: 52, lr: 1.23e-03
+2022-06-18 14:23:56,891 INFO [train.py:874] (2/4) Epoch 6, batch 1450, aishell_loss[loss=0.2037, simple_loss=0.2747, pruned_loss=0.06637, over 4867.00 frames.], tot_loss[loss=0.2069, simple_loss=0.2683, pruned_loss=0.07277, over 984998.51 frames.], batch size: 35, aishell_tot_loss[loss=0.2047, simple_loss=0.2733, pruned_loss=0.06808, over 959496.34 frames.], datatang_tot_loss[loss=0.2106, simple_loss=0.2636, pruned_loss=0.07878, over 959944.61 frames.], batch size: 35, lr: 1.23e-03
+2022-06-18 14:24:26,584 INFO [train.py:874] (2/4) Epoch 6, batch 1500, aishell_loss[loss=0.1786, simple_loss=0.2575, pruned_loss=0.04984, over 4935.00 frames.], tot_loss[loss=0.2083, simple_loss=0.2696, pruned_loss=0.07351, over 984886.03 frames.], batch size: 49, aishell_tot_loss[loss=0.2041, simple_loss=0.273, pruned_loss=0.06762, over 963046.73 frames.], datatang_tot_loss[loss=0.2127, simple_loss=0.265, pruned_loss=0.08017, over 962227.08 frames.], batch size: 49, lr: 1.23e-03
+2022-06-18 14:24:56,221 INFO [train.py:874] (2/4) Epoch 6, batch 1550, aishell_loss[loss=0.2149, simple_loss=0.2704, pruned_loss=0.07967, over 4914.00 frames.], tot_loss[loss=0.2081, simple_loss=0.2694, pruned_loss=0.07341, over 984831.67 frames.], batch size: 33, aishell_tot_loss[loss=0.204, simple_loss=0.2728, pruned_loss=0.06758, over 965862.51 frames.], datatang_tot_loss[loss=0.2128, simple_loss=0.2649, pruned_loss=0.08031, over 964559.08 frames.], batch size: 33, lr: 1.23e-03
+2022-06-18 14:25:27,121 INFO [train.py:874] (2/4) Epoch 6, batch 1600, aishell_loss[loss=0.1983, simple_loss=0.2705, pruned_loss=0.06304, over 4923.00 frames.], tot_loss[loss=0.2086, simple_loss=0.27, pruned_loss=0.07359, over 985260.72 frames.], batch size: 32, aishell_tot_loss[loss=0.2048, simple_loss=0.2734, pruned_loss=0.06804, over 968908.49 frames.], datatang_tot_loss[loss=0.2127, simple_loss=0.2648, pruned_loss=0.08027, over 966532.01 frames.], batch size: 32, lr: 1.23e-03
+2022-06-18 14:25:56,145 INFO [train.py:874] (2/4) Epoch 6, batch 1650, datatang_loss[loss=0.2104, simple_loss=0.2651, pruned_loss=0.07788, over 4913.00 frames.], tot_loss[loss=0.2082, simple_loss=0.2693, pruned_loss=0.07361, over 984659.30 frames.], batch size: 37, aishell_tot_loss[loss=0.2037, simple_loss=0.2725, pruned_loss=0.06751, over 969898.49 frames.], datatang_tot_loss[loss=0.2131, simple_loss=0.2654, pruned_loss=0.08043, over 969062.86 frames.], batch size: 37, lr: 1.23e-03
+2022-06-18 14:26:28,015 INFO [train.py:874] (2/4) Epoch 6, batch 1700, datatang_loss[loss=0.1864, simple_loss=0.2438, pruned_loss=0.0645, over 4957.00 frames.], tot_loss[loss=0.2069, simple_loss=0.2676, pruned_loss=0.07312, over 984162.81 frames.], batch size: 55, aishell_tot_loss[loss=0.2024, simple_loss=0.2709, pruned_loss=0.06696, over 970925.19 frames.], datatang_tot_loss[loss=0.213, simple_loss=0.2653, pruned_loss=0.08033, over 971095.78 frames.], batch size: 55, lr: 1.22e-03
+2022-06-18 14:26:57,264 INFO [train.py:874] (2/4) Epoch 6, batch 1750, aishell_loss[loss=0.2068, simple_loss=0.2793, pruned_loss=0.06712, over 4916.00 frames.], tot_loss[loss=0.2072, simple_loss=0.2679, pruned_loss=0.07327, over 983957.56 frames.], batch size: 41, aishell_tot_loss[loss=0.2029, simple_loss=0.2715, pruned_loss=0.06713, over 972268.07 frames.], datatang_tot_loss[loss=0.2126, simple_loss=0.2649, pruned_loss=0.08014, over 972629.32 frames.], batch size: 41, lr: 1.22e-03
+2022-06-18 14:27:27,451 INFO [train.py:874] (2/4) Epoch 6, batch 1800, aishell_loss[loss=0.1826, simple_loss=0.2602, pruned_loss=0.05246, over 4857.00 frames.], tot_loss[loss=0.2066, simple_loss=0.2673, pruned_loss=0.07296, over 984353.77 frames.], batch size: 38, aishell_tot_loss[loss=0.2024, simple_loss=0.2709, pruned_loss=0.06689, over 973828.34 frames.], datatang_tot_loss[loss=0.2124, simple_loss=0.2649, pruned_loss=0.07992, over 974204.06 frames.], batch size: 38, lr: 1.22e-03
+2022-06-18 14:27:59,152 INFO [train.py:874] (2/4) Epoch 6, batch 1850, aishell_loss[loss=0.2009, simple_loss=0.2727, pruned_loss=0.06448, over 4961.00 frames.], tot_loss[loss=0.207, simple_loss=0.2678, pruned_loss=0.07309, over 984517.06 frames.], batch size: 61, aishell_tot_loss[loss=0.2023, simple_loss=0.2711, pruned_loss=0.06675, over 975129.97 frames.], datatang_tot_loss[loss=0.2128, simple_loss=0.265, pruned_loss=0.08026, over 975495.12 frames.], batch size: 61, lr: 1.22e-03
+2022-06-18 14:28:29,045 INFO [train.py:874] (2/4) Epoch 6, batch 1900, aishell_loss[loss=0.2363, simple_loss=0.2894, pruned_loss=0.09162, over 4976.00 frames.], tot_loss[loss=0.2058, simple_loss=0.2674, pruned_loss=0.07206, over 985082.06 frames.], batch size: 51, aishell_tot_loss[loss=0.2017, simple_loss=0.2707, pruned_loss=0.06636, over 976829.87 frames.], datatang_tot_loss[loss=0.2123, simple_loss=0.2647, pruned_loss=0.07994, over 976562.70 frames.], batch size: 51, lr: 1.22e-03
+2022-06-18 14:28:59,012 INFO [train.py:874] (2/4) Epoch 6, batch 1950, aishell_loss[loss=0.222, simple_loss=0.277, pruned_loss=0.08351, over 4903.00 frames.], tot_loss[loss=0.2067, simple_loss=0.2678, pruned_loss=0.07284, over 984740.34 frames.], batch size: 34, aishell_tot_loss[loss=0.2019, simple_loss=0.2706, pruned_loss=0.06656, over 977442.93 frames.], datatang_tot_loss[loss=0.2127, simple_loss=0.2651, pruned_loss=0.08012, over 977567.49 frames.], batch size: 34, lr: 1.22e-03
+2022-06-18 14:29:30,643 INFO [train.py:874] (2/4) Epoch 6, batch 2000, datatang_loss[loss=0.2109, simple_loss=0.2699, pruned_loss=0.07595, over 4947.00 frames.], tot_loss[loss=0.2059, simple_loss=0.2675, pruned_loss=0.07215, over 985096.85 frames.], batch size: 88, aishell_tot_loss[loss=0.2016, simple_loss=0.2706, pruned_loss=0.06632, over 978523.70 frames.], datatang_tot_loss[loss=0.2119, simple_loss=0.2647, pruned_loss=0.07954, over 978566.33 frames.], batch size: 88, lr: 1.22e-03
+2022-06-18 14:29:30,644 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 14:29:46,436 INFO [train.py:914] (2/4) Epoch 6, validation: loss=0.1751, simple_loss=0.2557, pruned_loss=0.04723, over 1622729.00 frames.
+2022-06-18 14:30:18,806 INFO [train.py:874] (2/4) Epoch 6, batch 2050, aishell_loss[loss=0.202, simple_loss=0.2681, pruned_loss=0.06799, over 4978.00 frames.], tot_loss[loss=0.2074, simple_loss=0.2688, pruned_loss=0.07301, over 985410.82 frames.], batch size: 31, aishell_tot_loss[loss=0.2017, simple_loss=0.2709, pruned_loss=0.06625, over 979236.88 frames.], datatang_tot_loss[loss=0.2131, simple_loss=0.2659, pruned_loss=0.08017, over 979722.51 frames.], batch size: 31, lr: 1.22e-03
+2022-06-18 14:30:50,048 INFO [train.py:874] (2/4) Epoch 6, batch 2100, datatang_loss[loss=0.2175, simple_loss=0.2738, pruned_loss=0.08063, over 4939.00 frames.], tot_loss[loss=0.2075, simple_loss=0.2688, pruned_loss=0.07312, over 985364.04 frames.], batch size: 50, aishell_tot_loss[loss=0.202, simple_loss=0.2714, pruned_loss=0.06634, over 979700.97 frames.], datatang_tot_loss[loss=0.2127, simple_loss=0.2656, pruned_loss=0.07992, over 980583.68 frames.], batch size: 50, lr: 1.21e-03
+2022-06-18 14:31:18,748 INFO [train.py:874] (2/4) Epoch 6, batch 2150, datatang_loss[loss=0.2001, simple_loss=0.25, pruned_loss=0.07505, over 4923.00 frames.], tot_loss[loss=0.2071, simple_loss=0.2686, pruned_loss=0.07281, over 984912.58 frames.], batch size: 73, aishell_tot_loss[loss=0.202, simple_loss=0.2713, pruned_loss=0.06632, over 980273.09 frames.], datatang_tot_loss[loss=0.2126, simple_loss=0.2654, pruned_loss=0.07997, over 980770.72 frames.], batch size: 73, lr: 1.21e-03
+2022-06-18 14:31:50,740 INFO [train.py:874] (2/4) Epoch 6, batch 2200, aishell_loss[loss=0.2123, simple_loss=0.2715, pruned_loss=0.07661, over 4854.00 frames.], tot_loss[loss=0.2069, simple_loss=0.2685, pruned_loss=0.07272, over 985110.39 frames.], batch size: 28, aishell_tot_loss[loss=0.2019, simple_loss=0.2712, pruned_loss=0.06628, over 981211.50 frames.], datatang_tot_loss[loss=0.213, simple_loss=0.265, pruned_loss=0.08048, over 981077.42 frames.], batch size: 28, lr: 1.21e-03
+2022-06-18 14:32:20,685 INFO [train.py:874] (2/4) Epoch 6, batch 2250, datatang_loss[loss=0.2082, simple_loss=0.265, pruned_loss=0.07571, over 4921.00 frames.], tot_loss[loss=0.2066, simple_loss=0.268, pruned_loss=0.07264, over 985041.71 frames.], batch size: 73, aishell_tot_loss[loss=0.2017, simple_loss=0.271, pruned_loss=0.06619, over 981352.30 frames.], datatang_tot_loss[loss=0.2127, simple_loss=0.265, pruned_loss=0.08023, over 981794.52 frames.], batch size: 73, lr: 1.21e-03
+2022-06-18 14:32:50,358 INFO [train.py:874] (2/4) Epoch 6, batch 2300, datatang_loss[loss=0.1918, simple_loss=0.2604, pruned_loss=0.06161, over 4895.00 frames.], tot_loss[loss=0.2067, simple_loss=0.2685, pruned_loss=0.07246, over 985335.53 frames.], batch size: 25, aishell_tot_loss[loss=0.2005, simple_loss=0.2701, pruned_loss=0.06543, over 982131.28 frames.], datatang_tot_loss[loss=0.2143, simple_loss=0.2663, pruned_loss=0.08116, over 982150.55 frames.], batch size: 25, lr: 1.21e-03
+2022-06-18 14:33:22,407 INFO [train.py:874] (2/4) Epoch 6, batch 2350, datatang_loss[loss=0.1743, simple_loss=0.2413, pruned_loss=0.05365, over 4908.00 frames.], tot_loss[loss=0.2066, simple_loss=0.2681, pruned_loss=0.07254, over 985776.37 frames.], batch size: 47, aishell_tot_loss[loss=0.2008, simple_loss=0.2704, pruned_loss=0.06564, over 982932.65 frames.], datatang_tot_loss[loss=0.2139, simple_loss=0.2656, pruned_loss=0.08111, over 982565.38 frames.], batch size: 47, lr: 1.21e-03
+2022-06-18 14:33:50,370 INFO [train.py:874] (2/4) Epoch 6, batch 2400, aishell_loss[loss=0.1759, simple_loss=0.2538, pruned_loss=0.049, over 4923.00 frames.], tot_loss[loss=0.2064, simple_loss=0.2685, pruned_loss=0.07212, over 985656.20 frames.], batch size: 33, aishell_tot_loss[loss=0.2001, simple_loss=0.2701, pruned_loss=0.06505, over 983285.69 frames.], datatang_tot_loss[loss=0.2145, simple_loss=0.2662, pruned_loss=0.08141, over 982791.73 frames.], batch size: 33, lr: 1.21e-03
+2022-06-18 14:34:22,270 INFO [train.py:874] (2/4) Epoch 6, batch 2450, aishell_loss[loss=0.1987, simple_loss=0.2745, pruned_loss=0.06147, over 4937.00 frames.], tot_loss[loss=0.2057, simple_loss=0.2675, pruned_loss=0.07196, over 985692.19 frames.], batch size: 58, aishell_tot_loss[loss=0.1989, simple_loss=0.2688, pruned_loss=0.06453, over 983344.22 frames.], datatang_tot_loss[loss=0.2147, simple_loss=0.2666, pruned_loss=0.08142, over 983372.03 frames.], batch size: 58, lr: 1.21e-03
+2022-06-18 14:34:52,670 INFO [train.py:874] (2/4) Epoch 6, batch 2500, datatang_loss[loss=0.2254, simple_loss=0.2697, pruned_loss=0.09056, over 4918.00 frames.], tot_loss[loss=0.2043, simple_loss=0.266, pruned_loss=0.07131, over 985702.37 frames.], batch size: 79, aishell_tot_loss[loss=0.199, simple_loss=0.2689, pruned_loss=0.0646, over 983488.69 frames.], datatang_tot_loss[loss=0.2125, simple_loss=0.2649, pruned_loss=0.08011, over 983780.55 frames.], batch size: 79, lr: 1.20e-03
+2022-06-18 14:35:21,327 INFO [train.py:874] (2/4) Epoch 6, batch 2550, datatang_loss[loss=0.1918, simple_loss=0.2626, pruned_loss=0.06049, over 4952.00 frames.], tot_loss[loss=0.2039, simple_loss=0.2657, pruned_loss=0.07105, over 985876.08 frames.], batch size: 86, aishell_tot_loss[loss=0.199, simple_loss=0.2686, pruned_loss=0.06466, over 983811.88 frames.], datatang_tot_loss[loss=0.2118, simple_loss=0.2647, pruned_loss=0.07945, over 984124.17 frames.], batch size: 86, lr: 1.20e-03
+2022-06-18 14:35:53,587 INFO [train.py:874] (2/4) Epoch 6, batch 2600, aishell_loss[loss=0.2085, simple_loss=0.2811, pruned_loss=0.06792, over 4918.00 frames.], tot_loss[loss=0.205, simple_loss=0.2665, pruned_loss=0.07171, over 986081.49 frames.], batch size: 46, aishell_tot_loss[loss=0.2001, simple_loss=0.2697, pruned_loss=0.06523, over 984248.40 frames.], datatang_tot_loss[loss=0.2114, simple_loss=0.2643, pruned_loss=0.07931, over 984352.08 frames.], batch size: 46, lr: 1.20e-03
+2022-06-18 14:36:23,542 INFO [train.py:874] (2/4) Epoch 6, batch 2650, aishell_loss[loss=0.199, simple_loss=0.2736, pruned_loss=0.06221, over 4899.00 frames.], tot_loss[loss=0.2044, simple_loss=0.2662, pruned_loss=0.07129, over 986168.45 frames.], batch size: 60, aishell_tot_loss[loss=0.1994, simple_loss=0.2693, pruned_loss=0.06475, over 984401.19 frames.], datatang_tot_loss[loss=0.2115, simple_loss=0.264, pruned_loss=0.07946, over 984703.36 frames.], batch size: 60, lr: 1.20e-03
+2022-06-18 14:36:53,119 INFO [train.py:874] (2/4) Epoch 6, batch 2700, datatang_loss[loss=0.2103, simple_loss=0.2615, pruned_loss=0.07953, over 4931.00 frames.], tot_loss[loss=0.2033, simple_loss=0.2649, pruned_loss=0.07078, over 985753.79 frames.], batch size: 73, aishell_tot_loss[loss=0.199, simple_loss=0.2689, pruned_loss=0.06453, over 984207.53 frames.], datatang_tot_loss[loss=0.2103, simple_loss=0.2632, pruned_loss=0.07872, over 984846.69 frames.], batch size: 73, lr: 1.20e-03
+2022-06-18 14:37:24,165 INFO [train.py:874] (2/4) Epoch 6, batch 2750, datatang_loss[loss=0.2184, simple_loss=0.2717, pruned_loss=0.08257, over 4953.00 frames.], tot_loss[loss=0.2035, simple_loss=0.265, pruned_loss=0.07095, over 985677.87 frames.], batch size: 55, aishell_tot_loss[loss=0.1994, simple_loss=0.2689, pruned_loss=0.06489, over 984436.32 frames.], datatang_tot_loss[loss=0.2098, simple_loss=0.2629, pruned_loss=0.07831, over 984816.76 frames.], batch size: 55, lr: 1.20e-03
+2022-06-18 14:37:54,763 INFO [train.py:874] (2/4) Epoch 6, batch 2800, datatang_loss[loss=0.2275, simple_loss=0.2868, pruned_loss=0.08413, over 4950.00 frames.], tot_loss[loss=0.2037, simple_loss=0.2652, pruned_loss=0.0711, over 985965.87 frames.], batch size: 86, aishell_tot_loss[loss=0.1996, simple_loss=0.2692, pruned_loss=0.06501, over 984778.43 frames.], datatang_tot_loss[loss=0.2095, simple_loss=0.2627, pruned_loss=0.07815, over 985022.21 frames.], batch size: 86, lr: 1.20e-03
+2022-06-18 14:38:23,516 INFO [train.py:874] (2/4) Epoch 6, batch 2850, datatang_loss[loss=0.22, simple_loss=0.2727, pruned_loss=0.08365, over 4921.00 frames.], tot_loss[loss=0.2049, simple_loss=0.2657, pruned_loss=0.072, over 985512.04 frames.], batch size: 81, aishell_tot_loss[loss=0.1999, simple_loss=0.2692, pruned_loss=0.0653, over 984609.90 frames.], datatang_tot_loss[loss=0.2102, simple_loss=0.263, pruned_loss=0.07872, over 984968.81 frames.], batch size: 81, lr: 1.20e-03
+2022-06-18 14:38:55,051 INFO [train.py:874] (2/4) Epoch 6, batch 2900, aishell_loss[loss=0.1914, simple_loss=0.2633, pruned_loss=0.05972, over 4873.00 frames.], tot_loss[loss=0.2055, simple_loss=0.2668, pruned_loss=0.07211, over 985366.79 frames.], batch size: 35, aishell_tot_loss[loss=0.2008, simple_loss=0.2702, pruned_loss=0.06569, over 984661.11 frames.], datatang_tot_loss[loss=0.21, simple_loss=0.2632, pruned_loss=0.07834, over 984935.94 frames.], batch size: 35, lr: 1.19e-03
+2022-06-18 14:39:25,919 INFO [train.py:874] (2/4) Epoch 6, batch 2950, datatang_loss[loss=0.1965, simple_loss=0.2542, pruned_loss=0.06938, over 4950.00 frames.], tot_loss[loss=0.2056, simple_loss=0.2672, pruned_loss=0.07201, over 985641.40 frames.], batch size: 67, aishell_tot_loss[loss=0.2012, simple_loss=0.2708, pruned_loss=0.06575, over 984937.53 frames.], datatang_tot_loss[loss=0.2097, simple_loss=0.2631, pruned_loss=0.07813, over 985083.02 frames.], batch size: 67, lr: 1.19e-03
+2022-06-18 14:39:54,953 INFO [train.py:874] (2/4) Epoch 6, batch 3000, aishell_loss[loss=0.1404, simple_loss=0.2039, pruned_loss=0.0384, over 4928.00 frames.], tot_loss[loss=0.2047, simple_loss=0.2668, pruned_loss=0.07128, over 985876.44 frames.], batch size: 22, aishell_tot_loss[loss=0.2007, simple_loss=0.2706, pruned_loss=0.06541, over 985029.52 frames.], datatang_tot_loss[loss=0.2091, simple_loss=0.263, pruned_loss=0.07761, over 985375.21 frames.], batch size: 22, lr: 1.19e-03
+2022-06-18 14:39:54,954 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 14:40:11,299 INFO [train.py:914] (2/4) Epoch 6, validation: loss=0.1748, simple_loss=0.2561, pruned_loss=0.04677, over 1622729.00 frames.
+2022-06-18 14:40:41,206 INFO [train.py:874] (2/4) Epoch 6, batch 3050, aishell_loss[loss=0.2069, simple_loss=0.2785, pruned_loss=0.06764, over 4915.00 frames.], tot_loss[loss=0.2039, simple_loss=0.2662, pruned_loss=0.07079, over 986097.77 frames.], batch size: 52, aishell_tot_loss[loss=0.1995, simple_loss=0.2693, pruned_loss=0.06479, over 985225.37 frames.], datatang_tot_loss[loss=0.2096, simple_loss=0.2634, pruned_loss=0.07787, over 985582.46 frames.], batch size: 52, lr: 1.19e-03
+2022-06-18 14:41:11,400 INFO [train.py:874] (2/4) Epoch 6, batch 3100, aishell_loss[loss=0.2238, simple_loss=0.2958, pruned_loss=0.07593, over 4920.00 frames.], tot_loss[loss=0.2037, simple_loss=0.266, pruned_loss=0.07064, over 986058.61 frames.], batch size: 41, aishell_tot_loss[loss=0.1994, simple_loss=0.2694, pruned_loss=0.06467, over 985346.69 frames.], datatang_tot_loss[loss=0.2093, simple_loss=0.2631, pruned_loss=0.07779, over 985596.75 frames.], batch size: 41, lr: 1.19e-03
+2022-06-18 14:41:41,823 INFO [train.py:874] (2/4) Epoch 6, batch 3150, datatang_loss[loss=0.1921, simple_loss=0.238, pruned_loss=0.07308, over 4803.00 frames.], tot_loss[loss=0.2034, simple_loss=0.2658, pruned_loss=0.07054, over 985441.22 frames.], batch size: 23, aishell_tot_loss[loss=0.1996, simple_loss=0.2695, pruned_loss=0.06487, over 985084.46 frames.], datatang_tot_loss[loss=0.2088, simple_loss=0.2627, pruned_loss=0.07745, over 985365.65 frames.], batch size: 23, lr: 1.19e-03
+2022-06-18 14:42:15,238 INFO [train.py:874] (2/4) Epoch 6, batch 3200, aishell_loss[loss=0.1942, simple_loss=0.2676, pruned_loss=0.06046, over 4891.00 frames.], tot_loss[loss=0.2034, simple_loss=0.2664, pruned_loss=0.0702, over 985327.56 frames.], batch size: 42, aishell_tot_loss[loss=0.1999, simple_loss=0.2701, pruned_loss=0.06486, over 985035.58 frames.], datatang_tot_loss[loss=0.2085, simple_loss=0.2625, pruned_loss=0.07723, over 985367.19 frames.], batch size: 42, lr: 1.19e-03
+2022-06-18 14:42:46,473 INFO [train.py:874] (2/4) Epoch 6, batch 3250, datatang_loss[loss=0.22, simple_loss=0.2499, pruned_loss=0.09504, over 4869.00 frames.], tot_loss[loss=0.2026, simple_loss=0.2654, pruned_loss=0.06988, over 984571.39 frames.], batch size: 39, aishell_tot_loss[loss=0.1981, simple_loss=0.2684, pruned_loss=0.06389, over 984340.89 frames.], datatang_tot_loss[loss=0.2094, simple_loss=0.2631, pruned_loss=0.07783, over 985289.05 frames.], batch size: 39, lr: 1.19e-03
+2022-06-18 14:43:16,377 INFO [train.py:874] (2/4) Epoch 6, batch 3300, aishell_loss[loss=0.1889, simple_loss=0.2694, pruned_loss=0.05414, over 4897.00 frames.], tot_loss[loss=0.2028, simple_loss=0.2657, pruned_loss=0.06999, over 984597.34 frames.], batch size: 52, aishell_tot_loss[loss=0.1979, simple_loss=0.2684, pruned_loss=0.06369, over 984335.46 frames.], datatang_tot_loss[loss=0.2095, simple_loss=0.2634, pruned_loss=0.07785, over 985259.43 frames.], batch size: 52, lr: 1.18e-03
+2022-06-18 14:43:45,783 INFO [train.py:874] (2/4) Epoch 6, batch 3350, aishell_loss[loss=0.2004, simple_loss=0.282, pruned_loss=0.0594, over 4885.00 frames.], tot_loss[loss=0.2024, simple_loss=0.2654, pruned_loss=0.06969, over 985024.02 frames.], batch size: 35, aishell_tot_loss[loss=0.1977, simple_loss=0.2682, pruned_loss=0.06363, over 984569.65 frames.], datatang_tot_loss[loss=0.2092, simple_loss=0.2631, pruned_loss=0.07767, over 985443.60 frames.], batch size: 35, lr: 1.18e-03
+2022-06-18 14:44:17,236 INFO [train.py:874] (2/4) Epoch 6, batch 3400, aishell_loss[loss=0.202, simple_loss=0.282, pruned_loss=0.06098, over 4968.00 frames.], tot_loss[loss=0.2034, simple_loss=0.2666, pruned_loss=0.0701, over 985300.44 frames.], batch size: 51, aishell_tot_loss[loss=0.1979, simple_loss=0.2685, pruned_loss=0.06369, over 984684.84 frames.], datatang_tot_loss[loss=0.2101, simple_loss=0.2638, pruned_loss=0.0782, over 985648.02 frames.], batch size: 51, lr: 1.18e-03
+2022-06-18 14:44:46,245 INFO [train.py:874] (2/4) Epoch 6, batch 3450, aishell_loss[loss=0.2159, simple_loss=0.2827, pruned_loss=0.07457, over 4927.00 frames.], tot_loss[loss=0.2038, simple_loss=0.2667, pruned_loss=0.0704, over 984972.79 frames.], batch size: 68, aishell_tot_loss[loss=0.1979, simple_loss=0.2685, pruned_loss=0.06363, over 984690.27 frames.], datatang_tot_loss[loss=0.2105, simple_loss=0.264, pruned_loss=0.07848, over 985335.99 frames.], batch size: 68, lr: 1.18e-03
+2022-06-18 14:45:16,419 INFO [train.py:874] (2/4) Epoch 6, batch 3500, datatang_loss[loss=0.2948, simple_loss=0.326, pruned_loss=0.1318, over 4933.00 frames.], tot_loss[loss=0.2062, simple_loss=0.2693, pruned_loss=0.07153, over 985067.73 frames.], batch size: 108, aishell_tot_loss[loss=0.1986, simple_loss=0.2697, pruned_loss=0.06374, over 984884.33 frames.], datatang_tot_loss[loss=0.2125, simple_loss=0.2655, pruned_loss=0.07977, over 985246.81 frames.], batch size: 108, lr: 1.18e-03
+2022-06-18 14:45:47,004 INFO [train.py:874] (2/4) Epoch 6, batch 3550, aishell_loss[loss=0.1956, simple_loss=0.2767, pruned_loss=0.05725, over 4878.00 frames.], tot_loss[loss=0.2054, simple_loss=0.2692, pruned_loss=0.07087, over 985247.75 frames.], batch size: 42, aishell_tot_loss[loss=0.1985, simple_loss=0.2696, pruned_loss=0.06365, over 985082.03 frames.], datatang_tot_loss[loss=0.2126, simple_loss=0.2657, pruned_loss=0.07971, over 985248.64 frames.], batch size: 42, lr: 1.18e-03
+2022-06-18 14:46:15,901 INFO [train.py:874] (2/4) Epoch 6, batch 3600, aishell_loss[loss=0.1933, simple_loss=0.266, pruned_loss=0.0603, over 4938.00 frames.], tot_loss[loss=0.2043, simple_loss=0.2678, pruned_loss=0.07041, over 985555.48 frames.], batch size: 54, aishell_tot_loss[loss=0.1975, simple_loss=0.2688, pruned_loss=0.0631, over 985261.97 frames.], datatang_tot_loss[loss=0.2123, simple_loss=0.2654, pruned_loss=0.07961, over 985408.15 frames.], batch size: 54, lr: 1.18e-03
+2022-06-18 14:46:46,860 INFO [train.py:874] (2/4) Epoch 6, batch 3650, datatang_loss[loss=0.239, simple_loss=0.294, pruned_loss=0.092, over 4922.00 frames.], tot_loss[loss=0.2043, simple_loss=0.267, pruned_loss=0.07084, over 985978.60 frames.], batch size: 94, aishell_tot_loss[loss=0.197, simple_loss=0.2681, pruned_loss=0.06289, over 985560.82 frames.], datatang_tot_loss[loss=0.2124, simple_loss=0.2654, pruned_loss=0.07971, over 985612.00 frames.], batch size: 94, lr: 1.18e-03
+2022-06-18 14:47:16,891 INFO [train.py:874] (2/4) Epoch 6, batch 3700, datatang_loss[loss=0.2306, simple_loss=0.2932, pruned_loss=0.08397, over 4944.00 frames.], tot_loss[loss=0.2046, simple_loss=0.2672, pruned_loss=0.07101, over 986069.17 frames.], batch size: 99, aishell_tot_loss[loss=0.197, simple_loss=0.2681, pruned_loss=0.0629, over 985729.13 frames.], datatang_tot_loss[loss=0.2128, simple_loss=0.2658, pruned_loss=0.07988, over 985623.19 frames.], batch size: 99, lr: 1.18e-03
+2022-06-18 14:47:44,894 INFO [train.py:874] (2/4) Epoch 6, batch 3750, datatang_loss[loss=0.1755, simple_loss=0.2342, pruned_loss=0.0584, over 4887.00 frames.], tot_loss[loss=0.2039, simple_loss=0.2663, pruned_loss=0.07079, over 986190.99 frames.], batch size: 47, aishell_tot_loss[loss=0.1969, simple_loss=0.2682, pruned_loss=0.06286, over 985844.80 frames.], datatang_tot_loss[loss=0.2118, simple_loss=0.2648, pruned_loss=0.07936, over 985741.89 frames.], batch size: 47, lr: 1.17e-03
+2022-06-18 14:48:15,172 INFO [train.py:874] (2/4) Epoch 6, batch 3800, aishell_loss[loss=0.2079, simple_loss=0.2872, pruned_loss=0.06432, over 4875.00 frames.], tot_loss[loss=0.2036, simple_loss=0.266, pruned_loss=0.07064, over 985616.11 frames.], batch size: 42, aishell_tot_loss[loss=0.1981, simple_loss=0.2691, pruned_loss=0.06355, over 985537.50 frames.], datatang_tot_loss[loss=0.2101, simple_loss=0.2636, pruned_loss=0.07828, over 985548.21 frames.], batch size: 42, lr: 1.17e-03
+2022-06-18 14:48:43,814 INFO [train.py:874] (2/4) Epoch 6, batch 3850, datatang_loss[loss=0.1975, simple_loss=0.2535, pruned_loss=0.07077, over 4931.00 frames.], tot_loss[loss=0.2029, simple_loss=0.2654, pruned_loss=0.07017, over 985457.48 frames.], batch size: 79, aishell_tot_loss[loss=0.1973, simple_loss=0.2685, pruned_loss=0.06305, over 985300.81 frames.], datatang_tot_loss[loss=0.2098, simple_loss=0.2635, pruned_loss=0.07805, over 985633.54 frames.], batch size: 79, lr: 1.17e-03
+2022-06-18 14:49:12,061 INFO [train.py:874] (2/4) Epoch 6, batch 3900, aishell_loss[loss=0.2031, simple_loss=0.2756, pruned_loss=0.06535, over 4939.00 frames.], tot_loss[loss=0.2035, simple_loss=0.2658, pruned_loss=0.0706, over 985492.66 frames.], batch size: 58, aishell_tot_loss[loss=0.1983, simple_loss=0.2689, pruned_loss=0.06388, over 985361.47 frames.], datatang_tot_loss[loss=0.2095, simple_loss=0.2633, pruned_loss=0.07782, over 985631.03 frames.], batch size: 58, lr: 1.17e-03
+2022-06-18 14:49:41,277 INFO [train.py:874] (2/4) Epoch 6, batch 3950, datatang_loss[loss=0.2511, simple_loss=0.2935, pruned_loss=0.1044, over 4987.00 frames.], tot_loss[loss=0.2033, simple_loss=0.2653, pruned_loss=0.07067, over 985781.95 frames.], batch size: 40, aishell_tot_loss[loss=0.1982, simple_loss=0.2688, pruned_loss=0.06383, over 985319.59 frames.], datatang_tot_loss[loss=0.2093, simple_loss=0.2629, pruned_loss=0.07788, over 985984.17 frames.], batch size: 40, lr: 1.17e-03
+2022-06-18 14:50:10,157 INFO [train.py:874] (2/4) Epoch 6, batch 4000, aishell_loss[loss=0.1745, simple_loss=0.2463, pruned_loss=0.05128, over 4852.00 frames.], tot_loss[loss=0.2034, simple_loss=0.2658, pruned_loss=0.07056, over 985444.25 frames.], batch size: 28, aishell_tot_loss[loss=0.1987, simple_loss=0.2694, pruned_loss=0.06399, over 985168.34 frames.], datatang_tot_loss[loss=0.2089, simple_loss=0.2626, pruned_loss=0.07758, over 985812.08 frames.], batch size: 28, lr: 1.17e-03
+2022-06-18 14:50:10,157 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 14:50:26,075 INFO [train.py:914] (2/4) Epoch 6, validation: loss=0.1741, simple_loss=0.2559, pruned_loss=0.04619, over 1622729.00 frames.
+2022-06-18 14:51:43,915 INFO [train.py:874] (2/4) Epoch 7, batch 50, aishell_loss[loss=0.1934, simple_loss=0.2639, pruned_loss=0.06144, over 4942.00 frames.], tot_loss[loss=0.1897, simple_loss=0.2545, pruned_loss=0.06248, over 218177.22 frames.], batch size: 54, aishell_tot_loss[loss=0.1949, simple_loss=0.2649, pruned_loss=0.0624, over 115889.60 frames.], datatang_tot_loss[loss=0.1847, simple_loss=0.2441, pruned_loss=0.06261, over 115945.88 frames.], batch size: 54, lr: 1.12e-03
+2022-06-18 14:52:15,040 INFO [train.py:874] (2/4) Epoch 7, batch 100, datatang_loss[loss=0.2063, simple_loss=0.2605, pruned_loss=0.07603, over 4918.00 frames.], tot_loss[loss=0.1926, simple_loss=0.2575, pruned_loss=0.0638, over 388039.42 frames.], batch size: 81, aishell_tot_loss[loss=0.1961, simple_loss=0.2668, pruned_loss=0.0627, over 225569.30 frames.], datatang_tot_loss[loss=0.1885, simple_loss=0.2473, pruned_loss=0.06487, over 210781.49 frames.], batch size: 81, lr: 1.12e-03
+2022-06-18 14:52:44,466 INFO [train.py:874] (2/4) Epoch 7, batch 150, aishell_loss[loss=0.1986, simple_loss=0.2721, pruned_loss=0.06253, over 4943.00 frames.], tot_loss[loss=0.1968, simple_loss=0.2625, pruned_loss=0.0656, over 520341.68 frames.], batch size: 64, aishell_tot_loss[loss=0.1979, simple_loss=0.2687, pruned_loss=0.06352, over 344336.16 frames.], datatang_tot_loss[loss=0.1942, simple_loss=0.2525, pruned_loss=0.06793, over 270645.71 frames.], batch size: 64, lr: 1.12e-03
+2022-06-18 14:53:13,180 INFO [train.py:874] (2/4) Epoch 7, batch 200, datatang_loss[loss=0.2196, simple_loss=0.2677, pruned_loss=0.08577, over 4967.00 frames.], tot_loss[loss=0.1977, simple_loss=0.2626, pruned_loss=0.06642, over 623784.73 frames.], batch size: 60, aishell_tot_loss[loss=0.1973, simple_loss=0.2685, pruned_loss=0.06299, over 405707.88 frames.], datatang_tot_loss[loss=0.1971, simple_loss=0.255, pruned_loss=0.06958, over 370578.53 frames.], batch size: 60, lr: 1.12e-03
+2022-06-18 14:53:45,405 INFO [train.py:874] (2/4) Epoch 7, batch 250, aishell_loss[loss=0.1405, simple_loss=0.2048, pruned_loss=0.03812, over 4948.00 frames.], tot_loss[loss=0.1977, simple_loss=0.2622, pruned_loss=0.06662, over 703939.20 frames.], batch size: 25, aishell_tot_loss[loss=0.1966, simple_loss=0.268, pruned_loss=0.06258, over 473775.61 frames.], datatang_tot_loss[loss=0.1983, simple_loss=0.2556, pruned_loss=0.07053, over 443216.02 frames.], batch size: 25, lr: 1.11e-03
+2022-06-18 14:54:14,223 INFO [train.py:874] (2/4) Epoch 7, batch 300, aishell_loss[loss=0.2194, simple_loss=0.2922, pruned_loss=0.0733, over 4860.00 frames.], tot_loss[loss=0.198, simple_loss=0.2625, pruned_loss=0.0668, over 766373.18 frames.], batch size: 36, aishell_tot_loss[loss=0.1967, simple_loss=0.2676, pruned_loss=0.06286, over 545257.27 frames.], datatang_tot_loss[loss=0.1989, simple_loss=0.256, pruned_loss=0.07088, over 494927.54 frames.], batch size: 36, lr: 1.11e-03
+2022-06-18 14:54:43,151 INFO [train.py:874] (2/4) Epoch 7, batch 350, aishell_loss[loss=0.1994, simple_loss=0.2637, pruned_loss=0.06758, over 4964.00 frames.], tot_loss[loss=0.2002, simple_loss=0.2632, pruned_loss=0.06862, over 815102.33 frames.], batch size: 51, aishell_tot_loss[loss=0.1975, simple_loss=0.2674, pruned_loss=0.06373, over 593492.68 frames.], datatang_tot_loss[loss=0.2015, simple_loss=0.2578, pruned_loss=0.07258, over 556877.23 frames.], batch size: 51, lr: 1.11e-03
+2022-06-18 14:55:14,501 INFO [train.py:874] (2/4) Epoch 7, batch 400, aishell_loss[loss=0.2058, simple_loss=0.2732, pruned_loss=0.06915, over 4876.00 frames.], tot_loss[loss=0.1993, simple_loss=0.263, pruned_loss=0.06786, over 853123.56 frames.], batch size: 35, aishell_tot_loss[loss=0.1972, simple_loss=0.2675, pruned_loss=0.06342, over 646781.05 frames.], datatang_tot_loss[loss=0.2009, simple_loss=0.2573, pruned_loss=0.07225, over 599696.66 frames.], batch size: 35, lr: 1.11e-03
+2022-06-18 14:55:44,347 INFO [train.py:874] (2/4) Epoch 7, batch 450, datatang_loss[loss=0.2266, simple_loss=0.29, pruned_loss=0.08166, over 4954.00 frames.], tot_loss[loss=0.1992, simple_loss=0.2629, pruned_loss=0.06775, over 882596.44 frames.], batch size: 91, aishell_tot_loss[loss=0.1962, simple_loss=0.2668, pruned_loss=0.0628, over 678831.64 frames.], datatang_tot_loss[loss=0.2016, simple_loss=0.2584, pruned_loss=0.07241, over 653940.41 frames.], batch size: 91, lr: 1.11e-03
+2022-06-18 14:56:12,615 INFO [train.py:874] (2/4) Epoch 7, batch 500, aishell_loss[loss=0.198, simple_loss=0.2746, pruned_loss=0.0607, over 4977.00 frames.], tot_loss[loss=0.1993, simple_loss=0.2628, pruned_loss=0.06784, over 905565.77 frames.], batch size: 44, aishell_tot_loss[loss=0.1964, simple_loss=0.2668, pruned_loss=0.06297, over 716422.25 frames.], datatang_tot_loss[loss=0.2017, simple_loss=0.2584, pruned_loss=0.0725, over 691556.01 frames.], batch size: 44, lr: 1.11e-03
+2022-06-18 14:56:43,873 INFO [train.py:874] (2/4) Epoch 7, batch 550, aishell_loss[loss=0.189, simple_loss=0.2756, pruned_loss=0.05118, over 4887.00 frames.], tot_loss[loss=0.199, simple_loss=0.2626, pruned_loss=0.06772, over 923153.46 frames.], batch size: 50, aishell_tot_loss[loss=0.1963, simple_loss=0.267, pruned_loss=0.06281, over 741827.23 frames.], datatang_tot_loss[loss=0.2013, simple_loss=0.2582, pruned_loss=0.07223, over 732705.64 frames.], batch size: 50, lr: 1.11e-03
+2022-06-18 14:57:13,433 INFO [train.py:874] (2/4) Epoch 7, batch 600, datatang_loss[loss=0.1885, simple_loss=0.2483, pruned_loss=0.06435, over 4925.00 frames.], tot_loss[loss=0.1995, simple_loss=0.2626, pruned_loss=0.06816, over 936969.71 frames.], batch size: 73, aishell_tot_loss[loss=0.1957, simple_loss=0.2664, pruned_loss=0.0625, over 763813.83 frames.], datatang_tot_loss[loss=0.2024, simple_loss=0.2591, pruned_loss=0.0728, over 769222.36 frames.], batch size: 73, lr: 1.11e-03
+2022-06-18 14:57:42,476 INFO [train.py:874] (2/4) Epoch 7, batch 650, aishell_loss[loss=0.2242, simple_loss=0.2836, pruned_loss=0.08244, over 4871.00 frames.], tot_loss[loss=0.1991, simple_loss=0.2628, pruned_loss=0.06776, over 947025.94 frames.], batch size: 42, aishell_tot_loss[loss=0.1952, simple_loss=0.2662, pruned_loss=0.06208, over 792951.85 frames.], datatang_tot_loss[loss=0.2028, simple_loss=0.2593, pruned_loss=0.07315, over 790952.68 frames.], batch size: 42, lr: 1.11e-03
+2022-06-18 14:58:14,071 INFO [train.py:874] (2/4) Epoch 7, batch 700, datatang_loss[loss=0.1777, simple_loss=0.2442, pruned_loss=0.05562, over 4934.00 frames.], tot_loss[loss=0.199, simple_loss=0.2629, pruned_loss=0.06759, over 955125.01 frames.], batch size: 79, aishell_tot_loss[loss=0.1949, simple_loss=0.2662, pruned_loss=0.06183, over 815243.07 frames.], datatang_tot_loss[loss=0.2029, simple_loss=0.2594, pruned_loss=0.0732, over 813807.34 frames.], batch size: 79, lr: 1.11e-03
+2022-06-18 14:58:45,052 INFO [train.py:874] (2/4) Epoch 7, batch 750, datatang_loss[loss=0.1627, simple_loss=0.2403, pruned_loss=0.04252, over 4959.00 frames.], tot_loss[loss=0.199, simple_loss=0.2631, pruned_loss=0.06747, over 962065.67 frames.], batch size: 34, aishell_tot_loss[loss=0.1943, simple_loss=0.2661, pruned_loss=0.06126, over 833929.84 frames.], datatang_tot_loss[loss=0.2034, simple_loss=0.2599, pruned_loss=0.07345, over 835614.62 frames.], batch size: 34, lr: 1.10e-03
+2022-06-18 14:59:13,865 INFO [train.py:874] (2/4) Epoch 7, batch 800, datatang_loss[loss=0.1636, simple_loss=0.2346, pruned_loss=0.04629, over 4834.00 frames.], tot_loss[loss=0.1976, simple_loss=0.2621, pruned_loss=0.06656, over 967235.40 frames.], batch size: 30, aishell_tot_loss[loss=0.1942, simple_loss=0.2662, pruned_loss=0.06116, over 851830.76 frames.], datatang_tot_loss[loss=0.2018, simple_loss=0.2586, pruned_loss=0.07249, over 853198.27 frames.], batch size: 30, lr: 1.10e-03
+2022-06-18 14:59:44,733 INFO [train.py:874] (2/4) Epoch 7, batch 850, aishell_loss[loss=0.1528, simple_loss=0.2256, pruned_loss=0.04003, over 4947.00 frames.], tot_loss[loss=0.1977, simple_loss=0.2623, pruned_loss=0.06657, over 971265.03 frames.], batch size: 27, aishell_tot_loss[loss=0.1938, simple_loss=0.266, pruned_loss=0.06083, over 866313.69 frames.], datatang_tot_loss[loss=0.2021, simple_loss=0.259, pruned_loss=0.0726, over 869995.03 frames.], batch size: 27, lr: 1.10e-03
+2022-06-18 15:00:16,349 INFO [train.py:874] (2/4) Epoch 7, batch 900, datatang_loss[loss=0.255, simple_loss=0.2815, pruned_loss=0.1142, over 4973.00 frames.], tot_loss[loss=0.1976, simple_loss=0.2622, pruned_loss=0.06651, over 974532.28 frames.], batch size: 45, aishell_tot_loss[loss=0.1939, simple_loss=0.2662, pruned_loss=0.06081, over 878749.77 frames.], datatang_tot_loss[loss=0.2017, simple_loss=0.2588, pruned_loss=0.07234, over 885242.84 frames.], batch size: 45, lr: 1.10e-03
+2022-06-18 15:00:45,814 INFO [train.py:874] (2/4) Epoch 7, batch 950, aishell_loss[loss=0.2103, simple_loss=0.2864, pruned_loss=0.06714, over 4914.00 frames.], tot_loss[loss=0.1983, simple_loss=0.2627, pruned_loss=0.06691, over 977158.53 frames.], batch size: 52, aishell_tot_loss[loss=0.194, simple_loss=0.2662, pruned_loss=0.06089, over 891410.45 frames.], datatang_tot_loss[loss=0.2023, simple_loss=0.2592, pruned_loss=0.0727, over 897177.90 frames.], batch size: 52, lr: 1.10e-03
+2022-06-18 15:01:17,331 INFO [train.py:874] (2/4) Epoch 7, batch 1000, datatang_loss[loss=0.2039, simple_loss=0.2547, pruned_loss=0.07656, over 4979.00 frames.], tot_loss[loss=0.1985, simple_loss=0.2626, pruned_loss=0.06722, over 979019.83 frames.], batch size: 40, aishell_tot_loss[loss=0.1933, simple_loss=0.2655, pruned_loss=0.06059, over 901062.99 frames.], datatang_tot_loss[loss=0.2031, simple_loss=0.2599, pruned_loss=0.07315, over 908912.46 frames.], batch size: 40, lr: 1.10e-03
+2022-06-18 15:01:17,332 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 15:01:33,319 INFO [train.py:914] (2/4) Epoch 7, validation: loss=0.1728, simple_loss=0.2548, pruned_loss=0.04541, over 1622729.00 frames.
+2022-06-18 15:02:05,424 INFO [train.py:874] (2/4) Epoch 7, batch 1050, datatang_loss[loss=0.2208, simple_loss=0.2731, pruned_loss=0.08421, over 4928.00 frames.], tot_loss[loss=0.1986, simple_loss=0.2625, pruned_loss=0.06732, over 980430.34 frames.], batch size: 57, aishell_tot_loss[loss=0.1933, simple_loss=0.2654, pruned_loss=0.0606, over 909352.91 frames.], datatang_tot_loss[loss=0.2031, simple_loss=0.26, pruned_loss=0.07311, over 919377.39 frames.], batch size: 57, lr: 1.10e-03
+2022-06-18 15:02:35,722 INFO [train.py:874] (2/4) Epoch 7, batch 1100, datatang_loss[loss=0.1758, simple_loss=0.2408, pruned_loss=0.05543, over 4933.00 frames.], tot_loss[loss=0.1962, simple_loss=0.2616, pruned_loss=0.06539, over 981664.88 frames.], batch size: 79, aishell_tot_loss[loss=0.1917, simple_loss=0.2645, pruned_loss=0.05948, over 920746.20 frames.], datatang_tot_loss[loss=0.2026, simple_loss=0.2598, pruned_loss=0.07268, over 925098.29 frames.], batch size: 79, lr: 1.10e-03
+2022-06-18 15:03:04,433 INFO [train.py:874] (2/4) Epoch 7, batch 1150, aishell_loss[loss=0.2072, simple_loss=0.2836, pruned_loss=0.0654, over 4939.00 frames.], tot_loss[loss=0.1966, simple_loss=0.2617, pruned_loss=0.06581, over 982805.61 frames.], batch size: 54, aishell_tot_loss[loss=0.1927, simple_loss=0.2651, pruned_loss=0.0601, over 928228.96 frames.], datatang_tot_loss[loss=0.2018, simple_loss=0.259, pruned_loss=0.07229, over 932631.28 frames.], batch size: 54, lr: 1.10e-03
+2022-06-18 15:03:35,927 INFO [train.py:874] (2/4) Epoch 7, batch 1200, datatang_loss[loss=0.2154, simple_loss=0.2626, pruned_loss=0.08415, over 4965.00 frames.], tot_loss[loss=0.1963, simple_loss=0.2613, pruned_loss=0.06562, over 983545.97 frames.], batch size: 37, aishell_tot_loss[loss=0.1924, simple_loss=0.2649, pruned_loss=0.05989, over 935263.42 frames.], datatang_tot_loss[loss=0.2017, simple_loss=0.2587, pruned_loss=0.0723, over 938741.36 frames.], batch size: 37, lr: 1.10e-03
+2022-06-18 15:04:06,570 INFO [train.py:874] (2/4) Epoch 7, batch 1250, aishell_loss[loss=0.2031, simple_loss=0.2757, pruned_loss=0.06525, over 4951.00 frames.], tot_loss[loss=0.1963, simple_loss=0.2613, pruned_loss=0.06565, over 983963.09 frames.], batch size: 64, aishell_tot_loss[loss=0.1922, simple_loss=0.2647, pruned_loss=0.05984, over 942335.29 frames.], datatang_tot_loss[loss=0.2019, simple_loss=0.2586, pruned_loss=0.07255, over 943130.57 frames.], batch size: 64, lr: 1.09e-03
+2022-06-18 15:04:34,737 INFO [train.py:874] (2/4) Epoch 7, batch 1300, aishell_loss[loss=0.1999, simple_loss=0.2699, pruned_loss=0.06493, over 4931.00 frames.], tot_loss[loss=0.1971, simple_loss=0.2617, pruned_loss=0.06622, over 984608.89 frames.], batch size: 49, aishell_tot_loss[loss=0.1917, simple_loss=0.2643, pruned_loss=0.05956, over 947348.68 frames.], datatang_tot_loss[loss=0.203, simple_loss=0.2595, pruned_loss=0.07324, over 948500.10 frames.], batch size: 49, lr: 1.09e-03
+2022-06-18 15:05:04,913 INFO [train.py:874] (2/4) Epoch 7, batch 1350, aishell_loss[loss=0.1775, simple_loss=0.2514, pruned_loss=0.0518, over 4944.00 frames.], tot_loss[loss=0.1961, simple_loss=0.2609, pruned_loss=0.06564, over 984649.62 frames.], batch size: 32, aishell_tot_loss[loss=0.191, simple_loss=0.2635, pruned_loss=0.05921, over 951692.79 frames.], datatang_tot_loss[loss=0.2026, simple_loss=0.2593, pruned_loss=0.07291, over 952834.67 frames.], batch size: 32, lr: 1.09e-03
+2022-06-18 15:05:36,472 INFO [train.py:874] (2/4) Epoch 7, batch 1400, aishell_loss[loss=0.1835, simple_loss=0.2573, pruned_loss=0.05483, over 4922.00 frames.], tot_loss[loss=0.196, simple_loss=0.2609, pruned_loss=0.06551, over 984751.36 frames.], batch size: 41, aishell_tot_loss[loss=0.1908, simple_loss=0.2634, pruned_loss=0.05904, over 955577.45 frames.], datatang_tot_loss[loss=0.2025, simple_loss=0.2593, pruned_loss=0.07284, over 956687.59 frames.], batch size: 41, lr: 1.09e-03
+2022-06-18 15:06:05,397 INFO [train.py:874] (2/4) Epoch 7, batch 1450, aishell_loss[loss=0.2048, simple_loss=0.281, pruned_loss=0.06429, over 4938.00 frames.], tot_loss[loss=0.1961, simple_loss=0.2611, pruned_loss=0.06555, over 984483.11 frames.], batch size: 58, aishell_tot_loss[loss=0.1902, simple_loss=0.2631, pruned_loss=0.05868, over 958598.71 frames.], datatang_tot_loss[loss=0.2029, simple_loss=0.2598, pruned_loss=0.07304, over 960110.22 frames.], batch size: 58, lr: 1.09e-03
+2022-06-18 15:06:36,065 INFO [train.py:874] (2/4) Epoch 7, batch 1500, datatang_loss[loss=0.2147, simple_loss=0.2615, pruned_loss=0.08388, over 4919.00 frames.], tot_loss[loss=0.198, simple_loss=0.2629, pruned_loss=0.06654, over 984898.48 frames.], batch size: 64, aishell_tot_loss[loss=0.1916, simple_loss=0.2644, pruned_loss=0.05938, over 961523.39 frames.], datatang_tot_loss[loss=0.2034, simple_loss=0.2604, pruned_loss=0.07316, over 963536.33 frames.], batch size: 64, lr: 1.09e-03
+2022-06-18 15:07:07,638 INFO [train.py:874] (2/4) Epoch 7, batch 1550, aishell_loss[loss=0.197, simple_loss=0.282, pruned_loss=0.05601, over 4887.00 frames.], tot_loss[loss=0.1973, simple_loss=0.2625, pruned_loss=0.06607, over 985224.34 frames.], batch size: 50, aishell_tot_loss[loss=0.1922, simple_loss=0.265, pruned_loss=0.0597, over 964398.24 frames.], datatang_tot_loss[loss=0.2021, simple_loss=0.2594, pruned_loss=0.0724, over 966274.75 frames.], batch size: 50, lr: 1.09e-03
+2022-06-18 15:07:35,784 INFO [train.py:874] (2/4) Epoch 7, batch 1600, datatang_loss[loss=0.203, simple_loss=0.2521, pruned_loss=0.07693, over 4877.00 frames.], tot_loss[loss=0.1957, simple_loss=0.2618, pruned_loss=0.06477, over 985030.95 frames.], batch size: 39, aishell_tot_loss[loss=0.1916, simple_loss=0.2648, pruned_loss=0.05917, over 967444.62 frames.], datatang_tot_loss[loss=0.2014, simple_loss=0.2587, pruned_loss=0.07203, over 967742.97 frames.], batch size: 39, lr: 1.09e-03
+2022-06-18 15:08:07,453 INFO [train.py:874] (2/4) Epoch 7, batch 1650, datatang_loss[loss=0.1999, simple_loss=0.2596, pruned_loss=0.07004, over 4958.00 frames.], tot_loss[loss=0.1961, simple_loss=0.2615, pruned_loss=0.06535, over 985631.70 frames.], batch size: 67, aishell_tot_loss[loss=0.1915, simple_loss=0.2644, pruned_loss=0.05926, over 969391.29 frames.], datatang_tot_loss[loss=0.2015, simple_loss=0.2589, pruned_loss=0.07208, over 970527.33 frames.], batch size: 67, lr: 1.09e-03
+2022-06-18 15:08:38,375 INFO [train.py:874] (2/4) Epoch 7, batch 1700, datatang_loss[loss=0.2349, simple_loss=0.2889, pruned_loss=0.09052, over 4896.00 frames.], tot_loss[loss=0.1964, simple_loss=0.2617, pruned_loss=0.06555, over 985625.19 frames.], batch size: 47, aishell_tot_loss[loss=0.191, simple_loss=0.264, pruned_loss=0.05895, over 971088.61 frames.], datatang_tot_loss[loss=0.2022, simple_loss=0.2596, pruned_loss=0.07239, over 972490.72 frames.], batch size: 47, lr: 1.09e-03
+2022-06-18 15:09:08,393 INFO [train.py:874] (2/4) Epoch 7, batch 1750, aishell_loss[loss=0.1928, simple_loss=0.2764, pruned_loss=0.05462, over 4977.00 frames.], tot_loss[loss=0.1958, simple_loss=0.2608, pruned_loss=0.06535, over 985432.99 frames.], batch size: 39, aishell_tot_loss[loss=0.1909, simple_loss=0.2636, pruned_loss=0.05907, over 972255.38 frames.], datatang_tot_loss[loss=0.2013, simple_loss=0.259, pruned_loss=0.07178, over 974354.91 frames.], batch size: 39, lr: 1.08e-03
+2022-06-18 15:09:38,818 INFO [train.py:874] (2/4) Epoch 7, batch 1800, aishell_loss[loss=0.2229, simple_loss=0.2817, pruned_loss=0.08208, over 4986.00 frames.], tot_loss[loss=0.1973, simple_loss=0.2621, pruned_loss=0.06623, over 985648.15 frames.], batch size: 39, aishell_tot_loss[loss=0.1917, simple_loss=0.2642, pruned_loss=0.05955, over 973912.74 frames.], datatang_tot_loss[loss=0.2021, simple_loss=0.2596, pruned_loss=0.07235, over 975803.94 frames.], batch size: 39, lr: 1.08e-03
+2022-06-18 15:10:08,337 INFO [train.py:874] (2/4) Epoch 7, batch 1850, datatang_loss[loss=0.1913, simple_loss=0.253, pruned_loss=0.06477, over 4980.00 frames.], tot_loss[loss=0.198, simple_loss=0.2625, pruned_loss=0.06669, over 985572.34 frames.], batch size: 40, aishell_tot_loss[loss=0.1918, simple_loss=0.2643, pruned_loss=0.0596, over 975168.05 frames.], datatang_tot_loss[loss=0.2029, simple_loss=0.26, pruned_loss=0.07293, over 977018.49 frames.], batch size: 40, lr: 1.08e-03
+2022-06-18 15:10:38,751 INFO [train.py:874] (2/4) Epoch 7, batch 1900, aishell_loss[loss=0.2443, simple_loss=0.3092, pruned_loss=0.0897, over 4910.00 frames.], tot_loss[loss=0.1978, simple_loss=0.2627, pruned_loss=0.06647, over 985440.17 frames.], batch size: 68, aishell_tot_loss[loss=0.1928, simple_loss=0.2654, pruned_loss=0.06013, over 976177.59 frames.], datatang_tot_loss[loss=0.2018, simple_loss=0.2591, pruned_loss=0.07225, over 978117.05 frames.], batch size: 68, lr: 1.08e-03
+2022-06-18 15:11:10,027 INFO [train.py:874] (2/4) Epoch 7, batch 1950, aishell_loss[loss=0.1815, simple_loss=0.2514, pruned_loss=0.05582, over 4968.00 frames.], tot_loss[loss=0.198, simple_loss=0.2626, pruned_loss=0.06673, over 985638.97 frames.], batch size: 39, aishell_tot_loss[loss=0.1941, simple_loss=0.2664, pruned_loss=0.06092, over 976969.62 frames.], datatang_tot_loss[loss=0.2007, simple_loss=0.2584, pruned_loss=0.07147, over 979420.38 frames.], batch size: 39, lr: 1.08e-03
+2022-06-18 15:11:38,535 INFO [train.py:874] (2/4) Epoch 7, batch 2000, aishell_loss[loss=0.1898, simple_loss=0.2665, pruned_loss=0.05656, over 4958.00 frames.], tot_loss[loss=0.1972, simple_loss=0.2622, pruned_loss=0.06607, over 985486.01 frames.], batch size: 40, aishell_tot_loss[loss=0.1937, simple_loss=0.266, pruned_loss=0.06069, over 977836.96 frames.], datatang_tot_loss[loss=0.2004, simple_loss=0.2583, pruned_loss=0.07128, over 980176.22 frames.], batch size: 40, lr: 1.08e-03
+2022-06-18 15:11:38,536 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 15:11:55,039 INFO [train.py:914] (2/4) Epoch 7, validation: loss=0.1748, simple_loss=0.2567, pruned_loss=0.04647, over 1622729.00 frames.
+2022-06-18 15:12:25,168 INFO [train.py:874] (2/4) Epoch 7, batch 2050, aishell_loss[loss=0.2373, simple_loss=0.3058, pruned_loss=0.08438, over 4968.00 frames.], tot_loss[loss=0.1967, simple_loss=0.2624, pruned_loss=0.06553, over 985684.13 frames.], batch size: 81, aishell_tot_loss[loss=0.1936, simple_loss=0.2662, pruned_loss=0.06048, over 978819.69 frames.], datatang_tot_loss[loss=0.2001, simple_loss=0.2582, pruned_loss=0.07104, over 980964.58 frames.], batch size: 81, lr: 1.08e-03
+2022-06-18 15:12:54,798 INFO [train.py:874] (2/4) Epoch 7, batch 2100, aishell_loss[loss=0.201, simple_loss=0.264, pruned_loss=0.06901, over 4941.00 frames.], tot_loss[loss=0.1968, simple_loss=0.2629, pruned_loss=0.06538, over 985696.46 frames.], batch size: 58, aishell_tot_loss[loss=0.1937, simple_loss=0.2665, pruned_loss=0.06041, over 979763.48 frames.], datatang_tot_loss[loss=0.2003, simple_loss=0.2582, pruned_loss=0.07122, over 981435.96 frames.], batch size: 58, lr: 1.08e-03
+2022-06-18 15:13:25,724 INFO [train.py:874] (2/4) Epoch 7, batch 2150, datatang_loss[loss=0.2262, simple_loss=0.274, pruned_loss=0.08915, over 4922.00 frames.], tot_loss[loss=0.1963, simple_loss=0.2629, pruned_loss=0.06487, over 985904.73 frames.], batch size: 73, aishell_tot_loss[loss=0.1931, simple_loss=0.2662, pruned_loss=0.05996, over 980595.05 frames.], datatang_tot_loss[loss=0.2005, simple_loss=0.2585, pruned_loss=0.07128, over 982065.66 frames.], batch size: 73, lr: 1.08e-03
+2022-06-18 15:13:55,545 INFO [train.py:874] (2/4) Epoch 7, batch 2200, datatang_loss[loss=0.1751, simple_loss=0.2446, pruned_loss=0.05285, over 4932.00 frames.], tot_loss[loss=0.1958, simple_loss=0.2621, pruned_loss=0.06479, over 985389.92 frames.], batch size: 57, aishell_tot_loss[loss=0.1928, simple_loss=0.2661, pruned_loss=0.05973, over 980683.69 frames.], datatang_tot_loss[loss=0.1998, simple_loss=0.2581, pruned_loss=0.07079, over 982448.12 frames.], batch size: 57, lr: 1.08e-03
+2022-06-18 15:14:26,106 INFO [train.py:874] (2/4) Epoch 7, batch 2250, aishell_loss[loss=0.2182, simple_loss=0.2869, pruned_loss=0.07477, over 4917.00 frames.], tot_loss[loss=0.1966, simple_loss=0.2627, pruned_loss=0.06526, over 985754.19 frames.], batch size: 46, aishell_tot_loss[loss=0.1926, simple_loss=0.266, pruned_loss=0.05962, over 981614.48 frames.], datatang_tot_loss[loss=0.2009, simple_loss=0.2588, pruned_loss=0.07143, over 982830.24 frames.], batch size: 46, lr: 1.07e-03
+2022-06-18 15:14:56,305 INFO [train.py:874] (2/4) Epoch 7, batch 2300, aishell_loss[loss=0.1896, simple_loss=0.2684, pruned_loss=0.0554, over 4940.00 frames.], tot_loss[loss=0.1955, simple_loss=0.2624, pruned_loss=0.06425, over 985983.84 frames.], batch size: 31, aishell_tot_loss[loss=0.1922, simple_loss=0.2659, pruned_loss=0.05929, over 982385.45 frames.], datatang_tot_loss[loss=0.2002, simple_loss=0.2584, pruned_loss=0.07101, over 983167.71 frames.], batch size: 31, lr: 1.07e-03
+2022-06-18 15:15:26,201 INFO [train.py:874] (2/4) Epoch 7, batch 2350, datatang_loss[loss=0.1953, simple_loss=0.2539, pruned_loss=0.06829, over 4921.00 frames.], tot_loss[loss=0.1951, simple_loss=0.2616, pruned_loss=0.06429, over 985704.82 frames.], batch size: 75, aishell_tot_loss[loss=0.1914, simple_loss=0.2649, pruned_loss=0.0589, over 982407.56 frames.], datatang_tot_loss[loss=0.2004, simple_loss=0.2586, pruned_loss=0.07108, over 983576.99 frames.], batch size: 75, lr: 1.07e-03
+2022-06-18 15:15:56,417 INFO [train.py:874] (2/4) Epoch 7, batch 2400, datatang_loss[loss=0.1811, simple_loss=0.2538, pruned_loss=0.05421, over 4925.00 frames.], tot_loss[loss=0.1966, simple_loss=0.2628, pruned_loss=0.06522, over 985597.25 frames.], batch size: 83, aishell_tot_loss[loss=0.1921, simple_loss=0.2656, pruned_loss=0.0593, over 982656.38 frames.], datatang_tot_loss[loss=0.201, simple_loss=0.2592, pruned_loss=0.07144, over 983850.72 frames.], batch size: 83, lr: 1.07e-03
+2022-06-18 15:16:27,325 INFO [train.py:874] (2/4) Epoch 7, batch 2450, aishell_loss[loss=0.1641, simple_loss=0.2383, pruned_loss=0.04492, over 4864.00 frames.], tot_loss[loss=0.1946, simple_loss=0.2608, pruned_loss=0.06413, over 985370.03 frames.], batch size: 28, aishell_tot_loss[loss=0.1908, simple_loss=0.2643, pruned_loss=0.05864, over 982915.15 frames.], datatang_tot_loss[loss=0.2003, simple_loss=0.2585, pruned_loss=0.07103, over 983915.40 frames.], batch size: 28, lr: 1.07e-03
+2022-06-18 15:16:56,371 INFO [train.py:874] (2/4) Epoch 7, batch 2500, datatang_loss[loss=0.2272, simple_loss=0.281, pruned_loss=0.08676, over 4914.00 frames.], tot_loss[loss=0.1957, simple_loss=0.2617, pruned_loss=0.06489, over 985508.32 frames.], batch size: 98, aishell_tot_loss[loss=0.1914, simple_loss=0.2646, pruned_loss=0.05905, over 983183.45 frames.], datatang_tot_loss[loss=0.2008, simple_loss=0.259, pruned_loss=0.07131, over 984257.49 frames.], batch size: 98, lr: 1.07e-03
+2022-06-18 15:17:26,858 INFO [train.py:874] (2/4) Epoch 7, batch 2550, aishell_loss[loss=0.2128, simple_loss=0.2871, pruned_loss=0.06927, over 4946.00 frames.], tot_loss[loss=0.196, simple_loss=0.2623, pruned_loss=0.06486, over 985612.01 frames.], batch size: 64, aishell_tot_loss[loss=0.1921, simple_loss=0.2653, pruned_loss=0.05944, over 983485.81 frames.], datatang_tot_loss[loss=0.2005, simple_loss=0.2588, pruned_loss=0.07115, over 984512.14 frames.], batch size: 64, lr: 1.07e-03
+2022-06-18 15:17:58,012 INFO [train.py:874] (2/4) Epoch 7, batch 2600, aishell_loss[loss=0.1795, simple_loss=0.2649, pruned_loss=0.04701, over 4951.00 frames.], tot_loss[loss=0.1964, simple_loss=0.2625, pruned_loss=0.06519, over 985726.77 frames.], batch size: 56, aishell_tot_loss[loss=0.1928, simple_loss=0.2659, pruned_loss=0.05984, over 983830.43 frames.], datatang_tot_loss[loss=0.2003, simple_loss=0.2584, pruned_loss=0.07107, over 984679.65 frames.], batch size: 56, lr: 1.07e-03
+2022-06-18 15:18:28,166 INFO [train.py:874] (2/4) Epoch 7, batch 2650, aishell_loss[loss=0.186, simple_loss=0.2699, pruned_loss=0.05101, over 4909.00 frames.], tot_loss[loss=0.1957, simple_loss=0.2619, pruned_loss=0.06471, over 985848.45 frames.], batch size: 41, aishell_tot_loss[loss=0.1931, simple_loss=0.2661, pruned_loss=0.06007, over 984128.38 frames.], datatang_tot_loss[loss=0.1992, simple_loss=0.2577, pruned_loss=0.07033, over 984860.96 frames.], batch size: 41, lr: 1.07e-03
+2022-06-18 15:18:58,540 INFO [train.py:874] (2/4) Epoch 7, batch 2700, aishell_loss[loss=0.1912, simple_loss=0.2676, pruned_loss=0.05736, over 4918.00 frames.], tot_loss[loss=0.1973, simple_loss=0.2631, pruned_loss=0.06579, over 986399.95 frames.], batch size: 52, aishell_tot_loss[loss=0.1927, simple_loss=0.2657, pruned_loss=0.05983, over 984701.99 frames.], datatang_tot_loss[loss=0.2013, simple_loss=0.2592, pruned_loss=0.07167, over 985192.49 frames.], batch size: 52, lr: 1.07e-03
+2022-06-18 15:19:28,844 INFO [train.py:874] (2/4) Epoch 7, batch 2750, aishell_loss[loss=0.1479, simple_loss=0.2284, pruned_loss=0.03372, over 4865.00 frames.], tot_loss[loss=0.1965, simple_loss=0.2628, pruned_loss=0.06509, over 986089.82 frames.], batch size: 28, aishell_tot_loss[loss=0.1927, simple_loss=0.266, pruned_loss=0.05964, over 984641.16 frames.], datatang_tot_loss[loss=0.2006, simple_loss=0.2587, pruned_loss=0.07128, over 985269.56 frames.], batch size: 28, lr: 1.07e-03
+2022-06-18 15:19:58,434 INFO [train.py:874] (2/4) Epoch 7, batch 2800, aishell_loss[loss=0.1789, simple_loss=0.2489, pruned_loss=0.05445, over 4879.00 frames.], tot_loss[loss=0.1953, simple_loss=0.2613, pruned_loss=0.06463, over 985977.97 frames.], batch size: 28, aishell_tot_loss[loss=0.1926, simple_loss=0.2657, pruned_loss=0.05972, over 984726.64 frames.], datatang_tot_loss[loss=0.1993, simple_loss=0.2577, pruned_loss=0.07047, over 985322.42 frames.], batch size: 28, lr: 1.06e-03
+2022-06-18 15:20:29,256 INFO [train.py:874] (2/4) Epoch 7, batch 2850, aishell_loss[loss=0.1639, simple_loss=0.2462, pruned_loss=0.04074, over 4901.00 frames.], tot_loss[loss=0.1963, simple_loss=0.2624, pruned_loss=0.06513, over 985986.35 frames.], batch size: 34, aishell_tot_loss[loss=0.1927, simple_loss=0.266, pruned_loss=0.05971, over 984701.60 frames.], datatang_tot_loss[loss=0.2001, simple_loss=0.2585, pruned_loss=0.07085, over 985576.24 frames.], batch size: 34, lr: 1.06e-03
+2022-06-18 15:21:00,382 INFO [train.py:874] (2/4) Epoch 7, batch 2900, aishell_loss[loss=0.1985, simple_loss=0.2746, pruned_loss=0.06117, over 4966.00 frames.], tot_loss[loss=0.1964, simple_loss=0.2622, pruned_loss=0.06527, over 985949.45 frames.], batch size: 61, aishell_tot_loss[loss=0.1927, simple_loss=0.266, pruned_loss=0.05969, over 984759.04 frames.], datatang_tot_loss[loss=0.1999, simple_loss=0.2585, pruned_loss=0.07062, over 985644.78 frames.], batch size: 61, lr: 1.06e-03
+2022-06-18 15:21:30,576 INFO [train.py:874] (2/4) Epoch 7, batch 2950, aishell_loss[loss=0.1896, simple_loss=0.2598, pruned_loss=0.05972, over 4900.00 frames.], tot_loss[loss=0.1973, simple_loss=0.2626, pruned_loss=0.06603, over 985859.30 frames.], batch size: 34, aishell_tot_loss[loss=0.1932, simple_loss=0.2664, pruned_loss=0.06005, over 984727.87 frames.], datatang_tot_loss[loss=0.2002, simple_loss=0.2589, pruned_loss=0.07081, over 985731.05 frames.], batch size: 34, lr: 1.06e-03
+2022-06-18 15:22:01,257 INFO [train.py:874] (2/4) Epoch 7, batch 3000, datatang_loss[loss=0.1999, simple_loss=0.2554, pruned_loss=0.07216, over 4925.00 frames.], tot_loss[loss=0.1974, simple_loss=0.2625, pruned_loss=0.06611, over 985838.90 frames.], batch size: 71, aishell_tot_loss[loss=0.1936, simple_loss=0.2667, pruned_loss=0.06019, over 984878.43 frames.], datatang_tot_loss[loss=0.2, simple_loss=0.2586, pruned_loss=0.07068, over 985676.21 frames.], batch size: 71, lr: 1.06e-03
+2022-06-18 15:22:01,258 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 15:22:18,148 INFO [train.py:914] (2/4) Epoch 7, validation: loss=0.1737, simple_loss=0.2558, pruned_loss=0.0458, over 1622729.00 frames.
+2022-06-18 15:22:48,091 INFO [train.py:874] (2/4) Epoch 7, batch 3050, aishell_loss[loss=0.1983, simple_loss=0.2668, pruned_loss=0.06487, over 4857.00 frames.], tot_loss[loss=0.1953, simple_loss=0.2606, pruned_loss=0.06502, over 985954.65 frames.], batch size: 38, aishell_tot_loss[loss=0.1926, simple_loss=0.2657, pruned_loss=0.05976, over 984972.46 frames.], datatang_tot_loss[loss=0.1989, simple_loss=0.2575, pruned_loss=0.07013, over 985846.63 frames.], batch size: 38, lr: 1.06e-03
+2022-06-18 15:23:18,089 INFO [train.py:874] (2/4) Epoch 7, batch 3100, datatang_loss[loss=0.2359, simple_loss=0.289, pruned_loss=0.09145, over 4912.00 frames.], tot_loss[loss=0.1951, simple_loss=0.261, pruned_loss=0.06465, over 985691.18 frames.], batch size: 98, aishell_tot_loss[loss=0.1924, simple_loss=0.2658, pruned_loss=0.05948, over 984929.38 frames.], datatang_tot_loss[loss=0.1989, simple_loss=0.2575, pruned_loss=0.0702, over 985745.81 frames.], batch size: 98, lr: 1.06e-03
+2022-06-18 15:23:50,867 INFO [train.py:874] (2/4) Epoch 7, batch 3150, aishell_loss[loss=0.2006, simple_loss=0.2764, pruned_loss=0.06238, over 4932.00 frames.], tot_loss[loss=0.1939, simple_loss=0.2603, pruned_loss=0.06375, over 986091.07 frames.], batch size: 64, aishell_tot_loss[loss=0.1915, simple_loss=0.2651, pruned_loss=0.0589, over 985290.57 frames.], datatang_tot_loss[loss=0.1985, simple_loss=0.2571, pruned_loss=0.06992, over 985908.32 frames.], batch size: 64, lr: 1.06e-03
+2022-06-18 15:24:21,584 INFO [train.py:874] (2/4) Epoch 7, batch 3200, datatang_loss[loss=0.2044, simple_loss=0.2591, pruned_loss=0.07483, over 4983.00 frames.], tot_loss[loss=0.1931, simple_loss=0.2599, pruned_loss=0.06319, over 986166.82 frames.], batch size: 40, aishell_tot_loss[loss=0.1911, simple_loss=0.2651, pruned_loss=0.05854, over 985519.30 frames.], datatang_tot_loss[loss=0.1978, simple_loss=0.2567, pruned_loss=0.0695, over 985874.33 frames.], batch size: 40, lr: 1.06e-03
+2022-06-18 15:24:53,201 INFO [train.py:874] (2/4) Epoch 7, batch 3250, aishell_loss[loss=0.2, simple_loss=0.2749, pruned_loss=0.06251, over 4965.00 frames.], tot_loss[loss=0.1931, simple_loss=0.2604, pruned_loss=0.06293, over 985851.23 frames.], batch size: 44, aishell_tot_loss[loss=0.1912, simple_loss=0.2654, pruned_loss=0.05847, over 985429.03 frames.], datatang_tot_loss[loss=0.1977, simple_loss=0.2564, pruned_loss=0.06947, over 985735.56 frames.], batch size: 44, lr: 1.06e-03
+2022-06-18 15:25:21,867 INFO [train.py:874] (2/4) Epoch 7, batch 3300, datatang_loss[loss=0.2268, simple_loss=0.2773, pruned_loss=0.08821, over 4963.00 frames.], tot_loss[loss=0.1938, simple_loss=0.2607, pruned_loss=0.06342, over 985932.09 frames.], batch size: 86, aishell_tot_loss[loss=0.1903, simple_loss=0.2646, pruned_loss=0.058, over 985496.22 frames.], datatang_tot_loss[loss=0.1989, simple_loss=0.2574, pruned_loss=0.07016, over 985807.25 frames.], batch size: 86, lr: 1.05e-03
+2022-06-18 15:25:52,842 INFO [train.py:874] (2/4) Epoch 7, batch 3350, datatang_loss[loss=0.1958, simple_loss=0.232, pruned_loss=0.07979, over 4961.00 frames.], tot_loss[loss=0.193, simple_loss=0.2603, pruned_loss=0.06289, over 985793.05 frames.], batch size: 34, aishell_tot_loss[loss=0.19, simple_loss=0.2645, pruned_loss=0.05776, over 985545.80 frames.], datatang_tot_loss[loss=0.1984, simple_loss=0.2568, pruned_loss=0.06997, over 985669.18 frames.], batch size: 34, lr: 1.05e-03
+2022-06-18 15:26:24,279 INFO [train.py:874] (2/4) Epoch 7, batch 3400, aishell_loss[loss=0.2081, simple_loss=0.2849, pruned_loss=0.06559, over 4868.00 frames.], tot_loss[loss=0.1928, simple_loss=0.2598, pruned_loss=0.06292, over 986012.97 frames.], batch size: 37, aishell_tot_loss[loss=0.1896, simple_loss=0.264, pruned_loss=0.05763, over 985460.10 frames.], datatang_tot_loss[loss=0.1982, simple_loss=0.2567, pruned_loss=0.06985, over 986050.46 frames.], batch size: 37, lr: 1.05e-03
+2022-06-18 15:26:53,558 INFO [train.py:874] (2/4) Epoch 7, batch 3450, datatang_loss[loss=0.1932, simple_loss=0.2608, pruned_loss=0.06279, over 4904.00 frames.], tot_loss[loss=0.1943, simple_loss=0.2611, pruned_loss=0.06378, over 985847.80 frames.], batch size: 85, aishell_tot_loss[loss=0.1906, simple_loss=0.2646, pruned_loss=0.05827, over 985353.41 frames.], datatang_tot_loss[loss=0.1987, simple_loss=0.2572, pruned_loss=0.07007, over 986051.48 frames.], batch size: 85, lr: 1.05e-03
+2022-06-18 15:27:24,023 INFO [train.py:874] (2/4) Epoch 7, batch 3500, aishell_loss[loss=0.1759, simple_loss=0.2306, pruned_loss=0.0606, over 4769.00 frames.], tot_loss[loss=0.1947, simple_loss=0.2611, pruned_loss=0.06411, over 985515.14 frames.], batch size: 21, aishell_tot_loss[loss=0.19, simple_loss=0.264, pruned_loss=0.05803, over 985122.63 frames.], datatang_tot_loss[loss=0.1995, simple_loss=0.258, pruned_loss=0.07044, over 985966.39 frames.], batch size: 21, lr: 1.05e-03
+2022-06-18 15:27:53,565 INFO [train.py:874] (2/4) Epoch 7, batch 3550, aishell_loss[loss=0.1829, simple_loss=0.2609, pruned_loss=0.05244, over 4877.00 frames.], tot_loss[loss=0.1955, simple_loss=0.2617, pruned_loss=0.06462, over 985652.31 frames.], batch size: 35, aishell_tot_loss[loss=0.1907, simple_loss=0.2643, pruned_loss=0.0585, over 985117.14 frames.], datatang_tot_loss[loss=0.1998, simple_loss=0.2583, pruned_loss=0.0706, over 986115.62 frames.], batch size: 35, lr: 1.05e-03
+2022-06-18 15:28:23,403 INFO [train.py:874] (2/4) Epoch 7, batch 3600, aishell_loss[loss=0.2151, simple_loss=0.2732, pruned_loss=0.07848, over 4924.00 frames.], tot_loss[loss=0.1963, simple_loss=0.2625, pruned_loss=0.06501, over 985821.67 frames.], batch size: 41, aishell_tot_loss[loss=0.1916, simple_loss=0.265, pruned_loss=0.05907, over 985340.70 frames.], datatang_tot_loss[loss=0.1998, simple_loss=0.2585, pruned_loss=0.0706, over 986098.37 frames.], batch size: 41, lr: 1.05e-03
+2022-06-18 15:28:54,591 INFO [train.py:874] (2/4) Epoch 7, batch 3650, aishell_loss[loss=0.2354, simple_loss=0.2936, pruned_loss=0.08859, over 4881.00 frames.], tot_loss[loss=0.1964, simple_loss=0.2622, pruned_loss=0.06531, over 985828.79 frames.], batch size: 42, aishell_tot_loss[loss=0.1922, simple_loss=0.2654, pruned_loss=0.0595, over 985435.56 frames.], datatang_tot_loss[loss=0.1994, simple_loss=0.2581, pruned_loss=0.07041, over 986032.95 frames.], batch size: 42, lr: 1.05e-03
+2022-06-18 15:29:24,186 INFO [train.py:874] (2/4) Epoch 7, batch 3700, aishell_loss[loss=0.1662, simple_loss=0.2574, pruned_loss=0.03753, over 4951.00 frames.], tot_loss[loss=0.1962, simple_loss=0.2624, pruned_loss=0.06501, over 985790.34 frames.], batch size: 56, aishell_tot_loss[loss=0.192, simple_loss=0.2655, pruned_loss=0.05931, over 985472.34 frames.], datatang_tot_loss[loss=0.1995, simple_loss=0.2581, pruned_loss=0.07041, over 985981.79 frames.], batch size: 56, lr: 1.05e-03
+2022-06-18 15:29:54,972 INFO [train.py:874] (2/4) Epoch 7, batch 3750, aishell_loss[loss=0.1818, simple_loss=0.2334, pruned_loss=0.06506, over 4839.00 frames.], tot_loss[loss=0.1951, simple_loss=0.2617, pruned_loss=0.06427, over 985691.19 frames.], batch size: 21, aishell_tot_loss[loss=0.1916, simple_loss=0.2652, pruned_loss=0.05899, over 985351.49 frames.], datatang_tot_loss[loss=0.1989, simple_loss=0.2579, pruned_loss=0.06999, over 986020.00 frames.], batch size: 21, lr: 1.05e-03
+2022-06-18 15:30:23,977 INFO [train.py:874] (2/4) Epoch 7, batch 3800, aishell_loss[loss=0.2022, simple_loss=0.2748, pruned_loss=0.06484, over 4862.00 frames.], tot_loss[loss=0.1945, simple_loss=0.2613, pruned_loss=0.06381, over 985503.66 frames.], batch size: 37, aishell_tot_loss[loss=0.1911, simple_loss=0.2647, pruned_loss=0.05873, over 985119.99 frames.], datatang_tot_loss[loss=0.1988, simple_loss=0.258, pruned_loss=0.06978, over 986071.80 frames.], batch size: 37, lr: 1.05e-03
+2022-06-18 15:30:52,415 INFO [train.py:874] (2/4) Epoch 7, batch 3850, datatang_loss[loss=0.2274, simple_loss=0.2822, pruned_loss=0.08632, over 4909.00 frames.], tot_loss[loss=0.1942, simple_loss=0.261, pruned_loss=0.06375, over 984978.30 frames.], batch size: 52, aishell_tot_loss[loss=0.1911, simple_loss=0.2644, pruned_loss=0.05893, over 984667.49 frames.], datatang_tot_loss[loss=0.1983, simple_loss=0.2578, pruned_loss=0.06947, over 985974.47 frames.], batch size: 52, lr: 1.05e-03
+2022-06-18 15:31:22,652 INFO [train.py:874] (2/4) Epoch 7, batch 3900, aishell_loss[loss=0.157, simple_loss=0.2241, pruned_loss=0.0449, over 4967.00 frames.], tot_loss[loss=0.1942, simple_loss=0.261, pruned_loss=0.06372, over 985385.37 frames.], batch size: 25, aishell_tot_loss[loss=0.1909, simple_loss=0.2645, pruned_loss=0.05867, over 984933.72 frames.], datatang_tot_loss[loss=0.1983, simple_loss=0.2578, pruned_loss=0.06942, over 986025.04 frames.], batch size: 25, lr: 1.04e-03
+2022-06-18 15:31:51,713 INFO [train.py:874] (2/4) Epoch 7, batch 3950, aishell_loss[loss=0.1949, simple_loss=0.2723, pruned_loss=0.05874, over 4855.00 frames.], tot_loss[loss=0.1945, simple_loss=0.2612, pruned_loss=0.06394, over 985307.75 frames.], batch size: 36, aishell_tot_loss[loss=0.1922, simple_loss=0.2657, pruned_loss=0.05935, over 984864.07 frames.], datatang_tot_loss[loss=0.1973, simple_loss=0.2568, pruned_loss=0.06891, over 986017.40 frames.], batch size: 36, lr: 1.04e-03
+2022-06-18 15:32:19,176 INFO [train.py:874] (2/4) Epoch 7, batch 4000, aishell_loss[loss=0.1495, simple_loss=0.2072, pruned_loss=0.04587, over 4965.00 frames.], tot_loss[loss=0.1933, simple_loss=0.2605, pruned_loss=0.06309, over 985383.13 frames.], batch size: 21, aishell_tot_loss[loss=0.191, simple_loss=0.2647, pruned_loss=0.05867, over 984838.98 frames.], datatang_tot_loss[loss=0.1973, simple_loss=0.2568, pruned_loss=0.06888, over 986128.15 frames.], batch size: 21, lr: 1.04e-03
+2022-06-18 15:32:19,177 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 15:32:35,473 INFO [train.py:914] (2/4) Epoch 7, validation: loss=0.1718, simple_loss=0.2543, pruned_loss=0.04468, over 1622729.00 frames.
+2022-06-18 15:33:04,684 INFO [train.py:874] (2/4) Epoch 7, batch 4050, datatang_loss[loss=0.153, simple_loss=0.2161, pruned_loss=0.04497, over 4967.00 frames.], tot_loss[loss=0.1924, simple_loss=0.2595, pruned_loss=0.06261, over 985527.74 frames.], batch size: 31, aishell_tot_loss[loss=0.1904, simple_loss=0.2643, pruned_loss=0.05825, over 984812.15 frames.], datatang_tot_loss[loss=0.1966, simple_loss=0.2561, pruned_loss=0.06861, over 986275.80 frames.], batch size: 31, lr: 1.04e-03
+2022-06-18 15:33:33,292 INFO [train.py:874] (2/4) Epoch 7, batch 4100, aishell_loss[loss=0.1779, simple_loss=0.2595, pruned_loss=0.04809, over 4922.00 frames.], tot_loss[loss=0.1932, simple_loss=0.2603, pruned_loss=0.06298, over 985346.92 frames.], batch size: 41, aishell_tot_loss[loss=0.1909, simple_loss=0.2646, pruned_loss=0.05861, over 984488.86 frames.], datatang_tot_loss[loss=0.1967, simple_loss=0.2564, pruned_loss=0.06845, over 986422.62 frames.], batch size: 41, lr: 1.04e-03
+2022-06-18 15:34:37,390 INFO [train.py:874] (2/4) Epoch 8, batch 50, datatang_loss[loss=0.1905, simple_loss=0.259, pruned_loss=0.06096, over 4923.00 frames.], tot_loss[loss=0.1871, simple_loss=0.2554, pruned_loss=0.05942, over 218407.09 frames.], batch size: 94, aishell_tot_loss[loss=0.1968, simple_loss=0.2711, pruned_loss=0.06128, over 102748.42 frames.], datatang_tot_loss[loss=0.1794, simple_loss=0.2428, pruned_loss=0.05798, over 129121.35 frames.], batch size: 94, lr: 9.97e-04
+2022-06-18 15:35:06,761 INFO [train.py:874] (2/4) Epoch 8, batch 100, aishell_loss[loss=0.191, simple_loss=0.2555, pruned_loss=0.06325, over 4965.00 frames.], tot_loss[loss=0.1884, simple_loss=0.2573, pruned_loss=0.05973, over 388490.19 frames.], batch size: 40, aishell_tot_loss[loss=0.1915, simple_loss=0.266, pruned_loss=0.0585, over 229552.29 frames.], datatang_tot_loss[loss=0.1845, simple_loss=0.2469, pruned_loss=0.06105, over 207180.27 frames.], batch size: 40, lr: 9.97e-04
+2022-06-18 15:35:37,007 INFO [train.py:874] (2/4) Epoch 8, batch 150, aishell_loss[loss=0.2029, simple_loss=0.2644, pruned_loss=0.07065, over 4967.00 frames.], tot_loss[loss=0.187, simple_loss=0.2556, pruned_loss=0.05924, over 520836.33 frames.], batch size: 31, aishell_tot_loss[loss=0.1924, simple_loss=0.2663, pruned_loss=0.05921, over 318571.27 frames.], datatang_tot_loss[loss=0.1813, simple_loss=0.2439, pruned_loss=0.05931, over 298862.57 frames.], batch size: 31, lr: 9.96e-04
+2022-06-18 15:36:06,931 INFO [train.py:874] (2/4) Epoch 8, batch 200, datatang_loss[loss=0.2054, simple_loss=0.2669, pruned_loss=0.07197, over 4921.00 frames.], tot_loss[loss=0.1873, simple_loss=0.2554, pruned_loss=0.05961, over 623547.47 frames.], batch size: 81, aishell_tot_loss[loss=0.191, simple_loss=0.2647, pruned_loss=0.05867, over 399797.31 frames.], datatang_tot_loss[loss=0.1832, simple_loss=0.2453, pruned_loss=0.06057, over 376662.56 frames.], batch size: 81, lr: 9.95e-04
+2022-06-18 15:36:36,912 INFO [train.py:874] (2/4) Epoch 8, batch 250, aishell_loss[loss=0.2477, simple_loss=0.3275, pruned_loss=0.08396, over 4910.00 frames.], tot_loss[loss=0.1886, simple_loss=0.2563, pruned_loss=0.06048, over 703885.50 frames.], batch size: 78, aishell_tot_loss[loss=0.1917, simple_loss=0.2655, pruned_loss=0.05898, over 463527.98 frames.], datatang_tot_loss[loss=0.1849, simple_loss=0.2466, pruned_loss=0.06165, over 453882.24 frames.], batch size: 78, lr: 9.94e-04
+2022-06-18 15:37:06,796 INFO [train.py:874] (2/4) Epoch 8, batch 300, datatang_loss[loss=0.1883, simple_loss=0.2585, pruned_loss=0.05907, over 4951.00 frames.], tot_loss[loss=0.19, simple_loss=0.2578, pruned_loss=0.06105, over 766601.16 frames.], batch size: 91, aishell_tot_loss[loss=0.191, simple_loss=0.2651, pruned_loss=0.0584, over 525311.51 frames.], datatang_tot_loss[loss=0.1879, simple_loss=0.2495, pruned_loss=0.06316, over 516490.62 frames.], batch size: 91, lr: 9.93e-04
+2022-06-18 15:37:37,431 INFO [train.py:874] (2/4) Epoch 8, batch 350, datatang_loss[loss=0.1758, simple_loss=0.2438, pruned_loss=0.05388, over 4864.00 frames.], tot_loss[loss=0.191, simple_loss=0.2582, pruned_loss=0.06184, over 814987.10 frames.], batch size: 23, aishell_tot_loss[loss=0.1925, simple_loss=0.2667, pruned_loss=0.05915, over 564879.83 frames.], datatang_tot_loss[loss=0.1882, simple_loss=0.2496, pruned_loss=0.06341, over 585988.63 frames.], batch size: 23, lr: 9.93e-04
+2022-06-18 15:38:07,250 INFO [train.py:874] (2/4) Epoch 8, batch 400, aishell_loss[loss=0.2097, simple_loss=0.278, pruned_loss=0.07064, over 4921.00 frames.], tot_loss[loss=0.1896, simple_loss=0.257, pruned_loss=0.06104, over 852560.16 frames.], batch size: 80, aishell_tot_loss[loss=0.1918, simple_loss=0.2661, pruned_loss=0.05879, over 610423.60 frames.], datatang_tot_loss[loss=0.1872, simple_loss=0.2488, pruned_loss=0.0628, over 636566.04 frames.], batch size: 80, lr: 9.92e-04
+2022-06-18 15:38:37,478 INFO [train.py:874] (2/4) Epoch 8, batch 450, aishell_loss[loss=0.1456, simple_loss=0.226, pruned_loss=0.03262, over 4852.00 frames.], tot_loss[loss=0.1902, simple_loss=0.2581, pruned_loss=0.06116, over 881979.35 frames.], batch size: 28, aishell_tot_loss[loss=0.1922, simple_loss=0.2666, pruned_loss=0.05893, over 662668.22 frames.], datatang_tot_loss[loss=0.1876, simple_loss=0.2492, pruned_loss=0.06306, over 669920.02 frames.], batch size: 28, lr: 9.91e-04
+2022-06-18 15:39:07,771 INFO [train.py:874] (2/4) Epoch 8, batch 500, aishell_loss[loss=0.1966, simple_loss=0.2804, pruned_loss=0.05641, over 4945.00 frames.], tot_loss[loss=0.1905, simple_loss=0.2585, pruned_loss=0.06129, over 905049.08 frames.], batch size: 56, aishell_tot_loss[loss=0.1917, simple_loss=0.2662, pruned_loss=0.05859, over 697704.48 frames.], datatang_tot_loss[loss=0.1888, simple_loss=0.2503, pruned_loss=0.06358, over 710100.71 frames.], batch size: 56, lr: 9.90e-04
+2022-06-18 15:39:37,949 INFO [train.py:874] (2/4) Epoch 8, batch 550, datatang_loss[loss=0.1817, simple_loss=0.2485, pruned_loss=0.05749, over 4924.00 frames.], tot_loss[loss=0.1919, simple_loss=0.2594, pruned_loss=0.06217, over 923457.67 frames.], batch size: 79, aishell_tot_loss[loss=0.1922, simple_loss=0.2666, pruned_loss=0.05896, over 730593.82 frames.], datatang_tot_loss[loss=0.1901, simple_loss=0.2514, pruned_loss=0.06443, over 744103.54 frames.], batch size: 79, lr: 9.89e-04
+2022-06-18 15:40:08,001 INFO [train.py:874] (2/4) Epoch 8, batch 600, aishell_loss[loss=0.1894, simple_loss=0.279, pruned_loss=0.04985, over 4939.00 frames.], tot_loss[loss=0.1933, simple_loss=0.2605, pruned_loss=0.06303, over 937111.55 frames.], batch size: 33, aishell_tot_loss[loss=0.1917, simple_loss=0.2664, pruned_loss=0.05849, over 753597.00 frames.], datatang_tot_loss[loss=0.1927, simple_loss=0.2535, pruned_loss=0.06593, over 778886.99 frames.], batch size: 33, lr: 9.88e-04
+2022-06-18 15:40:37,298 INFO [train.py:874] (2/4) Epoch 8, batch 650, aishell_loss[loss=0.2286, simple_loss=0.3055, pruned_loss=0.07581, over 4876.00 frames.], tot_loss[loss=0.1928, simple_loss=0.2602, pruned_loss=0.06269, over 948103.53 frames.], batch size: 42, aishell_tot_loss[loss=0.1907, simple_loss=0.2655, pruned_loss=0.05793, over 785041.67 frames.], datatang_tot_loss[loss=0.1935, simple_loss=0.254, pruned_loss=0.06649, over 799747.25 frames.], batch size: 42, lr: 9.88e-04
+2022-06-18 15:41:07,622 INFO [train.py:874] (2/4) Epoch 8, batch 700, datatang_loss[loss=0.1724, simple_loss=0.2426, pruned_loss=0.05105, over 4930.00 frames.], tot_loss[loss=0.1917, simple_loss=0.2592, pruned_loss=0.06207, over 956498.50 frames.], batch size: 79, aishell_tot_loss[loss=0.1901, simple_loss=0.2649, pruned_loss=0.05762, over 807904.60 frames.], datatang_tot_loss[loss=0.1929, simple_loss=0.2535, pruned_loss=0.06611, over 822410.48 frames.], batch size: 79, lr: 9.87e-04
+2022-06-18 15:41:37,806 INFO [train.py:874] (2/4) Epoch 8, batch 750, aishell_loss[loss=0.1846, simple_loss=0.2616, pruned_loss=0.0538, over 4974.00 frames.], tot_loss[loss=0.1907, simple_loss=0.2586, pruned_loss=0.06136, over 963344.78 frames.], batch size: 61, aishell_tot_loss[loss=0.1888, simple_loss=0.2639, pruned_loss=0.05681, over 828246.09 frames.], datatang_tot_loss[loss=0.193, simple_loss=0.2538, pruned_loss=0.06605, over 842551.37 frames.], batch size: 61, lr: 9.86e-04
+2022-06-18 15:42:08,048 INFO [train.py:874] (2/4) Epoch 8, batch 800, datatang_loss[loss=0.1882, simple_loss=0.2618, pruned_loss=0.05733, over 4958.00 frames.], tot_loss[loss=0.1908, simple_loss=0.2588, pruned_loss=0.06141, over 968172.16 frames.], batch size: 86, aishell_tot_loss[loss=0.1889, simple_loss=0.264, pruned_loss=0.05688, over 845845.47 frames.], datatang_tot_loss[loss=0.193, simple_loss=0.254, pruned_loss=0.06595, over 860152.39 frames.], batch size: 86, lr: 9.85e-04
+2022-06-18 15:42:37,767 INFO [train.py:874] (2/4) Epoch 8, batch 850, aishell_loss[loss=0.1539, simple_loss=0.2259, pruned_loss=0.0409, over 4958.00 frames.], tot_loss[loss=0.1911, simple_loss=0.2592, pruned_loss=0.06149, over 972067.44 frames.], batch size: 27, aishell_tot_loss[loss=0.1889, simple_loss=0.2641, pruned_loss=0.05689, over 862798.82 frames.], datatang_tot_loss[loss=0.1932, simple_loss=0.2543, pruned_loss=0.06608, over 874526.19 frames.], batch size: 27, lr: 9.84e-04
+2022-06-18 15:43:07,861 INFO [train.py:874] (2/4) Epoch 8, batch 900, datatang_loss[loss=0.1636, simple_loss=0.2321, pruned_loss=0.04756, over 4915.00 frames.], tot_loss[loss=0.1906, simple_loss=0.2585, pruned_loss=0.06136, over 974651.74 frames.], batch size: 81, aishell_tot_loss[loss=0.1886, simple_loss=0.2634, pruned_loss=0.05692, over 877907.65 frames.], datatang_tot_loss[loss=0.193, simple_loss=0.2541, pruned_loss=0.06598, over 886626.65 frames.], batch size: 81, lr: 9.84e-04
+2022-06-18 15:43:39,063 INFO [train.py:874] (2/4) Epoch 8, batch 950, datatang_loss[loss=0.1965, simple_loss=0.244, pruned_loss=0.07448, over 4956.00 frames.], tot_loss[loss=0.1904, simple_loss=0.2579, pruned_loss=0.06145, over 977183.70 frames.], batch size: 45, aishell_tot_loss[loss=0.1882, simple_loss=0.2629, pruned_loss=0.05678, over 886173.68 frames.], datatang_tot_loss[loss=0.1929, simple_loss=0.2542, pruned_loss=0.06578, over 902267.63 frames.], batch size: 45, lr: 9.83e-04
+2022-06-18 15:44:08,431 INFO [train.py:874] (2/4) Epoch 8, batch 1000, datatang_loss[loss=0.1768, simple_loss=0.2374, pruned_loss=0.05814, over 4947.00 frames.], tot_loss[loss=0.1903, simple_loss=0.2579, pruned_loss=0.06139, over 979118.30 frames.], batch size: 37, aishell_tot_loss[loss=0.1885, simple_loss=0.2632, pruned_loss=0.05687, over 898692.07 frames.], datatang_tot_loss[loss=0.1926, simple_loss=0.2537, pruned_loss=0.06577, over 911482.55 frames.], batch size: 37, lr: 9.82e-04
+2022-06-18 15:44:08,432 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 15:44:24,590 INFO [train.py:914] (2/4) Epoch 8, validation: loss=0.1735, simple_loss=0.2544, pruned_loss=0.04634, over 1622729.00 frames.
+2022-06-18 15:44:54,414 INFO [train.py:874] (2/4) Epoch 8, batch 1050, datatang_loss[loss=0.1683, simple_loss=0.2322, pruned_loss=0.05219, over 4959.00 frames.], tot_loss[loss=0.19, simple_loss=0.2578, pruned_loss=0.06112, over 980557.48 frames.], batch size: 34, aishell_tot_loss[loss=0.1887, simple_loss=0.2635, pruned_loss=0.0569, over 908891.19 frames.], datatang_tot_loss[loss=0.192, simple_loss=0.2532, pruned_loss=0.06544, over 920246.68 frames.], batch size: 34, lr: 9.81e-04
+2022-06-18 15:45:25,068 INFO [train.py:874] (2/4) Epoch 8, batch 1100, aishell_loss[loss=0.1638, simple_loss=0.2383, pruned_loss=0.04463, over 4802.00 frames.], tot_loss[loss=0.1892, simple_loss=0.2574, pruned_loss=0.06056, over 981158.14 frames.], batch size: 26, aishell_tot_loss[loss=0.1881, simple_loss=0.2632, pruned_loss=0.05648, over 918489.01 frames.], datatang_tot_loss[loss=0.1917, simple_loss=0.2527, pruned_loss=0.06536, over 926979.81 frames.], batch size: 26, lr: 9.80e-04
+2022-06-18 15:45:54,909 INFO [train.py:874] (2/4) Epoch 8, batch 1150, datatang_loss[loss=0.1944, simple_loss=0.2586, pruned_loss=0.06512, over 4913.00 frames.], tot_loss[loss=0.1894, simple_loss=0.2577, pruned_loss=0.06054, over 982504.35 frames.], batch size: 71, aishell_tot_loss[loss=0.1882, simple_loss=0.2636, pruned_loss=0.05643, over 926237.89 frames.], datatang_tot_loss[loss=0.1916, simple_loss=0.2526, pruned_loss=0.06526, over 934383.85 frames.], batch size: 71, lr: 9.80e-04
+2022-06-18 15:46:24,544 INFO [train.py:874] (2/4) Epoch 8, batch 1200, aishell_loss[loss=0.2053, simple_loss=0.2891, pruned_loss=0.06073, over 4926.00 frames.], tot_loss[loss=0.1901, simple_loss=0.2586, pruned_loss=0.06081, over 983271.27 frames.], batch size: 78, aishell_tot_loss[loss=0.1888, simple_loss=0.264, pruned_loss=0.05683, over 934039.22 frames.], datatang_tot_loss[loss=0.1917, simple_loss=0.253, pruned_loss=0.06524, over 939786.09 frames.], batch size: 78, lr: 9.79e-04
+2022-06-18 15:46:55,064 INFO [train.py:874] (2/4) Epoch 8, batch 1250, datatang_loss[loss=0.1908, simple_loss=0.2496, pruned_loss=0.066, over 4974.00 frames.], tot_loss[loss=0.1904, simple_loss=0.2589, pruned_loss=0.06092, over 984063.65 frames.], batch size: 53, aishell_tot_loss[loss=0.1889, simple_loss=0.2641, pruned_loss=0.05689, over 940934.63 frames.], datatang_tot_loss[loss=0.1918, simple_loss=0.253, pruned_loss=0.06533, over 944712.64 frames.], batch size: 53, lr: 9.78e-04
+2022-06-18 15:47:24,851 INFO [train.py:874] (2/4) Epoch 8, batch 1300, aishell_loss[loss=0.1759, simple_loss=0.2639, pruned_loss=0.04395, over 4947.00 frames.], tot_loss[loss=0.1909, simple_loss=0.2594, pruned_loss=0.0612, over 984319.89 frames.], batch size: 56, aishell_tot_loss[loss=0.1892, simple_loss=0.2645, pruned_loss=0.05702, over 946249.69 frames.], datatang_tot_loss[loss=0.1921, simple_loss=0.2531, pruned_loss=0.06552, over 949405.87 frames.], batch size: 56, lr: 9.77e-04
+2022-06-18 15:47:55,251 INFO [train.py:874] (2/4) Epoch 8, batch 1350, datatang_loss[loss=0.1908, simple_loss=0.2489, pruned_loss=0.06636, over 4968.00 frames.], tot_loss[loss=0.1911, simple_loss=0.2599, pruned_loss=0.06115, over 984823.37 frames.], batch size: 45, aishell_tot_loss[loss=0.1894, simple_loss=0.2651, pruned_loss=0.05686, over 951009.71 frames.], datatang_tot_loss[loss=0.1922, simple_loss=0.2533, pruned_loss=0.06556, over 953756.89 frames.], batch size: 45, lr: 9.76e-04
+2022-06-18 15:48:24,915 INFO [train.py:874] (2/4) Epoch 8, batch 1400, aishell_loss[loss=0.202, simple_loss=0.2893, pruned_loss=0.05729, over 4927.00 frames.], tot_loss[loss=0.19, simple_loss=0.2594, pruned_loss=0.06029, over 985358.40 frames.], batch size: 68, aishell_tot_loss[loss=0.1887, simple_loss=0.2646, pruned_loss=0.05642, over 956016.90 frames.], datatang_tot_loss[loss=0.1918, simple_loss=0.253, pruned_loss=0.06525, over 956994.63 frames.], batch size: 68, lr: 9.76e-04
+2022-06-18 15:48:55,374 INFO [train.py:874] (2/4) Epoch 8, batch 1450, aishell_loss[loss=0.1625, simple_loss=0.2378, pruned_loss=0.04354, over 4959.00 frames.], tot_loss[loss=0.1892, simple_loss=0.2588, pruned_loss=0.05976, over 985294.85 frames.], batch size: 31, aishell_tot_loss[loss=0.1883, simple_loss=0.2644, pruned_loss=0.05608, over 959543.56 frames.], datatang_tot_loss[loss=0.1913, simple_loss=0.2528, pruned_loss=0.0649, over 960188.58 frames.], batch size: 31, lr: 9.75e-04
+2022-06-18 15:49:25,924 INFO [train.py:874] (2/4) Epoch 8, batch 1500, datatang_loss[loss=0.2003, simple_loss=0.2506, pruned_loss=0.07501, over 4924.00 frames.], tot_loss[loss=0.1893, simple_loss=0.2587, pruned_loss=0.05994, over 985334.78 frames.], batch size: 71, aishell_tot_loss[loss=0.1876, simple_loss=0.2638, pruned_loss=0.05574, over 962631.92 frames.], datatang_tot_loss[loss=0.192, simple_loss=0.2533, pruned_loss=0.06528, over 963130.76 frames.], batch size: 71, lr: 9.74e-04
+2022-06-18 15:49:55,559 INFO [train.py:874] (2/4) Epoch 8, batch 1550, datatang_loss[loss=0.1694, simple_loss=0.2321, pruned_loss=0.05332, over 4940.00 frames.], tot_loss[loss=0.1891, simple_loss=0.2583, pruned_loss=0.06001, over 985695.77 frames.], batch size: 62, aishell_tot_loss[loss=0.1874, simple_loss=0.2636, pruned_loss=0.05562, over 964439.25 frames.], datatang_tot_loss[loss=0.1917, simple_loss=0.2534, pruned_loss=0.06499, over 966929.07 frames.], batch size: 62, lr: 9.73e-04
+2022-06-18 15:50:25,888 INFO [train.py:874] (2/4) Epoch 8, batch 1600, datatang_loss[loss=0.1804, simple_loss=0.2431, pruned_loss=0.05886, over 4915.00 frames.], tot_loss[loss=0.1889, simple_loss=0.2577, pruned_loss=0.06003, over 985859.43 frames.], batch size: 64, aishell_tot_loss[loss=0.1866, simple_loss=0.2623, pruned_loss=0.05551, over 967701.53 frames.], datatang_tot_loss[loss=0.1922, simple_loss=0.2537, pruned_loss=0.06532, over 968597.66 frames.], batch size: 64, lr: 9.73e-04
+2022-06-18 15:50:55,445 INFO [train.py:874] (2/4) Epoch 8, batch 1650, aishell_loss[loss=0.2058, simple_loss=0.2812, pruned_loss=0.06515, over 4974.00 frames.], tot_loss[loss=0.1904, simple_loss=0.2592, pruned_loss=0.06077, over 985666.08 frames.], batch size: 44, aishell_tot_loss[loss=0.1876, simple_loss=0.2633, pruned_loss=0.05602, over 969702.97 frames.], datatang_tot_loss[loss=0.1927, simple_loss=0.2543, pruned_loss=0.06555, over 970558.71 frames.], batch size: 44, lr: 9.72e-04
+2022-06-18 15:51:24,656 INFO [train.py:874] (2/4) Epoch 8, batch 1700, datatang_loss[loss=0.1816, simple_loss=0.2653, pruned_loss=0.04894, over 4934.00 frames.], tot_loss[loss=0.1895, simple_loss=0.2587, pruned_loss=0.06017, over 985488.99 frames.], batch size: 42, aishell_tot_loss[loss=0.1874, simple_loss=0.2627, pruned_loss=0.05601, over 971889.22 frames.], datatang_tot_loss[loss=0.1922, simple_loss=0.2542, pruned_loss=0.06509, over 971831.12 frames.], batch size: 42, lr: 9.71e-04
+2022-06-18 15:51:53,607 INFO [train.py:874] (2/4) Epoch 8, batch 1750, aishell_loss[loss=0.1713, simple_loss=0.2518, pruned_loss=0.04541, over 4867.00 frames.], tot_loss[loss=0.1887, simple_loss=0.2581, pruned_loss=0.05969, over 985066.66 frames.], batch size: 28, aishell_tot_loss[loss=0.1864, simple_loss=0.2619, pruned_loss=0.05552, over 973304.47 frames.], datatang_tot_loss[loss=0.1924, simple_loss=0.2542, pruned_loss=0.06525, over 973193.48 frames.], batch size: 28, lr: 9.70e-04
+2022-06-18 15:52:24,174 INFO [train.py:874] (2/4) Epoch 8, batch 1800, datatang_loss[loss=0.1861, simple_loss=0.2388, pruned_loss=0.06665, over 4969.00 frames.], tot_loss[loss=0.1895, simple_loss=0.2582, pruned_loss=0.0604, over 984809.83 frames.], batch size: 60, aishell_tot_loss[loss=0.1867, simple_loss=0.2621, pruned_loss=0.05564, over 974072.15 frames.], datatang_tot_loss[loss=0.1925, simple_loss=0.2544, pruned_loss=0.06536, over 974916.32 frames.], batch size: 60, lr: 9.69e-04
+2022-06-18 15:52:53,916 INFO [train.py:874] (2/4) Epoch 8, batch 1850, datatang_loss[loss=0.1888, simple_loss=0.2325, pruned_loss=0.07256, over 4875.00 frames.], tot_loss[loss=0.1889, simple_loss=0.2575, pruned_loss=0.06015, over 985149.69 frames.], batch size: 44, aishell_tot_loss[loss=0.1856, simple_loss=0.261, pruned_loss=0.05512, over 975574.55 frames.], datatang_tot_loss[loss=0.193, simple_loss=0.2546, pruned_loss=0.06574, over 976223.17 frames.], batch size: 44, lr: 9.69e-04
+2022-06-18 15:53:23,133 INFO [train.py:874] (2/4) Epoch 8, batch 1900, aishell_loss[loss=0.1508, simple_loss=0.2248, pruned_loss=0.0384, over 4816.00 frames.], tot_loss[loss=0.1882, simple_loss=0.2566, pruned_loss=0.0599, over 985053.42 frames.], batch size: 26, aishell_tot_loss[loss=0.186, simple_loss=0.2615, pruned_loss=0.05525, over 976313.77 frames.], datatang_tot_loss[loss=0.1917, simple_loss=0.2533, pruned_loss=0.06501, over 977557.85 frames.], batch size: 26, lr: 9.68e-04
+2022-06-18 15:53:54,910 INFO [train.py:874] (2/4) Epoch 8, batch 1950, aishell_loss[loss=0.1843, simple_loss=0.2689, pruned_loss=0.0499, over 4950.00 frames.], tot_loss[loss=0.1882, simple_loss=0.2564, pruned_loss=0.06004, over 985152.58 frames.], batch size: 58, aishell_tot_loss[loss=0.1855, simple_loss=0.2611, pruned_loss=0.05495, over 977073.92 frames.], datatang_tot_loss[loss=0.1919, simple_loss=0.2534, pruned_loss=0.06521, over 978787.08 frames.], batch size: 58, lr: 9.67e-04
+2022-06-18 15:54:24,707 INFO [train.py:874] (2/4) Epoch 8, batch 2000, datatang_loss[loss=0.173, simple_loss=0.2323, pruned_loss=0.05691, over 4925.00 frames.], tot_loss[loss=0.1893, simple_loss=0.2573, pruned_loss=0.06059, over 985205.95 frames.], batch size: 57, aishell_tot_loss[loss=0.1866, simple_loss=0.262, pruned_loss=0.05562, over 978025.10 frames.], datatang_tot_loss[loss=0.1918, simple_loss=0.2533, pruned_loss=0.06517, over 979601.52 frames.], batch size: 57, lr: 9.66e-04
+2022-06-18 15:54:24,707 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 15:54:40,500 INFO [train.py:914] (2/4) Epoch 8, validation: loss=0.1711, simple_loss=0.254, pruned_loss=0.04413, over 1622729.00 frames.
+2022-06-18 15:55:10,325 INFO [train.py:874] (2/4) Epoch 8, batch 2050, aishell_loss[loss=0.2036, simple_loss=0.2849, pruned_loss=0.06117, over 4958.00 frames.], tot_loss[loss=0.1903, simple_loss=0.2582, pruned_loss=0.0612, over 985499.05 frames.], batch size: 64, aishell_tot_loss[loss=0.1867, simple_loss=0.2621, pruned_loss=0.05564, over 979187.18 frames.], datatang_tot_loss[loss=0.1929, simple_loss=0.254, pruned_loss=0.06593, over 980271.75 frames.], batch size: 64, lr: 9.66e-04
+2022-06-18 15:55:40,579 INFO [train.py:874] (2/4) Epoch 8, batch 2100, datatang_loss[loss=0.1576, simple_loss=0.2128, pruned_loss=0.05119, over 4923.00 frames.], tot_loss[loss=0.1903, simple_loss=0.258, pruned_loss=0.06127, over 985761.17 frames.], batch size: 37, aishell_tot_loss[loss=0.1871, simple_loss=0.2624, pruned_loss=0.05586, over 979877.56 frames.], datatang_tot_loss[loss=0.1925, simple_loss=0.2536, pruned_loss=0.0657, over 981205.88 frames.], batch size: 37, lr: 9.65e-04
+2022-06-18 15:56:10,905 INFO [train.py:874] (2/4) Epoch 8, batch 2150, datatang_loss[loss=0.2668, simple_loss=0.3168, pruned_loss=0.1084, over 4950.00 frames.], tot_loss[loss=0.1913, simple_loss=0.2591, pruned_loss=0.06181, over 985737.64 frames.], batch size: 108, aishell_tot_loss[loss=0.1864, simple_loss=0.2619, pruned_loss=0.05545, over 980575.26 frames.], datatang_tot_loss[loss=0.1943, simple_loss=0.2552, pruned_loss=0.06672, over 981703.55 frames.], batch size: 108, lr: 9.64e-04
+2022-06-18 15:56:40,056 INFO [train.py:874] (2/4) Epoch 8, batch 2200, datatang_loss[loss=0.1678, simple_loss=0.2372, pruned_loss=0.04926, over 4927.00 frames.], tot_loss[loss=0.1907, simple_loss=0.2586, pruned_loss=0.06141, over 985833.81 frames.], batch size: 79, aishell_tot_loss[loss=0.1869, simple_loss=0.2626, pruned_loss=0.05566, over 981209.10 frames.], datatang_tot_loss[loss=0.1934, simple_loss=0.2543, pruned_loss=0.06622, over 982243.40 frames.], batch size: 79, lr: 9.63e-04
+2022-06-18 15:57:10,685 INFO [train.py:874] (2/4) Epoch 8, batch 2250, aishell_loss[loss=0.1916, simple_loss=0.267, pruned_loss=0.05811, over 4957.00 frames.], tot_loss[loss=0.1907, simple_loss=0.259, pruned_loss=0.06123, over 985471.24 frames.], batch size: 40, aishell_tot_loss[loss=0.1864, simple_loss=0.2621, pruned_loss=0.05535, over 981628.58 frames.], datatang_tot_loss[loss=0.1942, simple_loss=0.2552, pruned_loss=0.06661, over 982406.36 frames.], batch size: 40, lr: 9.63e-04
+2022-06-18 15:57:40,689 INFO [train.py:874] (2/4) Epoch 8, batch 2300, datatang_loss[loss=0.2438, simple_loss=0.2937, pruned_loss=0.09693, over 4963.00 frames.], tot_loss[loss=0.1909, simple_loss=0.2595, pruned_loss=0.06109, over 985678.01 frames.], batch size: 91, aishell_tot_loss[loss=0.1861, simple_loss=0.2622, pruned_loss=0.05499, over 982267.32 frames.], datatang_tot_loss[loss=0.1949, simple_loss=0.2556, pruned_loss=0.06707, over 982801.62 frames.], batch size: 91, lr: 9.62e-04
+2022-06-18 15:58:10,146 INFO [train.py:874] (2/4) Epoch 8, batch 2350, aishell_loss[loss=0.2164, simple_loss=0.287, pruned_loss=0.0729, over 4911.00 frames.], tot_loss[loss=0.1904, simple_loss=0.2596, pruned_loss=0.06061, over 985885.07 frames.], batch size: 41, aishell_tot_loss[loss=0.1862, simple_loss=0.2622, pruned_loss=0.05506, over 982902.02 frames.], datatang_tot_loss[loss=0.1947, simple_loss=0.2555, pruned_loss=0.06696, over 983152.90 frames.], batch size: 41, lr: 9.61e-04
+2022-06-18 15:58:39,603 INFO [train.py:874] (2/4) Epoch 8, batch 2400, datatang_loss[loss=0.1919, simple_loss=0.2578, pruned_loss=0.063, over 4941.00 frames.], tot_loss[loss=0.1897, simple_loss=0.259, pruned_loss=0.06015, over 985732.77 frames.], batch size: 88, aishell_tot_loss[loss=0.1851, simple_loss=0.2612, pruned_loss=0.0545, over 982837.44 frames.], datatang_tot_loss[loss=0.1951, simple_loss=0.2559, pruned_loss=0.06715, over 983712.45 frames.], batch size: 88, lr: 9.60e-04
+2022-06-18 15:59:09,432 INFO [train.py:874] (2/4) Epoch 8, batch 2450, aishell_loss[loss=0.1833, simple_loss=0.2617, pruned_loss=0.05239, over 4968.00 frames.], tot_loss[loss=0.1902, simple_loss=0.2596, pruned_loss=0.06034, over 985829.38 frames.], batch size: 61, aishell_tot_loss[loss=0.1855, simple_loss=0.2617, pruned_loss=0.05465, over 983355.47 frames.], datatang_tot_loss[loss=0.1954, simple_loss=0.256, pruned_loss=0.06745, over 983902.44 frames.], batch size: 61, lr: 9.60e-04
+2022-06-18 15:59:39,452 INFO [train.py:874] (2/4) Epoch 8, batch 2500, aishell_loss[loss=0.1414, simple_loss=0.2191, pruned_loss=0.03191, over 4971.00 frames.], tot_loss[loss=0.1896, simple_loss=0.2591, pruned_loss=0.06007, over 986126.06 frames.], batch size: 27, aishell_tot_loss[loss=0.1846, simple_loss=0.261, pruned_loss=0.05412, over 983770.75 frames.], datatang_tot_loss[loss=0.1957, simple_loss=0.2564, pruned_loss=0.06749, over 984312.68 frames.], batch size: 27, lr: 9.59e-04
+2022-06-18 16:00:09,549 INFO [train.py:874] (2/4) Epoch 8, batch 2550, aishell_loss[loss=0.1981, simple_loss=0.276, pruned_loss=0.06004, over 4962.00 frames.], tot_loss[loss=0.1892, simple_loss=0.2582, pruned_loss=0.06009, over 986150.58 frames.], batch size: 40, aishell_tot_loss[loss=0.1847, simple_loss=0.261, pruned_loss=0.0542, over 984039.88 frames.], datatang_tot_loss[loss=0.1946, simple_loss=0.2557, pruned_loss=0.06672, over 984546.41 frames.], batch size: 40, lr: 9.58e-04
+2022-06-18 16:00:41,123 INFO [train.py:874] (2/4) Epoch 8, batch 2600, datatang_loss[loss=0.2174, simple_loss=0.2804, pruned_loss=0.07719, over 4955.00 frames.], tot_loss[loss=0.1892, simple_loss=0.2583, pruned_loss=0.06006, over 986477.40 frames.], batch size: 86, aishell_tot_loss[loss=0.1845, simple_loss=0.2611, pruned_loss=0.05397, over 984483.37 frames.], datatang_tot_loss[loss=0.1945, simple_loss=0.2557, pruned_loss=0.06664, over 984877.36 frames.], batch size: 86, lr: 9.57e-04
+2022-06-18 16:01:09,772 INFO [train.py:874] (2/4) Epoch 8, batch 2650, aishell_loss[loss=0.1921, simple_loss=0.2728, pruned_loss=0.0557, over 4971.00 frames.], tot_loss[loss=0.1904, simple_loss=0.259, pruned_loss=0.06091, over 986268.75 frames.], batch size: 44, aishell_tot_loss[loss=0.185, simple_loss=0.2615, pruned_loss=0.05429, over 984474.26 frames.], datatang_tot_loss[loss=0.1952, simple_loss=0.2561, pruned_loss=0.06711, over 985077.91 frames.], batch size: 44, lr: 9.57e-04
+2022-06-18 16:01:39,868 INFO [train.py:874] (2/4) Epoch 8, batch 2700, datatang_loss[loss=0.1558, simple_loss=0.2219, pruned_loss=0.04489, over 4925.00 frames.], tot_loss[loss=0.1904, simple_loss=0.259, pruned_loss=0.06089, over 986048.39 frames.], batch size: 73, aishell_tot_loss[loss=0.1853, simple_loss=0.2619, pruned_loss=0.05435, over 984619.05 frames.], datatang_tot_loss[loss=0.1951, simple_loss=0.2556, pruned_loss=0.06727, over 985060.64 frames.], batch size: 73, lr: 9.56e-04
+2022-06-18 16:02:09,342 INFO [train.py:874] (2/4) Epoch 8, batch 2750, datatang_loss[loss=0.2129, simple_loss=0.2697, pruned_loss=0.07807, over 4949.00 frames.], tot_loss[loss=0.1899, simple_loss=0.2583, pruned_loss=0.06071, over 985936.18 frames.], batch size: 99, aishell_tot_loss[loss=0.1855, simple_loss=0.2616, pruned_loss=0.05466, over 984686.90 frames.], datatang_tot_loss[loss=0.1945, simple_loss=0.2552, pruned_loss=0.0669, over 985167.62 frames.], batch size: 99, lr: 9.55e-04
+2022-06-18 16:02:39,819 INFO [train.py:874] (2/4) Epoch 8, batch 2800, datatang_loss[loss=0.1808, simple_loss=0.2448, pruned_loss=0.05839, over 4953.00 frames.], tot_loss[loss=0.1893, simple_loss=0.2578, pruned_loss=0.06038, over 985893.84 frames.], batch size: 86, aishell_tot_loss[loss=0.1856, simple_loss=0.2617, pruned_loss=0.05477, over 984764.28 frames.], datatang_tot_loss[loss=0.1937, simple_loss=0.2546, pruned_loss=0.06642, over 985289.48 frames.], batch size: 86, lr: 9.54e-04
+2022-06-18 16:03:10,493 INFO [train.py:874] (2/4) Epoch 8, batch 2850, aishell_loss[loss=0.195, simple_loss=0.2797, pruned_loss=0.05518, over 4942.00 frames.], tot_loss[loss=0.1886, simple_loss=0.2578, pruned_loss=0.05971, over 986069.52 frames.], batch size: 58, aishell_tot_loss[loss=0.1855, simple_loss=0.2619, pruned_loss=0.05454, over 985177.51 frames.], datatang_tot_loss[loss=0.193, simple_loss=0.2542, pruned_loss=0.06592, over 985276.93 frames.], batch size: 58, lr: 9.54e-04
+2022-06-18 16:03:39,755 INFO [train.py:874] (2/4) Epoch 8, batch 2900, aishell_loss[loss=0.1918, simple_loss=0.2648, pruned_loss=0.05937, over 4980.00 frames.], tot_loss[loss=0.1898, simple_loss=0.2585, pruned_loss=0.0606, over 986446.37 frames.], batch size: 44, aishell_tot_loss[loss=0.1857, simple_loss=0.2621, pruned_loss=0.05469, over 985261.56 frames.], datatang_tot_loss[loss=0.1938, simple_loss=0.2548, pruned_loss=0.06642, over 985769.47 frames.], batch size: 44, lr: 9.53e-04
+2022-06-18 16:04:10,428 INFO [train.py:874] (2/4) Epoch 8, batch 2950, aishell_loss[loss=0.2261, simple_loss=0.2999, pruned_loss=0.07616, over 4970.00 frames.], tot_loss[loss=0.1902, simple_loss=0.2591, pruned_loss=0.06064, over 986254.85 frames.], batch size: 44, aishell_tot_loss[loss=0.1863, simple_loss=0.2627, pruned_loss=0.05494, over 985323.26 frames.], datatang_tot_loss[loss=0.1937, simple_loss=0.2548, pruned_loss=0.06636, over 985724.13 frames.], batch size: 44, lr: 9.52e-04
+2022-06-18 16:04:40,903 INFO [train.py:874] (2/4) Epoch 8, batch 3000, aishell_loss[loss=0.2111, simple_loss=0.2796, pruned_loss=0.07129, over 4922.00 frames.], tot_loss[loss=0.1883, simple_loss=0.2572, pruned_loss=0.0597, over 986094.15 frames.], batch size: 41, aishell_tot_loss[loss=0.1855, simple_loss=0.2618, pruned_loss=0.05459, over 985439.40 frames.], datatang_tot_loss[loss=0.1925, simple_loss=0.2537, pruned_loss=0.06565, over 985606.03 frames.], batch size: 41, lr: 9.52e-04
+2022-06-18 16:04:40,904 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 16:04:57,746 INFO [train.py:914] (2/4) Epoch 8, validation: loss=0.1712, simple_loss=0.2536, pruned_loss=0.04441, over 1622729.00 frames.
+2022-06-18 16:05:32,050 INFO [train.py:874] (2/4) Epoch 8, batch 3050, datatang_loss[loss=0.2138, simple_loss=0.2728, pruned_loss=0.07736, over 4920.00 frames.], tot_loss[loss=0.1888, simple_loss=0.2576, pruned_loss=0.05998, over 985753.94 frames.], batch size: 42, aishell_tot_loss[loss=0.1852, simple_loss=0.2616, pruned_loss=0.05438, over 985308.56 frames.], datatang_tot_loss[loss=0.1931, simple_loss=0.2544, pruned_loss=0.06596, over 985527.79 frames.], batch size: 42, lr: 9.51e-04
+2022-06-18 16:06:02,993 INFO [train.py:874] (2/4) Epoch 8, batch 3100, aishell_loss[loss=0.1824, simple_loss=0.248, pruned_loss=0.05844, over 4956.00 frames.], tot_loss[loss=0.1874, simple_loss=0.2564, pruned_loss=0.05925, over 985746.58 frames.], batch size: 31, aishell_tot_loss[loss=0.1849, simple_loss=0.2614, pruned_loss=0.05426, over 985352.13 frames.], datatang_tot_loss[loss=0.1918, simple_loss=0.2533, pruned_loss=0.06509, over 985560.72 frames.], batch size: 31, lr: 9.50e-04
+2022-06-18 16:06:31,628 INFO [train.py:874] (2/4) Epoch 8, batch 3150, datatang_loss[loss=0.2159, simple_loss=0.2654, pruned_loss=0.08313, over 4924.00 frames.], tot_loss[loss=0.1868, simple_loss=0.2563, pruned_loss=0.05863, over 986057.54 frames.], batch size: 71, aishell_tot_loss[loss=0.1849, simple_loss=0.2615, pruned_loss=0.05416, over 985473.01 frames.], datatang_tot_loss[loss=0.1909, simple_loss=0.2527, pruned_loss=0.06448, over 985839.20 frames.], batch size: 71, lr: 9.49e-04
+2022-06-18 16:07:02,099 INFO [train.py:874] (2/4) Epoch 8, batch 3200, datatang_loss[loss=0.1888, simple_loss=0.2576, pruned_loss=0.05999, over 4947.00 frames.], tot_loss[loss=0.1855, simple_loss=0.255, pruned_loss=0.05797, over 985621.99 frames.], batch size: 69, aishell_tot_loss[loss=0.1838, simple_loss=0.2604, pruned_loss=0.05355, over 985077.40 frames.], datatang_tot_loss[loss=0.1902, simple_loss=0.2524, pruned_loss=0.06402, over 985880.82 frames.], batch size: 69, lr: 9.49e-04
+2022-06-18 16:07:32,560 INFO [train.py:874] (2/4) Epoch 8, batch 3250, datatang_loss[loss=0.1542, simple_loss=0.2287, pruned_loss=0.03988, over 4929.00 frames.], tot_loss[loss=0.1859, simple_loss=0.2557, pruned_loss=0.05806, over 985644.33 frames.], batch size: 71, aishell_tot_loss[loss=0.1846, simple_loss=0.2616, pruned_loss=0.05378, over 985013.37 frames.], datatang_tot_loss[loss=0.1897, simple_loss=0.2517, pruned_loss=0.06383, over 986018.99 frames.], batch size: 71, lr: 9.48e-04
+2022-06-18 16:08:02,003 INFO [train.py:874] (2/4) Epoch 8, batch 3300, datatang_loss[loss=0.1976, simple_loss=0.2641, pruned_loss=0.06552, over 4930.00 frames.], tot_loss[loss=0.1863, simple_loss=0.2561, pruned_loss=0.05827, over 985957.60 frames.], batch size: 57, aishell_tot_loss[loss=0.1851, simple_loss=0.2622, pruned_loss=0.05406, over 985360.57 frames.], datatang_tot_loss[loss=0.1893, simple_loss=0.2513, pruned_loss=0.0636, over 986025.02 frames.], batch size: 57, lr: 9.47e-04
+2022-06-18 16:08:32,930 INFO [train.py:874] (2/4) Epoch 8, batch 3350, aishell_loss[loss=0.2028, simple_loss=0.2818, pruned_loss=0.06193, over 4952.00 frames.], tot_loss[loss=0.187, simple_loss=0.2563, pruned_loss=0.05883, over 985876.18 frames.], batch size: 64, aishell_tot_loss[loss=0.1853, simple_loss=0.2623, pruned_loss=0.05418, over 985470.82 frames.], datatang_tot_loss[loss=0.1894, simple_loss=0.2514, pruned_loss=0.06369, over 985871.27 frames.], batch size: 64, lr: 9.46e-04
+2022-06-18 16:09:02,960 INFO [train.py:874] (2/4) Epoch 8, batch 3400, aishell_loss[loss=0.1804, simple_loss=0.254, pruned_loss=0.05341, over 4847.00 frames.], tot_loss[loss=0.187, simple_loss=0.2566, pruned_loss=0.0587, over 985984.01 frames.], batch size: 35, aishell_tot_loss[loss=0.1856, simple_loss=0.2626, pruned_loss=0.05426, over 985551.29 frames.], datatang_tot_loss[loss=0.1891, simple_loss=0.2514, pruned_loss=0.06337, over 985959.57 frames.], batch size: 35, lr: 9.46e-04
+2022-06-18 16:09:31,651 INFO [train.py:874] (2/4) Epoch 8, batch 3450, datatang_loss[loss=0.2294, simple_loss=0.2807, pruned_loss=0.08905, over 4925.00 frames.], tot_loss[loss=0.1863, simple_loss=0.2561, pruned_loss=0.05821, over 985947.48 frames.], batch size: 94, aishell_tot_loss[loss=0.185, simple_loss=0.2619, pruned_loss=0.05399, over 985423.00 frames.], datatang_tot_loss[loss=0.1888, simple_loss=0.2513, pruned_loss=0.06318, over 986110.51 frames.], batch size: 94, lr: 9.45e-04
+2022-06-18 16:10:01,621 INFO [train.py:874] (2/4) Epoch 8, batch 3500, datatang_loss[loss=0.1827, simple_loss=0.2486, pruned_loss=0.05846, over 4902.00 frames.], tot_loss[loss=0.1859, simple_loss=0.2562, pruned_loss=0.05785, over 986041.29 frames.], batch size: 47, aishell_tot_loss[loss=0.1847, simple_loss=0.2618, pruned_loss=0.05384, over 985586.06 frames.], datatang_tot_loss[loss=0.1886, simple_loss=0.2513, pruned_loss=0.06297, over 986117.08 frames.], batch size: 47, lr: 9.44e-04
+2022-06-18 16:10:31,541 INFO [train.py:874] (2/4) Epoch 8, batch 3550, datatang_loss[loss=0.1863, simple_loss=0.2532, pruned_loss=0.05969, over 4938.00 frames.], tot_loss[loss=0.1858, simple_loss=0.256, pruned_loss=0.05781, over 986066.96 frames.], batch size: 50, aishell_tot_loss[loss=0.1846, simple_loss=0.2616, pruned_loss=0.05384, over 985675.10 frames.], datatang_tot_loss[loss=0.1883, simple_loss=0.2512, pruned_loss=0.06274, over 986096.83 frames.], batch size: 50, lr: 9.44e-04
+2022-06-18 16:11:02,116 INFO [train.py:874] (2/4) Epoch 8, batch 3600, datatang_loss[loss=0.2501, simple_loss=0.2814, pruned_loss=0.1094, over 4933.00 frames.], tot_loss[loss=0.1866, simple_loss=0.2564, pruned_loss=0.05845, over 985907.65 frames.], batch size: 73, aishell_tot_loss[loss=0.1839, simple_loss=0.2611, pruned_loss=0.05337, over 985793.49 frames.], datatang_tot_loss[loss=0.1896, simple_loss=0.2522, pruned_loss=0.06355, over 985856.45 frames.], batch size: 73, lr: 9.43e-04
+2022-06-18 16:11:32,843 INFO [train.py:874] (2/4) Epoch 8, batch 3650, datatang_loss[loss=0.1974, simple_loss=0.2433, pruned_loss=0.07577, over 4857.00 frames.], tot_loss[loss=0.1872, simple_loss=0.2569, pruned_loss=0.0587, over 985792.34 frames.], batch size: 24, aishell_tot_loss[loss=0.1849, simple_loss=0.2619, pruned_loss=0.05392, over 985699.50 frames.], datatang_tot_loss[loss=0.1892, simple_loss=0.2519, pruned_loss=0.06326, over 985857.99 frames.], batch size: 24, lr: 9.42e-04
+2022-06-18 16:12:03,746 INFO [train.py:874] (2/4) Epoch 8, batch 3700, aishell_loss[loss=0.169, simple_loss=0.2461, pruned_loss=0.04589, over 4980.00 frames.], tot_loss[loss=0.1881, simple_loss=0.2576, pruned_loss=0.0593, over 985542.33 frames.], batch size: 51, aishell_tot_loss[loss=0.1847, simple_loss=0.2619, pruned_loss=0.05373, over 985501.41 frames.], datatang_tot_loss[loss=0.1903, simple_loss=0.2529, pruned_loss=0.06387, over 985801.60 frames.], batch size: 51, lr: 9.42e-04
+2022-06-18 16:12:32,577 INFO [train.py:874] (2/4) Epoch 8, batch 3750, datatang_loss[loss=0.2101, simple_loss=0.2714, pruned_loss=0.07436, over 4954.00 frames.], tot_loss[loss=0.1877, simple_loss=0.2574, pruned_loss=0.059, over 985514.13 frames.], batch size: 50, aishell_tot_loss[loss=0.1843, simple_loss=0.2613, pruned_loss=0.0536, over 985484.03 frames.], datatang_tot_loss[loss=0.1905, simple_loss=0.2532, pruned_loss=0.06393, over 985770.16 frames.], batch size: 50, lr: 9.41e-04
+2022-06-18 16:13:02,545 INFO [train.py:874] (2/4) Epoch 8, batch 3800, datatang_loss[loss=0.1793, simple_loss=0.2473, pruned_loss=0.05564, over 4963.00 frames.], tot_loss[loss=0.1878, simple_loss=0.2572, pruned_loss=0.05917, over 985412.05 frames.], batch size: 67, aishell_tot_loss[loss=0.1852, simple_loss=0.2621, pruned_loss=0.05415, over 985387.43 frames.], datatang_tot_loss[loss=0.1897, simple_loss=0.2524, pruned_loss=0.06356, over 985733.24 frames.], batch size: 67, lr: 9.40e-04
+2022-06-18 16:13:31,932 INFO [train.py:874] (2/4) Epoch 8, batch 3850, aishell_loss[loss=0.1519, simple_loss=0.2307, pruned_loss=0.03649, over 4904.00 frames.], tot_loss[loss=0.1876, simple_loss=0.2572, pruned_loss=0.05906, over 985266.32 frames.], batch size: 28, aishell_tot_loss[loss=0.1856, simple_loss=0.2622, pruned_loss=0.05448, over 985298.80 frames.], datatang_tot_loss[loss=0.1892, simple_loss=0.2521, pruned_loss=0.06318, over 985629.36 frames.], batch size: 28, lr: 9.39e-04
+2022-06-18 16:14:00,683 INFO [train.py:874] (2/4) Epoch 8, batch 3900, aishell_loss[loss=0.1844, simple_loss=0.2628, pruned_loss=0.05297, over 4925.00 frames.], tot_loss[loss=0.1871, simple_loss=0.2572, pruned_loss=0.0585, over 985637.40 frames.], batch size: 49, aishell_tot_loss[loss=0.1851, simple_loss=0.2621, pruned_loss=0.05403, over 985680.83 frames.], datatang_tot_loss[loss=0.1894, simple_loss=0.2521, pruned_loss=0.06331, over 985594.95 frames.], batch size: 49, lr: 9.39e-04
+2022-06-18 16:14:30,200 INFO [train.py:874] (2/4) Epoch 8, batch 3950, datatang_loss[loss=0.1699, simple_loss=0.2355, pruned_loss=0.05212, over 4947.00 frames.], tot_loss[loss=0.1849, simple_loss=0.255, pruned_loss=0.05737, over 985531.05 frames.], batch size: 25, aishell_tot_loss[loss=0.1842, simple_loss=0.2612, pruned_loss=0.05357, over 985472.10 frames.], datatang_tot_loss[loss=0.1878, simple_loss=0.251, pruned_loss=0.06227, over 985697.13 frames.], batch size: 25, lr: 9.38e-04
+2022-06-18 16:14:59,691 INFO [train.py:874] (2/4) Epoch 8, batch 4000, datatang_loss[loss=0.176, simple_loss=0.2478, pruned_loss=0.05214, over 4919.00 frames.], tot_loss[loss=0.1856, simple_loss=0.2556, pruned_loss=0.05779, over 985756.46 frames.], batch size: 83, aishell_tot_loss[loss=0.1845, simple_loss=0.2614, pruned_loss=0.05381, over 985636.78 frames.], datatang_tot_loss[loss=0.1879, simple_loss=0.2512, pruned_loss=0.06227, over 985755.62 frames.], batch size: 83, lr: 9.37e-04
+2022-06-18 16:14:59,692 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 16:15:16,495 INFO [train.py:914] (2/4) Epoch 8, validation: loss=0.1678, simple_loss=0.2518, pruned_loss=0.0419, over 1622729.00 frames.
+2022-06-18 16:15:46,523 INFO [train.py:874] (2/4) Epoch 8, batch 4050, datatang_loss[loss=0.1561, simple_loss=0.2198, pruned_loss=0.04613, over 4914.00 frames.], tot_loss[loss=0.1863, simple_loss=0.2562, pruned_loss=0.05822, over 985766.31 frames.], batch size: 57, aishell_tot_loss[loss=0.1858, simple_loss=0.2624, pruned_loss=0.05462, over 985668.71 frames.], datatang_tot_loss[loss=0.1872, simple_loss=0.2508, pruned_loss=0.06178, over 985757.38 frames.], batch size: 57, lr: 9.37e-04
+2022-06-18 16:16:15,518 INFO [train.py:874] (2/4) Epoch 8, batch 4100, datatang_loss[loss=0.1589, simple_loss=0.2302, pruned_loss=0.04382, over 4905.00 frames.], tot_loss[loss=0.1863, simple_loss=0.2559, pruned_loss=0.05834, over 986042.11 frames.], batch size: 85, aishell_tot_loss[loss=0.1861, simple_loss=0.2627, pruned_loss=0.05475, over 985729.05 frames.], datatang_tot_loss[loss=0.1868, simple_loss=0.2503, pruned_loss=0.06166, over 985997.16 frames.], batch size: 85, lr: 9.36e-04
+2022-06-18 16:16:43,398 INFO [train.py:874] (2/4) Epoch 8, batch 4150, aishell_loss[loss=0.1785, simple_loss=0.2572, pruned_loss=0.04993, over 4917.00 frames.], tot_loss[loss=0.1871, simple_loss=0.2564, pruned_loss=0.05894, over 985627.06 frames.], batch size: 41, aishell_tot_loss[loss=0.186, simple_loss=0.2623, pruned_loss=0.0548, over 985246.24 frames.], datatang_tot_loss[loss=0.1878, simple_loss=0.251, pruned_loss=0.06227, over 986074.06 frames.], batch size: 41, lr: 9.35e-04
+2022-06-18 16:18:01,442 INFO [train.py:874] (2/4) Epoch 9, batch 50, datatang_loss[loss=0.1562, simple_loss=0.2185, pruned_loss=0.04699, over 4925.00 frames.], tot_loss[loss=0.18, simple_loss=0.252, pruned_loss=0.05398, over 218498.23 frames.], batch size: 34, aishell_tot_loss[loss=0.1845, simple_loss=0.2623, pruned_loss=0.05331, over 133333.42 frames.], datatang_tot_loss[loss=0.1742, simple_loss=0.2383, pruned_loss=0.05502, over 98490.73 frames.], batch size: 34, lr: 8.97e-04
+2022-06-18 16:18:32,026 INFO [train.py:874] (2/4) Epoch 9, batch 100, datatang_loss[loss=0.1475, simple_loss=0.2182, pruned_loss=0.03837, over 4916.00 frames.], tot_loss[loss=0.1803, simple_loss=0.2511, pruned_loss=0.05472, over 388570.06 frames.], batch size: 57, aishell_tot_loss[loss=0.1864, simple_loss=0.2636, pruned_loss=0.05459, over 218461.74 frames.], datatang_tot_loss[loss=0.1746, simple_loss=0.2393, pruned_loss=0.05491, over 218549.05 frames.], batch size: 57, lr: 8.96e-04
+2022-06-18 16:19:01,238 INFO [train.py:874] (2/4) Epoch 9, batch 150, datatang_loss[loss=0.1518, simple_loss=0.2199, pruned_loss=0.04186, over 4925.00 frames.], tot_loss[loss=0.1786, simple_loss=0.2491, pruned_loss=0.05409, over 520820.02 frames.], batch size: 73, aishell_tot_loss[loss=0.1853, simple_loss=0.2626, pruned_loss=0.05404, over 294802.48 frames.], datatang_tot_loss[loss=0.1732, simple_loss=0.2379, pruned_loss=0.05429, over 322523.13 frames.], batch size: 73, lr: 8.96e-04
+2022-06-18 16:19:31,587 INFO [train.py:874] (2/4) Epoch 9, batch 200, datatang_loss[loss=0.1489, simple_loss=0.2217, pruned_loss=0.03807, over 4922.00 frames.], tot_loss[loss=0.1779, simple_loss=0.2491, pruned_loss=0.05331, over 623757.80 frames.], batch size: 73, aishell_tot_loss[loss=0.1825, simple_loss=0.2598, pruned_loss=0.05262, over 397011.56 frames.], datatang_tot_loss[loss=0.1735, simple_loss=0.2381, pruned_loss=0.05444, over 379791.60 frames.], batch size: 73, lr: 8.95e-04
+2022-06-18 16:20:02,107 INFO [train.py:874] (2/4) Epoch 9, batch 250, datatang_loss[loss=0.1477, simple_loss=0.2216, pruned_loss=0.03691, over 4925.00 frames.], tot_loss[loss=0.1774, simple_loss=0.2486, pruned_loss=0.05314, over 703941.62 frames.], batch size: 83, aishell_tot_loss[loss=0.1827, simple_loss=0.2598, pruned_loss=0.0528, over 466180.01 frames.], datatang_tot_loss[loss=0.1726, simple_loss=0.2373, pruned_loss=0.0539, over 451258.35 frames.], batch size: 83, lr: 8.94e-04
+2022-06-18 16:20:30,900 INFO [train.py:874] (2/4) Epoch 9, batch 300, datatang_loss[loss=0.1758, simple_loss=0.2459, pruned_loss=0.05285, over 4930.00 frames.], tot_loss[loss=0.1774, simple_loss=0.2485, pruned_loss=0.05316, over 766183.70 frames.], batch size: 50, aishell_tot_loss[loss=0.1821, simple_loss=0.2589, pruned_loss=0.05262, over 532031.09 frames.], datatang_tot_loss[loss=0.1729, simple_loss=0.2377, pruned_loss=0.05401, over 509147.03 frames.], batch size: 50, lr: 8.94e-04
+2022-06-18 16:21:02,306 INFO [train.py:874] (2/4) Epoch 9, batch 350, aishell_loss[loss=0.1298, simple_loss=0.1972, pruned_loss=0.0312, over 4889.00 frames.], tot_loss[loss=0.1771, simple_loss=0.2484, pruned_loss=0.05289, over 814820.81 frames.], batch size: 21, aishell_tot_loss[loss=0.1807, simple_loss=0.258, pruned_loss=0.05169, over 589300.51 frames.], datatang_tot_loss[loss=0.1736, simple_loss=0.2382, pruned_loss=0.05453, over 561178.69 frames.], batch size: 21, lr: 8.93e-04
+2022-06-18 16:21:31,254 INFO [train.py:874] (2/4) Epoch 9, batch 400, datatang_loss[loss=0.1816, simple_loss=0.2429, pruned_loss=0.06016, over 4922.00 frames.], tot_loss[loss=0.1779, simple_loss=0.2494, pruned_loss=0.05318, over 853000.41 frames.], batch size: 81, aishell_tot_loss[loss=0.1794, simple_loss=0.2567, pruned_loss=0.05111, over 649835.79 frames.], datatang_tot_loss[loss=0.1758, simple_loss=0.2402, pruned_loss=0.05574, over 596058.03 frames.], batch size: 81, lr: 8.92e-04
+2022-06-18 16:22:01,507 INFO [train.py:874] (2/4) Epoch 9, batch 450, datatang_loss[loss=0.1899, simple_loss=0.2503, pruned_loss=0.06474, over 4970.00 frames.], tot_loss[loss=0.1793, simple_loss=0.2507, pruned_loss=0.0539, over 881992.42 frames.], batch size: 37, aishell_tot_loss[loss=0.1804, simple_loss=0.2576, pruned_loss=0.05158, over 685955.45 frames.], datatang_tot_loss[loss=0.1769, simple_loss=0.2417, pruned_loss=0.05611, over 645389.65 frames.], batch size: 37, lr: 8.92e-04
+2022-06-18 16:22:32,026 INFO [train.py:874] (2/4) Epoch 9, batch 500, datatang_loss[loss=0.166, simple_loss=0.2352, pruned_loss=0.04837, over 4912.00 frames.], tot_loss[loss=0.1808, simple_loss=0.252, pruned_loss=0.05476, over 905189.40 frames.], batch size: 75, aishell_tot_loss[loss=0.18, simple_loss=0.2576, pruned_loss=0.05123, over 713152.24 frames.], datatang_tot_loss[loss=0.1796, simple_loss=0.2444, pruned_loss=0.05742, over 694595.28 frames.], batch size: 75, lr: 8.91e-04
+2022-06-18 16:23:02,676 INFO [train.py:874] (2/4) Epoch 9, batch 550, datatang_loss[loss=0.1647, simple_loss=0.243, pruned_loss=0.04324, over 4932.00 frames.], tot_loss[loss=0.1818, simple_loss=0.2528, pruned_loss=0.0554, over 923151.59 frames.], batch size: 79, aishell_tot_loss[loss=0.1806, simple_loss=0.2582, pruned_loss=0.05149, over 739168.88 frames.], datatang_tot_loss[loss=0.1807, simple_loss=0.2456, pruned_loss=0.05795, over 735352.88 frames.], batch size: 79, lr: 8.90e-04
+2022-06-18 16:23:32,309 INFO [train.py:874] (2/4) Epoch 9, batch 600, aishell_loss[loss=0.1787, simple_loss=0.2536, pruned_loss=0.05195, over 4923.00 frames.], tot_loss[loss=0.1824, simple_loss=0.253, pruned_loss=0.05591, over 936682.88 frames.], batch size: 41, aishell_tot_loss[loss=0.1812, simple_loss=0.2584, pruned_loss=0.052, over 770299.80 frames.], datatang_tot_loss[loss=0.1812, simple_loss=0.2457, pruned_loss=0.05839, over 762329.10 frames.], batch size: 41, lr: 8.90e-04
+2022-06-18 16:24:02,267 INFO [train.py:874] (2/4) Epoch 9, batch 650, datatang_loss[loss=0.1707, simple_loss=0.2337, pruned_loss=0.05386, over 4875.00 frames.], tot_loss[loss=0.1844, simple_loss=0.2548, pruned_loss=0.05698, over 947622.90 frames.], batch size: 39, aishell_tot_loss[loss=0.1818, simple_loss=0.259, pruned_loss=0.0523, over 796598.08 frames.], datatang_tot_loss[loss=0.1835, simple_loss=0.2477, pruned_loss=0.05969, over 787752.36 frames.], batch size: 39, lr: 8.89e-04
+2022-06-18 16:24:31,728 INFO [train.py:874] (2/4) Epoch 9, batch 700, aishell_loss[loss=0.1647, simple_loss=0.2448, pruned_loss=0.04233, over 4863.00 frames.], tot_loss[loss=0.184, simple_loss=0.2551, pruned_loss=0.05644, over 956192.45 frames.], batch size: 36, aishell_tot_loss[loss=0.1817, simple_loss=0.2591, pruned_loss=0.05216, over 821497.52 frames.], datatang_tot_loss[loss=0.1836, simple_loss=0.2481, pruned_loss=0.05956, over 808410.59 frames.], batch size: 36, lr: 8.88e-04
+2022-06-18 16:25:02,543 INFO [train.py:874] (2/4) Epoch 9, batch 750, aishell_loss[loss=0.1951, simple_loss=0.2658, pruned_loss=0.06223, over 4958.00 frames.], tot_loss[loss=0.1842, simple_loss=0.2556, pruned_loss=0.05645, over 962497.39 frames.], batch size: 61, aishell_tot_loss[loss=0.1827, simple_loss=0.2602, pruned_loss=0.05259, over 844189.97 frames.], datatang_tot_loss[loss=0.1834, simple_loss=0.2477, pruned_loss=0.05957, over 825303.85 frames.], batch size: 61, lr: 8.88e-04
+2022-06-18 16:25:33,512 INFO [train.py:874] (2/4) Epoch 9, batch 800, aishell_loss[loss=0.1737, simple_loss=0.257, pruned_loss=0.04518, over 4889.00 frames.], tot_loss[loss=0.1833, simple_loss=0.2547, pruned_loss=0.05589, over 967559.70 frames.], batch size: 34, aishell_tot_loss[loss=0.1817, simple_loss=0.2595, pruned_loss=0.05195, over 860808.88 frames.], datatang_tot_loss[loss=0.1836, simple_loss=0.2478, pruned_loss=0.05966, over 844180.04 frames.], batch size: 34, lr: 8.87e-04
+2022-06-18 16:26:02,654 INFO [train.py:874] (2/4) Epoch 9, batch 850, aishell_loss[loss=0.2299, simple_loss=0.2926, pruned_loss=0.08365, over 4915.00 frames.], tot_loss[loss=0.1847, simple_loss=0.2558, pruned_loss=0.05678, over 971440.21 frames.], batch size: 68, aishell_tot_loss[loss=0.1823, simple_loss=0.2602, pruned_loss=0.05222, over 874484.32 frames.], datatang_tot_loss[loss=0.1848, simple_loss=0.2489, pruned_loss=0.06037, over 861864.98 frames.], batch size: 68, lr: 8.87e-04
+2022-06-18 16:26:33,075 INFO [train.py:874] (2/4) Epoch 9, batch 900, datatang_loss[loss=0.2027, simple_loss=0.2684, pruned_loss=0.06851, over 4931.00 frames.], tot_loss[loss=0.1855, simple_loss=0.2567, pruned_loss=0.05714, over 974711.16 frames.], batch size: 94, aishell_tot_loss[loss=0.1834, simple_loss=0.2606, pruned_loss=0.05314, over 892800.47 frames.], datatang_tot_loss[loss=0.185, simple_loss=0.2492, pruned_loss=0.06045, over 870427.50 frames.], batch size: 94, lr: 8.86e-04
+2022-06-18 16:27:02,539 INFO [train.py:874] (2/4) Epoch 9, batch 950, aishell_loss[loss=0.1762, simple_loss=0.2596, pruned_loss=0.04637, over 4904.00 frames.], tot_loss[loss=0.1859, simple_loss=0.257, pruned_loss=0.05737, over 977149.39 frames.], batch size: 52, aishell_tot_loss[loss=0.1837, simple_loss=0.261, pruned_loss=0.05317, over 906130.02 frames.], datatang_tot_loss[loss=0.1856, simple_loss=0.2493, pruned_loss=0.06099, over 880928.84 frames.], batch size: 52, lr: 8.85e-04
+2022-06-18 16:27:32,411 INFO [train.py:874] (2/4) Epoch 9, batch 1000, datatang_loss[loss=0.1838, simple_loss=0.2503, pruned_loss=0.05868, over 4922.00 frames.], tot_loss[loss=0.1861, simple_loss=0.2572, pruned_loss=0.05756, over 978903.33 frames.], batch size: 77, aishell_tot_loss[loss=0.1835, simple_loss=0.261, pruned_loss=0.05303, over 916505.88 frames.], datatang_tot_loss[loss=0.1865, simple_loss=0.2498, pruned_loss=0.06157, over 891760.20 frames.], batch size: 77, lr: 8.85e-04
+2022-06-18 16:27:32,412 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 16:27:48,780 INFO [train.py:914] (2/4) Epoch 9, validation: loss=0.1688, simple_loss=0.2514, pruned_loss=0.04309, over 1622729.00 frames.
+2022-06-18 16:28:18,765 INFO [train.py:874] (2/4) Epoch 9, batch 1050, aishell_loss[loss=0.1857, simple_loss=0.2694, pruned_loss=0.05097, over 4932.00 frames.], tot_loss[loss=0.1856, simple_loss=0.2567, pruned_loss=0.05727, over 980533.75 frames.], batch size: 64, aishell_tot_loss[loss=0.1834, simple_loss=0.261, pruned_loss=0.05293, over 925044.07 frames.], datatang_tot_loss[loss=0.1863, simple_loss=0.2496, pruned_loss=0.06144, over 902448.35 frames.], batch size: 64, lr: 8.84e-04
+2022-06-18 16:28:49,626 INFO [train.py:874] (2/4) Epoch 9, batch 1100, datatang_loss[loss=0.1886, simple_loss=0.247, pruned_loss=0.06508, over 4946.00 frames.], tot_loss[loss=0.1849, simple_loss=0.2563, pruned_loss=0.05679, over 981530.03 frames.], batch size: 69, aishell_tot_loss[loss=0.183, simple_loss=0.2606, pruned_loss=0.05272, over 932089.02 frames.], datatang_tot_loss[loss=0.1861, simple_loss=0.2499, pruned_loss=0.06116, over 912086.44 frames.], batch size: 69, lr: 8.83e-04
+2022-06-18 16:29:19,118 INFO [train.py:874] (2/4) Epoch 9, batch 1150, datatang_loss[loss=0.1974, simple_loss=0.2586, pruned_loss=0.06812, over 4970.00 frames.], tot_loss[loss=0.1854, simple_loss=0.2568, pruned_loss=0.05701, over 982252.66 frames.], batch size: 45, aishell_tot_loss[loss=0.1833, simple_loss=0.2609, pruned_loss=0.05289, over 938370.04 frames.], datatang_tot_loss[loss=0.1865, simple_loss=0.2505, pruned_loss=0.06123, over 920564.75 frames.], batch size: 45, lr: 8.83e-04
+2022-06-18 16:29:49,953 INFO [train.py:874] (2/4) Epoch 9, batch 1200, aishell_loss[loss=0.1801, simple_loss=0.2536, pruned_loss=0.05327, over 4869.00 frames.], tot_loss[loss=0.1856, simple_loss=0.2572, pruned_loss=0.05703, over 983100.41 frames.], batch size: 35, aishell_tot_loss[loss=0.1835, simple_loss=0.2611, pruned_loss=0.05299, over 944751.53 frames.], datatang_tot_loss[loss=0.1867, simple_loss=0.2508, pruned_loss=0.06133, over 927189.99 frames.], batch size: 35, lr: 8.82e-04
+2022-06-18 16:30:20,544 INFO [train.py:874] (2/4) Epoch 9, batch 1250, datatang_loss[loss=0.2481, simple_loss=0.2917, pruned_loss=0.1023, over 4943.00 frames.], tot_loss[loss=0.1848, simple_loss=0.2568, pruned_loss=0.05644, over 983703.22 frames.], batch size: 69, aishell_tot_loss[loss=0.1824, simple_loss=0.2604, pruned_loss=0.05218, over 950913.33 frames.], datatang_tot_loss[loss=0.1873, simple_loss=0.251, pruned_loss=0.0618, over 932073.59 frames.], batch size: 69, lr: 8.82e-04
+2022-06-18 16:30:49,218 INFO [train.py:874] (2/4) Epoch 9, batch 1300, aishell_loss[loss=0.1974, simple_loss=0.2821, pruned_loss=0.05632, over 4935.00 frames.], tot_loss[loss=0.1854, simple_loss=0.2571, pruned_loss=0.05685, over 984256.75 frames.], batch size: 80, aishell_tot_loss[loss=0.1821, simple_loss=0.2603, pruned_loss=0.05197, over 955616.19 frames.], datatang_tot_loss[loss=0.1883, simple_loss=0.2516, pruned_loss=0.06251, over 937641.47 frames.], batch size: 80, lr: 8.81e-04
+2022-06-18 16:31:19,908 INFO [train.py:874] (2/4) Epoch 9, batch 1350, aishell_loss[loss=0.1654, simple_loss=0.2477, pruned_loss=0.04149, over 4915.00 frames.], tot_loss[loss=0.1846, simple_loss=0.2563, pruned_loss=0.05642, over 984357.04 frames.], batch size: 52, aishell_tot_loss[loss=0.1816, simple_loss=0.2596, pruned_loss=0.05182, over 959359.50 frames.], datatang_tot_loss[loss=0.188, simple_loss=0.2515, pruned_loss=0.06223, over 942628.16 frames.], batch size: 52, lr: 8.80e-04
+2022-06-18 16:31:50,461 INFO [train.py:874] (2/4) Epoch 9, batch 1400, aishell_loss[loss=0.1652, simple_loss=0.2452, pruned_loss=0.04261, over 4895.00 frames.], tot_loss[loss=0.185, simple_loss=0.2566, pruned_loss=0.0567, over 984337.41 frames.], batch size: 42, aishell_tot_loss[loss=0.1817, simple_loss=0.2597, pruned_loss=0.05188, over 962510.77 frames.], datatang_tot_loss[loss=0.1884, simple_loss=0.2518, pruned_loss=0.06252, over 947089.57 frames.], batch size: 42, lr: 8.80e-04
+2022-06-18 16:32:21,354 INFO [train.py:874] (2/4) Epoch 9, batch 1450, datatang_loss[loss=0.1755, simple_loss=0.2467, pruned_loss=0.05218, over 4953.00 frames.], tot_loss[loss=0.1837, simple_loss=0.255, pruned_loss=0.0562, over 984236.81 frames.], batch size: 86, aishell_tot_loss[loss=0.1812, simple_loss=0.2588, pruned_loss=0.05182, over 965147.15 frames.], datatang_tot_loss[loss=0.1875, simple_loss=0.2511, pruned_loss=0.06194, over 951227.42 frames.], batch size: 86, lr: 8.79e-04
+2022-06-18 16:32:52,179 INFO [train.py:874] (2/4) Epoch 9, batch 1500, datatang_loss[loss=0.2002, simple_loss=0.2563, pruned_loss=0.07198, over 4974.00 frames.], tot_loss[loss=0.183, simple_loss=0.2544, pruned_loss=0.05585, over 984318.50 frames.], batch size: 55, aishell_tot_loss[loss=0.1811, simple_loss=0.2587, pruned_loss=0.05175, over 967018.60 frames.], datatang_tot_loss[loss=0.1867, simple_loss=0.2506, pruned_loss=0.06143, over 955639.56 frames.], batch size: 55, lr: 8.78e-04
+2022-06-18 16:33:21,716 INFO [train.py:874] (2/4) Epoch 9, batch 1550, aishell_loss[loss=0.1576, simple_loss=0.2296, pruned_loss=0.04278, over 4971.00 frames.], tot_loss[loss=0.1841, simple_loss=0.2548, pruned_loss=0.05671, over 984321.36 frames.], batch size: 27, aishell_tot_loss[loss=0.1808, simple_loss=0.2581, pruned_loss=0.05179, over 968723.01 frames.], datatang_tot_loss[loss=0.1879, simple_loss=0.2516, pruned_loss=0.06205, over 959457.01 frames.], batch size: 27, lr: 8.78e-04
+2022-06-18 16:33:52,466 INFO [train.py:874] (2/4) Epoch 9, batch 1600, datatang_loss[loss=0.1591, simple_loss=0.2374, pruned_loss=0.04036, over 4926.00 frames.], tot_loss[loss=0.1852, simple_loss=0.2559, pruned_loss=0.05728, over 984952.29 frames.], batch size: 71, aishell_tot_loss[loss=0.1817, simple_loss=0.2589, pruned_loss=0.05225, over 970989.45 frames.], datatang_tot_loss[loss=0.1882, simple_loss=0.2519, pruned_loss=0.0623, over 962494.55 frames.], batch size: 71, lr: 8.77e-04
+2022-06-18 16:34:23,055 INFO [train.py:874] (2/4) Epoch 9, batch 1650, datatang_loss[loss=0.1894, simple_loss=0.2485, pruned_loss=0.06516, over 4967.00 frames.], tot_loss[loss=0.1851, simple_loss=0.2557, pruned_loss=0.05723, over 985158.71 frames.], batch size: 45, aishell_tot_loss[loss=0.1817, simple_loss=0.259, pruned_loss=0.05216, over 972852.26 frames.], datatang_tot_loss[loss=0.1882, simple_loss=0.2516, pruned_loss=0.06234, over 965100.51 frames.], batch size: 45, lr: 8.77e-04
+2022-06-18 16:34:54,130 INFO [train.py:874] (2/4) Epoch 9, batch 1700, datatang_loss[loss=0.16, simple_loss=0.2365, pruned_loss=0.04173, over 4969.00 frames.], tot_loss[loss=0.1845, simple_loss=0.2551, pruned_loss=0.05691, over 985291.70 frames.], batch size: 45, aishell_tot_loss[loss=0.1808, simple_loss=0.2586, pruned_loss=0.05157, over 974235.68 frames.], datatang_tot_loss[loss=0.1883, simple_loss=0.2517, pruned_loss=0.06247, over 967744.59 frames.], batch size: 45, lr: 8.76e-04
+2022-06-18 16:35:25,092 INFO [train.py:874] (2/4) Epoch 9, batch 1750, aishell_loss[loss=0.165, simple_loss=0.2401, pruned_loss=0.04495, over 4981.00 frames.], tot_loss[loss=0.1844, simple_loss=0.2546, pruned_loss=0.05715, over 985114.65 frames.], batch size: 51, aishell_tot_loss[loss=0.1804, simple_loss=0.2579, pruned_loss=0.05142, over 975108.91 frames.], datatang_tot_loss[loss=0.1885, simple_loss=0.2518, pruned_loss=0.06258, over 970180.15 frames.], batch size: 51, lr: 8.75e-04
+2022-06-18 16:35:56,450 INFO [train.py:874] (2/4) Epoch 9, batch 1800, aishell_loss[loss=0.1937, simple_loss=0.2615, pruned_loss=0.06299, over 4951.00 frames.], tot_loss[loss=0.1848, simple_loss=0.2548, pruned_loss=0.05743, over 985402.32 frames.], batch size: 58, aishell_tot_loss[loss=0.1811, simple_loss=0.2586, pruned_loss=0.05183, over 976277.25 frames.], datatang_tot_loss[loss=0.1881, simple_loss=0.2514, pruned_loss=0.06237, over 972302.82 frames.], batch size: 58, lr: 8.75e-04
+2022-06-18 16:36:25,539 INFO [train.py:874] (2/4) Epoch 9, batch 1850, aishell_loss[loss=0.2006, simple_loss=0.2794, pruned_loss=0.06087, over 4961.00 frames.], tot_loss[loss=0.1855, simple_loss=0.2556, pruned_loss=0.05771, over 985485.56 frames.], batch size: 44, aishell_tot_loss[loss=0.1814, simple_loss=0.2589, pruned_loss=0.05191, over 977513.67 frames.], datatang_tot_loss[loss=0.1887, simple_loss=0.2518, pruned_loss=0.06278, over 973728.86 frames.], batch size: 44, lr: 8.74e-04
+2022-06-18 16:36:56,409 INFO [train.py:874] (2/4) Epoch 9, batch 1900, datatang_loss[loss=0.2553, simple_loss=0.3061, pruned_loss=0.1022, over 4923.00 frames.], tot_loss[loss=0.1851, simple_loss=0.2553, pruned_loss=0.0574, over 986042.53 frames.], batch size: 108, aishell_tot_loss[loss=0.181, simple_loss=0.2583, pruned_loss=0.05187, over 978931.26 frames.], datatang_tot_loss[loss=0.1886, simple_loss=0.2521, pruned_loss=0.06259, over 975214.13 frames.], batch size: 108, lr: 8.73e-04
+2022-06-18 16:37:28,046 INFO [train.py:874] (2/4) Epoch 9, batch 1950, aishell_loss[loss=0.1831, simple_loss=0.2621, pruned_loss=0.05207, over 4912.00 frames.], tot_loss[loss=0.1835, simple_loss=0.2542, pruned_loss=0.05639, over 985990.36 frames.], batch size: 46, aishell_tot_loss[loss=0.1808, simple_loss=0.2583, pruned_loss=0.05168, over 979674.68 frames.], datatang_tot_loss[loss=0.1872, simple_loss=0.251, pruned_loss=0.0617, over 976543.01 frames.], batch size: 46, lr: 8.73e-04
+2022-06-18 16:37:57,034 INFO [train.py:874] (2/4) Epoch 9, batch 2000, datatang_loss[loss=0.1804, simple_loss=0.255, pruned_loss=0.05287, over 4819.00 frames.], tot_loss[loss=0.1847, simple_loss=0.2559, pruned_loss=0.05674, over 986084.52 frames.], batch size: 30, aishell_tot_loss[loss=0.1815, simple_loss=0.2591, pruned_loss=0.05193, over 980697.64 frames.], datatang_tot_loss[loss=0.1877, simple_loss=0.2518, pruned_loss=0.06175, over 977476.38 frames.], batch size: 30, lr: 8.72e-04
+2022-06-18 16:37:57,035 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 16:38:13,636 INFO [train.py:914] (2/4) Epoch 9, validation: loss=0.1694, simple_loss=0.2525, pruned_loss=0.04315, over 1622729.00 frames.
+2022-06-18 16:38:43,924 INFO [train.py:874] (2/4) Epoch 9, batch 2050, datatang_loss[loss=0.1579, simple_loss=0.2314, pruned_loss=0.04225, over 4921.00 frames.], tot_loss[loss=0.1842, simple_loss=0.255, pruned_loss=0.05672, over 985748.17 frames.], batch size: 83, aishell_tot_loss[loss=0.1808, simple_loss=0.2582, pruned_loss=0.0517, over 980983.64 frames.], datatang_tot_loss[loss=0.1878, simple_loss=0.2519, pruned_loss=0.06184, over 978513.86 frames.], batch size: 83, lr: 8.72e-04
+2022-06-18 16:39:14,761 INFO [train.py:874] (2/4) Epoch 9, batch 2100, aishell_loss[loss=0.1639, simple_loss=0.2482, pruned_loss=0.03985, over 4913.00 frames.], tot_loss[loss=0.1841, simple_loss=0.2546, pruned_loss=0.0568, over 986190.80 frames.], batch size: 52, aishell_tot_loss[loss=0.1805, simple_loss=0.2577, pruned_loss=0.0516, over 981704.16 frames.], datatang_tot_loss[loss=0.1879, simple_loss=0.252, pruned_loss=0.06187, over 979707.80 frames.], batch size: 52, lr: 8.71e-04
+2022-06-18 16:39:46,126 INFO [train.py:874] (2/4) Epoch 9, batch 2150, aishell_loss[loss=0.1916, simple_loss=0.2677, pruned_loss=0.05776, over 4935.00 frames.], tot_loss[loss=0.1833, simple_loss=0.2542, pruned_loss=0.05619, over 985760.57 frames.], batch size: 58, aishell_tot_loss[loss=0.1801, simple_loss=0.2576, pruned_loss=0.05133, over 982132.83 frames.], datatang_tot_loss[loss=0.1873, simple_loss=0.2517, pruned_loss=0.06148, over 980124.47 frames.], batch size: 58, lr: 8.70e-04
+2022-06-18 16:40:16,371 INFO [train.py:874] (2/4) Epoch 9, batch 2200, aishell_loss[loss=0.1448, simple_loss=0.2318, pruned_loss=0.0289, over 4866.00 frames.], tot_loss[loss=0.1826, simple_loss=0.2537, pruned_loss=0.0558, over 985664.71 frames.], batch size: 35, aishell_tot_loss[loss=0.1802, simple_loss=0.2578, pruned_loss=0.05133, over 982462.78 frames.], datatang_tot_loss[loss=0.1864, simple_loss=0.251, pruned_loss=0.06095, over 980822.43 frames.], batch size: 35, lr: 8.70e-04
+2022-06-18 16:40:46,545 INFO [train.py:874] (2/4) Epoch 9, batch 2250, datatang_loss[loss=0.1838, simple_loss=0.2577, pruned_loss=0.05493, over 4911.00 frames.], tot_loss[loss=0.1827, simple_loss=0.2537, pruned_loss=0.05586, over 985666.78 frames.], batch size: 42, aishell_tot_loss[loss=0.1803, simple_loss=0.2578, pruned_loss=0.05135, over 982948.24 frames.], datatang_tot_loss[loss=0.1862, simple_loss=0.2508, pruned_loss=0.06083, over 981297.57 frames.], batch size: 42, lr: 8.69e-04
+2022-06-18 16:41:17,729 INFO [train.py:874] (2/4) Epoch 9, batch 2300, datatang_loss[loss=0.187, simple_loss=0.2548, pruned_loss=0.0596, over 4952.00 frames.], tot_loss[loss=0.1818, simple_loss=0.253, pruned_loss=0.05528, over 985799.70 frames.], batch size: 67, aishell_tot_loss[loss=0.1794, simple_loss=0.257, pruned_loss=0.05087, over 983258.00 frames.], datatang_tot_loss[loss=0.1861, simple_loss=0.2506, pruned_loss=0.06075, over 981929.79 frames.], batch size: 67, lr: 8.69e-04
+2022-06-18 16:41:48,084 INFO [train.py:874] (2/4) Epoch 9, batch 2350, aishell_loss[loss=0.1904, simple_loss=0.2684, pruned_loss=0.05616, over 4884.00 frames.], tot_loss[loss=0.1818, simple_loss=0.2529, pruned_loss=0.05532, over 985567.63 frames.], batch size: 34, aishell_tot_loss[loss=0.1795, simple_loss=0.2571, pruned_loss=0.05101, over 983476.74 frames.], datatang_tot_loss[loss=0.1857, simple_loss=0.2503, pruned_loss=0.06052, over 982218.36 frames.], batch size: 34, lr: 8.68e-04
+2022-06-18 16:42:20,331 INFO [train.py:874] (2/4) Epoch 9, batch 2400, aishell_loss[loss=0.1805, simple_loss=0.2637, pruned_loss=0.04868, over 4978.00 frames.], tot_loss[loss=0.1817, simple_loss=0.253, pruned_loss=0.05518, over 985953.41 frames.], batch size: 61, aishell_tot_loss[loss=0.1802, simple_loss=0.2577, pruned_loss=0.05135, over 983936.74 frames.], datatang_tot_loss[loss=0.1847, simple_loss=0.2495, pruned_loss=0.05996, over 982790.68 frames.], batch size: 61, lr: 8.67e-04
+2022-06-18 16:42:51,901 INFO [train.py:874] (2/4) Epoch 9, batch 2450, datatang_loss[loss=0.1812, simple_loss=0.2547, pruned_loss=0.05384, over 4953.00 frames.], tot_loss[loss=0.1833, simple_loss=0.2549, pruned_loss=0.05589, over 985674.85 frames.], batch size: 62, aishell_tot_loss[loss=0.1813, simple_loss=0.2588, pruned_loss=0.0519, over 983920.69 frames.], datatang_tot_loss[loss=0.1853, simple_loss=0.2501, pruned_loss=0.0603, over 983068.33 frames.], batch size: 62, lr: 8.67e-04
+2022-06-18 16:43:22,136 INFO [train.py:874] (2/4) Epoch 9, batch 2500, datatang_loss[loss=0.1965, simple_loss=0.2686, pruned_loss=0.06223, over 4939.00 frames.], tot_loss[loss=0.1823, simple_loss=0.2537, pruned_loss=0.05542, over 985813.04 frames.], batch size: 88, aishell_tot_loss[loss=0.1808, simple_loss=0.2583, pruned_loss=0.05166, over 984261.13 frames.], datatang_tot_loss[loss=0.1847, simple_loss=0.2495, pruned_loss=0.05996, over 983385.63 frames.], batch size: 88, lr: 8.66e-04
+2022-06-18 16:43:51,852 INFO [train.py:874] (2/4) Epoch 9, batch 2550, aishell_loss[loss=0.1716, simple_loss=0.2551, pruned_loss=0.04403, over 4859.00 frames.], tot_loss[loss=0.1828, simple_loss=0.2543, pruned_loss=0.05565, over 986087.12 frames.], batch size: 37, aishell_tot_loss[loss=0.1808, simple_loss=0.2586, pruned_loss=0.05145, over 984485.19 frames.], datatang_tot_loss[loss=0.1852, simple_loss=0.2499, pruned_loss=0.06023, over 983923.17 frames.], batch size: 37, lr: 8.66e-04
+2022-06-18 16:44:23,861 INFO [train.py:874] (2/4) Epoch 9, batch 2600, datatang_loss[loss=0.1668, simple_loss=0.2389, pruned_loss=0.04733, over 4952.00 frames.], tot_loss[loss=0.1829, simple_loss=0.2539, pruned_loss=0.0559, over 985823.30 frames.], batch size: 86, aishell_tot_loss[loss=0.1808, simple_loss=0.2584, pruned_loss=0.05161, over 984352.80 frames.], datatang_tot_loss[loss=0.1852, simple_loss=0.2498, pruned_loss=0.06026, over 984212.44 frames.], batch size: 86, lr: 8.65e-04
+2022-06-18 16:44:54,186 INFO [train.py:874] (2/4) Epoch 9, batch 2650, aishell_loss[loss=0.1855, simple_loss=0.2706, pruned_loss=0.05022, over 4969.00 frames.], tot_loss[loss=0.1825, simple_loss=0.2536, pruned_loss=0.05571, over 985777.08 frames.], batch size: 40, aishell_tot_loss[loss=0.1802, simple_loss=0.2579, pruned_loss=0.05121, over 984525.16 frames.], datatang_tot_loss[loss=0.1854, simple_loss=0.2498, pruned_loss=0.06048, over 984349.57 frames.], batch size: 40, lr: 8.64e-04
+2022-06-18 16:45:24,647 INFO [train.py:874] (2/4) Epoch 9, batch 2700, aishell_loss[loss=0.1732, simple_loss=0.2477, pruned_loss=0.04933, over 4966.00 frames.], tot_loss[loss=0.1822, simple_loss=0.2531, pruned_loss=0.05566, over 985604.41 frames.], batch size: 31, aishell_tot_loss[loss=0.1802, simple_loss=0.2578, pruned_loss=0.05124, over 984576.40 frames.], datatang_tot_loss[loss=0.1848, simple_loss=0.2496, pruned_loss=0.06005, over 984423.51 frames.], batch size: 31, lr: 8.64e-04
+2022-06-18 16:45:55,554 INFO [train.py:874] (2/4) Epoch 9, batch 2750, datatang_loss[loss=0.2175, simple_loss=0.2692, pruned_loss=0.08288, over 4926.00 frames.], tot_loss[loss=0.1823, simple_loss=0.2535, pruned_loss=0.05554, over 985165.98 frames.], batch size: 81, aishell_tot_loss[loss=0.181, simple_loss=0.2586, pruned_loss=0.0517, over 984516.79 frames.], datatang_tot_loss[loss=0.1841, simple_loss=0.249, pruned_loss=0.05958, over 984274.31 frames.], batch size: 81, lr: 8.63e-04
+2022-06-18 16:46:25,637 INFO [train.py:874] (2/4) Epoch 9, batch 2800, datatang_loss[loss=0.1681, simple_loss=0.2365, pruned_loss=0.04988, over 4932.00 frames.], tot_loss[loss=0.182, simple_loss=0.2534, pruned_loss=0.05533, over 985513.05 frames.], batch size: 34, aishell_tot_loss[loss=0.1805, simple_loss=0.258, pruned_loss=0.05147, over 984795.52 frames.], datatang_tot_loss[loss=0.1842, simple_loss=0.2493, pruned_loss=0.05961, over 984535.82 frames.], batch size: 34, lr: 8.63e-04
+2022-06-18 16:46:55,361 INFO [train.py:874] (2/4) Epoch 9, batch 2850, datatang_loss[loss=0.1639, simple_loss=0.231, pruned_loss=0.04841, over 4921.00 frames.], tot_loss[loss=0.1821, simple_loss=0.2535, pruned_loss=0.05531, over 985960.13 frames.], batch size: 75, aishell_tot_loss[loss=0.1798, simple_loss=0.2578, pruned_loss=0.05094, over 985128.42 frames.], datatang_tot_loss[loss=0.1849, simple_loss=0.2495, pruned_loss=0.06013, over 984870.71 frames.], batch size: 75, lr: 8.62e-04
+2022-06-18 16:47:30,664 INFO [train.py:874] (2/4) Epoch 9, batch 2900, datatang_loss[loss=0.2462, simple_loss=0.2947, pruned_loss=0.09888, over 4874.00 frames.], tot_loss[loss=0.1818, simple_loss=0.2535, pruned_loss=0.05504, over 985800.53 frames.], batch size: 39, aishell_tot_loss[loss=0.1803, simple_loss=0.2584, pruned_loss=0.05105, over 985210.98 frames.], datatang_tot_loss[loss=0.1842, simple_loss=0.2489, pruned_loss=0.05976, over 984858.92 frames.], batch size: 39, lr: 8.61e-04
+2022-06-18 16:48:01,970 INFO [train.py:874] (2/4) Epoch 9, batch 2950, datatang_loss[loss=0.1695, simple_loss=0.2345, pruned_loss=0.05219, over 4942.00 frames.], tot_loss[loss=0.1819, simple_loss=0.2533, pruned_loss=0.05527, over 985849.97 frames.], batch size: 69, aishell_tot_loss[loss=0.1804, simple_loss=0.2587, pruned_loss=0.05106, over 985262.52 frames.], datatang_tot_loss[loss=0.184, simple_loss=0.2486, pruned_loss=0.05968, over 985054.54 frames.], batch size: 69, lr: 8.61e-04
+2022-06-18 16:48:32,283 INFO [train.py:874] (2/4) Epoch 9, batch 3000, datatang_loss[loss=0.1708, simple_loss=0.2368, pruned_loss=0.05239, over 4962.00 frames.], tot_loss[loss=0.1812, simple_loss=0.2527, pruned_loss=0.0548, over 986057.64 frames.], batch size: 37, aishell_tot_loss[loss=0.1797, simple_loss=0.258, pruned_loss=0.05066, over 985480.44 frames.], datatang_tot_loss[loss=0.1839, simple_loss=0.2486, pruned_loss=0.05963, over 985207.23 frames.], batch size: 37, lr: 8.60e-04
+2022-06-18 16:48:32,283 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 16:48:48,781 INFO [train.py:914] (2/4) Epoch 9, validation: loss=0.169, simple_loss=0.2511, pruned_loss=0.04348, over 1622729.00 frames.
+2022-06-18 16:49:18,616 INFO [train.py:874] (2/4) Epoch 9, batch 3050, datatang_loss[loss=0.1589, simple_loss=0.2294, pruned_loss=0.04419, over 4929.00 frames.], tot_loss[loss=0.1808, simple_loss=0.253, pruned_loss=0.05428, over 985726.94 frames.], batch size: 71, aishell_tot_loss[loss=0.1797, simple_loss=0.2584, pruned_loss=0.05045, over 985524.23 frames.], datatang_tot_loss[loss=0.1834, simple_loss=0.2481, pruned_loss=0.05936, over 984968.30 frames.], batch size: 71, lr: 8.60e-04
+2022-06-18 16:49:51,930 INFO [train.py:874] (2/4) Epoch 9, batch 3100, datatang_loss[loss=0.1959, simple_loss=0.2568, pruned_loss=0.06751, over 4922.00 frames.], tot_loss[loss=0.1809, simple_loss=0.253, pruned_loss=0.05436, over 985899.28 frames.], batch size: 81, aishell_tot_loss[loss=0.1795, simple_loss=0.2582, pruned_loss=0.05041, over 985595.07 frames.], datatang_tot_loss[loss=0.1834, simple_loss=0.2484, pruned_loss=0.05925, over 985196.62 frames.], batch size: 81, lr: 8.59e-04
+2022-06-18 16:50:22,216 INFO [train.py:874] (2/4) Epoch 9, batch 3150, datatang_loss[loss=0.1968, simple_loss=0.2619, pruned_loss=0.06584, over 4952.00 frames.], tot_loss[loss=0.1811, simple_loss=0.253, pruned_loss=0.05462, over 985616.18 frames.], batch size: 86, aishell_tot_loss[loss=0.1787, simple_loss=0.2575, pruned_loss=0.04996, over 985246.10 frames.], datatang_tot_loss[loss=0.1844, simple_loss=0.2491, pruned_loss=0.05979, over 985354.47 frames.], batch size: 86, lr: 8.59e-04
+2022-06-18 16:50:52,582 INFO [train.py:874] (2/4) Epoch 9, batch 3200, datatang_loss[loss=0.1749, simple_loss=0.2451, pruned_loss=0.05241, over 4953.00 frames.], tot_loss[loss=0.1809, simple_loss=0.2531, pruned_loss=0.05432, over 985775.68 frames.], batch size: 67, aishell_tot_loss[loss=0.1781, simple_loss=0.2572, pruned_loss=0.04953, over 985345.02 frames.], datatang_tot_loss[loss=0.1845, simple_loss=0.2495, pruned_loss=0.05982, over 985510.06 frames.], batch size: 67, lr: 8.58e-04
+2022-06-18 16:51:25,240 INFO [train.py:874] (2/4) Epoch 9, batch 3250, datatang_loss[loss=0.164, simple_loss=0.2396, pruned_loss=0.0442, over 4938.00 frames.], tot_loss[loss=0.1812, simple_loss=0.2531, pruned_loss=0.05471, over 985694.77 frames.], batch size: 69, aishell_tot_loss[loss=0.1785, simple_loss=0.2575, pruned_loss=0.04979, over 985361.39 frames.], datatang_tot_loss[loss=0.1843, simple_loss=0.2493, pruned_loss=0.0596, over 985488.62 frames.], batch size: 69, lr: 8.57e-04
+2022-06-18 16:51:55,891 INFO [train.py:874] (2/4) Epoch 9, batch 3300, aishell_loss[loss=0.197, simple_loss=0.2717, pruned_loss=0.06115, over 4955.00 frames.], tot_loss[loss=0.1812, simple_loss=0.2528, pruned_loss=0.05476, over 985589.11 frames.], batch size: 56, aishell_tot_loss[loss=0.1788, simple_loss=0.2578, pruned_loss=0.04991, over 985135.33 frames.], datatang_tot_loss[loss=0.1838, simple_loss=0.2489, pruned_loss=0.05933, over 985677.80 frames.], batch size: 56, lr: 8.57e-04
+2022-06-18 16:52:26,647 INFO [train.py:874] (2/4) Epoch 9, batch 3350, datatang_loss[loss=0.2106, simple_loss=0.2775, pruned_loss=0.07187, over 4954.00 frames.], tot_loss[loss=0.1801, simple_loss=0.2523, pruned_loss=0.05398, over 985522.56 frames.], batch size: 86, aishell_tot_loss[loss=0.1787, simple_loss=0.2578, pruned_loss=0.04977, over 984923.17 frames.], datatang_tot_loss[loss=0.1827, simple_loss=0.2483, pruned_loss=0.05861, over 985851.79 frames.], batch size: 86, lr: 8.56e-04
+2022-06-18 16:52:57,985 INFO [train.py:874] (2/4) Epoch 9, batch 3400, datatang_loss[loss=0.2009, simple_loss=0.2719, pruned_loss=0.065, over 4895.00 frames.], tot_loss[loss=0.1792, simple_loss=0.2514, pruned_loss=0.05344, over 984790.73 frames.], batch size: 42, aishell_tot_loss[loss=0.1781, simple_loss=0.2572, pruned_loss=0.04953, over 984603.02 frames.], datatang_tot_loss[loss=0.1821, simple_loss=0.2477, pruned_loss=0.05822, over 985417.98 frames.], batch size: 42, lr: 8.56e-04
+2022-06-18 16:53:28,217 INFO [train.py:874] (2/4) Epoch 9, batch 3450, datatang_loss[loss=0.1698, simple_loss=0.2379, pruned_loss=0.05089, over 4951.00 frames.], tot_loss[loss=0.1785, simple_loss=0.2509, pruned_loss=0.05301, over 984892.06 frames.], batch size: 67, aishell_tot_loss[loss=0.1776, simple_loss=0.2567, pruned_loss=0.04926, over 984590.66 frames.], datatang_tot_loss[loss=0.1817, simple_loss=0.2474, pruned_loss=0.05798, over 985494.71 frames.], batch size: 67, lr: 8.55e-04
+2022-06-18 16:53:59,569 INFO [train.py:874] (2/4) Epoch 9, batch 3500, aishell_loss[loss=0.1836, simple_loss=0.269, pruned_loss=0.04915, over 4956.00 frames.], tot_loss[loss=0.1796, simple_loss=0.2516, pruned_loss=0.05386, over 985307.58 frames.], batch size: 64, aishell_tot_loss[loss=0.1781, simple_loss=0.2572, pruned_loss=0.04948, over 984726.81 frames.], datatang_tot_loss[loss=0.1821, simple_loss=0.2476, pruned_loss=0.05835, over 985747.52 frames.], batch size: 64, lr: 8.55e-04
+2022-06-18 16:54:30,570 INFO [train.py:874] (2/4) Epoch 9, batch 3550, aishell_loss[loss=0.1646, simple_loss=0.237, pruned_loss=0.04613, over 4817.00 frames.], tot_loss[loss=0.1805, simple_loss=0.2524, pruned_loss=0.05426, over 985021.06 frames.], batch size: 24, aishell_tot_loss[loss=0.1785, simple_loss=0.2574, pruned_loss=0.04978, over 984529.83 frames.], datatang_tot_loss[loss=0.1825, simple_loss=0.2479, pruned_loss=0.05855, over 985673.19 frames.], batch size: 24, lr: 8.54e-04
+2022-06-18 16:55:00,705 INFO [train.py:874] (2/4) Epoch 9, batch 3600, datatang_loss[loss=0.1915, simple_loss=0.2555, pruned_loss=0.06374, over 4934.00 frames.], tot_loss[loss=0.1808, simple_loss=0.2527, pruned_loss=0.05442, over 984983.89 frames.], batch size: 37, aishell_tot_loss[loss=0.1791, simple_loss=0.258, pruned_loss=0.05008, over 984353.09 frames.], datatang_tot_loss[loss=0.1822, simple_loss=0.2476, pruned_loss=0.05841, over 985793.28 frames.], batch size: 37, lr: 8.53e-04
+2022-06-18 16:55:30,666 INFO [train.py:874] (2/4) Epoch 9, batch 3650, aishell_loss[loss=0.2079, simple_loss=0.2769, pruned_loss=0.06944, over 4911.00 frames.], tot_loss[loss=0.1809, simple_loss=0.2534, pruned_loss=0.05417, over 985078.06 frames.], batch size: 34, aishell_tot_loss[loss=0.1797, simple_loss=0.2588, pruned_loss=0.05031, over 984286.88 frames.], datatang_tot_loss[loss=0.1818, simple_loss=0.2473, pruned_loss=0.05818, over 985981.94 frames.], batch size: 34, lr: 8.53e-04
+2022-06-18 16:56:03,410 INFO [train.py:874] (2/4) Epoch 9, batch 3700, datatang_loss[loss=0.191, simple_loss=0.2444, pruned_loss=0.06877, over 4958.00 frames.], tot_loss[loss=0.1814, simple_loss=0.2539, pruned_loss=0.05442, over 985158.11 frames.], batch size: 37, aishell_tot_loss[loss=0.1799, simple_loss=0.2589, pruned_loss=0.05043, over 984541.71 frames.], datatang_tot_loss[loss=0.1822, simple_loss=0.2475, pruned_loss=0.05848, over 985821.62 frames.], batch size: 37, lr: 8.52e-04
+2022-06-18 16:56:32,385 INFO [train.py:874] (2/4) Epoch 9, batch 3750, datatang_loss[loss=0.1763, simple_loss=0.2524, pruned_loss=0.05009, over 4912.00 frames.], tot_loss[loss=0.1814, simple_loss=0.2541, pruned_loss=0.05433, over 984855.92 frames.], batch size: 81, aishell_tot_loss[loss=0.1795, simple_loss=0.2589, pruned_loss=0.0501, over 984366.92 frames.], datatang_tot_loss[loss=0.1827, simple_loss=0.248, pruned_loss=0.05865, over 985662.55 frames.], batch size: 81, lr: 8.52e-04
+2022-06-18 16:57:03,081 INFO [train.py:874] (2/4) Epoch 9, batch 3800, aishell_loss[loss=0.1767, simple_loss=0.2569, pruned_loss=0.04822, over 4979.00 frames.], tot_loss[loss=0.1825, simple_loss=0.2547, pruned_loss=0.05512, over 985124.59 frames.], batch size: 51, aishell_tot_loss[loss=0.1797, simple_loss=0.2592, pruned_loss=0.05017, over 984518.42 frames.], datatang_tot_loss[loss=0.1836, simple_loss=0.2488, pruned_loss=0.05918, over 985734.56 frames.], batch size: 51, lr: 8.51e-04
+2022-06-18 16:57:32,528 INFO [train.py:874] (2/4) Epoch 9, batch 3850, aishell_loss[loss=0.1991, simple_loss=0.2762, pruned_loss=0.06096, over 4953.00 frames.], tot_loss[loss=0.1826, simple_loss=0.2543, pruned_loss=0.05541, over 985016.17 frames.], batch size: 31, aishell_tot_loss[loss=0.1806, simple_loss=0.2596, pruned_loss=0.05083, over 984606.51 frames.], datatang_tot_loss[loss=0.183, simple_loss=0.2483, pruned_loss=0.05883, over 985517.19 frames.], batch size: 31, lr: 8.51e-04
+2022-06-18 16:58:01,281 INFO [train.py:874] (2/4) Epoch 9, batch 3900, aishell_loss[loss=0.1727, simple_loss=0.2568, pruned_loss=0.04428, over 4949.00 frames.], tot_loss[loss=0.1832, simple_loss=0.2554, pruned_loss=0.05552, over 984979.78 frames.], batch size: 45, aishell_tot_loss[loss=0.1814, simple_loss=0.2604, pruned_loss=0.05118, over 984587.26 frames.], datatang_tot_loss[loss=0.1832, simple_loss=0.2486, pruned_loss=0.05888, over 985514.48 frames.], batch size: 45, lr: 8.50e-04
+2022-06-18 16:58:29,653 INFO [train.py:874] (2/4) Epoch 9, batch 3950, datatang_loss[loss=0.1879, simple_loss=0.2495, pruned_loss=0.06313, over 4915.00 frames.], tot_loss[loss=0.1817, simple_loss=0.2546, pruned_loss=0.05442, over 985112.51 frames.], batch size: 83, aishell_tot_loss[loss=0.1807, simple_loss=0.2599, pruned_loss=0.05073, over 984643.99 frames.], datatang_tot_loss[loss=0.1825, simple_loss=0.2483, pruned_loss=0.05839, over 985583.41 frames.], batch size: 83, lr: 8.49e-04
+2022-06-18 16:58:57,893 INFO [train.py:874] (2/4) Epoch 9, batch 4000, aishell_loss[loss=0.168, simple_loss=0.2558, pruned_loss=0.04015, over 4866.00 frames.], tot_loss[loss=0.1818, simple_loss=0.255, pruned_loss=0.05432, over 985165.84 frames.], batch size: 37, aishell_tot_loss[loss=0.1809, simple_loss=0.2602, pruned_loss=0.05085, over 984675.67 frames.], datatang_tot_loss[loss=0.1824, simple_loss=0.2482, pruned_loss=0.0583, over 985643.38 frames.], batch size: 37, lr: 8.49e-04
+2022-06-18 16:58:57,894 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 16:59:14,870 INFO [train.py:914] (2/4) Epoch 9, validation: loss=0.1682, simple_loss=0.2517, pruned_loss=0.04232, over 1622729.00 frames.
+2022-06-18 16:59:43,930 INFO [train.py:874] (2/4) Epoch 9, batch 4050, datatang_loss[loss=0.1945, simple_loss=0.2591, pruned_loss=0.06498, over 4939.00 frames.], tot_loss[loss=0.1824, simple_loss=0.255, pruned_loss=0.05484, over 985104.87 frames.], batch size: 42, aishell_tot_loss[loss=0.1808, simple_loss=0.2601, pruned_loss=0.05079, over 984523.95 frames.], datatang_tot_loss[loss=0.1831, simple_loss=0.2489, pruned_loss=0.05865, over 985724.02 frames.], batch size: 42, lr: 8.48e-04
+2022-06-18 17:00:57,159 INFO [train.py:874] (2/4) Epoch 10, batch 50, datatang_loss[loss=0.1563, simple_loss=0.2321, pruned_loss=0.04023, over 4965.00 frames.], tot_loss[loss=0.169, simple_loss=0.2416, pruned_loss=0.04824, over 218618.72 frames.], batch size: 67, aishell_tot_loss[loss=0.1728, simple_loss=0.2492, pruned_loss=0.04825, over 111709.39 frames.], datatang_tot_loss[loss=0.1653, simple_loss=0.2345, pruned_loss=0.04806, over 120571.16 frames.], batch size: 67, lr: 8.15e-04
+2022-06-18 17:01:24,346 INFO [train.py:874] (2/4) Epoch 10, batch 100, aishell_loss[loss=0.1569, simple_loss=0.2498, pruned_loss=0.03196, over 4978.00 frames.], tot_loss[loss=0.1723, simple_loss=0.2463, pruned_loss=0.04911, over 388833.48 frames.], batch size: 51, aishell_tot_loss[loss=0.1767, simple_loss=0.2541, pruned_loss=0.04967, over 244904.16 frames.], datatang_tot_loss[loss=0.1657, simple_loss=0.2352, pruned_loss=0.04809, over 191457.43 frames.], batch size: 51, lr: 8.14e-04
+2022-06-18 17:01:55,761 INFO [train.py:874] (2/4) Epoch 10, batch 150, aishell_loss[loss=0.1686, simple_loss=0.2539, pruned_loss=0.04166, over 4899.00 frames.], tot_loss[loss=0.1718, simple_loss=0.2459, pruned_loss=0.04886, over 521037.03 frames.], batch size: 34, aishell_tot_loss[loss=0.1751, simple_loss=0.2531, pruned_loss=0.04858, over 348174.85 frames.], datatang_tot_loss[loss=0.1668, simple_loss=0.2354, pruned_loss=0.04909, over 267258.92 frames.], batch size: 34, lr: 8.14e-04
+2022-06-18 17:02:27,248 INFO [train.py:874] (2/4) Epoch 10, batch 200, datatang_loss[loss=0.1513, simple_loss=0.2153, pruned_loss=0.04363, over 4892.00 frames.], tot_loss[loss=0.1737, simple_loss=0.2475, pruned_loss=0.04993, over 623928.35 frames.], batch size: 42, aishell_tot_loss[loss=0.1771, simple_loss=0.2551, pruned_loss=0.04951, over 431424.92 frames.], datatang_tot_loss[loss=0.1683, simple_loss=0.2365, pruned_loss=0.05001, over 342370.75 frames.], batch size: 42, lr: 8.13e-04
+2022-06-18 17:02:55,478 INFO [train.py:874] (2/4) Epoch 10, batch 250, datatang_loss[loss=0.1663, simple_loss=0.2242, pruned_loss=0.05423, over 4841.00 frames.], tot_loss[loss=0.1749, simple_loss=0.2486, pruned_loss=0.0506, over 703779.65 frames.], batch size: 30, aishell_tot_loss[loss=0.1788, simple_loss=0.2573, pruned_loss=0.0502, over 499036.51 frames.], datatang_tot_loss[loss=0.1689, simple_loss=0.2366, pruned_loss=0.05056, over 415047.26 frames.], batch size: 30, lr: 8.13e-04
+2022-06-18 17:03:27,369 INFO [train.py:874] (2/4) Epoch 10, batch 300, aishell_loss[loss=0.1551, simple_loss=0.2306, pruned_loss=0.03978, over 4982.00 frames.], tot_loss[loss=0.1751, simple_loss=0.2483, pruned_loss=0.05095, over 766563.75 frames.], batch size: 27, aishell_tot_loss[loss=0.1777, simple_loss=0.2562, pruned_loss=0.04959, over 556545.76 frames.], datatang_tot_loss[loss=0.171, simple_loss=0.2382, pruned_loss=0.05195, over 482347.73 frames.], batch size: 27, lr: 8.12e-04
+2022-06-18 17:03:58,465 INFO [train.py:874] (2/4) Epoch 10, batch 350, datatang_loss[loss=0.177, simple_loss=0.2519, pruned_loss=0.05102, over 4937.00 frames.], tot_loss[loss=0.1757, simple_loss=0.2486, pruned_loss=0.05135, over 815088.99 frames.], batch size: 94, aishell_tot_loss[loss=0.1763, simple_loss=0.2548, pruned_loss=0.04892, over 605068.69 frames.], datatang_tot_loss[loss=0.1737, simple_loss=0.2407, pruned_loss=0.05329, over 543919.94 frames.], batch size: 94, lr: 8.12e-04
+2022-06-18 17:04:26,543 INFO [train.py:874] (2/4) Epoch 10, batch 400, datatang_loss[loss=0.1601, simple_loss=0.2315, pruned_loss=0.04437, over 4926.00 frames.], tot_loss[loss=0.1781, simple_loss=0.2509, pruned_loss=0.05263, over 852873.28 frames.], batch size: 79, aishell_tot_loss[loss=0.1775, simple_loss=0.2562, pruned_loss=0.04938, over 651414.64 frames.], datatang_tot_loss[loss=0.1763, simple_loss=0.2428, pruned_loss=0.05487, over 594128.41 frames.], batch size: 79, lr: 8.11e-04
+2022-06-18 17:04:57,205 INFO [train.py:874] (2/4) Epoch 10, batch 450, aishell_loss[loss=0.2053, simple_loss=0.2724, pruned_loss=0.06913, over 4876.00 frames.], tot_loss[loss=0.1787, simple_loss=0.252, pruned_loss=0.05272, over 881932.40 frames.], batch size: 35, aishell_tot_loss[loss=0.1787, simple_loss=0.2578, pruned_loss=0.04982, over 699484.62 frames.], datatang_tot_loss[loss=0.1763, simple_loss=0.2427, pruned_loss=0.05496, over 629328.10 frames.], batch size: 35, lr: 8.11e-04
+2022-06-18 17:05:26,438 INFO [train.py:874] (2/4) Epoch 10, batch 500, aishell_loss[loss=0.1792, simple_loss=0.2508, pruned_loss=0.05377, over 4887.00 frames.], tot_loss[loss=0.1789, simple_loss=0.252, pruned_loss=0.05288, over 905004.76 frames.], batch size: 35, aishell_tot_loss[loss=0.1785, simple_loss=0.2572, pruned_loss=0.04991, over 739455.33 frames.], datatang_tot_loss[loss=0.1771, simple_loss=0.2435, pruned_loss=0.05536, over 663394.85 frames.], batch size: 35, lr: 8.10e-04
+2022-06-18 17:05:55,038 INFO [train.py:874] (2/4) Epoch 10, batch 550, datatang_loss[loss=0.1674, simple_loss=0.2377, pruned_loss=0.04858, over 4918.00 frames.], tot_loss[loss=0.1793, simple_loss=0.2524, pruned_loss=0.05308, over 922902.40 frames.], batch size: 77, aishell_tot_loss[loss=0.1778, simple_loss=0.2566, pruned_loss=0.04948, over 773601.51 frames.], datatang_tot_loss[loss=0.1788, simple_loss=0.2448, pruned_loss=0.0564, over 694393.78 frames.], batch size: 77, lr: 8.10e-04
+2022-06-18 17:06:26,478 INFO [train.py:874] (2/4) Epoch 10, batch 600, aishell_loss[loss=0.1642, simple_loss=0.2499, pruned_loss=0.03918, over 4978.00 frames.], tot_loss[loss=0.1793, simple_loss=0.2522, pruned_loss=0.05319, over 936950.62 frames.], batch size: 51, aishell_tot_loss[loss=0.1778, simple_loss=0.2566, pruned_loss=0.04947, over 800323.48 frames.], datatang_tot_loss[loss=0.1792, simple_loss=0.2449, pruned_loss=0.05672, over 726391.67 frames.], batch size: 51, lr: 8.09e-04
+2022-06-18 17:06:55,503 INFO [train.py:874] (2/4) Epoch 10, batch 650, datatang_loss[loss=0.2132, simple_loss=0.2739, pruned_loss=0.07622, over 4934.00 frames.], tot_loss[loss=0.178, simple_loss=0.2509, pruned_loss=0.05254, over 948159.67 frames.], batch size: 98, aishell_tot_loss[loss=0.177, simple_loss=0.2561, pruned_loss=0.04893, over 819020.49 frames.], datatang_tot_loss[loss=0.1785, simple_loss=0.2445, pruned_loss=0.05629, over 761711.07 frames.], batch size: 98, lr: 8.09e-04
+2022-06-18 17:07:24,952 INFO [train.py:874] (2/4) Epoch 10, batch 700, aishell_loss[loss=0.1695, simple_loss=0.2537, pruned_loss=0.04265, over 4871.00 frames.], tot_loss[loss=0.1787, simple_loss=0.2518, pruned_loss=0.05282, over 956210.84 frames.], batch size: 36, aishell_tot_loss[loss=0.1776, simple_loss=0.2567, pruned_loss=0.0493, over 839873.09 frames.], datatang_tot_loss[loss=0.1788, simple_loss=0.2451, pruned_loss=0.05631, over 786075.78 frames.], batch size: 36, lr: 8.08e-04
+2022-06-18 17:07:56,720 INFO [train.py:874] (2/4) Epoch 10, batch 750, aishell_loss[loss=0.1784, simple_loss=0.2624, pruned_loss=0.04724, over 4932.00 frames.], tot_loss[loss=0.1801, simple_loss=0.2528, pruned_loss=0.05364, over 962862.52 frames.], batch size: 68, aishell_tot_loss[loss=0.1778, simple_loss=0.2568, pruned_loss=0.04938, over 854989.31 frames.], datatang_tot_loss[loss=0.1804, simple_loss=0.2466, pruned_loss=0.05713, over 812444.56 frames.], batch size: 68, lr: 8.08e-04
+2022-06-18 17:08:27,665 INFO [train.py:874] (2/4) Epoch 10, batch 800, aishell_loss[loss=0.1448, simple_loss=0.2117, pruned_loss=0.03893, over 4821.00 frames.], tot_loss[loss=0.1799, simple_loss=0.2525, pruned_loss=0.05359, over 967358.22 frames.], batch size: 24, aishell_tot_loss[loss=0.1768, simple_loss=0.2561, pruned_loss=0.04873, over 868118.04 frames.], datatang_tot_loss[loss=0.1814, simple_loss=0.2475, pruned_loss=0.0577, over 835040.32 frames.], batch size: 24, lr: 8.07e-04
+2022-06-18 17:08:57,234 INFO [train.py:874] (2/4) Epoch 10, batch 850, datatang_loss[loss=0.1958, simple_loss=0.2469, pruned_loss=0.07237, over 4906.00 frames.], tot_loss[loss=0.1803, simple_loss=0.2529, pruned_loss=0.05384, over 971272.99 frames.], batch size: 47, aishell_tot_loss[loss=0.177, simple_loss=0.2564, pruned_loss=0.04876, over 880200.52 frames.], datatang_tot_loss[loss=0.1819, simple_loss=0.248, pruned_loss=0.0579, over 854807.07 frames.], batch size: 47, lr: 8.07e-04
+2022-06-18 17:09:29,300 INFO [train.py:874] (2/4) Epoch 10, batch 900, aishell_loss[loss=0.2001, simple_loss=0.287, pruned_loss=0.05665, over 4912.00 frames.], tot_loss[loss=0.1789, simple_loss=0.2518, pruned_loss=0.05302, over 973969.78 frames.], batch size: 46, aishell_tot_loss[loss=0.1767, simple_loss=0.2561, pruned_loss=0.04863, over 892792.29 frames.], datatang_tot_loss[loss=0.1808, simple_loss=0.2471, pruned_loss=0.05724, over 869413.34 frames.], batch size: 46, lr: 8.06e-04
+2022-06-18 17:09:58,847 INFO [train.py:874] (2/4) Epoch 10, batch 950, datatang_loss[loss=0.1635, simple_loss=0.2366, pruned_loss=0.04518, over 4845.00 frames.], tot_loss[loss=0.1797, simple_loss=0.2528, pruned_loss=0.05334, over 976538.76 frames.], batch size: 24, aishell_tot_loss[loss=0.1776, simple_loss=0.2572, pruned_loss=0.04899, over 904109.51 frames.], datatang_tot_loss[loss=0.1809, simple_loss=0.2471, pruned_loss=0.0573, over 882597.44 frames.], batch size: 24, lr: 8.05e-04
+2022-06-18 17:10:29,038 INFO [train.py:874] (2/4) Epoch 10, batch 1000, aishell_loss[loss=0.1519, simple_loss=0.226, pruned_loss=0.03891, over 4957.00 frames.], tot_loss[loss=0.1793, simple_loss=0.2525, pruned_loss=0.05305, over 978318.08 frames.], batch size: 27, aishell_tot_loss[loss=0.1775, simple_loss=0.2572, pruned_loss=0.04889, over 912933.54 frames.], datatang_tot_loss[loss=0.1806, simple_loss=0.2471, pruned_loss=0.05701, over 895452.52 frames.], batch size: 27, lr: 8.05e-04
+2022-06-18 17:10:29,039 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 17:10:46,215 INFO [train.py:914] (2/4) Epoch 10, validation: loss=0.1689, simple_loss=0.2533, pruned_loss=0.04224, over 1622729.00 frames.
+2022-06-18 17:11:16,219 INFO [train.py:874] (2/4) Epoch 10, batch 1050, aishell_loss[loss=0.1844, simple_loss=0.2607, pruned_loss=0.05405, over 4863.00 frames.], tot_loss[loss=0.18, simple_loss=0.2525, pruned_loss=0.05371, over 980235.19 frames.], batch size: 37, aishell_tot_loss[loss=0.1774, simple_loss=0.257, pruned_loss=0.04894, over 918484.25 frames.], datatang_tot_loss[loss=0.1813, simple_loss=0.2479, pruned_loss=0.05732, over 910007.91 frames.], batch size: 37, lr: 8.04e-04
+2022-06-18 17:11:46,552 INFO [train.py:874] (2/4) Epoch 10, batch 1100, aishell_loss[loss=0.1671, simple_loss=0.2543, pruned_loss=0.03999, over 4928.00 frames.], tot_loss[loss=0.1801, simple_loss=0.2526, pruned_loss=0.05378, over 981574.87 frames.], batch size: 52, aishell_tot_loss[loss=0.1774, simple_loss=0.2569, pruned_loss=0.04901, over 927615.74 frames.], datatang_tot_loss[loss=0.1817, simple_loss=0.248, pruned_loss=0.05768, over 917672.20 frames.], batch size: 52, lr: 8.04e-04
+2022-06-18 17:12:16,114 INFO [train.py:874] (2/4) Epoch 10, batch 1150, datatang_loss[loss=0.1412, simple_loss=0.2117, pruned_loss=0.03531, over 4970.00 frames.], tot_loss[loss=0.1793, simple_loss=0.2524, pruned_loss=0.05315, over 982656.80 frames.], batch size: 67, aishell_tot_loss[loss=0.1765, simple_loss=0.2561, pruned_loss=0.04842, over 934816.52 frames.], datatang_tot_loss[loss=0.1819, simple_loss=0.2484, pruned_loss=0.05769, over 925447.61 frames.], batch size: 67, lr: 8.03e-04
+2022-06-18 17:12:47,062 INFO [train.py:874] (2/4) Epoch 10, batch 1200, aishell_loss[loss=0.189, simple_loss=0.2632, pruned_loss=0.05736, over 4872.00 frames.], tot_loss[loss=0.1795, simple_loss=0.252, pruned_loss=0.05349, over 983262.24 frames.], batch size: 36, aishell_tot_loss[loss=0.1768, simple_loss=0.2563, pruned_loss=0.04862, over 939717.48 frames.], datatang_tot_loss[loss=0.1817, simple_loss=0.2481, pruned_loss=0.05765, over 933733.92 frames.], batch size: 36, lr: 8.03e-04
+2022-06-18 17:13:18,849 INFO [train.py:874] (2/4) Epoch 10, batch 1250, aishell_loss[loss=0.2065, simple_loss=0.2897, pruned_loss=0.06169, over 4928.00 frames.], tot_loss[loss=0.1802, simple_loss=0.2525, pruned_loss=0.0539, over 983965.22 frames.], batch size: 80, aishell_tot_loss[loss=0.1776, simple_loss=0.2571, pruned_loss=0.04908, over 945339.59 frames.], datatang_tot_loss[loss=0.1817, simple_loss=0.2478, pruned_loss=0.05776, over 939790.38 frames.], batch size: 80, lr: 8.02e-04
+2022-06-18 17:13:47,295 INFO [train.py:874] (2/4) Epoch 10, batch 1300, datatang_loss[loss=0.1706, simple_loss=0.2355, pruned_loss=0.05285, over 4936.00 frames.], tot_loss[loss=0.1786, simple_loss=0.2513, pruned_loss=0.05294, over 984640.23 frames.], batch size: 50, aishell_tot_loss[loss=0.1768, simple_loss=0.2565, pruned_loss=0.04854, over 949797.10 frames.], datatang_tot_loss[loss=0.1808, simple_loss=0.2471, pruned_loss=0.05728, over 945833.57 frames.], batch size: 50, lr: 8.02e-04
+2022-06-18 17:14:20,325 INFO [train.py:874] (2/4) Epoch 10, batch 1350, datatang_loss[loss=0.2582, simple_loss=0.3137, pruned_loss=0.1014, over 4962.00 frames.], tot_loss[loss=0.1774, simple_loss=0.25, pruned_loss=0.05245, over 985034.31 frames.], batch size: 99, aishell_tot_loss[loss=0.1767, simple_loss=0.2565, pruned_loss=0.04847, over 952253.88 frames.], datatang_tot_loss[loss=0.1794, simple_loss=0.246, pruned_loss=0.05638, over 952561.89 frames.], batch size: 99, lr: 8.01e-04
+2022-06-18 17:14:52,510 INFO [train.py:874] (2/4) Epoch 10, batch 1400, datatang_loss[loss=0.172, simple_loss=0.239, pruned_loss=0.05254, over 4955.00 frames.], tot_loss[loss=0.1778, simple_loss=0.2507, pruned_loss=0.05245, over 985141.18 frames.], batch size: 86, aishell_tot_loss[loss=0.1773, simple_loss=0.257, pruned_loss=0.04882, over 957025.71 frames.], datatang_tot_loss[loss=0.1791, simple_loss=0.2457, pruned_loss=0.05627, over 955579.63 frames.], batch size: 86, lr: 8.01e-04
+2022-06-18 17:15:21,177 INFO [train.py:874] (2/4) Epoch 10, batch 1450, datatang_loss[loss=0.1509, simple_loss=0.2264, pruned_loss=0.0377, over 4955.00 frames.], tot_loss[loss=0.1772, simple_loss=0.2498, pruned_loss=0.05229, over 984900.45 frames.], batch size: 69, aishell_tot_loss[loss=0.1774, simple_loss=0.257, pruned_loss=0.04888, over 959682.76 frames.], datatang_tot_loss[loss=0.1783, simple_loss=0.2448, pruned_loss=0.05591, over 959457.82 frames.], batch size: 69, lr: 8.00e-04
+2022-06-18 17:15:52,704 INFO [train.py:874] (2/4) Epoch 10, batch 1500, datatang_loss[loss=0.1293, simple_loss=0.2099, pruned_loss=0.02434, over 4967.00 frames.], tot_loss[loss=0.1771, simple_loss=0.2495, pruned_loss=0.05231, over 985475.16 frames.], batch size: 40, aishell_tot_loss[loss=0.1772, simple_loss=0.2567, pruned_loss=0.0489, over 962613.23 frames.], datatang_tot_loss[loss=0.1782, simple_loss=0.2447, pruned_loss=0.05584, over 963100.40 frames.], batch size: 40, lr: 8.00e-04
+2022-06-18 17:16:23,733 INFO [train.py:874] (2/4) Epoch 10, batch 1550, datatang_loss[loss=0.1694, simple_loss=0.2329, pruned_loss=0.05299, over 4975.00 frames.], tot_loss[loss=0.177, simple_loss=0.2499, pruned_loss=0.05201, over 985050.20 frames.], batch size: 37, aishell_tot_loss[loss=0.1768, simple_loss=0.2565, pruned_loss=0.04854, over 965430.98 frames.], datatang_tot_loss[loss=0.1784, simple_loss=0.2447, pruned_loss=0.05604, over 965162.48 frames.], batch size: 37, lr: 7.99e-04
+2022-06-18 17:16:51,682 INFO [train.py:874] (2/4) Epoch 10, batch 1600, aishell_loss[loss=0.2016, simple_loss=0.2812, pruned_loss=0.06102, over 4958.00 frames.], tot_loss[loss=0.1768, simple_loss=0.2496, pruned_loss=0.05201, over 985353.81 frames.], batch size: 40, aishell_tot_loss[loss=0.177, simple_loss=0.2568, pruned_loss=0.04862, over 967935.60 frames.], datatang_tot_loss[loss=0.1778, simple_loss=0.2439, pruned_loss=0.05589, over 967623.79 frames.], batch size: 40, lr: 7.99e-04
+2022-06-18 17:17:24,164 INFO [train.py:874] (2/4) Epoch 10, batch 1650, datatang_loss[loss=0.2042, simple_loss=0.2589, pruned_loss=0.07473, over 4946.00 frames.], tot_loss[loss=0.1768, simple_loss=0.2494, pruned_loss=0.05207, over 985127.55 frames.], batch size: 88, aishell_tot_loss[loss=0.1764, simple_loss=0.2562, pruned_loss=0.0483, over 969799.78 frames.], datatang_tot_loss[loss=0.1782, simple_loss=0.2443, pruned_loss=0.05611, over 969666.14 frames.], batch size: 88, lr: 7.98e-04
+2022-06-18 17:17:56,412 INFO [train.py:874] (2/4) Epoch 10, batch 1700, aishell_loss[loss=0.1809, simple_loss=0.2672, pruned_loss=0.04728, over 4889.00 frames.], tot_loss[loss=0.1774, simple_loss=0.2506, pruned_loss=0.0521, over 985327.46 frames.], batch size: 34, aishell_tot_loss[loss=0.1769, simple_loss=0.257, pruned_loss=0.04838, over 971727.94 frames.], datatang_tot_loss[loss=0.1784, simple_loss=0.2445, pruned_loss=0.05616, over 971583.62 frames.], batch size: 34, lr: 7.98e-04
+2022-06-18 17:18:24,504 INFO [train.py:874] (2/4) Epoch 10, batch 1750, datatang_loss[loss=0.1837, simple_loss=0.2514, pruned_loss=0.05805, over 4941.00 frames.], tot_loss[loss=0.1775, simple_loss=0.2508, pruned_loss=0.05206, over 985367.69 frames.], batch size: 88, aishell_tot_loss[loss=0.1765, simple_loss=0.2566, pruned_loss=0.04818, over 973284.69 frames.], datatang_tot_loss[loss=0.1788, simple_loss=0.245, pruned_loss=0.05629, over 973293.10 frames.], batch size: 88, lr: 7.97e-04
+2022-06-18 17:18:55,216 INFO [train.py:874] (2/4) Epoch 10, batch 1800, datatang_loss[loss=0.171, simple_loss=0.2409, pruned_loss=0.05055, over 4974.00 frames.], tot_loss[loss=0.1775, simple_loss=0.2505, pruned_loss=0.05222, over 985439.13 frames.], batch size: 45, aishell_tot_loss[loss=0.1767, simple_loss=0.2567, pruned_loss=0.04833, over 974537.63 frames.], datatang_tot_loss[loss=0.1786, simple_loss=0.2446, pruned_loss=0.05625, over 974957.89 frames.], batch size: 45, lr: 7.97e-04
+2022-06-18 17:19:27,418 INFO [train.py:874] (2/4) Epoch 10, batch 1850, datatang_loss[loss=0.1668, simple_loss=0.2331, pruned_loss=0.0502, over 4955.00 frames.], tot_loss[loss=0.1785, simple_loss=0.2513, pruned_loss=0.05287, over 985489.89 frames.], batch size: 55, aishell_tot_loss[loss=0.1771, simple_loss=0.2571, pruned_loss=0.04856, over 975812.07 frames.], datatang_tot_loss[loss=0.1792, simple_loss=0.245, pruned_loss=0.05667, over 976264.79 frames.], batch size: 55, lr: 7.96e-04
+2022-06-18 17:19:55,245 INFO [train.py:874] (2/4) Epoch 10, batch 1900, aishell_loss[loss=0.1853, simple_loss=0.2719, pruned_loss=0.04936, over 4958.00 frames.], tot_loss[loss=0.1783, simple_loss=0.251, pruned_loss=0.05284, over 985621.69 frames.], batch size: 58, aishell_tot_loss[loss=0.1772, simple_loss=0.2571, pruned_loss=0.04867, over 976621.81 frames.], datatang_tot_loss[loss=0.1789, simple_loss=0.2452, pruned_loss=0.0563, over 977798.94 frames.], batch size: 58, lr: 7.96e-04
+2022-06-18 17:20:27,318 INFO [train.py:874] (2/4) Epoch 10, batch 1950, datatang_loss[loss=0.2085, simple_loss=0.2692, pruned_loss=0.07394, over 4946.00 frames.], tot_loss[loss=0.1797, simple_loss=0.2524, pruned_loss=0.05347, over 985377.82 frames.], batch size: 86, aishell_tot_loss[loss=0.1785, simple_loss=0.2582, pruned_loss=0.04942, over 977297.72 frames.], datatang_tot_loss[loss=0.1791, simple_loss=0.2455, pruned_loss=0.05638, over 978858.04 frames.], batch size: 86, lr: 7.95e-04
+2022-06-18 17:20:59,403 INFO [train.py:874] (2/4) Epoch 10, batch 2000, aishell_loss[loss=0.1741, simple_loss=0.2562, pruned_loss=0.046, over 4928.00 frames.], tot_loss[loss=0.179, simple_loss=0.2519, pruned_loss=0.05308, over 985138.12 frames.], batch size: 33, aishell_tot_loss[loss=0.1776, simple_loss=0.2572, pruned_loss=0.04903, over 977942.41 frames.], datatang_tot_loss[loss=0.1795, simple_loss=0.246, pruned_loss=0.05646, over 979683.00 frames.], batch size: 33, lr: 7.95e-04
+2022-06-18 17:20:59,404 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 17:21:15,565 INFO [train.py:914] (2/4) Epoch 10, validation: loss=0.1694, simple_loss=0.2523, pruned_loss=0.04323, over 1622729.00 frames.
+2022-06-18 17:21:47,946 INFO [train.py:874] (2/4) Epoch 10, batch 2050, aishell_loss[loss=0.1876, simple_loss=0.2717, pruned_loss=0.05169, over 4939.00 frames.], tot_loss[loss=0.1792, simple_loss=0.2523, pruned_loss=0.05308, over 985433.04 frames.], batch size: 49, aishell_tot_loss[loss=0.1779, simple_loss=0.2576, pruned_loss=0.04914, over 978957.98 frames.], datatang_tot_loss[loss=0.1795, simple_loss=0.2459, pruned_loss=0.05657, over 980501.39 frames.], batch size: 49, lr: 7.94e-04
+2022-06-18 17:22:16,703 INFO [train.py:874] (2/4) Epoch 10, batch 2100, aishell_loss[loss=0.1959, simple_loss=0.2706, pruned_loss=0.06058, over 4976.00 frames.], tot_loss[loss=0.1791, simple_loss=0.2526, pruned_loss=0.05277, over 985265.04 frames.], batch size: 39, aishell_tot_loss[loss=0.1785, simple_loss=0.2583, pruned_loss=0.04938, over 979435.25 frames.], datatang_tot_loss[loss=0.1789, simple_loss=0.2456, pruned_loss=0.05611, over 981181.23 frames.], batch size: 39, lr: 7.94e-04
+2022-06-18 17:22:48,237 INFO [train.py:874] (2/4) Epoch 10, batch 2150, aishell_loss[loss=0.1534, simple_loss=0.2352, pruned_loss=0.03577, over 4973.00 frames.], tot_loss[loss=0.1776, simple_loss=0.2514, pruned_loss=0.05192, over 985110.52 frames.], batch size: 30, aishell_tot_loss[loss=0.1782, simple_loss=0.2579, pruned_loss=0.04927, over 979905.33 frames.], datatang_tot_loss[loss=0.1777, simple_loss=0.245, pruned_loss=0.05526, over 981701.33 frames.], batch size: 30, lr: 7.93e-04
+2022-06-18 17:23:19,885 INFO [train.py:874] (2/4) Epoch 10, batch 2200, aishell_loss[loss=0.1689, simple_loss=0.2616, pruned_loss=0.03814, over 4956.00 frames.], tot_loss[loss=0.178, simple_loss=0.252, pruned_loss=0.05203, over 985070.98 frames.], batch size: 30, aishell_tot_loss[loss=0.1782, simple_loss=0.2584, pruned_loss=0.04897, over 980255.67 frames.], datatang_tot_loss[loss=0.1781, simple_loss=0.2451, pruned_loss=0.05558, over 982333.19 frames.], batch size: 30, lr: 7.93e-04
+2022-06-18 17:23:47,885 INFO [train.py:874] (2/4) Epoch 10, batch 2250, datatang_loss[loss=0.186, simple_loss=0.254, pruned_loss=0.05903, over 4924.00 frames.], tot_loss[loss=0.1772, simple_loss=0.2515, pruned_loss=0.05142, over 984723.71 frames.], batch size: 77, aishell_tot_loss[loss=0.1772, simple_loss=0.2574, pruned_loss=0.04846, over 980723.66 frames.], datatang_tot_loss[loss=0.1781, simple_loss=0.2454, pruned_loss=0.05545, over 982395.75 frames.], batch size: 77, lr: 7.92e-04
+2022-06-18 17:24:17,566 INFO [train.py:874] (2/4) Epoch 10, batch 2300, aishell_loss[loss=0.1724, simple_loss=0.2507, pruned_loss=0.047, over 4881.00 frames.], tot_loss[loss=0.1763, simple_loss=0.2512, pruned_loss=0.05069, over 984913.60 frames.], batch size: 47, aishell_tot_loss[loss=0.1765, simple_loss=0.257, pruned_loss=0.048, over 981471.52 frames.], datatang_tot_loss[loss=0.1778, simple_loss=0.2452, pruned_loss=0.05521, over 982640.22 frames.], batch size: 47, lr: 7.92e-04
+2022-06-18 17:24:49,699 INFO [train.py:874] (2/4) Epoch 10, batch 2350, datatang_loss[loss=0.2151, simple_loss=0.2661, pruned_loss=0.08204, over 4958.00 frames.], tot_loss[loss=0.1747, simple_loss=0.2493, pruned_loss=0.05009, over 985467.57 frames.], batch size: 55, aishell_tot_loss[loss=0.1755, simple_loss=0.2561, pruned_loss=0.04746, over 982302.51 frames.], datatang_tot_loss[loss=0.1769, simple_loss=0.244, pruned_loss=0.0549, over 983074.38 frames.], batch size: 55, lr: 7.91e-04
+2022-06-18 17:25:17,484 INFO [train.py:874] (2/4) Epoch 10, batch 2400, datatang_loss[loss=0.1999, simple_loss=0.2592, pruned_loss=0.07023, over 4870.00 frames.], tot_loss[loss=0.1753, simple_loss=0.2499, pruned_loss=0.05035, over 985809.43 frames.], batch size: 39, aishell_tot_loss[loss=0.1758, simple_loss=0.2564, pruned_loss=0.04757, over 982754.48 frames.], datatang_tot_loss[loss=0.1769, simple_loss=0.2444, pruned_loss=0.05465, over 983617.75 frames.], batch size: 39, lr: 7.91e-04
+2022-06-18 17:25:49,001 INFO [train.py:874] (2/4) Epoch 10, batch 2450, datatang_loss[loss=0.1809, simple_loss=0.2314, pruned_loss=0.06521, over 4977.00 frames.], tot_loss[loss=0.1756, simple_loss=0.2502, pruned_loss=0.05051, over 985799.72 frames.], batch size: 60, aishell_tot_loss[loss=0.1756, simple_loss=0.2565, pruned_loss=0.04736, over 982788.06 frames.], datatang_tot_loss[loss=0.1771, simple_loss=0.2445, pruned_loss=0.05485, over 984198.50 frames.], batch size: 60, lr: 7.90e-04
+2022-06-18 17:26:19,965 INFO [train.py:874] (2/4) Epoch 10, batch 2500, aishell_loss[loss=0.1509, simple_loss=0.2167, pruned_loss=0.0425, over 4961.00 frames.], tot_loss[loss=0.1744, simple_loss=0.2487, pruned_loss=0.05002, over 985172.02 frames.], batch size: 21, aishell_tot_loss[loss=0.1739, simple_loss=0.2545, pruned_loss=0.04662, over 982508.60 frames.], datatang_tot_loss[loss=0.1775, simple_loss=0.2448, pruned_loss=0.05508, over 984386.85 frames.], batch size: 21, lr: 7.90e-04
+2022-06-18 17:26:50,240 INFO [train.py:874] (2/4) Epoch 10, batch 2550, datatang_loss[loss=0.1983, simple_loss=0.2669, pruned_loss=0.0649, over 4962.00 frames.], tot_loss[loss=0.1747, simple_loss=0.2491, pruned_loss=0.05015, over 985359.59 frames.], batch size: 99, aishell_tot_loss[loss=0.174, simple_loss=0.2547, pruned_loss=0.04666, over 982966.64 frames.], datatang_tot_loss[loss=0.1773, simple_loss=0.2447, pruned_loss=0.05498, over 984521.30 frames.], batch size: 99, lr: 7.89e-04
+2022-06-18 17:27:22,110 INFO [train.py:874] (2/4) Epoch 10, batch 2600, datatang_loss[loss=0.1756, simple_loss=0.2512, pruned_loss=0.04996, over 4851.00 frames.], tot_loss[loss=0.1743, simple_loss=0.2489, pruned_loss=0.04981, over 985405.68 frames.], batch size: 30, aishell_tot_loss[loss=0.1737, simple_loss=0.2544, pruned_loss=0.04652, over 983345.96 frames.], datatang_tot_loss[loss=0.177, simple_loss=0.2446, pruned_loss=0.05466, over 984574.19 frames.], batch size: 30, lr: 7.89e-04
+2022-06-18 17:27:53,225 INFO [train.py:874] (2/4) Epoch 10, batch 2650, aishell_loss[loss=0.1818, simple_loss=0.2588, pruned_loss=0.05242, over 4958.00 frames.], tot_loss[loss=0.1759, simple_loss=0.2496, pruned_loss=0.05115, over 985785.82 frames.], batch size: 31, aishell_tot_loss[loss=0.1744, simple_loss=0.2547, pruned_loss=0.04703, over 983920.11 frames.], datatang_tot_loss[loss=0.1777, simple_loss=0.245, pruned_loss=0.05525, over 984724.92 frames.], batch size: 31, lr: 7.88e-04
+2022-06-18 17:28:23,886 INFO [train.py:874] (2/4) Epoch 10, batch 2700, datatang_loss[loss=0.1799, simple_loss=0.2429, pruned_loss=0.05845, over 4937.00 frames.], tot_loss[loss=0.1764, simple_loss=0.2493, pruned_loss=0.05171, over 985867.70 frames.], batch size: 88, aishell_tot_loss[loss=0.1739, simple_loss=0.2542, pruned_loss=0.04677, over 984083.01 frames.], datatang_tot_loss[loss=0.1785, simple_loss=0.2456, pruned_loss=0.0557, over 984964.40 frames.], batch size: 88, lr: 7.88e-04
+2022-06-18 17:28:56,001 INFO [train.py:874] (2/4) Epoch 10, batch 2750, aishell_loss[loss=0.1703, simple_loss=0.2645, pruned_loss=0.03803, over 4956.00 frames.], tot_loss[loss=0.1762, simple_loss=0.249, pruned_loss=0.05172, over 985697.42 frames.], batch size: 40, aishell_tot_loss[loss=0.1751, simple_loss=0.2552, pruned_loss=0.04752, over 984148.31 frames.], datatang_tot_loss[loss=0.1772, simple_loss=0.2443, pruned_loss=0.05503, over 985029.09 frames.], batch size: 40, lr: 7.87e-04
+2022-06-18 17:29:26,466 INFO [train.py:874] (2/4) Epoch 10, batch 2800, datatang_loss[loss=0.1934, simple_loss=0.2616, pruned_loss=0.06256, over 4927.00 frames.], tot_loss[loss=0.1758, simple_loss=0.2488, pruned_loss=0.05139, over 985390.96 frames.], batch size: 75, aishell_tot_loss[loss=0.1755, simple_loss=0.2554, pruned_loss=0.04776, over 984321.97 frames.], datatang_tot_loss[loss=0.1764, simple_loss=0.2436, pruned_loss=0.05458, over 984800.09 frames.], batch size: 75, lr: 7.87e-04
+2022-06-18 17:30:00,713 INFO [train.py:874] (2/4) Epoch 10, batch 2850, datatang_loss[loss=0.1655, simple_loss=0.2345, pruned_loss=0.04821, over 4923.00 frames.], tot_loss[loss=0.1766, simple_loss=0.2499, pruned_loss=0.05165, over 985547.06 frames.], batch size: 73, aishell_tot_loss[loss=0.1759, simple_loss=0.2561, pruned_loss=0.04783, over 984526.20 frames.], datatang_tot_loss[loss=0.1768, simple_loss=0.2439, pruned_loss=0.05488, over 984966.85 frames.], batch size: 73, lr: 7.86e-04
+2022-06-18 17:30:32,966 INFO [train.py:874] (2/4) Epoch 10, batch 2900, datatang_loss[loss=0.1569, simple_loss=0.2294, pruned_loss=0.04217, over 4958.00 frames.], tot_loss[loss=0.1775, simple_loss=0.2508, pruned_loss=0.05211, over 985339.21 frames.], batch size: 67, aishell_tot_loss[loss=0.1771, simple_loss=0.257, pruned_loss=0.04853, over 984548.60 frames.], datatang_tot_loss[loss=0.1767, simple_loss=0.244, pruned_loss=0.05471, over 984915.03 frames.], batch size: 67, lr: 7.86e-04
+2022-06-18 17:31:03,359 INFO [train.py:874] (2/4) Epoch 10, batch 2950, aishell_loss[loss=0.1837, simple_loss=0.266, pruned_loss=0.05071, over 4940.00 frames.], tot_loss[loss=0.1782, simple_loss=0.2512, pruned_loss=0.05264, over 985797.61 frames.], batch size: 68, aishell_tot_loss[loss=0.1769, simple_loss=0.2567, pruned_loss=0.04859, over 984916.68 frames.], datatang_tot_loss[loss=0.1777, simple_loss=0.2446, pruned_loss=0.0554, over 985175.96 frames.], batch size: 68, lr: 7.85e-04
+2022-06-18 17:31:32,443 INFO [train.py:874] (2/4) Epoch 10, batch 3000, aishell_loss[loss=0.1648, simple_loss=0.2491, pruned_loss=0.04022, over 4952.00 frames.], tot_loss[loss=0.1772, simple_loss=0.2502, pruned_loss=0.05215, over 985607.97 frames.], batch size: 40, aishell_tot_loss[loss=0.1764, simple_loss=0.2561, pruned_loss=0.04833, over 984898.62 frames.], datatang_tot_loss[loss=0.1774, simple_loss=0.2443, pruned_loss=0.05525, over 985173.46 frames.], batch size: 40, lr: 7.85e-04
+2022-06-18 17:31:32,444 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 17:31:49,157 INFO [train.py:914] (2/4) Epoch 10, validation: loss=0.1672, simple_loss=0.2512, pruned_loss=0.04165, over 1622729.00 frames.
+2022-06-18 17:32:18,362 INFO [train.py:874] (2/4) Epoch 10, batch 3050, aishell_loss[loss=0.1581, simple_loss=0.2404, pruned_loss=0.03793, over 4953.00 frames.], tot_loss[loss=0.1756, simple_loss=0.2487, pruned_loss=0.05127, over 985262.03 frames.], batch size: 56, aishell_tot_loss[loss=0.1755, simple_loss=0.2555, pruned_loss=0.04773, over 984752.26 frames.], datatang_tot_loss[loss=0.1767, simple_loss=0.2435, pruned_loss=0.05497, over 985089.48 frames.], batch size: 56, lr: 7.85e-04
+2022-06-18 17:32:48,941 INFO [train.py:874] (2/4) Epoch 10, batch 3100, aishell_loss[loss=0.1828, simple_loss=0.2717, pruned_loss=0.04692, over 4925.00 frames.], tot_loss[loss=0.1761, simple_loss=0.2495, pruned_loss=0.05133, over 985092.08 frames.], batch size: 41, aishell_tot_loss[loss=0.1757, simple_loss=0.2557, pruned_loss=0.04785, over 984665.10 frames.], datatang_tot_loss[loss=0.1768, simple_loss=0.2438, pruned_loss=0.05496, over 985084.30 frames.], batch size: 41, lr: 7.84e-04
+2022-06-18 17:33:21,081 INFO [train.py:874] (2/4) Epoch 10, batch 3150, aishell_loss[loss=0.1908, simple_loss=0.2723, pruned_loss=0.05464, over 4940.00 frames.], tot_loss[loss=0.1754, simple_loss=0.2488, pruned_loss=0.05106, over 985151.87 frames.], batch size: 32, aishell_tot_loss[loss=0.1753, simple_loss=0.2552, pruned_loss=0.04769, over 984673.49 frames.], datatang_tot_loss[loss=0.1766, simple_loss=0.2434, pruned_loss=0.05484, over 985179.16 frames.], batch size: 32, lr: 7.84e-04
+2022-06-18 17:33:49,712 INFO [train.py:874] (2/4) Epoch 10, batch 3200, aishell_loss[loss=0.1702, simple_loss=0.2602, pruned_loss=0.04011, over 4941.00 frames.], tot_loss[loss=0.1763, simple_loss=0.2496, pruned_loss=0.05147, over 985491.30 frames.], batch size: 58, aishell_tot_loss[loss=0.1757, simple_loss=0.2556, pruned_loss=0.0479, over 984960.24 frames.], datatang_tot_loss[loss=0.1769, simple_loss=0.2437, pruned_loss=0.05505, over 985305.69 frames.], batch size: 58, lr: 7.83e-04
+2022-06-18 17:34:21,039 INFO [train.py:874] (2/4) Epoch 10, batch 3250, datatang_loss[loss=0.1918, simple_loss=0.2639, pruned_loss=0.0598, over 4953.00 frames.], tot_loss[loss=0.1765, simple_loss=0.2498, pruned_loss=0.0516, over 986035.03 frames.], batch size: 86, aishell_tot_loss[loss=0.1753, simple_loss=0.2549, pruned_loss=0.04779, over 985458.60 frames.], datatang_tot_loss[loss=0.1776, simple_loss=0.2444, pruned_loss=0.05544, over 985455.54 frames.], batch size: 86, lr: 7.83e-04
+2022-06-18 17:34:51,782 INFO [train.py:874] (2/4) Epoch 10, batch 3300, datatang_loss[loss=0.166, simple_loss=0.236, pruned_loss=0.04794, over 4957.00 frames.], tot_loss[loss=0.1755, simple_loss=0.2485, pruned_loss=0.05121, over 985909.13 frames.], batch size: 86, aishell_tot_loss[loss=0.175, simple_loss=0.2546, pruned_loss=0.04773, over 985298.13 frames.], datatang_tot_loss[loss=0.1768, simple_loss=0.2436, pruned_loss=0.05497, over 985623.95 frames.], batch size: 86, lr: 7.82e-04
+2022-06-18 17:35:21,929 INFO [train.py:874] (2/4) Epoch 10, batch 3350, aishell_loss[loss=0.1648, simple_loss=0.2516, pruned_loss=0.03902, over 4950.00 frames.], tot_loss[loss=0.1765, simple_loss=0.2499, pruned_loss=0.05157, over 985928.06 frames.], batch size: 56, aishell_tot_loss[loss=0.1759, simple_loss=0.2557, pruned_loss=0.04802, over 985337.85 frames.], datatang_tot_loss[loss=0.177, simple_loss=0.2438, pruned_loss=0.05509, over 985712.43 frames.], batch size: 56, lr: 7.82e-04
+2022-06-18 17:35:53,647 INFO [train.py:874] (2/4) Epoch 10, batch 3400, aishell_loss[loss=0.1822, simple_loss=0.2653, pruned_loss=0.04951, over 4868.00 frames.], tot_loss[loss=0.1761, simple_loss=0.2503, pruned_loss=0.05089, over 986036.10 frames.], batch size: 38, aishell_tot_loss[loss=0.1762, simple_loss=0.2563, pruned_loss=0.04799, over 985453.25 frames.], datatang_tot_loss[loss=0.1762, simple_loss=0.2434, pruned_loss=0.05454, over 985812.14 frames.], batch size: 38, lr: 7.81e-04
+2022-06-18 17:36:23,921 INFO [train.py:874] (2/4) Epoch 10, batch 3450, datatang_loss[loss=0.1749, simple_loss=0.2429, pruned_loss=0.05341, over 4913.00 frames.], tot_loss[loss=0.1774, simple_loss=0.2517, pruned_loss=0.05159, over 985906.99 frames.], batch size: 75, aishell_tot_loss[loss=0.1766, simple_loss=0.2569, pruned_loss=0.04817, over 985308.90 frames.], datatang_tot_loss[loss=0.1772, simple_loss=0.2442, pruned_loss=0.05511, over 985926.33 frames.], batch size: 75, lr: 7.81e-04
+2022-06-18 17:36:53,798 INFO [train.py:874] (2/4) Epoch 10, batch 3500, aishell_loss[loss=0.189, simple_loss=0.2708, pruned_loss=0.05359, over 4960.00 frames.], tot_loss[loss=0.1777, simple_loss=0.2519, pruned_loss=0.05179, over 986003.53 frames.], batch size: 61, aishell_tot_loss[loss=0.1764, simple_loss=0.2567, pruned_loss=0.04808, over 985420.07 frames.], datatang_tot_loss[loss=0.1778, simple_loss=0.2447, pruned_loss=0.05544, over 985985.38 frames.], batch size: 61, lr: 7.80e-04
+2022-06-18 17:37:26,109 INFO [train.py:874] (2/4) Epoch 10, batch 3550, datatang_loss[loss=0.1436, simple_loss=0.2144, pruned_loss=0.03639, over 4927.00 frames.], tot_loss[loss=0.178, simple_loss=0.252, pruned_loss=0.05196, over 986122.69 frames.], batch size: 26, aishell_tot_loss[loss=0.1764, simple_loss=0.2567, pruned_loss=0.04808, over 985471.63 frames.], datatang_tot_loss[loss=0.1782, simple_loss=0.2453, pruned_loss=0.05552, over 986132.48 frames.], batch size: 26, lr: 7.80e-04
+2022-06-18 17:37:55,040 INFO [train.py:874] (2/4) Epoch 10, batch 3600, datatang_loss[loss=0.1603, simple_loss=0.2225, pruned_loss=0.04902, over 4954.00 frames.], tot_loss[loss=0.1783, simple_loss=0.2521, pruned_loss=0.05219, over 986020.73 frames.], batch size: 55, aishell_tot_loss[loss=0.1765, simple_loss=0.2565, pruned_loss=0.04829, over 985470.59 frames.], datatang_tot_loss[loss=0.1786, simple_loss=0.2457, pruned_loss=0.05578, over 986110.33 frames.], batch size: 55, lr: 7.79e-04
+2022-06-18 17:38:26,097 INFO [train.py:874] (2/4) Epoch 10, batch 3650, datatang_loss[loss=0.1673, simple_loss=0.2223, pruned_loss=0.05612, over 4980.00 frames.], tot_loss[loss=0.1787, simple_loss=0.2523, pruned_loss=0.05255, over 986140.43 frames.], batch size: 37, aishell_tot_loss[loss=0.1774, simple_loss=0.2572, pruned_loss=0.04879, over 985439.48 frames.], datatang_tot_loss[loss=0.1784, simple_loss=0.2455, pruned_loss=0.05565, over 986307.79 frames.], batch size: 37, lr: 7.79e-04
+2022-06-18 17:38:57,466 INFO [train.py:874] (2/4) Epoch 10, batch 3700, aishell_loss[loss=0.1718, simple_loss=0.2581, pruned_loss=0.04276, over 4955.00 frames.], tot_loss[loss=0.1796, simple_loss=0.253, pruned_loss=0.0531, over 986267.17 frames.], batch size: 64, aishell_tot_loss[loss=0.1774, simple_loss=0.2573, pruned_loss=0.04873, over 985577.39 frames.], datatang_tot_loss[loss=0.1795, simple_loss=0.2465, pruned_loss=0.05627, over 986353.20 frames.], batch size: 64, lr: 7.78e-04
+2022-06-18 17:39:26,931 INFO [train.py:874] (2/4) Epoch 10, batch 3750, datatang_loss[loss=0.1673, simple_loss=0.2416, pruned_loss=0.04651, over 4914.00 frames.], tot_loss[loss=0.1785, simple_loss=0.2521, pruned_loss=0.05247, over 986267.28 frames.], batch size: 81, aishell_tot_loss[loss=0.1764, simple_loss=0.2563, pruned_loss=0.04824, over 985676.16 frames.], datatang_tot_loss[loss=0.1797, simple_loss=0.2467, pruned_loss=0.05632, over 986337.84 frames.], batch size: 81, lr: 7.78e-04
+2022-06-18 17:39:57,895 INFO [train.py:874] (2/4) Epoch 10, batch 3800, datatang_loss[loss=0.1679, simple_loss=0.2393, pruned_loss=0.04831, over 4957.00 frames.], tot_loss[loss=0.1776, simple_loss=0.2515, pruned_loss=0.05184, over 986240.01 frames.], batch size: 67, aishell_tot_loss[loss=0.1772, simple_loss=0.2571, pruned_loss=0.04863, over 985712.65 frames.], datatang_tot_loss[loss=0.178, simple_loss=0.2454, pruned_loss=0.0553, over 986341.13 frames.], batch size: 67, lr: 7.77e-04
+2022-06-18 17:40:26,657 INFO [train.py:874] (2/4) Epoch 10, batch 3850, aishell_loss[loss=0.1982, simple_loss=0.2703, pruned_loss=0.063, over 4906.00 frames.], tot_loss[loss=0.1767, simple_loss=0.2504, pruned_loss=0.05149, over 986329.66 frames.], batch size: 46, aishell_tot_loss[loss=0.1764, simple_loss=0.2563, pruned_loss=0.04823, over 985684.00 frames.], datatang_tot_loss[loss=0.1778, simple_loss=0.2451, pruned_loss=0.05526, over 986523.27 frames.], batch size: 46, lr: 7.77e-04
+2022-06-18 17:40:56,136 INFO [train.py:874] (2/4) Epoch 10, batch 3900, datatang_loss[loss=0.1694, simple_loss=0.2393, pruned_loss=0.04975, over 4920.00 frames.], tot_loss[loss=0.1784, simple_loss=0.2515, pruned_loss=0.05261, over 986305.59 frames.], batch size: 81, aishell_tot_loss[loss=0.1765, simple_loss=0.2564, pruned_loss=0.04824, over 985702.90 frames.], datatang_tot_loss[loss=0.1794, simple_loss=0.2463, pruned_loss=0.0562, over 986518.38 frames.], batch size: 81, lr: 7.76e-04
+2022-06-18 17:41:24,503 INFO [train.py:874] (2/4) Epoch 10, batch 3950, datatang_loss[loss=0.167, simple_loss=0.2377, pruned_loss=0.04813, over 4957.00 frames.], tot_loss[loss=0.1785, simple_loss=0.2519, pruned_loss=0.05259, over 985978.10 frames.], batch size: 86, aishell_tot_loss[loss=0.1774, simple_loss=0.2572, pruned_loss=0.04876, over 985659.94 frames.], datatang_tot_loss[loss=0.1788, simple_loss=0.2459, pruned_loss=0.05585, over 986267.43 frames.], batch size: 86, lr: 7.76e-04
+2022-06-18 17:41:54,472 INFO [train.py:874] (2/4) Epoch 10, batch 4000, aishell_loss[loss=0.1929, simple_loss=0.2662, pruned_loss=0.05985, over 4942.00 frames.], tot_loss[loss=0.1791, simple_loss=0.2524, pruned_loss=0.05287, over 986042.45 frames.], batch size: 49, aishell_tot_loss[loss=0.1776, simple_loss=0.2574, pruned_loss=0.0489, over 985803.79 frames.], datatang_tot_loss[loss=0.1792, simple_loss=0.2461, pruned_loss=0.05612, over 986201.74 frames.], batch size: 49, lr: 7.76e-04
+2022-06-18 17:41:54,473 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 17:42:11,515 INFO [train.py:914] (2/4) Epoch 10, validation: loss=0.169, simple_loss=0.2525, pruned_loss=0.04271, over 1622729.00 frames.
+2022-06-18 17:42:40,784 INFO [train.py:874] (2/4) Epoch 10, batch 4050, aishell_loss[loss=0.1675, simple_loss=0.2553, pruned_loss=0.03988, over 4971.00 frames.], tot_loss[loss=0.1778, simple_loss=0.2512, pruned_loss=0.05218, over 985573.97 frames.], batch size: 51, aishell_tot_loss[loss=0.1762, simple_loss=0.256, pruned_loss=0.04822, over 985423.79 frames.], datatang_tot_loss[loss=0.1794, simple_loss=0.2463, pruned_loss=0.05625, over 986106.38 frames.], batch size: 51, lr: 7.75e-04
+2022-06-18 17:43:07,905 INFO [train.py:874] (2/4) Epoch 10, batch 4100, aishell_loss[loss=0.1996, simple_loss=0.2743, pruned_loss=0.06251, over 4912.00 frames.], tot_loss[loss=0.1775, simple_loss=0.2513, pruned_loss=0.05182, over 985291.84 frames.], batch size: 41, aishell_tot_loss[loss=0.1769, simple_loss=0.2569, pruned_loss=0.04839, over 985147.85 frames.], datatang_tot_loss[loss=0.1785, simple_loss=0.2453, pruned_loss=0.05581, over 986061.03 frames.], batch size: 41, lr: 7.75e-04
+2022-06-18 17:44:17,190 INFO [train.py:874] (2/4) Epoch 11, batch 50, aishell_loss[loss=0.1625, simple_loss=0.2453, pruned_loss=0.03983, over 4956.00 frames.], tot_loss[loss=0.1737, simple_loss=0.25, pruned_loss=0.04872, over 218593.24 frames.], batch size: 61, aishell_tot_loss[loss=0.1796, simple_loss=0.2618, pruned_loss=0.0487, over 137705.26 frames.], datatang_tot_loss[loss=0.1648, simple_loss=0.2321, pruned_loss=0.04876, over 94015.08 frames.], batch size: 61, lr: 7.46e-04
+2022-06-18 17:44:48,046 INFO [train.py:874] (2/4) Epoch 11, batch 100, datatang_loss[loss=0.1596, simple_loss=0.2345, pruned_loss=0.04234, over 4941.00 frames.], tot_loss[loss=0.1708, simple_loss=0.2464, pruned_loss=0.04759, over 388588.65 frames.], batch size: 69, aishell_tot_loss[loss=0.1754, simple_loss=0.257, pruned_loss=0.0469, over 233652.27 frames.], datatang_tot_loss[loss=0.1659, simple_loss=0.2346, pruned_loss=0.04861, over 203065.60 frames.], batch size: 69, lr: 7.45e-04
+2022-06-18 17:45:16,810 INFO [train.py:874] (2/4) Epoch 11, batch 150, aishell_loss[loss=0.1753, simple_loss=0.2614, pruned_loss=0.04462, over 4884.00 frames.], tot_loss[loss=0.169, simple_loss=0.2425, pruned_loss=0.04774, over 520407.64 frames.], batch size: 42, aishell_tot_loss[loss=0.1728, simple_loss=0.2537, pruned_loss=0.04593, over 298228.57 frames.], datatang_tot_loss[loss=0.1663, simple_loss=0.2337, pruned_loss=0.04947, over 318768.73 frames.], batch size: 42, lr: 7.45e-04
+2022-06-18 17:45:48,301 INFO [train.py:874] (2/4) Epoch 11, batch 200, datatang_loss[loss=0.1384, simple_loss=0.2182, pruned_loss=0.02932, over 4922.00 frames.], tot_loss[loss=0.1708, simple_loss=0.2446, pruned_loss=0.04848, over 623718.83 frames.], batch size: 83, aishell_tot_loss[loss=0.1757, simple_loss=0.2558, pruned_loss=0.0478, over 385447.01 frames.], datatang_tot_loss[loss=0.1659, simple_loss=0.2339, pruned_loss=0.04898, over 391370.52 frames.], batch size: 83, lr: 7.44e-04
+2022-06-18 17:46:17,711 INFO [train.py:874] (2/4) Epoch 11, batch 250, datatang_loss[loss=0.1944, simple_loss=0.2563, pruned_loss=0.06624, over 4978.00 frames.], tot_loss[loss=0.1712, simple_loss=0.2452, pruned_loss=0.04862, over 703695.84 frames.], batch size: 60, aishell_tot_loss[loss=0.1747, simple_loss=0.2545, pruned_loss=0.04745, over 466426.28 frames.], datatang_tot_loss[loss=0.1672, simple_loss=0.2351, pruned_loss=0.04967, over 450703.15 frames.], batch size: 60, lr: 7.44e-04
+2022-06-18 17:46:47,524 INFO [train.py:874] (2/4) Epoch 11, batch 300, datatang_loss[loss=0.1485, simple_loss=0.2235, pruned_loss=0.03678, over 4923.00 frames.], tot_loss[loss=0.1714, simple_loss=0.2451, pruned_loss=0.04887, over 766220.81 frames.], batch size: 77, aishell_tot_loss[loss=0.174, simple_loss=0.2539, pruned_loss=0.04711, over 516043.40 frames.], datatang_tot_loss[loss=0.1686, simple_loss=0.2366, pruned_loss=0.05034, over 525320.77 frames.], batch size: 77, lr: 7.43e-04
+2022-06-18 17:47:17,424 INFO [train.py:874] (2/4) Epoch 11, batch 350, aishell_loss[loss=0.1794, simple_loss=0.2637, pruned_loss=0.04758, over 4948.00 frames.], tot_loss[loss=0.1734, simple_loss=0.2482, pruned_loss=0.0493, over 814909.74 frames.], batch size: 54, aishell_tot_loss[loss=0.1754, simple_loss=0.2557, pruned_loss=0.0475, over 585613.98 frames.], datatang_tot_loss[loss=0.17, simple_loss=0.2384, pruned_loss=0.05083, over 565101.99 frames.], batch size: 54, lr: 7.43e-04
+2022-06-18 17:47:47,201 INFO [train.py:874] (2/4) Epoch 11, batch 400, aishell_loss[loss=0.1676, simple_loss=0.2432, pruned_loss=0.04605, over 4988.00 frames.], tot_loss[loss=0.1733, simple_loss=0.2478, pruned_loss=0.04942, over 853085.16 frames.], batch size: 25, aishell_tot_loss[loss=0.1747, simple_loss=0.255, pruned_loss=0.04719, over 631285.50 frames.], datatang_tot_loss[loss=0.1709, simple_loss=0.2392, pruned_loss=0.05134, over 616496.33 frames.], batch size: 25, lr: 7.42e-04
+2022-06-18 17:48:17,508 INFO [train.py:874] (2/4) Epoch 11, batch 450, datatang_loss[loss=0.1865, simple_loss=0.2583, pruned_loss=0.05738, over 4937.00 frames.], tot_loss[loss=0.1732, simple_loss=0.2477, pruned_loss=0.04937, over 882538.30 frames.], batch size: 69, aishell_tot_loss[loss=0.1743, simple_loss=0.2548, pruned_loss=0.04691, over 658539.11 frames.], datatang_tot_loss[loss=0.1716, simple_loss=0.2404, pruned_loss=0.05134, over 674456.68 frames.], batch size: 69, lr: 7.42e-04
+2022-06-18 17:48:48,099 INFO [train.py:874] (2/4) Epoch 11, batch 500, datatang_loss[loss=0.1846, simple_loss=0.2525, pruned_loss=0.05831, over 4959.00 frames.], tot_loss[loss=0.174, simple_loss=0.2488, pruned_loss=0.04964, over 905167.64 frames.], batch size: 91, aishell_tot_loss[loss=0.1745, simple_loss=0.2555, pruned_loss=0.04676, over 697097.01 frames.], datatang_tot_loss[loss=0.1725, simple_loss=0.2411, pruned_loss=0.05191, over 710838.32 frames.], batch size: 91, lr: 7.42e-04
+2022-06-18 17:49:17,348 INFO [train.py:874] (2/4) Epoch 11, batch 550, datatang_loss[loss=0.1635, simple_loss=0.2334, pruned_loss=0.04682, over 4913.00 frames.], tot_loss[loss=0.1735, simple_loss=0.2481, pruned_loss=0.04949, over 922842.51 frames.], batch size: 75, aishell_tot_loss[loss=0.1744, simple_loss=0.2553, pruned_loss=0.04676, over 730943.77 frames.], datatang_tot_loss[loss=0.1722, simple_loss=0.2406, pruned_loss=0.05184, over 743173.45 frames.], batch size: 75, lr: 7.41e-04
+2022-06-18 17:49:48,433 INFO [train.py:874] (2/4) Epoch 11, batch 600, datatang_loss[loss=0.1568, simple_loss=0.2303, pruned_loss=0.04159, over 4921.00 frames.], tot_loss[loss=0.1739, simple_loss=0.2482, pruned_loss=0.04977, over 936549.46 frames.], batch size: 25, aishell_tot_loss[loss=0.1747, simple_loss=0.2556, pruned_loss=0.04684, over 756249.73 frames.], datatang_tot_loss[loss=0.1724, simple_loss=0.2408, pruned_loss=0.05204, over 775886.54 frames.], batch size: 25, lr: 7.41e-04
+2022-06-18 17:50:18,130 INFO [train.py:874] (2/4) Epoch 11, batch 650, aishell_loss[loss=0.1643, simple_loss=0.2447, pruned_loss=0.04199, over 4908.00 frames.], tot_loss[loss=0.1736, simple_loss=0.2483, pruned_loss=0.04943, over 947293.48 frames.], batch size: 46, aishell_tot_loss[loss=0.1745, simple_loss=0.2553, pruned_loss=0.04684, over 788029.61 frames.], datatang_tot_loss[loss=0.1723, simple_loss=0.2408, pruned_loss=0.05188, over 796011.99 frames.], batch size: 46, lr: 7.40e-04
+2022-06-18 17:50:47,911 INFO [train.py:874] (2/4) Epoch 11, batch 700, aishell_loss[loss=0.1617, simple_loss=0.2453, pruned_loss=0.03905, over 4938.00 frames.], tot_loss[loss=0.1744, simple_loss=0.2491, pruned_loss=0.04979, over 955803.17 frames.], batch size: 49, aishell_tot_loss[loss=0.1741, simple_loss=0.2549, pruned_loss=0.0466, over 813053.56 frames.], datatang_tot_loss[loss=0.1737, simple_loss=0.2422, pruned_loss=0.05266, over 816650.99 frames.], batch size: 49, lr: 7.40e-04
+2022-06-18 17:51:18,372 INFO [train.py:874] (2/4) Epoch 11, batch 750, aishell_loss[loss=0.1639, simple_loss=0.2484, pruned_loss=0.03969, over 4970.00 frames.], tot_loss[loss=0.1752, simple_loss=0.2498, pruned_loss=0.05035, over 962371.21 frames.], batch size: 44, aishell_tot_loss[loss=0.1745, simple_loss=0.2552, pruned_loss=0.04691, over 832544.84 frames.], datatang_tot_loss[loss=0.1745, simple_loss=0.2429, pruned_loss=0.05305, over 837324.22 frames.], batch size: 44, lr: 7.39e-04
+2022-06-18 17:51:48,627 INFO [train.py:874] (2/4) Epoch 11, batch 800, datatang_loss[loss=0.1974, simple_loss=0.2587, pruned_loss=0.06809, over 4977.00 frames.], tot_loss[loss=0.1745, simple_loss=0.2489, pruned_loss=0.05004, over 967416.29 frames.], batch size: 60, aishell_tot_loss[loss=0.174, simple_loss=0.2545, pruned_loss=0.04679, over 851568.82 frames.], datatang_tot_loss[loss=0.1742, simple_loss=0.2426, pruned_loss=0.05294, over 853685.95 frames.], batch size: 60, lr: 7.39e-04
+2022-06-18 17:52:17,417 INFO [train.py:874] (2/4) Epoch 11, batch 850, aishell_loss[loss=0.1633, simple_loss=0.2515, pruned_loss=0.03752, over 4970.00 frames.], tot_loss[loss=0.1746, simple_loss=0.2491, pruned_loss=0.05001, over 971140.71 frames.], batch size: 39, aishell_tot_loss[loss=0.1745, simple_loss=0.2549, pruned_loss=0.047, over 868771.13 frames.], datatang_tot_loss[loss=0.174, simple_loss=0.2423, pruned_loss=0.05281, over 867467.32 frames.], batch size: 39, lr: 7.39e-04
+2022-06-18 17:52:48,390 INFO [train.py:874] (2/4) Epoch 11, batch 900, datatang_loss[loss=0.1895, simple_loss=0.2424, pruned_loss=0.06832, over 4902.00 frames.], tot_loss[loss=0.1749, simple_loss=0.2493, pruned_loss=0.05026, over 974435.98 frames.], batch size: 47, aishell_tot_loss[loss=0.1742, simple_loss=0.2547, pruned_loss=0.04685, over 880933.58 frames.], datatang_tot_loss[loss=0.1748, simple_loss=0.2431, pruned_loss=0.05318, over 883070.81 frames.], batch size: 47, lr: 7.38e-04
+2022-06-18 17:53:18,615 INFO [train.py:874] (2/4) Epoch 11, batch 950, datatang_loss[loss=0.1685, simple_loss=0.2332, pruned_loss=0.05186, over 4924.00 frames.], tot_loss[loss=0.1758, simple_loss=0.2501, pruned_loss=0.0507, over 977314.63 frames.], batch size: 73, aishell_tot_loss[loss=0.1741, simple_loss=0.2546, pruned_loss=0.04682, over 892919.16 frames.], datatang_tot_loss[loss=0.1759, simple_loss=0.2443, pruned_loss=0.05378, over 895910.45 frames.], batch size: 73, lr: 7.38e-04
+2022-06-18 17:53:46,776 INFO [train.py:874] (2/4) Epoch 11, batch 1000, aishell_loss[loss=0.1945, simple_loss=0.274, pruned_loss=0.05754, over 4882.00 frames.], tot_loss[loss=0.1771, simple_loss=0.2515, pruned_loss=0.05132, over 979244.26 frames.], batch size: 47, aishell_tot_loss[loss=0.1747, simple_loss=0.2553, pruned_loss=0.04708, over 906306.49 frames.], datatang_tot_loss[loss=0.177, simple_loss=0.245, pruned_loss=0.05454, over 904100.36 frames.], batch size: 47, lr: 7.37e-04
+2022-06-18 17:53:46,777 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 17:54:02,789 INFO [train.py:914] (2/4) Epoch 11, validation: loss=0.1676, simple_loss=0.2521, pruned_loss=0.04158, over 1622729.00 frames.
+2022-06-18 17:54:32,112 INFO [train.py:874] (2/4) Epoch 11, batch 1050, aishell_loss[loss=0.1463, simple_loss=0.2333, pruned_loss=0.02965, over 4873.00 frames.], tot_loss[loss=0.1759, simple_loss=0.251, pruned_loss=0.05041, over 980630.04 frames.], batch size: 28, aishell_tot_loss[loss=0.1742, simple_loss=0.2549, pruned_loss=0.04673, over 917947.92 frames.], datatang_tot_loss[loss=0.1767, simple_loss=0.2448, pruned_loss=0.05425, over 911217.51 frames.], batch size: 28, lr: 7.37e-04
+2022-06-18 17:55:02,907 INFO [train.py:874] (2/4) Epoch 11, batch 1100, datatang_loss[loss=0.1672, simple_loss=0.2403, pruned_loss=0.04704, over 4964.00 frames.], tot_loss[loss=0.1751, simple_loss=0.2504, pruned_loss=0.04991, over 981620.59 frames.], batch size: 25, aishell_tot_loss[loss=0.1741, simple_loss=0.2548, pruned_loss=0.04668, over 926607.58 frames.], datatang_tot_loss[loss=0.176, simple_loss=0.2443, pruned_loss=0.05382, over 919069.12 frames.], batch size: 25, lr: 7.36e-04
+2022-06-18 17:55:32,142 INFO [train.py:874] (2/4) Epoch 11, batch 1150, datatang_loss[loss=0.1483, simple_loss=0.2122, pruned_loss=0.04216, over 4912.00 frames.], tot_loss[loss=0.1743, simple_loss=0.2493, pruned_loss=0.04968, over 982652.58 frames.], batch size: 64, aishell_tot_loss[loss=0.174, simple_loss=0.2547, pruned_loss=0.04665, over 932648.92 frames.], datatang_tot_loss[loss=0.1752, simple_loss=0.2437, pruned_loss=0.05342, over 928072.88 frames.], batch size: 64, lr: 7.36e-04
+2022-06-18 17:56:02,932 INFO [train.py:874] (2/4) Epoch 11, batch 1200, aishell_loss[loss=0.2016, simple_loss=0.28, pruned_loss=0.06158, over 4987.00 frames.], tot_loss[loss=0.1743, simple_loss=0.2488, pruned_loss=0.04995, over 983505.05 frames.], batch size: 39, aishell_tot_loss[loss=0.1739, simple_loss=0.2543, pruned_loss=0.04673, over 937936.46 frames.], datatang_tot_loss[loss=0.1752, simple_loss=0.2437, pruned_loss=0.0534, over 936061.66 frames.], batch size: 39, lr: 7.36e-04
+2022-06-18 17:56:33,867 INFO [train.py:874] (2/4) Epoch 11, batch 1250, datatang_loss[loss=0.1933, simple_loss=0.2451, pruned_loss=0.07073, over 4986.00 frames.], tot_loss[loss=0.1734, simple_loss=0.2473, pruned_loss=0.04975, over 984209.73 frames.], batch size: 26, aishell_tot_loss[loss=0.1731, simple_loss=0.2535, pruned_loss=0.04636, over 942875.88 frames.], datatang_tot_loss[loss=0.1749, simple_loss=0.243, pruned_loss=0.05336, over 942830.06 frames.], batch size: 26, lr: 7.35e-04
+2022-06-18 17:57:03,619 INFO [train.py:874] (2/4) Epoch 11, batch 1300, datatang_loss[loss=0.1703, simple_loss=0.2361, pruned_loss=0.05226, over 4942.00 frames.], tot_loss[loss=0.172, simple_loss=0.246, pruned_loss=0.04901, over 984735.72 frames.], batch size: 50, aishell_tot_loss[loss=0.1721, simple_loss=0.2526, pruned_loss=0.04584, over 947806.05 frames.], datatang_tot_loss[loss=0.1741, simple_loss=0.2423, pruned_loss=0.053, over 948207.13 frames.], batch size: 50, lr: 7.35e-04
+2022-06-18 17:57:34,404 INFO [train.py:874] (2/4) Epoch 11, batch 1350, datatang_loss[loss=0.1691, simple_loss=0.2349, pruned_loss=0.05159, over 4916.00 frames.], tot_loss[loss=0.1719, simple_loss=0.2462, pruned_loss=0.0488, over 985016.74 frames.], batch size: 57, aishell_tot_loss[loss=0.1722, simple_loss=0.2528, pruned_loss=0.04575, over 952125.71 frames.], datatang_tot_loss[loss=0.1737, simple_loss=0.2421, pruned_loss=0.0527, over 952824.85 frames.], batch size: 57, lr: 7.34e-04
+2022-06-18 17:58:05,155 INFO [train.py:874] (2/4) Epoch 11, batch 1400, datatang_loss[loss=0.1681, simple_loss=0.2471, pruned_loss=0.0445, over 4900.00 frames.], tot_loss[loss=0.1718, simple_loss=0.2467, pruned_loss=0.04847, over 985378.34 frames.], batch size: 24, aishell_tot_loss[loss=0.1719, simple_loss=0.2527, pruned_loss=0.04553, over 956800.34 frames.], datatang_tot_loss[loss=0.1736, simple_loss=0.2421, pruned_loss=0.05257, over 956204.46 frames.], batch size: 24, lr: 7.34e-04
+2022-06-18 17:58:33,316 INFO [train.py:874] (2/4) Epoch 11, batch 1450, datatang_loss[loss=0.1561, simple_loss=0.2281, pruned_loss=0.042, over 4960.00 frames.], tot_loss[loss=0.1713, simple_loss=0.2461, pruned_loss=0.04828, over 985524.39 frames.], batch size: 67, aishell_tot_loss[loss=0.1711, simple_loss=0.2517, pruned_loss=0.04525, over 960775.10 frames.], datatang_tot_loss[loss=0.1736, simple_loss=0.242, pruned_loss=0.05266, over 959149.36 frames.], batch size: 67, lr: 7.33e-04
+2022-06-18 17:59:04,228 INFO [train.py:874] (2/4) Epoch 11, batch 1500, datatang_loss[loss=0.1697, simple_loss=0.2307, pruned_loss=0.05432, over 4938.00 frames.], tot_loss[loss=0.1721, simple_loss=0.247, pruned_loss=0.04856, over 985573.86 frames.], batch size: 42, aishell_tot_loss[loss=0.1715, simple_loss=0.2522, pruned_loss=0.04541, over 963678.30 frames.], datatang_tot_loss[loss=0.1738, simple_loss=0.2424, pruned_loss=0.05265, over 962319.15 frames.], batch size: 42, lr: 7.33e-04
+2022-06-18 17:59:35,778 INFO [train.py:874] (2/4) Epoch 11, batch 1550, aishell_loss[loss=0.1447, simple_loss=0.2116, pruned_loss=0.0389, over 4975.00 frames.], tot_loss[loss=0.1724, simple_loss=0.2474, pruned_loss=0.04867, over 985472.98 frames.], batch size: 25, aishell_tot_loss[loss=0.1716, simple_loss=0.2522, pruned_loss=0.04543, over 965978.34 frames.], datatang_tot_loss[loss=0.174, simple_loss=0.2427, pruned_loss=0.05264, over 965244.67 frames.], batch size: 25, lr: 7.33e-04
+2022-06-18 18:00:05,424 INFO [train.py:874] (2/4) Epoch 11, batch 1600, datatang_loss[loss=0.1598, simple_loss=0.2359, pruned_loss=0.04178, over 4958.00 frames.], tot_loss[loss=0.1732, simple_loss=0.2479, pruned_loss=0.04931, over 985812.89 frames.], batch size: 31, aishell_tot_loss[loss=0.1711, simple_loss=0.2517, pruned_loss=0.04527, over 968386.36 frames.], datatang_tot_loss[loss=0.1752, simple_loss=0.2438, pruned_loss=0.05336, over 967877.59 frames.], batch size: 31, lr: 7.32e-04
+2022-06-18 18:00:35,332 INFO [train.py:874] (2/4) Epoch 11, batch 1650, aishell_loss[loss=0.181, simple_loss=0.2642, pruned_loss=0.04887, over 4930.00 frames.], tot_loss[loss=0.1736, simple_loss=0.248, pruned_loss=0.04965, over 985479.29 frames.], batch size: 32, aishell_tot_loss[loss=0.1714, simple_loss=0.2518, pruned_loss=0.04548, over 969662.07 frames.], datatang_tot_loss[loss=0.1754, simple_loss=0.244, pruned_loss=0.05337, over 970404.63 frames.], batch size: 32, lr: 7.32e-04
+2022-06-18 18:01:06,989 INFO [train.py:874] (2/4) Epoch 11, batch 1700, aishell_loss[loss=0.1768, simple_loss=0.2642, pruned_loss=0.04469, over 4930.00 frames.], tot_loss[loss=0.1741, simple_loss=0.2486, pruned_loss=0.04982, over 985345.02 frames.], batch size: 54, aishell_tot_loss[loss=0.1721, simple_loss=0.2527, pruned_loss=0.04578, over 970979.20 frames.], datatang_tot_loss[loss=0.1752, simple_loss=0.2439, pruned_loss=0.05326, over 972571.42 frames.], batch size: 54, lr: 7.31e-04
+2022-06-18 18:01:36,119 INFO [train.py:874] (2/4) Epoch 11, batch 1750, aishell_loss[loss=0.1859, simple_loss=0.2646, pruned_loss=0.05354, over 4883.00 frames.], tot_loss[loss=0.175, simple_loss=0.2492, pruned_loss=0.05044, over 985733.74 frames.], batch size: 42, aishell_tot_loss[loss=0.1728, simple_loss=0.253, pruned_loss=0.04628, over 973054.87 frames.], datatang_tot_loss[loss=0.1757, simple_loss=0.2443, pruned_loss=0.0536, over 974119.57 frames.], batch size: 42, lr: 7.31e-04
+2022-06-18 18:02:06,153 INFO [train.py:874] (2/4) Epoch 11, batch 1800, datatang_loss[loss=0.1423, simple_loss=0.2138, pruned_loss=0.03543, over 4950.00 frames.], tot_loss[loss=0.1738, simple_loss=0.2485, pruned_loss=0.04958, over 985532.10 frames.], batch size: 45, aishell_tot_loss[loss=0.1725, simple_loss=0.253, pruned_loss=0.04599, over 974312.47 frames.], datatang_tot_loss[loss=0.1749, simple_loss=0.2436, pruned_loss=0.05312, over 975528.77 frames.], batch size: 45, lr: 7.30e-04
+2022-06-18 18:02:37,427 INFO [train.py:874] (2/4) Epoch 11, batch 1850, aishell_loss[loss=0.1338, simple_loss=0.2011, pruned_loss=0.03328, over 4802.00 frames.], tot_loss[loss=0.1739, simple_loss=0.2487, pruned_loss=0.04951, over 985381.76 frames.], batch size: 21, aishell_tot_loss[loss=0.1721, simple_loss=0.2527, pruned_loss=0.04578, over 975452.39 frames.], datatang_tot_loss[loss=0.1753, simple_loss=0.2441, pruned_loss=0.05327, over 976742.95 frames.], batch size: 21, lr: 7.30e-04
+2022-06-18 18:03:06,837 INFO [train.py:874] (2/4) Epoch 11, batch 1900, datatang_loss[loss=0.1711, simple_loss=0.2392, pruned_loss=0.05149, over 4935.00 frames.], tot_loss[loss=0.1729, simple_loss=0.2478, pruned_loss=0.04898, over 985233.30 frames.], batch size: 79, aishell_tot_loss[loss=0.1717, simple_loss=0.2525, pruned_loss=0.04549, over 976217.22 frames.], datatang_tot_loss[loss=0.1746, simple_loss=0.2436, pruned_loss=0.05285, over 977985.98 frames.], batch size: 79, lr: 7.30e-04
+2022-06-18 18:03:36,011 INFO [train.py:874] (2/4) Epoch 11, batch 1950, datatang_loss[loss=0.1559, simple_loss=0.2284, pruned_loss=0.04171, over 4920.00 frames.], tot_loss[loss=0.1735, simple_loss=0.2485, pruned_loss=0.04918, over 985126.30 frames.], batch size: 73, aishell_tot_loss[loss=0.1725, simple_loss=0.2533, pruned_loss=0.04584, over 977267.90 frames.], datatang_tot_loss[loss=0.1745, simple_loss=0.2434, pruned_loss=0.05282, over 978774.83 frames.], batch size: 73, lr: 7.29e-04
+2022-06-18 18:04:07,290 INFO [train.py:874] (2/4) Epoch 11, batch 2000, aishell_loss[loss=0.1735, simple_loss=0.2559, pruned_loss=0.04562, over 4883.00 frames.], tot_loss[loss=0.1726, simple_loss=0.2476, pruned_loss=0.04885, over 984969.18 frames.], batch size: 35, aishell_tot_loss[loss=0.1719, simple_loss=0.2528, pruned_loss=0.0455, over 977829.21 frames.], datatang_tot_loss[loss=0.1742, simple_loss=0.2429, pruned_loss=0.05275, over 979719.15 frames.], batch size: 35, lr: 7.29e-04
+2022-06-18 18:04:07,291 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 18:04:24,293 INFO [train.py:914] (2/4) Epoch 11, validation: loss=0.167, simple_loss=0.2509, pruned_loss=0.04156, over 1622729.00 frames.
+2022-06-18 18:04:53,779 INFO [train.py:874] (2/4) Epoch 11, batch 2050, aishell_loss[loss=0.1661, simple_loss=0.2443, pruned_loss=0.04394, over 4981.00 frames.], tot_loss[loss=0.1728, simple_loss=0.2478, pruned_loss=0.04893, over 985367.44 frames.], batch size: 48, aishell_tot_loss[loss=0.1715, simple_loss=0.2524, pruned_loss=0.04533, over 979094.07 frames.], datatang_tot_loss[loss=0.1747, simple_loss=0.2433, pruned_loss=0.05311, over 980356.61 frames.], batch size: 48, lr: 7.28e-04
+2022-06-18 18:05:23,745 INFO [train.py:874] (2/4) Epoch 11, batch 2100, aishell_loss[loss=0.1498, simple_loss=0.2381, pruned_loss=0.03077, over 4920.00 frames.], tot_loss[loss=0.1727, simple_loss=0.2482, pruned_loss=0.04865, over 985296.98 frames.], batch size: 33, aishell_tot_loss[loss=0.1713, simple_loss=0.2523, pruned_loss=0.04512, over 979800.13 frames.], datatang_tot_loss[loss=0.1749, simple_loss=0.2435, pruned_loss=0.05316, over 980933.91 frames.], batch size: 33, lr: 7.28e-04
+2022-06-18 18:05:53,426 INFO [train.py:874] (2/4) Epoch 11, batch 2150, datatang_loss[loss=0.158, simple_loss=0.2352, pruned_loss=0.04039, over 4951.00 frames.], tot_loss[loss=0.174, simple_loss=0.2488, pruned_loss=0.04958, over 985486.63 frames.], batch size: 86, aishell_tot_loss[loss=0.1714, simple_loss=0.2526, pruned_loss=0.04512, over 980080.14 frames.], datatang_tot_loss[loss=0.1758, simple_loss=0.2443, pruned_loss=0.05363, over 981929.83 frames.], batch size: 86, lr: 7.28e-04
+2022-06-18 18:06:24,917 INFO [train.py:874] (2/4) Epoch 11, batch 2200, datatang_loss[loss=0.1637, simple_loss=0.2342, pruned_loss=0.04662, over 4920.00 frames.], tot_loss[loss=0.173, simple_loss=0.248, pruned_loss=0.04905, over 985897.72 frames.], batch size: 75, aishell_tot_loss[loss=0.1719, simple_loss=0.2533, pruned_loss=0.04522, over 980695.90 frames.], datatang_tot_loss[loss=0.1744, simple_loss=0.2432, pruned_loss=0.0528, over 982752.92 frames.], batch size: 75, lr: 7.27e-04
+2022-06-18 18:06:54,554 INFO [train.py:874] (2/4) Epoch 11, batch 2250, aishell_loss[loss=0.1726, simple_loss=0.2549, pruned_loss=0.04515, over 4955.00 frames.], tot_loss[loss=0.174, simple_loss=0.249, pruned_loss=0.04948, over 985806.57 frames.], batch size: 54, aishell_tot_loss[loss=0.1721, simple_loss=0.2536, pruned_loss=0.0453, over 981173.99 frames.], datatang_tot_loss[loss=0.1751, simple_loss=0.244, pruned_loss=0.05308, over 983132.76 frames.], batch size: 54, lr: 7.27e-04
+2022-06-18 18:07:24,135 INFO [train.py:874] (2/4) Epoch 11, batch 2300, datatang_loss[loss=0.1511, simple_loss=0.2284, pruned_loss=0.03687, over 4958.00 frames.], tot_loss[loss=0.1729, simple_loss=0.248, pruned_loss=0.0489, over 985533.68 frames.], batch size: 86, aishell_tot_loss[loss=0.172, simple_loss=0.2533, pruned_loss=0.04535, over 981315.20 frames.], datatang_tot_loss[loss=0.1742, simple_loss=0.2433, pruned_loss=0.05253, over 983563.67 frames.], batch size: 86, lr: 7.26e-04
+2022-06-18 18:07:55,326 INFO [train.py:874] (2/4) Epoch 11, batch 2350, datatang_loss[loss=0.1881, simple_loss=0.2498, pruned_loss=0.06323, over 4963.00 frames.], tot_loss[loss=0.1727, simple_loss=0.2478, pruned_loss=0.04884, over 985980.05 frames.], batch size: 55, aishell_tot_loss[loss=0.1717, simple_loss=0.2528, pruned_loss=0.04529, over 982028.08 frames.], datatang_tot_loss[loss=0.1742, simple_loss=0.2432, pruned_loss=0.05264, over 984103.02 frames.], batch size: 55, lr: 7.26e-04
+2022-06-18 18:08:25,218 INFO [train.py:874] (2/4) Epoch 11, batch 2400, aishell_loss[loss=0.2006, simple_loss=0.2834, pruned_loss=0.05891, over 4949.00 frames.], tot_loss[loss=0.1735, simple_loss=0.2488, pruned_loss=0.04908, over 986200.43 frames.], batch size: 45, aishell_tot_loss[loss=0.1719, simple_loss=0.2534, pruned_loss=0.04524, over 982755.75 frames.], datatang_tot_loss[loss=0.1748, simple_loss=0.2435, pruned_loss=0.05301, over 984304.98 frames.], batch size: 45, lr: 7.25e-04
+2022-06-18 18:08:55,148 INFO [train.py:874] (2/4) Epoch 11, batch 2450, datatang_loss[loss=0.1627, simple_loss=0.2346, pruned_loss=0.04543, over 4922.00 frames.], tot_loss[loss=0.1738, simple_loss=0.249, pruned_loss=0.04935, over 986247.07 frames.], batch size: 47, aishell_tot_loss[loss=0.1721, simple_loss=0.2534, pruned_loss=0.04535, over 983098.44 frames.], datatang_tot_loss[loss=0.175, simple_loss=0.2438, pruned_loss=0.05312, over 984630.37 frames.], batch size: 47, lr: 7.25e-04
+2022-06-18 18:09:26,583 INFO [train.py:874] (2/4) Epoch 11, batch 2500, datatang_loss[loss=0.1628, simple_loss=0.2279, pruned_loss=0.04889, over 4905.00 frames.], tot_loss[loss=0.1735, simple_loss=0.2486, pruned_loss=0.04917, over 986371.84 frames.], batch size: 52, aishell_tot_loss[loss=0.1722, simple_loss=0.2536, pruned_loss=0.04542, over 983597.01 frames.], datatang_tot_loss[loss=0.1745, simple_loss=0.2434, pruned_loss=0.05276, over 984809.73 frames.], batch size: 52, lr: 7.25e-04
+2022-06-18 18:09:56,446 INFO [train.py:874] (2/4) Epoch 11, batch 2550, datatang_loss[loss=0.1353, simple_loss=0.1983, pruned_loss=0.03614, over 4971.00 frames.], tot_loss[loss=0.1729, simple_loss=0.2481, pruned_loss=0.04879, over 986283.24 frames.], batch size: 45, aishell_tot_loss[loss=0.1724, simple_loss=0.2539, pruned_loss=0.0454, over 983828.54 frames.], datatang_tot_loss[loss=0.1737, simple_loss=0.2427, pruned_loss=0.05239, over 984995.17 frames.], batch size: 45, lr: 7.24e-04
+2022-06-18 18:10:26,143 INFO [train.py:874] (2/4) Epoch 11, batch 2600, datatang_loss[loss=0.1936, simple_loss=0.253, pruned_loss=0.06709, over 4972.00 frames.], tot_loss[loss=0.1746, simple_loss=0.2499, pruned_loss=0.04961, over 986248.10 frames.], batch size: 37, aishell_tot_loss[loss=0.1722, simple_loss=0.2539, pruned_loss=0.04529, over 984054.39 frames.], datatang_tot_loss[loss=0.1757, simple_loss=0.2444, pruned_loss=0.05344, over 985183.63 frames.], batch size: 37, lr: 7.24e-04
+2022-06-18 18:10:56,969 INFO [train.py:874] (2/4) Epoch 11, batch 2650, datatang_loss[loss=0.2068, simple_loss=0.2685, pruned_loss=0.07258, over 4925.00 frames.], tot_loss[loss=0.1733, simple_loss=0.2486, pruned_loss=0.04897, over 985840.59 frames.], batch size: 83, aishell_tot_loss[loss=0.1712, simple_loss=0.2528, pruned_loss=0.04478, over 983778.65 frames.], datatang_tot_loss[loss=0.1755, simple_loss=0.2441, pruned_loss=0.05347, over 985433.09 frames.], batch size: 83, lr: 7.23e-04
+2022-06-18 18:11:31,785 INFO [train.py:874] (2/4) Epoch 11, batch 2700, aishell_loss[loss=0.1812, simple_loss=0.2635, pruned_loss=0.04948, over 4960.00 frames.], tot_loss[loss=0.1732, simple_loss=0.2485, pruned_loss=0.049, over 985361.11 frames.], batch size: 61, aishell_tot_loss[loss=0.1717, simple_loss=0.2534, pruned_loss=0.045, over 983520.25 frames.], datatang_tot_loss[loss=0.175, simple_loss=0.2435, pruned_loss=0.05321, over 985449.90 frames.], batch size: 61, lr: 7.23e-04
+2022-06-18 18:12:01,112 INFO [train.py:874] (2/4) Epoch 11, batch 2750, aishell_loss[loss=0.1422, simple_loss=0.2253, pruned_loss=0.02959, over 4979.00 frames.], tot_loss[loss=0.1731, simple_loss=0.2479, pruned_loss=0.0492, over 985535.41 frames.], batch size: 30, aishell_tot_loss[loss=0.1715, simple_loss=0.2533, pruned_loss=0.04488, over 983706.82 frames.], datatang_tot_loss[loss=0.175, simple_loss=0.2431, pruned_loss=0.05342, over 985623.97 frames.], batch size: 30, lr: 7.23e-04
+2022-06-18 18:12:33,124 INFO [train.py:874] (2/4) Epoch 11, batch 2800, datatang_loss[loss=0.1586, simple_loss=0.2274, pruned_loss=0.04488, over 4920.00 frames.], tot_loss[loss=0.174, simple_loss=0.2489, pruned_loss=0.04954, over 985528.03 frames.], batch size: 64, aishell_tot_loss[loss=0.1721, simple_loss=0.2538, pruned_loss=0.04519, over 983823.44 frames.], datatang_tot_loss[loss=0.1753, simple_loss=0.2436, pruned_loss=0.05347, over 985708.53 frames.], batch size: 64, lr: 7.22e-04
+2022-06-18 18:13:03,708 INFO [train.py:874] (2/4) Epoch 11, batch 2850, datatang_loss[loss=0.1881, simple_loss=0.2486, pruned_loss=0.06373, over 4959.00 frames.], tot_loss[loss=0.1732, simple_loss=0.2479, pruned_loss=0.04924, over 985393.94 frames.], batch size: 60, aishell_tot_loss[loss=0.1717, simple_loss=0.2531, pruned_loss=0.04514, over 983767.12 frames.], datatang_tot_loss[loss=0.1749, simple_loss=0.2433, pruned_loss=0.0532, over 985785.42 frames.], batch size: 60, lr: 7.22e-04
+2022-06-18 18:13:33,064 INFO [train.py:874] (2/4) Epoch 11, batch 2900, datatang_loss[loss=0.1439, simple_loss=0.2201, pruned_loss=0.03388, over 4920.00 frames.], tot_loss[loss=0.172, simple_loss=0.2467, pruned_loss=0.04867, over 985541.71 frames.], batch size: 83, aishell_tot_loss[loss=0.1708, simple_loss=0.252, pruned_loss=0.04482, over 983743.20 frames.], datatang_tot_loss[loss=0.1745, simple_loss=0.2431, pruned_loss=0.0529, over 986089.58 frames.], batch size: 83, lr: 7.21e-04
+2022-06-18 18:14:04,182 INFO [train.py:874] (2/4) Epoch 11, batch 2950, datatang_loss[loss=0.1704, simple_loss=0.2314, pruned_loss=0.05474, over 4914.00 frames.], tot_loss[loss=0.1724, simple_loss=0.2466, pruned_loss=0.04908, over 985627.40 frames.], batch size: 42, aishell_tot_loss[loss=0.1716, simple_loss=0.2525, pruned_loss=0.04533, over 983945.99 frames.], datatang_tot_loss[loss=0.1739, simple_loss=0.2424, pruned_loss=0.05267, over 986089.03 frames.], batch size: 42, lr: 7.21e-04
+2022-06-18 18:14:33,557 INFO [train.py:874] (2/4) Epoch 11, batch 3000, datatang_loss[loss=0.2514, simple_loss=0.3032, pruned_loss=0.09979, over 4959.00 frames.], tot_loss[loss=0.1733, simple_loss=0.2472, pruned_loss=0.0497, over 985861.11 frames.], batch size: 99, aishell_tot_loss[loss=0.1718, simple_loss=0.2525, pruned_loss=0.0455, over 984401.08 frames.], datatang_tot_loss[loss=0.1746, simple_loss=0.2428, pruned_loss=0.05321, over 986015.07 frames.], batch size: 99, lr: 7.21e-04
+2022-06-18 18:14:33,558 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 18:14:50,207 INFO [train.py:914] (2/4) Epoch 11, validation: loss=0.1662, simple_loss=0.2504, pruned_loss=0.04101, over 1622729.00 frames.
+2022-06-18 18:15:20,341 INFO [train.py:874] (2/4) Epoch 11, batch 3050, aishell_loss[loss=0.1827, simple_loss=0.2715, pruned_loss=0.04696, over 4930.00 frames.], tot_loss[loss=0.1732, simple_loss=0.2477, pruned_loss=0.04936, over 985568.22 frames.], batch size: 78, aishell_tot_loss[loss=0.1722, simple_loss=0.2532, pruned_loss=0.0456, over 984321.49 frames.], datatang_tot_loss[loss=0.1741, simple_loss=0.2425, pruned_loss=0.05286, over 985944.71 frames.], batch size: 78, lr: 7.20e-04
+2022-06-18 18:15:50,756 INFO [train.py:874] (2/4) Epoch 11, batch 3100, aishell_loss[loss=0.1787, simple_loss=0.2613, pruned_loss=0.04803, over 4957.00 frames.], tot_loss[loss=0.1737, simple_loss=0.2479, pruned_loss=0.04974, over 986293.15 frames.], batch size: 61, aishell_tot_loss[loss=0.1725, simple_loss=0.2533, pruned_loss=0.04586, over 985021.12 frames.], datatang_tot_loss[loss=0.1743, simple_loss=0.2425, pruned_loss=0.05303, over 986112.47 frames.], batch size: 61, lr: 7.20e-04
+2022-06-18 18:16:23,154 INFO [train.py:874] (2/4) Epoch 11, batch 3150, datatang_loss[loss=0.1661, simple_loss=0.2349, pruned_loss=0.0487, over 4936.00 frames.], tot_loss[loss=0.172, simple_loss=0.2463, pruned_loss=0.04885, over 986077.76 frames.], batch size: 50, aishell_tot_loss[loss=0.1724, simple_loss=0.2531, pruned_loss=0.0458, over 984828.60 frames.], datatang_tot_loss[loss=0.1727, simple_loss=0.2412, pruned_loss=0.05213, over 986229.92 frames.], batch size: 50, lr: 7.19e-04
+2022-06-18 18:16:53,746 INFO [train.py:874] (2/4) Epoch 11, batch 3200, aishell_loss[loss=0.1916, simple_loss=0.2653, pruned_loss=0.05888, over 4855.00 frames.], tot_loss[loss=0.1729, simple_loss=0.2476, pruned_loss=0.04914, over 985611.36 frames.], batch size: 35, aishell_tot_loss[loss=0.1722, simple_loss=0.2533, pruned_loss=0.04558, over 984664.98 frames.], datatang_tot_loss[loss=0.1738, simple_loss=0.2421, pruned_loss=0.05269, over 986046.32 frames.], batch size: 35, lr: 7.19e-04
+2022-06-18 18:17:23,941 INFO [train.py:874] (2/4) Epoch 11, batch 3250, datatang_loss[loss=0.2005, simple_loss=0.2693, pruned_loss=0.06588, over 4917.00 frames.], tot_loss[loss=0.173, simple_loss=0.2478, pruned_loss=0.04914, over 985591.38 frames.], batch size: 81, aishell_tot_loss[loss=0.1722, simple_loss=0.2532, pruned_loss=0.04564, over 984775.05 frames.], datatang_tot_loss[loss=0.1738, simple_loss=0.2422, pruned_loss=0.05268, over 985979.26 frames.], batch size: 81, lr: 7.19e-04
+2022-06-18 18:17:56,613 INFO [train.py:874] (2/4) Epoch 11, batch 3300, aishell_loss[loss=0.1696, simple_loss=0.255, pruned_loss=0.04214, over 4968.00 frames.], tot_loss[loss=0.1738, simple_loss=0.2484, pruned_loss=0.04954, over 985927.27 frames.], batch size: 61, aishell_tot_loss[loss=0.1731, simple_loss=0.2539, pruned_loss=0.04617, over 985302.67 frames.], datatang_tot_loss[loss=0.1736, simple_loss=0.2421, pruned_loss=0.05258, over 985867.79 frames.], batch size: 61, lr: 7.18e-04
+2022-06-18 18:18:26,802 INFO [train.py:874] (2/4) Epoch 11, batch 3350, datatang_loss[loss=0.1591, simple_loss=0.231, pruned_loss=0.04362, over 4933.00 frames.], tot_loss[loss=0.1739, simple_loss=0.2486, pruned_loss=0.04958, over 985923.61 frames.], batch size: 83, aishell_tot_loss[loss=0.1733, simple_loss=0.2539, pruned_loss=0.04631, over 985450.62 frames.], datatang_tot_loss[loss=0.1737, simple_loss=0.2422, pruned_loss=0.05261, over 985797.90 frames.], batch size: 83, lr: 7.18e-04
+2022-06-18 18:18:57,231 INFO [train.py:874] (2/4) Epoch 11, batch 3400, aishell_loss[loss=0.1888, simple_loss=0.2659, pruned_loss=0.05591, over 4913.00 frames.], tot_loss[loss=0.1746, simple_loss=0.2493, pruned_loss=0.04997, over 985905.28 frames.], batch size: 41, aishell_tot_loss[loss=0.1728, simple_loss=0.2536, pruned_loss=0.04599, over 985431.86 frames.], datatang_tot_loss[loss=0.175, simple_loss=0.2435, pruned_loss=0.05326, over 985863.22 frames.], batch size: 41, lr: 7.17e-04
+2022-06-18 18:19:29,615 INFO [train.py:874] (2/4) Epoch 11, batch 3450, datatang_loss[loss=0.1517, simple_loss=0.218, pruned_loss=0.0427, over 4926.00 frames.], tot_loss[loss=0.1736, simple_loss=0.248, pruned_loss=0.04964, over 985667.22 frames.], batch size: 71, aishell_tot_loss[loss=0.1728, simple_loss=0.2537, pruned_loss=0.04595, over 985362.72 frames.], datatang_tot_loss[loss=0.1741, simple_loss=0.2423, pruned_loss=0.05292, over 985738.12 frames.], batch size: 71, lr: 7.17e-04
+2022-06-18 18:20:00,332 INFO [train.py:874] (2/4) Epoch 11, batch 3500, datatang_loss[loss=0.1454, simple_loss=0.2193, pruned_loss=0.03571, over 4943.00 frames.], tot_loss[loss=0.1734, simple_loss=0.2482, pruned_loss=0.04934, over 985529.63 frames.], batch size: 42, aishell_tot_loss[loss=0.1732, simple_loss=0.2543, pruned_loss=0.04606, over 985266.51 frames.], datatang_tot_loss[loss=0.1735, simple_loss=0.242, pruned_loss=0.05252, over 985711.97 frames.], batch size: 42, lr: 7.17e-04
+2022-06-18 18:20:30,444 INFO [train.py:874] (2/4) Epoch 11, batch 3550, aishell_loss[loss=0.1703, simple_loss=0.2582, pruned_loss=0.04115, over 4948.00 frames.], tot_loss[loss=0.1727, simple_loss=0.2471, pruned_loss=0.04919, over 985701.59 frames.], batch size: 64, aishell_tot_loss[loss=0.1723, simple_loss=0.2531, pruned_loss=0.04574, over 985382.96 frames.], datatang_tot_loss[loss=0.1737, simple_loss=0.242, pruned_loss=0.05276, over 985787.10 frames.], batch size: 64, lr: 7.16e-04
+2022-06-18 18:21:01,553 INFO [train.py:874] (2/4) Epoch 11, batch 3600, datatang_loss[loss=0.1664, simple_loss=0.2418, pruned_loss=0.04555, over 4952.00 frames.], tot_loss[loss=0.1719, simple_loss=0.2467, pruned_loss=0.04852, over 985812.85 frames.], batch size: 62, aishell_tot_loss[loss=0.1718, simple_loss=0.2526, pruned_loss=0.04548, over 985349.03 frames.], datatang_tot_loss[loss=0.1733, simple_loss=0.2417, pruned_loss=0.05242, over 985969.51 frames.], batch size: 62, lr: 7.16e-04
+2022-06-18 18:21:30,587 INFO [train.py:874] (2/4) Epoch 11, batch 3650, aishell_loss[loss=0.1496, simple_loss=0.2313, pruned_loss=0.03398, over 4986.00 frames.], tot_loss[loss=0.1715, simple_loss=0.2464, pruned_loss=0.04833, over 985599.18 frames.], batch size: 30, aishell_tot_loss[loss=0.1716, simple_loss=0.2524, pruned_loss=0.04539, over 985382.84 frames.], datatang_tot_loss[loss=0.173, simple_loss=0.2413, pruned_loss=0.05237, over 985744.52 frames.], batch size: 30, lr: 7.15e-04
+2022-06-18 18:22:02,511 INFO [train.py:874] (2/4) Epoch 11, batch 3700, aishell_loss[loss=0.1952, simple_loss=0.2761, pruned_loss=0.0572, over 4934.00 frames.], tot_loss[loss=0.1728, simple_loss=0.2475, pruned_loss=0.04908, over 985418.68 frames.], batch size: 58, aishell_tot_loss[loss=0.1723, simple_loss=0.253, pruned_loss=0.04581, over 985304.46 frames.], datatang_tot_loss[loss=0.1735, simple_loss=0.2417, pruned_loss=0.05263, over 985641.22 frames.], batch size: 58, lr: 7.15e-04
+2022-06-18 18:22:32,566 INFO [train.py:874] (2/4) Epoch 11, batch 3750, aishell_loss[loss=0.204, simple_loss=0.273, pruned_loss=0.06749, over 4862.00 frames.], tot_loss[loss=0.1726, simple_loss=0.2479, pruned_loss=0.0486, over 985792.01 frames.], batch size: 36, aishell_tot_loss[loss=0.172, simple_loss=0.253, pruned_loss=0.04547, over 985676.51 frames.], datatang_tot_loss[loss=0.1736, simple_loss=0.2419, pruned_loss=0.05262, over 985656.96 frames.], batch size: 36, lr: 7.15e-04
+2022-06-18 18:23:02,887 INFO [train.py:874] (2/4) Epoch 11, batch 3800, aishell_loss[loss=0.1501, simple_loss=0.2381, pruned_loss=0.03107, over 4989.00 frames.], tot_loss[loss=0.1718, simple_loss=0.247, pruned_loss=0.04831, over 985487.47 frames.], batch size: 27, aishell_tot_loss[loss=0.1713, simple_loss=0.2524, pruned_loss=0.04511, over 985516.68 frames.], datatang_tot_loss[loss=0.1734, simple_loss=0.2417, pruned_loss=0.05249, over 985514.84 frames.], batch size: 27, lr: 7.14e-04
+2022-06-18 18:23:32,300 INFO [train.py:874] (2/4) Epoch 11, batch 3850, datatang_loss[loss=0.1879, simple_loss=0.256, pruned_loss=0.05993, over 4932.00 frames.], tot_loss[loss=0.1717, simple_loss=0.2472, pruned_loss=0.04812, over 985407.36 frames.], batch size: 81, aishell_tot_loss[loss=0.171, simple_loss=0.2521, pruned_loss=0.045, over 985423.72 frames.], datatang_tot_loss[loss=0.1735, simple_loss=0.242, pruned_loss=0.05247, over 985527.49 frames.], batch size: 81, lr: 7.14e-04
+2022-06-18 18:24:01,226 INFO [train.py:874] (2/4) Epoch 11, batch 3900, aishell_loss[loss=0.1906, simple_loss=0.2761, pruned_loss=0.05254, over 4949.00 frames.], tot_loss[loss=0.172, simple_loss=0.2472, pruned_loss=0.04838, over 985869.53 frames.], batch size: 54, aishell_tot_loss[loss=0.1709, simple_loss=0.2518, pruned_loss=0.04502, over 985776.99 frames.], datatang_tot_loss[loss=0.1738, simple_loss=0.2422, pruned_loss=0.05266, over 985636.78 frames.], batch size: 54, lr: 7.14e-04
+2022-06-18 18:24:28,627 INFO [train.py:874] (2/4) Epoch 11, batch 3950, datatang_loss[loss=0.1646, simple_loss=0.2198, pruned_loss=0.05471, over 4961.00 frames.], tot_loss[loss=0.1718, simple_loss=0.2474, pruned_loss=0.04813, over 985727.52 frames.], batch size: 37, aishell_tot_loss[loss=0.1705, simple_loss=0.2517, pruned_loss=0.04462, over 985612.72 frames.], datatang_tot_loss[loss=0.1741, simple_loss=0.2424, pruned_loss=0.05287, over 985693.51 frames.], batch size: 37, lr: 7.13e-04
+2022-06-18 18:24:59,120 INFO [train.py:874] (2/4) Epoch 11, batch 4000, aishell_loss[loss=0.1471, simple_loss=0.2308, pruned_loss=0.03174, over 4895.00 frames.], tot_loss[loss=0.1708, simple_loss=0.2459, pruned_loss=0.04786, over 985925.11 frames.], batch size: 34, aishell_tot_loss[loss=0.1694, simple_loss=0.2505, pruned_loss=0.04412, over 985667.13 frames.], datatang_tot_loss[loss=0.174, simple_loss=0.2423, pruned_loss=0.05284, over 985863.87 frames.], batch size: 34, lr: 7.13e-04
+2022-06-18 18:24:59,120 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 18:25:16,905 INFO [train.py:914] (2/4) Epoch 11, validation: loss=0.1668, simple_loss=0.2498, pruned_loss=0.04186, over 1622729.00 frames.
+2022-06-18 18:25:45,363 INFO [train.py:874] (2/4) Epoch 11, batch 4050, aishell_loss[loss=0.1574, simple_loss=0.2414, pruned_loss=0.03674, over 4887.00 frames.], tot_loss[loss=0.1715, simple_loss=0.2466, pruned_loss=0.04822, over 985971.35 frames.], batch size: 28, aishell_tot_loss[loss=0.1699, simple_loss=0.2509, pruned_loss=0.04447, over 985670.01 frames.], datatang_tot_loss[loss=0.174, simple_loss=0.2424, pruned_loss=0.05281, over 985955.75 frames.], batch size: 28, lr: 7.12e-04
+2022-06-18 18:26:14,015 INFO [train.py:874] (2/4) Epoch 11, batch 4100, datatang_loss[loss=0.1374, simple_loss=0.2091, pruned_loss=0.03286, over 4924.00 frames.], tot_loss[loss=0.1707, simple_loss=0.246, pruned_loss=0.04765, over 985457.03 frames.], batch size: 26, aishell_tot_loss[loss=0.1697, simple_loss=0.2509, pruned_loss=0.04426, over 985109.52 frames.], datatang_tot_loss[loss=0.1733, simple_loss=0.2419, pruned_loss=0.05232, over 986031.91 frames.], batch size: 26, lr: 7.12e-04
+2022-06-18 18:27:19,851 INFO [train.py:874] (2/4) Epoch 12, batch 50, datatang_loss[loss=0.1625, simple_loss=0.2341, pruned_loss=0.04546, over 4925.00 frames.], tot_loss[loss=0.1682, simple_loss=0.2446, pruned_loss=0.04594, over 218789.91 frames.], batch size: 94, aishell_tot_loss[loss=0.1718, simple_loss=0.2543, pruned_loss=0.04464, over 129252.59 frames.], datatang_tot_loss[loss=0.164, simple_loss=0.2327, pruned_loss=0.04764, over 103038.30 frames.], batch size: 94, lr: 6.86e-04
+2022-06-18 18:27:51,806 INFO [train.py:874] (2/4) Epoch 12, batch 100, datatang_loss[loss=0.142, simple_loss=0.2173, pruned_loss=0.03339, over 4927.00 frames.], tot_loss[loss=0.1676, simple_loss=0.2441, pruned_loss=0.04553, over 388701.27 frames.], batch size: 73, aishell_tot_loss[loss=0.1713, simple_loss=0.254, pruned_loss=0.04427, over 233708.32 frames.], datatang_tot_loss[loss=0.1637, simple_loss=0.233, pruned_loss=0.04713, over 203172.73 frames.], batch size: 73, lr: 6.86e-04
+2022-06-18 18:28:21,784 INFO [train.py:874] (2/4) Epoch 12, batch 150, datatang_loss[loss=0.1748, simple_loss=0.2365, pruned_loss=0.05659, over 4948.00 frames.], tot_loss[loss=0.1673, simple_loss=0.2428, pruned_loss=0.04593, over 521122.84 frames.], batch size: 55, aishell_tot_loss[loss=0.1716, simple_loss=0.2541, pruned_loss=0.04453, over 302069.34 frames.], datatang_tot_loss[loss=0.1637, simple_loss=0.2331, pruned_loss=0.04717, over 315830.00 frames.], batch size: 55, lr: 6.86e-04
+2022-06-18 18:28:52,333 INFO [train.py:874] (2/4) Epoch 12, batch 200, aishell_loss[loss=0.189, simple_loss=0.2621, pruned_loss=0.05793, over 4928.00 frames.], tot_loss[loss=0.1692, simple_loss=0.2442, pruned_loss=0.04705, over 624078.23 frames.], batch size: 33, aishell_tot_loss[loss=0.1736, simple_loss=0.2546, pruned_loss=0.04628, over 385650.43 frames.], datatang_tot_loss[loss=0.1645, simple_loss=0.2343, pruned_loss=0.04739, over 391685.98 frames.], batch size: 33, lr: 6.85e-04
+2022-06-18 18:29:23,910 INFO [train.py:874] (2/4) Epoch 12, batch 250, datatang_loss[loss=0.1864, simple_loss=0.2487, pruned_loss=0.06209, over 4883.00 frames.], tot_loss[loss=0.169, simple_loss=0.2453, pruned_loss=0.0464, over 704012.57 frames.], batch size: 25, aishell_tot_loss[loss=0.1728, simple_loss=0.2544, pruned_loss=0.04557, over 476761.22 frames.], datatang_tot_loss[loss=0.1645, simple_loss=0.2345, pruned_loss=0.04723, over 440336.74 frames.], batch size: 25, lr: 6.85e-04
+2022-06-18 18:29:54,014 INFO [train.py:874] (2/4) Epoch 12, batch 300, aishell_loss[loss=0.1953, simple_loss=0.2681, pruned_loss=0.06119, over 4925.00 frames.], tot_loss[loss=0.1677, simple_loss=0.2438, pruned_loss=0.04583, over 766615.54 frames.], batch size: 58, aishell_tot_loss[loss=0.1709, simple_loss=0.2525, pruned_loss=0.04468, over 536587.87 frames.], datatang_tot_loss[loss=0.1645, simple_loss=0.2346, pruned_loss=0.04722, over 504853.72 frames.], batch size: 58, lr: 6.85e-04
+2022-06-18 18:30:23,414 INFO [train.py:874] (2/4) Epoch 12, batch 350, aishell_loss[loss=0.1784, simple_loss=0.2576, pruned_loss=0.04962, over 4983.00 frames.], tot_loss[loss=0.1683, simple_loss=0.2441, pruned_loss=0.04625, over 815066.31 frames.], batch size: 51, aishell_tot_loss[loss=0.17, simple_loss=0.2517, pruned_loss=0.04415, over 587346.60 frames.], datatang_tot_loss[loss=0.1664, simple_loss=0.236, pruned_loss=0.04835, over 563616.55 frames.], batch size: 51, lr: 6.84e-04
+2022-06-18 18:30:56,356 INFO [train.py:874] (2/4) Epoch 12, batch 400, datatang_loss[loss=0.1483, simple_loss=0.2225, pruned_loss=0.0371, over 4934.00 frames.], tot_loss[loss=0.1672, simple_loss=0.2428, pruned_loss=0.04577, over 852529.88 frames.], batch size: 69, aishell_tot_loss[loss=0.1685, simple_loss=0.2501, pruned_loss=0.04345, over 623300.87 frames.], datatang_tot_loss[loss=0.1665, simple_loss=0.2367, pruned_loss=0.04814, over 624202.61 frames.], batch size: 69, lr: 6.84e-04
+2022-06-18 18:31:26,878 INFO [train.py:874] (2/4) Epoch 12, batch 450, aishell_loss[loss=0.141, simple_loss=0.2159, pruned_loss=0.03301, over 4979.00 frames.], tot_loss[loss=0.1669, simple_loss=0.2429, pruned_loss=0.04546, over 882091.91 frames.], batch size: 27, aishell_tot_loss[loss=0.1685, simple_loss=0.2502, pruned_loss=0.04339, over 669176.09 frames.], datatang_tot_loss[loss=0.1659, simple_loss=0.2362, pruned_loss=0.04779, over 663603.54 frames.], batch size: 27, lr: 6.84e-04
+2022-06-18 18:31:57,604 INFO [train.py:874] (2/4) Epoch 12, batch 500, aishell_loss[loss=0.1652, simple_loss=0.2497, pruned_loss=0.04034, over 4909.00 frames.], tot_loss[loss=0.1673, simple_loss=0.2438, pruned_loss=0.04543, over 904740.87 frames.], batch size: 68, aishell_tot_loss[loss=0.1688, simple_loss=0.2509, pruned_loss=0.04332, over 711815.43 frames.], datatang_tot_loss[loss=0.166, simple_loss=0.2363, pruned_loss=0.04788, over 695622.70 frames.], batch size: 68, lr: 6.83e-04
+2022-06-18 18:32:29,680 INFO [train.py:874] (2/4) Epoch 12, batch 550, aishell_loss[loss=0.1756, simple_loss=0.2527, pruned_loss=0.04932, over 4912.00 frames.], tot_loss[loss=0.1685, simple_loss=0.2445, pruned_loss=0.04626, over 922643.28 frames.], batch size: 52, aishell_tot_loss[loss=0.1694, simple_loss=0.2513, pruned_loss=0.04378, over 740365.78 frames.], datatang_tot_loss[loss=0.1669, simple_loss=0.2371, pruned_loss=0.0484, over 733604.30 frames.], batch size: 52, lr: 6.83e-04
+2022-06-18 18:33:00,934 INFO [train.py:874] (2/4) Epoch 12, batch 600, aishell_loss[loss=0.1905, simple_loss=0.2673, pruned_loss=0.05682, over 4912.00 frames.], tot_loss[loss=0.1683, simple_loss=0.2442, pruned_loss=0.04621, over 936361.67 frames.], batch size: 52, aishell_tot_loss[loss=0.1684, simple_loss=0.2504, pruned_loss=0.04317, over 762464.93 frames.], datatang_tot_loss[loss=0.1679, simple_loss=0.2382, pruned_loss=0.04879, over 769793.08 frames.], batch size: 52, lr: 6.82e-04
+2022-06-18 18:33:32,101 INFO [train.py:874] (2/4) Epoch 12, batch 650, aishell_loss[loss=0.175, simple_loss=0.2379, pruned_loss=0.05606, over 4943.00 frames.], tot_loss[loss=0.1701, simple_loss=0.2457, pruned_loss=0.04724, over 947166.12 frames.], batch size: 32, aishell_tot_loss[loss=0.1683, simple_loss=0.2502, pruned_loss=0.04318, over 787526.95 frames.], datatang_tot_loss[loss=0.1702, simple_loss=0.2403, pruned_loss=0.05008, over 796269.44 frames.], batch size: 32, lr: 6.82e-04
+2022-06-18 18:34:02,833 INFO [train.py:874] (2/4) Epoch 12, batch 700, datatang_loss[loss=0.2321, simple_loss=0.2946, pruned_loss=0.08486, over 4943.00 frames.], tot_loss[loss=0.1709, simple_loss=0.247, pruned_loss=0.04737, over 955972.11 frames.], batch size: 109, aishell_tot_loss[loss=0.1692, simple_loss=0.2512, pruned_loss=0.04359, over 817885.07 frames.], datatang_tot_loss[loss=0.1706, simple_loss=0.2406, pruned_loss=0.05029, over 811872.71 frames.], batch size: 109, lr: 6.82e-04
+2022-06-18 18:34:31,746 INFO [train.py:874] (2/4) Epoch 12, batch 750, datatang_loss[loss=0.1624, simple_loss=0.2403, pruned_loss=0.04222, over 4923.00 frames.], tot_loss[loss=0.1707, simple_loss=0.2471, pruned_loss=0.04712, over 962496.16 frames.], batch size: 64, aishell_tot_loss[loss=0.1688, simple_loss=0.251, pruned_loss=0.0433, over 839750.00 frames.], datatang_tot_loss[loss=0.1711, simple_loss=0.2411, pruned_loss=0.05051, over 830058.65 frames.], batch size: 64, lr: 6.81e-04
+2022-06-18 18:35:04,178 INFO [train.py:874] (2/4) Epoch 12, batch 800, aishell_loss[loss=0.1509, simple_loss=0.2315, pruned_loss=0.03517, over 4963.00 frames.], tot_loss[loss=0.1705, simple_loss=0.247, pruned_loss=0.04695, over 967887.14 frames.], batch size: 44, aishell_tot_loss[loss=0.1684, simple_loss=0.2506, pruned_loss=0.04309, over 858218.49 frames.], datatang_tot_loss[loss=0.1714, simple_loss=0.2416, pruned_loss=0.05065, over 847274.43 frames.], batch size: 44, lr: 6.81e-04
+2022-06-18 18:35:34,612 INFO [train.py:874] (2/4) Epoch 12, batch 850, datatang_loss[loss=0.1838, simple_loss=0.2553, pruned_loss=0.05618, over 4970.00 frames.], tot_loss[loss=0.1705, simple_loss=0.2463, pruned_loss=0.04738, over 972196.15 frames.], batch size: 34, aishell_tot_loss[loss=0.1693, simple_loss=0.251, pruned_loss=0.04379, over 872292.75 frames.], datatang_tot_loss[loss=0.1708, simple_loss=0.2407, pruned_loss=0.0504, over 864986.83 frames.], batch size: 34, lr: 6.81e-04
+2022-06-18 18:36:04,252 INFO [train.py:874] (2/4) Epoch 12, batch 900, aishell_loss[loss=0.162, simple_loss=0.2533, pruned_loss=0.03534, over 4923.00 frames.], tot_loss[loss=0.1698, simple_loss=0.2458, pruned_loss=0.04692, over 975310.47 frames.], batch size: 52, aishell_tot_loss[loss=0.1687, simple_loss=0.2507, pruned_loss=0.0434, over 885084.79 frames.], datatang_tot_loss[loss=0.1706, simple_loss=0.2406, pruned_loss=0.05027, over 879921.94 frames.], batch size: 52, lr: 6.80e-04
+2022-06-18 18:36:36,373 INFO [train.py:874] (2/4) Epoch 12, batch 950, datatang_loss[loss=0.2198, simple_loss=0.2883, pruned_loss=0.07569, over 4926.00 frames.], tot_loss[loss=0.1711, simple_loss=0.2469, pruned_loss=0.04767, over 977562.50 frames.], batch size: 107, aishell_tot_loss[loss=0.1689, simple_loss=0.2507, pruned_loss=0.04349, over 897741.89 frames.], datatang_tot_loss[loss=0.172, simple_loss=0.2418, pruned_loss=0.0511, over 891475.38 frames.], batch size: 107, lr: 6.80e-04
+2022-06-18 18:37:06,849 INFO [train.py:874] (2/4) Epoch 12, batch 1000, aishell_loss[loss=0.1896, simple_loss=0.274, pruned_loss=0.05261, over 4951.00 frames.], tot_loss[loss=0.1725, simple_loss=0.2483, pruned_loss=0.0484, over 979212.41 frames.], batch size: 54, aishell_tot_loss[loss=0.1698, simple_loss=0.2517, pruned_loss=0.04393, over 908638.73 frames.], datatang_tot_loss[loss=0.1729, simple_loss=0.2424, pruned_loss=0.05167, over 901799.33 frames.], batch size: 54, lr: 6.79e-04
+2022-06-18 18:37:06,850 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 18:37:23,199 INFO [train.py:914] (2/4) Epoch 12, validation: loss=0.1654, simple_loss=0.2501, pruned_loss=0.04036, over 1622729.00 frames.
+2022-06-18 18:37:54,284 INFO [train.py:874] (2/4) Epoch 12, batch 1050, datatang_loss[loss=0.1754, simple_loss=0.2524, pruned_loss=0.0492, over 4934.00 frames.], tot_loss[loss=0.1725, simple_loss=0.2482, pruned_loss=0.04838, over 980406.90 frames.], batch size: 79, aishell_tot_loss[loss=0.1695, simple_loss=0.2515, pruned_loss=0.04381, over 915672.84 frames.], datatang_tot_loss[loss=0.1733, simple_loss=0.2431, pruned_loss=0.05179, over 913578.44 frames.], batch size: 79, lr: 6.79e-04
+2022-06-18 18:38:26,525 INFO [train.py:874] (2/4) Epoch 12, batch 1100, aishell_loss[loss=0.1653, simple_loss=0.2501, pruned_loss=0.04028, over 4918.00 frames.], tot_loss[loss=0.1723, simple_loss=0.2479, pruned_loss=0.04837, over 981775.63 frames.], batch size: 52, aishell_tot_loss[loss=0.1701, simple_loss=0.2519, pruned_loss=0.04421, over 923567.02 frames.], datatang_tot_loss[loss=0.1728, simple_loss=0.2427, pruned_loss=0.05145, over 922642.18 frames.], batch size: 52, lr: 6.79e-04
+2022-06-18 18:38:56,143 INFO [train.py:874] (2/4) Epoch 12, batch 1150, aishell_loss[loss=0.1827, simple_loss=0.2593, pruned_loss=0.05301, over 4926.00 frames.], tot_loss[loss=0.172, simple_loss=0.2477, pruned_loss=0.04816, over 982795.22 frames.], batch size: 33, aishell_tot_loss[loss=0.1698, simple_loss=0.2517, pruned_loss=0.04395, over 931373.94 frames.], datatang_tot_loss[loss=0.173, simple_loss=0.2427, pruned_loss=0.05162, over 929747.35 frames.], batch size: 33, lr: 6.78e-04
+2022-06-18 18:39:27,821 INFO [train.py:874] (2/4) Epoch 12, batch 1200, aishell_loss[loss=0.1702, simple_loss=0.2524, pruned_loss=0.044, over 4919.00 frames.], tot_loss[loss=0.1707, simple_loss=0.2467, pruned_loss=0.04737, over 983319.73 frames.], batch size: 41, aishell_tot_loss[loss=0.1693, simple_loss=0.2515, pruned_loss=0.04353, over 938079.26 frames.], datatang_tot_loss[loss=0.1722, simple_loss=0.2419, pruned_loss=0.05127, over 935890.46 frames.], batch size: 41, lr: 6.78e-04
+2022-06-18 18:39:58,227 INFO [train.py:874] (2/4) Epoch 12, batch 1250, aishell_loss[loss=0.1647, simple_loss=0.2426, pruned_loss=0.04338, over 4961.00 frames.], tot_loss[loss=0.17, simple_loss=0.2461, pruned_loss=0.04691, over 983222.02 frames.], batch size: 31, aishell_tot_loss[loss=0.169, simple_loss=0.2512, pruned_loss=0.04338, over 942949.78 frames.], datatang_tot_loss[loss=0.1717, simple_loss=0.2416, pruned_loss=0.05088, over 941835.97 frames.], batch size: 31, lr: 6.78e-04
+2022-06-18 18:40:30,425 INFO [train.py:874] (2/4) Epoch 12, batch 1300, aishell_loss[loss=0.1574, simple_loss=0.2371, pruned_loss=0.03883, over 4940.00 frames.], tot_loss[loss=0.1697, simple_loss=0.2459, pruned_loss=0.04671, over 983510.81 frames.], batch size: 32, aishell_tot_loss[loss=0.1689, simple_loss=0.2511, pruned_loss=0.0433, over 947275.83 frames.], datatang_tot_loss[loss=0.1714, simple_loss=0.2415, pruned_loss=0.05063, over 947430.63 frames.], batch size: 32, lr: 6.77e-04
+2022-06-18 18:41:01,634 INFO [train.py:874] (2/4) Epoch 12, batch 1350, datatang_loss[loss=0.1704, simple_loss=0.2333, pruned_loss=0.05376, over 4956.00 frames.], tot_loss[loss=0.1692, simple_loss=0.2452, pruned_loss=0.04655, over 983369.48 frames.], batch size: 55, aishell_tot_loss[loss=0.1687, simple_loss=0.2509, pruned_loss=0.04327, over 951143.34 frames.], datatang_tot_loss[loss=0.1708, simple_loss=0.2409, pruned_loss=0.05035, over 951928.28 frames.], batch size: 55, lr: 6.77e-04
+2022-06-18 18:41:32,118 INFO [train.py:874] (2/4) Epoch 12, batch 1400, datatang_loss[loss=0.1895, simple_loss=0.2548, pruned_loss=0.06212, over 4947.00 frames.], tot_loss[loss=0.1698, simple_loss=0.2457, pruned_loss=0.04694, over 983909.45 frames.], batch size: 91, aishell_tot_loss[loss=0.1686, simple_loss=0.2509, pruned_loss=0.04319, over 954923.51 frames.], datatang_tot_loss[loss=0.1714, simple_loss=0.2414, pruned_loss=0.05069, over 956205.22 frames.], batch size: 91, lr: 6.77e-04
+2022-06-18 18:42:02,772 INFO [train.py:874] (2/4) Epoch 12, batch 1450, datatang_loss[loss=0.1552, simple_loss=0.2265, pruned_loss=0.04193, over 4958.00 frames.], tot_loss[loss=0.1694, simple_loss=0.246, pruned_loss=0.04645, over 984263.35 frames.], batch size: 55, aishell_tot_loss[loss=0.1691, simple_loss=0.2512, pruned_loss=0.04343, over 959473.56 frames.], datatang_tot_loss[loss=0.1706, simple_loss=0.2408, pruned_loss=0.05018, over 958717.79 frames.], batch size: 55, lr: 6.76e-04
+2022-06-18 18:42:33,340 INFO [train.py:874] (2/4) Epoch 12, batch 1500, datatang_loss[loss=0.1528, simple_loss=0.2245, pruned_loss=0.04054, over 4918.00 frames.], tot_loss[loss=0.1689, simple_loss=0.2456, pruned_loss=0.04611, over 984054.08 frames.], batch size: 77, aishell_tot_loss[loss=0.1685, simple_loss=0.2508, pruned_loss=0.04313, over 962393.15 frames.], datatang_tot_loss[loss=0.1705, simple_loss=0.2408, pruned_loss=0.05009, over 961501.60 frames.], batch size: 77, lr: 6.76e-04
+2022-06-18 18:43:02,292 INFO [train.py:874] (2/4) Epoch 12, batch 1550, aishell_loss[loss=0.1743, simple_loss=0.2566, pruned_loss=0.04597, over 4971.00 frames.], tot_loss[loss=0.1697, simple_loss=0.2463, pruned_loss=0.04653, over 984361.92 frames.], batch size: 44, aishell_tot_loss[loss=0.1686, simple_loss=0.2509, pruned_loss=0.04311, over 965492.19 frames.], datatang_tot_loss[loss=0.1711, simple_loss=0.2411, pruned_loss=0.05051, over 963921.99 frames.], batch size: 44, lr: 6.76e-04
+2022-06-18 18:43:34,348 INFO [train.py:874] (2/4) Epoch 12, batch 1600, aishell_loss[loss=0.1569, simple_loss=0.2378, pruned_loss=0.03798, over 4871.00 frames.], tot_loss[loss=0.1696, simple_loss=0.2464, pruned_loss=0.04638, over 984701.30 frames.], batch size: 35, aishell_tot_loss[loss=0.1681, simple_loss=0.2505, pruned_loss=0.04284, over 968014.97 frames.], datatang_tot_loss[loss=0.1714, simple_loss=0.2416, pruned_loss=0.05061, over 966377.52 frames.], batch size: 35, lr: 6.75e-04
+2022-06-18 18:44:05,015 INFO [train.py:874] (2/4) Epoch 12, batch 1650, aishell_loss[loss=0.159, simple_loss=0.2402, pruned_loss=0.03888, over 4975.00 frames.], tot_loss[loss=0.1684, simple_loss=0.2449, pruned_loss=0.04594, over 984871.58 frames.], batch size: 39, aishell_tot_loss[loss=0.1674, simple_loss=0.2497, pruned_loss=0.04254, over 969728.82 frames.], datatang_tot_loss[loss=0.1708, simple_loss=0.241, pruned_loss=0.05029, over 968970.35 frames.], batch size: 39, lr: 6.75e-04
+2022-06-18 18:44:35,460 INFO [train.py:874] (2/4) Epoch 12, batch 1700, aishell_loss[loss=0.1829, simple_loss=0.2697, pruned_loss=0.04806, over 4963.00 frames.], tot_loss[loss=0.169, simple_loss=0.2453, pruned_loss=0.04637, over 985019.17 frames.], batch size: 51, aishell_tot_loss[loss=0.1681, simple_loss=0.2503, pruned_loss=0.04296, over 971517.53 frames.], datatang_tot_loss[loss=0.1706, simple_loss=0.2408, pruned_loss=0.05018, over 970988.46 frames.], batch size: 51, lr: 6.74e-04
+2022-06-18 18:45:06,614 INFO [train.py:874] (2/4) Epoch 12, batch 1750, aishell_loss[loss=0.176, simple_loss=0.2742, pruned_loss=0.0389, over 4977.00 frames.], tot_loss[loss=0.1691, simple_loss=0.2454, pruned_loss=0.04641, over 984524.72 frames.], batch size: 44, aishell_tot_loss[loss=0.1676, simple_loss=0.2497, pruned_loss=0.04272, over 972677.97 frames.], datatang_tot_loss[loss=0.1711, simple_loss=0.2411, pruned_loss=0.05049, over 972539.60 frames.], batch size: 44, lr: 6.74e-04
+2022-06-18 18:45:38,043 INFO [train.py:874] (2/4) Epoch 12, batch 1800, aishell_loss[loss=0.1568, simple_loss=0.2502, pruned_loss=0.03171, over 4929.00 frames.], tot_loss[loss=0.1696, simple_loss=0.246, pruned_loss=0.04657, over 984641.56 frames.], batch size: 49, aishell_tot_loss[loss=0.1679, simple_loss=0.2503, pruned_loss=0.04279, over 974030.89 frames.], datatang_tot_loss[loss=0.1712, simple_loss=0.2413, pruned_loss=0.05049, over 974121.07 frames.], batch size: 49, lr: 6.74e-04
+2022-06-18 18:46:08,164 INFO [train.py:874] (2/4) Epoch 12, batch 1850, aishell_loss[loss=0.1941, simple_loss=0.27, pruned_loss=0.05913, over 4921.00 frames.], tot_loss[loss=0.1702, simple_loss=0.2464, pruned_loss=0.04696, over 984734.28 frames.], batch size: 52, aishell_tot_loss[loss=0.1682, simple_loss=0.2506, pruned_loss=0.04289, over 974970.54 frames.], datatang_tot_loss[loss=0.1715, simple_loss=0.2414, pruned_loss=0.05077, over 975765.78 frames.], batch size: 52, lr: 6.73e-04
+2022-06-18 18:46:38,137 INFO [train.py:874] (2/4) Epoch 12, batch 1900, datatang_loss[loss=0.1822, simple_loss=0.2386, pruned_loss=0.06289, over 4910.00 frames.], tot_loss[loss=0.1691, simple_loss=0.2455, pruned_loss=0.04634, over 984377.27 frames.], batch size: 52, aishell_tot_loss[loss=0.168, simple_loss=0.2506, pruned_loss=0.04271, over 976036.07 frames.], datatang_tot_loss[loss=0.1707, simple_loss=0.2404, pruned_loss=0.05046, over 976519.90 frames.], batch size: 52, lr: 6.73e-04
+2022-06-18 18:47:08,331 INFO [train.py:874] (2/4) Epoch 12, batch 1950, datatang_loss[loss=0.1602, simple_loss=0.2308, pruned_loss=0.04486, over 4921.00 frames.], tot_loss[loss=0.169, simple_loss=0.2459, pruned_loss=0.04604, over 984900.36 frames.], batch size: 83, aishell_tot_loss[loss=0.1678, simple_loss=0.2507, pruned_loss=0.0425, over 977567.95 frames.], datatang_tot_loss[loss=0.1707, simple_loss=0.2405, pruned_loss=0.05046, over 977459.91 frames.], batch size: 83, lr: 6.73e-04
+2022-06-18 18:47:38,737 INFO [train.py:874] (2/4) Epoch 12, batch 2000, datatang_loss[loss=0.1667, simple_loss=0.247, pruned_loss=0.0432, over 4956.00 frames.], tot_loss[loss=0.1681, simple_loss=0.2448, pruned_loss=0.04572, over 985101.98 frames.], batch size: 86, aishell_tot_loss[loss=0.1673, simple_loss=0.2501, pruned_loss=0.04224, over 978234.72 frames.], datatang_tot_loss[loss=0.1702, simple_loss=0.2401, pruned_loss=0.05013, over 978739.45 frames.], batch size: 86, lr: 6.72e-04
+2022-06-18 18:47:38,738 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 18:47:55,401 INFO [train.py:914] (2/4) Epoch 12, validation: loss=0.1674, simple_loss=0.25, pruned_loss=0.04245, over 1622729.00 frames.
+2022-06-18 18:48:25,431 INFO [train.py:874] (2/4) Epoch 12, batch 2050, datatang_loss[loss=0.1575, simple_loss=0.2429, pruned_loss=0.03605, over 4985.00 frames.], tot_loss[loss=0.1688, simple_loss=0.2454, pruned_loss=0.04605, over 985374.06 frames.], batch size: 31, aishell_tot_loss[loss=0.1679, simple_loss=0.2507, pruned_loss=0.04259, over 979182.18 frames.], datatang_tot_loss[loss=0.1701, simple_loss=0.2402, pruned_loss=0.04999, over 979627.24 frames.], batch size: 31, lr: 6.72e-04
+2022-06-18 18:48:56,079 INFO [train.py:874] (2/4) Epoch 12, batch 2100, aishell_loss[loss=0.1446, simple_loss=0.2363, pruned_loss=0.02642, over 4933.00 frames.], tot_loss[loss=0.1684, simple_loss=0.2455, pruned_loss=0.04568, over 985414.23 frames.], batch size: 49, aishell_tot_loss[loss=0.1678, simple_loss=0.2507, pruned_loss=0.0425, over 980012.29 frames.], datatang_tot_loss[loss=0.1698, simple_loss=0.24, pruned_loss=0.04978, over 980242.72 frames.], batch size: 49, lr: 6.72e-04
+2022-06-18 18:49:26,400 INFO [train.py:874] (2/4) Epoch 12, batch 2150, datatang_loss[loss=0.1792, simple_loss=0.2303, pruned_loss=0.06399, over 4896.00 frames.], tot_loss[loss=0.169, simple_loss=0.246, pruned_loss=0.04597, over 985585.38 frames.], batch size: 30, aishell_tot_loss[loss=0.1681, simple_loss=0.2509, pruned_loss=0.04265, over 980743.20 frames.], datatang_tot_loss[loss=0.1701, simple_loss=0.2404, pruned_loss=0.0499, over 980937.26 frames.], batch size: 30, lr: 6.71e-04
+2022-06-18 18:49:55,761 INFO [train.py:874] (2/4) Epoch 12, batch 2200, datatang_loss[loss=0.1554, simple_loss=0.2284, pruned_loss=0.04115, over 4950.00 frames.], tot_loss[loss=0.1701, simple_loss=0.2466, pruned_loss=0.04684, over 985790.91 frames.], batch size: 86, aishell_tot_loss[loss=0.1689, simple_loss=0.2514, pruned_loss=0.04319, over 981400.50 frames.], datatang_tot_loss[loss=0.1705, simple_loss=0.2407, pruned_loss=0.05014, over 981606.95 frames.], batch size: 86, lr: 6.71e-04
+2022-06-18 18:50:25,833 INFO [train.py:874] (2/4) Epoch 12, batch 2250, datatang_loss[loss=0.2346, simple_loss=0.2923, pruned_loss=0.08845, over 4917.00 frames.], tot_loss[loss=0.1714, simple_loss=0.2471, pruned_loss=0.04788, over 985127.08 frames.], batch size: 108, aishell_tot_loss[loss=0.1695, simple_loss=0.2515, pruned_loss=0.04377, over 981379.55 frames.], datatang_tot_loss[loss=0.1713, simple_loss=0.2413, pruned_loss=0.05067, over 981945.13 frames.], batch size: 108, lr: 6.71e-04
+2022-06-18 18:50:57,132 INFO [train.py:874] (2/4) Epoch 12, batch 2300, datatang_loss[loss=0.1803, simple_loss=0.2556, pruned_loss=0.0525, over 4952.00 frames.], tot_loss[loss=0.171, simple_loss=0.2464, pruned_loss=0.04778, over 985359.94 frames.], batch size: 91, aishell_tot_loss[loss=0.1694, simple_loss=0.2513, pruned_loss=0.04377, over 981877.76 frames.], datatang_tot_loss[loss=0.1711, simple_loss=0.2411, pruned_loss=0.05057, over 982497.01 frames.], batch size: 91, lr: 6.70e-04
+2022-06-18 18:51:26,959 INFO [train.py:874] (2/4) Epoch 12, batch 2350, datatang_loss[loss=0.1642, simple_loss=0.2334, pruned_loss=0.04751, over 4929.00 frames.], tot_loss[loss=0.171, simple_loss=0.247, pruned_loss=0.04751, over 985115.67 frames.], batch size: 62, aishell_tot_loss[loss=0.1701, simple_loss=0.2522, pruned_loss=0.04406, over 982081.52 frames.], datatang_tot_loss[loss=0.1706, simple_loss=0.2407, pruned_loss=0.05025, over 982783.37 frames.], batch size: 62, lr: 6.70e-04
+2022-06-18 18:51:58,525 INFO [train.py:874] (2/4) Epoch 12, batch 2400, datatang_loss[loss=0.1728, simple_loss=0.2511, pruned_loss=0.04725, over 4941.00 frames.], tot_loss[loss=0.1717, simple_loss=0.2475, pruned_loss=0.04797, over 985324.48 frames.], batch size: 88, aishell_tot_loss[loss=0.1705, simple_loss=0.2525, pruned_loss=0.04425, over 982319.26 frames.], datatang_tot_loss[loss=0.1712, simple_loss=0.2412, pruned_loss=0.05057, over 983391.07 frames.], batch size: 88, lr: 6.70e-04
+2022-06-18 18:52:28,274 INFO [train.py:874] (2/4) Epoch 12, batch 2450, aishell_loss[loss=0.1949, simple_loss=0.2788, pruned_loss=0.05553, over 4913.00 frames.], tot_loss[loss=0.1711, simple_loss=0.2471, pruned_loss=0.04755, over 985839.19 frames.], batch size: 34, aishell_tot_loss[loss=0.1705, simple_loss=0.2524, pruned_loss=0.04427, over 983085.25 frames.], datatang_tot_loss[loss=0.1708, simple_loss=0.2408, pruned_loss=0.05038, over 983769.99 frames.], batch size: 34, lr: 6.69e-04
+2022-06-18 18:52:59,292 INFO [train.py:874] (2/4) Epoch 12, batch 2500, datatang_loss[loss=0.1423, simple_loss=0.2142, pruned_loss=0.03521, over 4903.00 frames.], tot_loss[loss=0.1705, simple_loss=0.2461, pruned_loss=0.04748, over 985867.07 frames.], batch size: 64, aishell_tot_loss[loss=0.1708, simple_loss=0.253, pruned_loss=0.04433, over 983226.82 frames.], datatang_tot_loss[loss=0.1699, simple_loss=0.2397, pruned_loss=0.05004, over 984188.86 frames.], batch size: 64, lr: 6.69e-04
+2022-06-18 18:53:29,791 INFO [train.py:874] (2/4) Epoch 12, batch 2550, aishell_loss[loss=0.1458, simple_loss=0.2295, pruned_loss=0.031, over 4943.00 frames.], tot_loss[loss=0.1702, simple_loss=0.2459, pruned_loss=0.04724, over 986017.90 frames.], batch size: 49, aishell_tot_loss[loss=0.1705, simple_loss=0.2526, pruned_loss=0.04424, over 983767.48 frames.], datatang_tot_loss[loss=0.1699, simple_loss=0.2396, pruned_loss=0.0501, over 984342.95 frames.], batch size: 49, lr: 6.69e-04
+2022-06-18 18:54:06,340 INFO [train.py:874] (2/4) Epoch 12, batch 2600, datatang_loss[loss=0.177, simple_loss=0.2518, pruned_loss=0.05106, over 4952.00 frames.], tot_loss[loss=0.1691, simple_loss=0.2447, pruned_loss=0.04676, over 986338.71 frames.], batch size: 86, aishell_tot_loss[loss=0.1693, simple_loss=0.2515, pruned_loss=0.0436, over 984348.70 frames.], datatang_tot_loss[loss=0.1699, simple_loss=0.2395, pruned_loss=0.05011, over 984564.47 frames.], batch size: 86, lr: 6.68e-04
+2022-06-18 18:54:36,798 INFO [train.py:874] (2/4) Epoch 12, batch 2650, datatang_loss[loss=0.1925, simple_loss=0.2645, pruned_loss=0.06022, over 4918.00 frames.], tot_loss[loss=0.17, simple_loss=0.2459, pruned_loss=0.04706, over 986298.32 frames.], batch size: 98, aishell_tot_loss[loss=0.1691, simple_loss=0.2514, pruned_loss=0.04342, over 984468.38 frames.], datatang_tot_loss[loss=0.1709, simple_loss=0.2406, pruned_loss=0.05057, over 984837.07 frames.], batch size: 98, lr: 6.68e-04
+2022-06-18 18:55:07,910 INFO [train.py:874] (2/4) Epoch 12, batch 2700, datatang_loss[loss=0.1391, simple_loss=0.2119, pruned_loss=0.03315, over 4917.00 frames.], tot_loss[loss=0.1689, simple_loss=0.2449, pruned_loss=0.04639, over 986068.16 frames.], batch size: 75, aishell_tot_loss[loss=0.1691, simple_loss=0.2511, pruned_loss=0.04357, over 984607.80 frames.], datatang_tot_loss[loss=0.1698, simple_loss=0.2397, pruned_loss=0.04992, over 984841.78 frames.], batch size: 75, lr: 6.68e-04
+2022-06-18 18:55:38,327 INFO [train.py:874] (2/4) Epoch 12, batch 2750, datatang_loss[loss=0.1556, simple_loss=0.2291, pruned_loss=0.04103, over 4962.00 frames.], tot_loss[loss=0.1693, simple_loss=0.2452, pruned_loss=0.04668, over 986123.19 frames.], batch size: 67, aishell_tot_loss[loss=0.1699, simple_loss=0.2519, pruned_loss=0.0439, over 984917.02 frames.], datatang_tot_loss[loss=0.1693, simple_loss=0.2391, pruned_loss=0.04975, over 984896.45 frames.], batch size: 67, lr: 6.67e-04
+2022-06-18 18:56:09,574 INFO [train.py:874] (2/4) Epoch 12, batch 2800, aishell_loss[loss=0.1732, simple_loss=0.2507, pruned_loss=0.04779, over 4881.00 frames.], tot_loss[loss=0.169, simple_loss=0.2448, pruned_loss=0.0466, over 986077.65 frames.], batch size: 47, aishell_tot_loss[loss=0.1697, simple_loss=0.2517, pruned_loss=0.04381, over 984936.19 frames.], datatang_tot_loss[loss=0.1692, simple_loss=0.239, pruned_loss=0.04972, over 985127.50 frames.], batch size: 47, lr: 6.67e-04
+2022-06-18 18:56:40,337 INFO [train.py:874] (2/4) Epoch 12, batch 2850, datatang_loss[loss=0.1866, simple_loss=0.2547, pruned_loss=0.05922, over 4872.00 frames.], tot_loss[loss=0.1692, simple_loss=0.2452, pruned_loss=0.04666, over 986117.04 frames.], batch size: 39, aishell_tot_loss[loss=0.1698, simple_loss=0.252, pruned_loss=0.04386, over 985144.25 frames.], datatang_tot_loss[loss=0.1691, simple_loss=0.2391, pruned_loss=0.04959, over 985201.17 frames.], batch size: 39, lr: 6.66e-04
+2022-06-18 18:57:09,891 INFO [train.py:874] (2/4) Epoch 12, batch 2900, aishell_loss[loss=0.1861, simple_loss=0.2621, pruned_loss=0.05507, over 4979.00 frames.], tot_loss[loss=0.1693, simple_loss=0.2454, pruned_loss=0.04659, over 985742.25 frames.], batch size: 48, aishell_tot_loss[loss=0.1691, simple_loss=0.2513, pruned_loss=0.04346, over 984973.56 frames.], datatang_tot_loss[loss=0.1699, simple_loss=0.2397, pruned_loss=0.05001, over 985183.33 frames.], batch size: 48, lr: 6.66e-04
+2022-06-18 18:57:41,736 INFO [train.py:874] (2/4) Epoch 12, batch 2950, aishell_loss[loss=0.1294, simple_loss=0.2141, pruned_loss=0.02236, over 4956.00 frames.], tot_loss[loss=0.17, simple_loss=0.2454, pruned_loss=0.04732, over 985584.70 frames.], batch size: 30, aishell_tot_loss[loss=0.1692, simple_loss=0.2512, pruned_loss=0.04361, over 984769.53 frames.], datatang_tot_loss[loss=0.1705, simple_loss=0.2401, pruned_loss=0.05045, over 985357.98 frames.], batch size: 30, lr: 6.66e-04
+2022-06-18 18:58:12,915 INFO [train.py:874] (2/4) Epoch 12, batch 3000, aishell_loss[loss=0.1256, simple_loss=0.2026, pruned_loss=0.02426, over 4959.00 frames.], tot_loss[loss=0.1691, simple_loss=0.2446, pruned_loss=0.0468, over 985990.91 frames.], batch size: 25, aishell_tot_loss[loss=0.1689, simple_loss=0.2506, pruned_loss=0.04358, over 985069.54 frames.], datatang_tot_loss[loss=0.1699, simple_loss=0.2397, pruned_loss=0.05004, over 985620.16 frames.], batch size: 25, lr: 6.65e-04
+2022-06-18 18:58:12,916 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 18:58:29,560 INFO [train.py:914] (2/4) Epoch 12, validation: loss=0.1665, simple_loss=0.2506, pruned_loss=0.04121, over 1622729.00 frames.
+2022-06-18 18:59:00,854 INFO [train.py:874] (2/4) Epoch 12, batch 3050, aishell_loss[loss=0.187, simple_loss=0.2721, pruned_loss=0.05099, over 4930.00 frames.], tot_loss[loss=0.1681, simple_loss=0.2441, pruned_loss=0.04611, over 986251.29 frames.], batch size: 64, aishell_tot_loss[loss=0.1686, simple_loss=0.2505, pruned_loss=0.04336, over 985220.10 frames.], datatang_tot_loss[loss=0.1691, simple_loss=0.2392, pruned_loss=0.04944, over 985895.61 frames.], batch size: 64, lr: 6.65e-04
+2022-06-18 18:59:32,406 INFO [train.py:874] (2/4) Epoch 12, batch 3100, aishell_loss[loss=0.1856, simple_loss=0.2645, pruned_loss=0.0534, over 4871.00 frames.], tot_loss[loss=0.1674, simple_loss=0.2434, pruned_loss=0.04569, over 985938.70 frames.], batch size: 35, aishell_tot_loss[loss=0.1681, simple_loss=0.25, pruned_loss=0.04308, over 985120.74 frames.], datatang_tot_loss[loss=0.1687, simple_loss=0.2388, pruned_loss=0.04925, over 985854.22 frames.], batch size: 35, lr: 6.65e-04
+2022-06-18 19:00:02,795 INFO [train.py:874] (2/4) Epoch 12, batch 3150, aishell_loss[loss=0.162, simple_loss=0.2451, pruned_loss=0.03947, over 4924.00 frames.], tot_loss[loss=0.1669, simple_loss=0.243, pruned_loss=0.04539, over 985872.70 frames.], batch size: 58, aishell_tot_loss[loss=0.1679, simple_loss=0.2499, pruned_loss=0.04293, over 985018.19 frames.], datatang_tot_loss[loss=0.1681, simple_loss=0.2385, pruned_loss=0.04882, over 985974.38 frames.], batch size: 58, lr: 6.64e-04
+2022-06-18 19:00:33,839 INFO [train.py:874] (2/4) Epoch 12, batch 3200, aishell_loss[loss=0.1699, simple_loss=0.2578, pruned_loss=0.04097, over 4895.00 frames.], tot_loss[loss=0.1662, simple_loss=0.2427, pruned_loss=0.0449, over 985718.69 frames.], batch size: 50, aishell_tot_loss[loss=0.1672, simple_loss=0.2492, pruned_loss=0.04257, over 984787.61 frames.], datatang_tot_loss[loss=0.1678, simple_loss=0.2384, pruned_loss=0.04864, over 986140.56 frames.], batch size: 50, lr: 6.64e-04
+2022-06-18 19:01:04,141 INFO [train.py:874] (2/4) Epoch 12, batch 3250, aishell_loss[loss=0.1732, simple_loss=0.2576, pruned_loss=0.04437, over 4918.00 frames.], tot_loss[loss=0.1657, simple_loss=0.2422, pruned_loss=0.04465, over 985644.33 frames.], batch size: 52, aishell_tot_loss[loss=0.1664, simple_loss=0.2487, pruned_loss=0.04202, over 984979.51 frames.], datatang_tot_loss[loss=0.1678, simple_loss=0.2382, pruned_loss=0.04874, over 985939.53 frames.], batch size: 52, lr: 6.64e-04
+2022-06-18 19:01:35,674 INFO [train.py:874] (2/4) Epoch 12, batch 3300, datatang_loss[loss=0.1481, simple_loss=0.2187, pruned_loss=0.03875, over 4967.00 frames.], tot_loss[loss=0.1653, simple_loss=0.2416, pruned_loss=0.04449, over 985896.09 frames.], batch size: 60, aishell_tot_loss[loss=0.1662, simple_loss=0.2485, pruned_loss=0.04198, over 985213.43 frames.], datatang_tot_loss[loss=0.1672, simple_loss=0.2374, pruned_loss=0.04843, over 986020.04 frames.], batch size: 60, lr: 6.63e-04
+2022-06-18 19:02:06,908 INFO [train.py:874] (2/4) Epoch 12, batch 3350, aishell_loss[loss=0.1638, simple_loss=0.2506, pruned_loss=0.0385, over 4975.00 frames.], tot_loss[loss=0.1659, simple_loss=0.2428, pruned_loss=0.04452, over 985665.75 frames.], batch size: 51, aishell_tot_loss[loss=0.1668, simple_loss=0.2493, pruned_loss=0.04218, over 985058.92 frames.], datatang_tot_loss[loss=0.1669, simple_loss=0.2373, pruned_loss=0.04828, over 986014.79 frames.], batch size: 51, lr: 6.63e-04
+2022-06-18 19:02:37,644 INFO [train.py:874] (2/4) Epoch 12, batch 3400, datatang_loss[loss=0.1533, simple_loss=0.2222, pruned_loss=0.04218, over 4926.00 frames.], tot_loss[loss=0.1661, simple_loss=0.2432, pruned_loss=0.04453, over 985522.56 frames.], batch size: 83, aishell_tot_loss[loss=0.167, simple_loss=0.2496, pruned_loss=0.04219, over 984861.11 frames.], datatang_tot_loss[loss=0.1668, simple_loss=0.2372, pruned_loss=0.0482, over 986119.98 frames.], batch size: 83, lr: 6.63e-04
+2022-06-18 19:03:08,071 INFO [train.py:874] (2/4) Epoch 12, batch 3450, aishell_loss[loss=0.203, simple_loss=0.283, pruned_loss=0.0615, over 4952.00 frames.], tot_loss[loss=0.1672, simple_loss=0.2442, pruned_loss=0.04509, over 985806.93 frames.], batch size: 58, aishell_tot_loss[loss=0.1675, simple_loss=0.2502, pruned_loss=0.04243, over 985213.58 frames.], datatang_tot_loss[loss=0.1672, simple_loss=0.2375, pruned_loss=0.04846, over 986082.76 frames.], batch size: 58, lr: 6.62e-04
+2022-06-18 19:03:38,668 INFO [train.py:874] (2/4) Epoch 12, batch 3500, aishell_loss[loss=0.1805, simple_loss=0.2667, pruned_loss=0.04716, over 4949.00 frames.], tot_loss[loss=0.167, simple_loss=0.2438, pruned_loss=0.04505, over 985757.46 frames.], batch size: 54, aishell_tot_loss[loss=0.1672, simple_loss=0.2497, pruned_loss=0.04239, over 985331.13 frames.], datatang_tot_loss[loss=0.1672, simple_loss=0.2375, pruned_loss=0.04843, over 985945.75 frames.], batch size: 54, lr: 6.62e-04
+2022-06-18 19:04:07,460 INFO [train.py:874] (2/4) Epoch 12, batch 3550, datatang_loss[loss=0.1641, simple_loss=0.2353, pruned_loss=0.04644, over 4897.00 frames.], tot_loss[loss=0.1667, simple_loss=0.2435, pruned_loss=0.04495, over 985887.23 frames.], batch size: 30, aishell_tot_loss[loss=0.1669, simple_loss=0.2494, pruned_loss=0.04222, over 985568.56 frames.], datatang_tot_loss[loss=0.1672, simple_loss=0.2376, pruned_loss=0.0484, over 985871.85 frames.], batch size: 30, lr: 6.62e-04
+2022-06-18 19:04:39,572 INFO [train.py:874] (2/4) Epoch 12, batch 3600, datatang_loss[loss=0.1853, simple_loss=0.2513, pruned_loss=0.05966, over 4928.00 frames.], tot_loss[loss=0.1668, simple_loss=0.2437, pruned_loss=0.04494, over 985589.33 frames.], batch size: 81, aishell_tot_loss[loss=0.1671, simple_loss=0.2495, pruned_loss=0.04233, over 985388.25 frames.], datatang_tot_loss[loss=0.1671, simple_loss=0.2376, pruned_loss=0.04826, over 985781.35 frames.], batch size: 81, lr: 6.61e-04
+2022-06-18 19:05:11,050 INFO [train.py:874] (2/4) Epoch 12, batch 3650, aishell_loss[loss=0.1533, simple_loss=0.2433, pruned_loss=0.0317, over 4962.00 frames.], tot_loss[loss=0.1668, simple_loss=0.2437, pruned_loss=0.04493, over 985584.12 frames.], batch size: 56, aishell_tot_loss[loss=0.1669, simple_loss=0.2494, pruned_loss=0.04222, over 985353.89 frames.], datatang_tot_loss[loss=0.1671, simple_loss=0.2376, pruned_loss=0.04825, over 985823.85 frames.], batch size: 56, lr: 6.61e-04
+2022-06-18 19:05:40,303 INFO [train.py:874] (2/4) Epoch 12, batch 3700, datatang_loss[loss=0.1625, simple_loss=0.2338, pruned_loss=0.0456, over 4973.00 frames.], tot_loss[loss=0.167, simple_loss=0.2435, pruned_loss=0.04528, over 985051.97 frames.], batch size: 37, aishell_tot_loss[loss=0.167, simple_loss=0.2493, pruned_loss=0.04237, over 984981.67 frames.], datatang_tot_loss[loss=0.1672, simple_loss=0.2378, pruned_loss=0.04831, over 985613.99 frames.], batch size: 37, lr: 6.61e-04
+2022-06-18 19:06:10,952 INFO [train.py:874] (2/4) Epoch 12, batch 3750, aishell_loss[loss=0.1428, simple_loss=0.2263, pruned_loss=0.02962, over 4859.00 frames.], tot_loss[loss=0.1664, simple_loss=0.2433, pruned_loss=0.04476, over 984899.84 frames.], batch size: 37, aishell_tot_loss[loss=0.1667, simple_loss=0.2489, pruned_loss=0.04229, over 985005.13 frames.], datatang_tot_loss[loss=0.1668, simple_loss=0.2374, pruned_loss=0.0481, over 985400.20 frames.], batch size: 37, lr: 6.60e-04
+2022-06-18 19:06:39,855 INFO [train.py:874] (2/4) Epoch 12, batch 3800, datatang_loss[loss=0.1358, simple_loss=0.2174, pruned_loss=0.02707, over 4926.00 frames.], tot_loss[loss=0.1662, simple_loss=0.2431, pruned_loss=0.04467, over 984996.47 frames.], batch size: 81, aishell_tot_loss[loss=0.1668, simple_loss=0.2491, pruned_loss=0.04222, over 985071.43 frames.], datatang_tot_loss[loss=0.1665, simple_loss=0.237, pruned_loss=0.04799, over 985364.63 frames.], batch size: 81, lr: 6.60e-04
+2022-06-18 19:07:09,916 INFO [train.py:874] (2/4) Epoch 12, batch 3850, aishell_loss[loss=0.1707, simple_loss=0.2676, pruned_loss=0.03691, over 4974.00 frames.], tot_loss[loss=0.1666, simple_loss=0.2433, pruned_loss=0.04495, over 985378.09 frames.], batch size: 61, aishell_tot_loss[loss=0.1675, simple_loss=0.2499, pruned_loss=0.04259, over 985263.92 frames.], datatang_tot_loss[loss=0.166, simple_loss=0.2365, pruned_loss=0.0478, over 985519.51 frames.], batch size: 61, lr: 6.60e-04
+2022-06-18 19:07:38,510 INFO [train.py:874] (2/4) Epoch 12, batch 3900, aishell_loss[loss=0.1515, simple_loss=0.2392, pruned_loss=0.0319, over 4899.00 frames.], tot_loss[loss=0.166, simple_loss=0.2424, pruned_loss=0.04476, over 985067.45 frames.], batch size: 28, aishell_tot_loss[loss=0.1672, simple_loss=0.2496, pruned_loss=0.0424, over 984703.05 frames.], datatang_tot_loss[loss=0.1658, simple_loss=0.2363, pruned_loss=0.04758, over 985726.37 frames.], batch size: 28, lr: 6.59e-04
+2022-06-18 19:08:09,099 INFO [train.py:874] (2/4) Epoch 12, batch 3950, aishell_loss[loss=0.1917, simple_loss=0.2708, pruned_loss=0.0563, over 4943.00 frames.], tot_loss[loss=0.1663, simple_loss=0.2423, pruned_loss=0.04518, over 985073.01 frames.], batch size: 45, aishell_tot_loss[loss=0.1678, simple_loss=0.25, pruned_loss=0.04276, over 984654.28 frames.], datatang_tot_loss[loss=0.1654, simple_loss=0.2359, pruned_loss=0.04747, over 985742.37 frames.], batch size: 45, lr: 6.59e-04
+2022-06-18 19:08:37,851 INFO [train.py:874] (2/4) Epoch 12, batch 4000, datatang_loss[loss=0.1428, simple_loss=0.2025, pruned_loss=0.04158, over 4923.00 frames.], tot_loss[loss=0.1652, simple_loss=0.2416, pruned_loss=0.04443, over 984930.45 frames.], batch size: 42, aishell_tot_loss[loss=0.1676, simple_loss=0.25, pruned_loss=0.04255, over 984455.52 frames.], datatang_tot_loss[loss=0.1644, simple_loss=0.2349, pruned_loss=0.04693, over 985776.62 frames.], batch size: 42, lr: 6.59e-04
+2022-06-18 19:08:37,852 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 19:08:54,415 INFO [train.py:914] (2/4) Epoch 12, validation: loss=0.1659, simple_loss=0.2501, pruned_loss=0.0409, over 1622729.00 frames.
+2022-06-18 19:09:24,430 INFO [train.py:874] (2/4) Epoch 12, batch 4050, datatang_loss[loss=0.1533, simple_loss=0.2355, pruned_loss=0.03556, over 4922.00 frames.], tot_loss[loss=0.1645, simple_loss=0.2405, pruned_loss=0.04428, over 984877.36 frames.], batch size: 83, aishell_tot_loss[loss=0.1667, simple_loss=0.2491, pruned_loss=0.04217, over 984143.47 frames.], datatang_tot_loss[loss=0.1645, simple_loss=0.2349, pruned_loss=0.047, over 985971.25 frames.], batch size: 83, lr: 6.58e-04
+2022-06-18 19:09:52,463 INFO [train.py:874] (2/4) Epoch 12, batch 4100, datatang_loss[loss=0.1604, simple_loss=0.2258, pruned_loss=0.04746, over 4903.00 frames.], tot_loss[loss=0.1649, simple_loss=0.2414, pruned_loss=0.04416, over 984598.74 frames.], batch size: 47, aishell_tot_loss[loss=0.1665, simple_loss=0.2491, pruned_loss=0.04195, over 983848.38 frames.], datatang_tot_loss[loss=0.1647, simple_loss=0.2352, pruned_loss=0.04712, over 985976.78 frames.], batch size: 47, lr: 6.58e-04
+2022-06-18 19:11:11,422 INFO [train.py:874] (2/4) Epoch 13, batch 50, datatang_loss[loss=0.1453, simple_loss=0.2316, pruned_loss=0.02947, over 4920.00 frames.], tot_loss[loss=0.1654, simple_loss=0.2432, pruned_loss=0.04383, over 218270.50 frames.], batch size: 83, aishell_tot_loss[loss=0.1737, simple_loss=0.2584, pruned_loss=0.04448, over 107010.25 frames.], datatang_tot_loss[loss=0.1588, simple_loss=0.2307, pruned_loss=0.04346, over 124825.88 frames.], batch size: 83, lr: 6.36e-04
+2022-06-18 19:11:41,572 INFO [train.py:874] (2/4) Epoch 13, batch 100, aishell_loss[loss=0.1504, simple_loss=0.2397, pruned_loss=0.03059, over 4878.00 frames.], tot_loss[loss=0.1649, simple_loss=0.243, pruned_loss=0.04338, over 388317.44 frames.], batch size: 42, aishell_tot_loss[loss=0.1711, simple_loss=0.2553, pruned_loss=0.04348, over 229703.39 frames.], datatang_tot_loss[loss=0.158, simple_loss=0.2291, pruned_loss=0.04346, over 206852.18 frames.], batch size: 42, lr: 6.35e-04
+2022-06-18 19:12:12,255 INFO [train.py:874] (2/4) Epoch 13, batch 150, datatang_loss[loss=0.1534, simple_loss=0.2222, pruned_loss=0.04228, over 4955.00 frames.], tot_loss[loss=0.1642, simple_loss=0.2412, pruned_loss=0.04354, over 520601.57 frames.], batch size: 55, aishell_tot_loss[loss=0.1693, simple_loss=0.2527, pruned_loss=0.04298, over 308650.27 frames.], datatang_tot_loss[loss=0.1594, simple_loss=0.2304, pruned_loss=0.04416, over 308694.07 frames.], batch size: 55, lr: 6.35e-04
+2022-06-18 19:12:44,623 INFO [train.py:874] (2/4) Epoch 13, batch 200, datatang_loss[loss=0.145, simple_loss=0.2205, pruned_loss=0.0348, over 4918.00 frames.], tot_loss[loss=0.1626, simple_loss=0.2397, pruned_loss=0.0427, over 623720.95 frames.], batch size: 64, aishell_tot_loss[loss=0.1689, simple_loss=0.2524, pruned_loss=0.04272, over 379204.64 frames.], datatang_tot_loss[loss=0.1574, simple_loss=0.2289, pruned_loss=0.04297, over 397480.50 frames.], batch size: 64, lr: 6.35e-04
+2022-06-18 19:13:14,986 INFO [train.py:874] (2/4) Epoch 13, batch 250, datatang_loss[loss=0.1396, simple_loss=0.2134, pruned_loss=0.03296, over 4890.00 frames.], tot_loss[loss=0.1626, simple_loss=0.2395, pruned_loss=0.04285, over 703813.04 frames.], batch size: 52, aishell_tot_loss[loss=0.1679, simple_loss=0.2513, pruned_loss=0.04232, over 445134.18 frames.], datatang_tot_loss[loss=0.1585, simple_loss=0.2298, pruned_loss=0.04358, over 471887.33 frames.], batch size: 52, lr: 6.34e-04
+2022-06-18 19:13:46,324 INFO [train.py:874] (2/4) Epoch 13, batch 300, datatang_loss[loss=0.1602, simple_loss=0.2227, pruned_loss=0.04887, over 4953.00 frames.], tot_loss[loss=0.1625, simple_loss=0.2394, pruned_loss=0.04279, over 766245.80 frames.], batch size: 45, aishell_tot_loss[loss=0.1669, simple_loss=0.2503, pruned_loss=0.04173, over 499169.47 frames.], datatang_tot_loss[loss=0.1595, simple_loss=0.2309, pruned_loss=0.04399, over 541323.34 frames.], batch size: 45, lr: 6.34e-04
+2022-06-18 19:14:18,044 INFO [train.py:874] (2/4) Epoch 13, batch 350, aishell_loss[loss=0.1494, simple_loss=0.2371, pruned_loss=0.03081, over 4971.00 frames.], tot_loss[loss=0.1625, simple_loss=0.2392, pruned_loss=0.04294, over 814910.11 frames.], batch size: 44, aishell_tot_loss[loss=0.167, simple_loss=0.25, pruned_loss=0.04203, over 549877.45 frames.], datatang_tot_loss[loss=0.1593, simple_loss=0.231, pruned_loss=0.04386, over 599607.76 frames.], batch size: 44, lr: 6.34e-04
+2022-06-18 19:14:49,252 INFO [train.py:874] (2/4) Epoch 13, batch 400, aishell_loss[loss=0.1688, simple_loss=0.2477, pruned_loss=0.0449, over 4935.00 frames.], tot_loss[loss=0.1641, simple_loss=0.2407, pruned_loss=0.04374, over 852434.63 frames.], batch size: 58, aishell_tot_loss[loss=0.1675, simple_loss=0.2504, pruned_loss=0.04228, over 598889.33 frames.], datatang_tot_loss[loss=0.1609, simple_loss=0.2324, pruned_loss=0.04474, over 646810.78 frames.], batch size: 58, lr: 6.33e-04
+2022-06-18 19:15:19,249 INFO [train.py:874] (2/4) Epoch 13, batch 450, aishell_loss[loss=0.1483, simple_loss=0.2178, pruned_loss=0.03939, over 4984.00 frames.], tot_loss[loss=0.1649, simple_loss=0.2413, pruned_loss=0.04426, over 882141.10 frames.], batch size: 25, aishell_tot_loss[loss=0.167, simple_loss=0.2499, pruned_loss=0.04204, over 635762.30 frames.], datatang_tot_loss[loss=0.1626, simple_loss=0.2339, pruned_loss=0.04564, over 694315.50 frames.], batch size: 25, lr: 6.33e-04
+2022-06-18 19:15:51,456 INFO [train.py:874] (2/4) Epoch 13, batch 500, aishell_loss[loss=0.1443, simple_loss=0.2222, pruned_loss=0.03316, over 4811.00 frames.], tot_loss[loss=0.1657, simple_loss=0.2422, pruned_loss=0.04458, over 905169.36 frames.], batch size: 26, aishell_tot_loss[loss=0.1666, simple_loss=0.2493, pruned_loss=0.04191, over 680067.21 frames.], datatang_tot_loss[loss=0.164, simple_loss=0.2353, pruned_loss=0.04637, over 726129.59 frames.], batch size: 26, lr: 6.33e-04
+2022-06-18 19:16:22,047 INFO [train.py:874] (2/4) Epoch 13, batch 550, datatang_loss[loss=0.1656, simple_loss=0.237, pruned_loss=0.04712, over 4924.00 frames.], tot_loss[loss=0.1674, simple_loss=0.2437, pruned_loss=0.04561, over 922856.15 frames.], batch size: 81, aishell_tot_loss[loss=0.166, simple_loss=0.2489, pruned_loss=0.04161, over 717208.17 frames.], datatang_tot_loss[loss=0.1669, simple_loss=0.2375, pruned_loss=0.04812, over 755581.93 frames.], batch size: 81, lr: 6.32e-04
+2022-06-18 19:16:51,676 INFO [train.py:874] (2/4) Epoch 13, batch 600, datatang_loss[loss=0.1782, simple_loss=0.2489, pruned_loss=0.05376, over 4929.00 frames.], tot_loss[loss=0.1678, simple_loss=0.2442, pruned_loss=0.04568, over 936913.41 frames.], batch size: 57, aishell_tot_loss[loss=0.1673, simple_loss=0.2499, pruned_loss=0.04233, over 750168.66 frames.], datatang_tot_loss[loss=0.1664, simple_loss=0.2373, pruned_loss=0.04778, over 781645.74 frames.], batch size: 57, lr: 6.32e-04
+2022-06-18 19:17:23,496 INFO [train.py:874] (2/4) Epoch 13, batch 650, aishell_loss[loss=0.1803, simple_loss=0.2761, pruned_loss=0.04227, over 4963.00 frames.], tot_loss[loss=0.1675, simple_loss=0.2443, pruned_loss=0.04534, over 947802.80 frames.], batch size: 44, aishell_tot_loss[loss=0.1669, simple_loss=0.2494, pruned_loss=0.04216, over 781977.17 frames.], datatang_tot_loss[loss=0.1667, simple_loss=0.2378, pruned_loss=0.04784, over 802168.96 frames.], batch size: 44, lr: 6.32e-04
+2022-06-18 19:17:54,563 INFO [train.py:874] (2/4) Epoch 13, batch 700, aishell_loss[loss=0.168, simple_loss=0.2575, pruned_loss=0.03926, over 4936.00 frames.], tot_loss[loss=0.1677, simple_loss=0.2446, pruned_loss=0.04543, over 956190.56 frames.], batch size: 58, aishell_tot_loss[loss=0.1672, simple_loss=0.2498, pruned_loss=0.04231, over 807706.67 frames.], datatang_tot_loss[loss=0.1669, simple_loss=0.2378, pruned_loss=0.04797, over 822198.60 frames.], batch size: 58, lr: 6.31e-04
+2022-06-18 19:18:24,368 INFO [train.py:874] (2/4) Epoch 13, batch 750, aishell_loss[loss=0.1559, simple_loss=0.246, pruned_loss=0.03291, over 4905.00 frames.], tot_loss[loss=0.1675, simple_loss=0.2446, pruned_loss=0.04525, over 962576.47 frames.], batch size: 41, aishell_tot_loss[loss=0.167, simple_loss=0.25, pruned_loss=0.04204, over 829162.51 frames.], datatang_tot_loss[loss=0.1671, simple_loss=0.2378, pruned_loss=0.04817, over 840868.56 frames.], batch size: 41, lr: 6.31e-04
+2022-06-18 19:18:56,664 INFO [train.py:874] (2/4) Epoch 13, batch 800, aishell_loss[loss=0.1609, simple_loss=0.2475, pruned_loss=0.03717, over 4968.00 frames.], tot_loss[loss=0.1679, simple_loss=0.2445, pruned_loss=0.04567, over 967850.84 frames.], batch size: 61, aishell_tot_loss[loss=0.1669, simple_loss=0.2495, pruned_loss=0.04213, over 848253.60 frames.], datatang_tot_loss[loss=0.1678, simple_loss=0.2384, pruned_loss=0.04864, over 857464.90 frames.], batch size: 61, lr: 6.31e-04
+2022-06-18 19:19:27,013 INFO [train.py:874] (2/4) Epoch 13, batch 850, aishell_loss[loss=0.1437, simple_loss=0.2308, pruned_loss=0.0283, over 4977.00 frames.], tot_loss[loss=0.167, simple_loss=0.2436, pruned_loss=0.04518, over 971717.69 frames.], batch size: 51, aishell_tot_loss[loss=0.1666, simple_loss=0.2491, pruned_loss=0.04203, over 862435.43 frames.], datatang_tot_loss[loss=0.1671, simple_loss=0.238, pruned_loss=0.04813, over 874315.04 frames.], batch size: 51, lr: 6.31e-04
+2022-06-18 19:19:57,498 INFO [train.py:874] (2/4) Epoch 13, batch 900, aishell_loss[loss=0.1844, simple_loss=0.2504, pruned_loss=0.05926, over 4962.00 frames.], tot_loss[loss=0.1653, simple_loss=0.242, pruned_loss=0.04433, over 974541.18 frames.], batch size: 44, aishell_tot_loss[loss=0.1655, simple_loss=0.2479, pruned_loss=0.04152, over 877189.41 frames.], datatang_tot_loss[loss=0.1664, simple_loss=0.2373, pruned_loss=0.04769, over 886967.03 frames.], batch size: 44, lr: 6.30e-04
+2022-06-18 19:20:28,920 INFO [train.py:874] (2/4) Epoch 13, batch 950, aishell_loss[loss=0.1632, simple_loss=0.2548, pruned_loss=0.03578, over 4913.00 frames.], tot_loss[loss=0.1649, simple_loss=0.2421, pruned_loss=0.04385, over 976937.78 frames.], batch size: 52, aishell_tot_loss[loss=0.1653, simple_loss=0.2478, pruned_loss=0.04137, over 893336.49 frames.], datatang_tot_loss[loss=0.166, simple_loss=0.237, pruned_loss=0.04745, over 895340.72 frames.], batch size: 52, lr: 6.30e-04
+2022-06-18 19:20:58,023 INFO [train.py:874] (2/4) Epoch 13, batch 1000, aishell_loss[loss=0.1666, simple_loss=0.2539, pruned_loss=0.03967, over 4978.00 frames.], tot_loss[loss=0.1662, simple_loss=0.2432, pruned_loss=0.04459, over 978783.78 frames.], batch size: 44, aishell_tot_loss[loss=0.1659, simple_loss=0.2482, pruned_loss=0.04178, over 904598.83 frames.], datatang_tot_loss[loss=0.1667, simple_loss=0.2378, pruned_loss=0.04778, over 905507.57 frames.], batch size: 44, lr: 6.30e-04
+2022-06-18 19:20:58,024 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 19:21:15,123 INFO [train.py:914] (2/4) Epoch 13, validation: loss=0.1646, simple_loss=0.2486, pruned_loss=0.04032, over 1622729.00 frames.
+2022-06-18 19:21:45,793 INFO [train.py:874] (2/4) Epoch 13, batch 1050, datatang_loss[loss=0.1358, simple_loss=0.2049, pruned_loss=0.03338, over 4887.00 frames.], tot_loss[loss=0.1661, simple_loss=0.2436, pruned_loss=0.04425, over 980181.45 frames.], batch size: 52, aishell_tot_loss[loss=0.1657, simple_loss=0.2486, pruned_loss=0.04144, over 914787.13 frames.], datatang_tot_loss[loss=0.1667, simple_loss=0.2378, pruned_loss=0.04774, over 914172.20 frames.], batch size: 52, lr: 6.29e-04
+2022-06-18 19:22:16,621 INFO [train.py:874] (2/4) Epoch 13, batch 1100, datatang_loss[loss=0.1369, simple_loss=0.2096, pruned_loss=0.03213, over 4936.00 frames.], tot_loss[loss=0.1659, simple_loss=0.243, pruned_loss=0.0444, over 981089.65 frames.], batch size: 57, aishell_tot_loss[loss=0.1653, simple_loss=0.2482, pruned_loss=0.04124, over 922146.65 frames.], datatang_tot_loss[loss=0.1669, simple_loss=0.2378, pruned_loss=0.04798, over 923248.74 frames.], batch size: 57, lr: 6.29e-04
+2022-06-18 19:22:46,970 INFO [train.py:874] (2/4) Epoch 13, batch 1150, aishell_loss[loss=0.1548, simple_loss=0.2345, pruned_loss=0.03751, over 4921.00 frames.], tot_loss[loss=0.1665, simple_loss=0.2435, pruned_loss=0.04469, over 982059.53 frames.], batch size: 33, aishell_tot_loss[loss=0.1656, simple_loss=0.2486, pruned_loss=0.04131, over 930480.68 frames.], datatang_tot_loss[loss=0.1672, simple_loss=0.2378, pruned_loss=0.04828, over 929687.80 frames.], batch size: 33, lr: 6.29e-04
+2022-06-18 19:23:17,706 INFO [train.py:874] (2/4) Epoch 13, batch 1200, datatang_loss[loss=0.1511, simple_loss=0.2221, pruned_loss=0.04003, over 4926.00 frames.], tot_loss[loss=0.1669, simple_loss=0.2439, pruned_loss=0.04497, over 982797.12 frames.], batch size: 57, aishell_tot_loss[loss=0.1659, simple_loss=0.2488, pruned_loss=0.04144, over 936724.69 frames.], datatang_tot_loss[loss=0.1674, simple_loss=0.2381, pruned_loss=0.04834, over 936477.25 frames.], batch size: 57, lr: 6.28e-04
+2022-06-18 19:23:48,002 INFO [train.py:874] (2/4) Epoch 13, batch 1250, datatang_loss[loss=0.1719, simple_loss=0.2426, pruned_loss=0.0506, over 4959.00 frames.], tot_loss[loss=0.1667, simple_loss=0.2435, pruned_loss=0.04493, over 983531.72 frames.], batch size: 26, aishell_tot_loss[loss=0.1655, simple_loss=0.2484, pruned_loss=0.04131, over 942733.18 frames.], datatang_tot_loss[loss=0.1675, simple_loss=0.2381, pruned_loss=0.04849, over 942129.18 frames.], batch size: 26, lr: 6.28e-04
+2022-06-18 19:24:19,115 INFO [train.py:874] (2/4) Epoch 13, batch 1300, datatang_loss[loss=0.157, simple_loss=0.2282, pruned_loss=0.04286, over 4936.00 frames.], tot_loss[loss=0.1666, simple_loss=0.2429, pruned_loss=0.04517, over 983750.24 frames.], batch size: 50, aishell_tot_loss[loss=0.1655, simple_loss=0.2484, pruned_loss=0.04125, over 946690.35 frames.], datatang_tot_loss[loss=0.1675, simple_loss=0.2378, pruned_loss=0.04863, over 948084.10 frames.], batch size: 50, lr: 6.28e-04
+2022-06-18 19:24:49,538 INFO [train.py:874] (2/4) Epoch 13, batch 1350, aishell_loss[loss=0.1637, simple_loss=0.2492, pruned_loss=0.0391, over 4945.00 frames.], tot_loss[loss=0.1656, simple_loss=0.2424, pruned_loss=0.04436, over 984134.53 frames.], batch size: 54, aishell_tot_loss[loss=0.1656, simple_loss=0.2487, pruned_loss=0.04122, over 950750.52 frames.], datatang_tot_loss[loss=0.1663, simple_loss=0.2372, pruned_loss=0.04768, over 952978.42 frames.], batch size: 54, lr: 6.27e-04
+2022-06-18 19:25:19,801 INFO [train.py:874] (2/4) Epoch 13, batch 1400, datatang_loss[loss=0.1812, simple_loss=0.2481, pruned_loss=0.05717, over 4916.00 frames.], tot_loss[loss=0.1662, simple_loss=0.2429, pruned_loss=0.04474, over 984215.80 frames.], batch size: 64, aishell_tot_loss[loss=0.1655, simple_loss=0.2486, pruned_loss=0.04118, over 954202.37 frames.], datatang_tot_loss[loss=0.1669, simple_loss=0.2379, pruned_loss=0.04798, over 957170.45 frames.], batch size: 64, lr: 6.27e-04
+2022-06-18 19:25:50,258 INFO [train.py:874] (2/4) Epoch 13, batch 1450, aishell_loss[loss=0.1331, simple_loss=0.2151, pruned_loss=0.02554, over 4970.00 frames.], tot_loss[loss=0.1661, simple_loss=0.2424, pruned_loss=0.0449, over 984635.48 frames.], batch size: 27, aishell_tot_loss[loss=0.1652, simple_loss=0.2479, pruned_loss=0.04125, over 957604.21 frames.], datatang_tot_loss[loss=0.1671, simple_loss=0.238, pruned_loss=0.04808, over 960917.42 frames.], batch size: 27, lr: 6.27e-04
+2022-06-18 19:26:20,089 INFO [train.py:874] (2/4) Epoch 13, batch 1500, datatang_loss[loss=0.1665, simple_loss=0.2359, pruned_loss=0.04857, over 4931.00 frames.], tot_loss[loss=0.1662, simple_loss=0.2421, pruned_loss=0.04516, over 984678.22 frames.], batch size: 79, aishell_tot_loss[loss=0.1657, simple_loss=0.2481, pruned_loss=0.04171, over 960688.74 frames.], datatang_tot_loss[loss=0.1667, simple_loss=0.2375, pruned_loss=0.04799, over 963870.50 frames.], batch size: 79, lr: 6.27e-04
+2022-06-18 19:26:50,902 INFO [train.py:874] (2/4) Epoch 13, batch 1550, aishell_loss[loss=0.1724, simple_loss=0.261, pruned_loss=0.04191, over 4867.00 frames.], tot_loss[loss=0.1669, simple_loss=0.2432, pruned_loss=0.0453, over 984804.23 frames.], batch size: 36, aishell_tot_loss[loss=0.1662, simple_loss=0.2488, pruned_loss=0.04183, over 963838.38 frames.], datatang_tot_loss[loss=0.1669, simple_loss=0.2375, pruned_loss=0.04815, over 966153.63 frames.], batch size: 36, lr: 6.26e-04
+2022-06-18 19:27:20,338 INFO [train.py:874] (2/4) Epoch 13, batch 1600, datatang_loss[loss=0.1596, simple_loss=0.2252, pruned_loss=0.04699, over 4975.00 frames.], tot_loss[loss=0.1666, simple_loss=0.2431, pruned_loss=0.04505, over 984860.33 frames.], batch size: 34, aishell_tot_loss[loss=0.166, simple_loss=0.2485, pruned_loss=0.0417, over 966104.02 frames.], datatang_tot_loss[loss=0.167, simple_loss=0.2378, pruned_loss=0.04809, over 968614.03 frames.], batch size: 34, lr: 6.26e-04
+2022-06-18 19:27:49,904 INFO [train.py:874] (2/4) Epoch 13, batch 1650, datatang_loss[loss=0.1779, simple_loss=0.2501, pruned_loss=0.05283, over 4959.00 frames.], tot_loss[loss=0.1669, simple_loss=0.2435, pruned_loss=0.04509, over 984431.59 frames.], batch size: 37, aishell_tot_loss[loss=0.1665, simple_loss=0.2491, pruned_loss=0.042, over 967770.51 frames.], datatang_tot_loss[loss=0.1667, simple_loss=0.2376, pruned_loss=0.04789, over 970623.32 frames.], batch size: 37, lr: 6.26e-04
+2022-06-18 19:28:22,049 INFO [train.py:874] (2/4) Epoch 13, batch 1700, aishell_loss[loss=0.173, simple_loss=0.2563, pruned_loss=0.0449, over 4939.00 frames.], tot_loss[loss=0.1664, simple_loss=0.243, pruned_loss=0.04491, over 984563.54 frames.], batch size: 49, aishell_tot_loss[loss=0.1663, simple_loss=0.2487, pruned_loss=0.04195, over 969197.19 frames.], datatang_tot_loss[loss=0.1665, simple_loss=0.2377, pruned_loss=0.04764, over 972851.93 frames.], batch size: 49, lr: 6.25e-04
+2022-06-18 19:28:51,100 INFO [train.py:874] (2/4) Epoch 13, batch 1750, datatang_loss[loss=0.1933, simple_loss=0.2613, pruned_loss=0.06271, over 4969.00 frames.], tot_loss[loss=0.1661, simple_loss=0.243, pruned_loss=0.04466, over 984602.21 frames.], batch size: 60, aishell_tot_loss[loss=0.1661, simple_loss=0.2485, pruned_loss=0.04187, over 971135.65 frames.], datatang_tot_loss[loss=0.1664, simple_loss=0.2376, pruned_loss=0.0476, over 974205.69 frames.], batch size: 60, lr: 6.25e-04
+2022-06-18 19:29:20,515 INFO [train.py:874] (2/4) Epoch 13, batch 1800, aishell_loss[loss=0.1753, simple_loss=0.2596, pruned_loss=0.04546, over 4873.00 frames.], tot_loss[loss=0.1653, simple_loss=0.2419, pruned_loss=0.04429, over 984275.78 frames.], batch size: 35, aishell_tot_loss[loss=0.1656, simple_loss=0.2477, pruned_loss=0.0417, over 972536.79 frames.], datatang_tot_loss[loss=0.166, simple_loss=0.237, pruned_loss=0.04744, over 975321.07 frames.], batch size: 35, lr: 6.25e-04
+2022-06-18 19:29:52,027 INFO [train.py:874] (2/4) Epoch 13, batch 1850, datatang_loss[loss=0.1693, simple_loss=0.2356, pruned_loss=0.05147, over 4930.00 frames.], tot_loss[loss=0.1665, simple_loss=0.2425, pruned_loss=0.04521, over 984643.94 frames.], batch size: 79, aishell_tot_loss[loss=0.166, simple_loss=0.2481, pruned_loss=0.04194, over 974111.40 frames.], datatang_tot_loss[loss=0.1667, simple_loss=0.2373, pruned_loss=0.0481, over 976558.43 frames.], batch size: 79, lr: 6.24e-04
+2022-06-18 19:30:21,947 INFO [train.py:874] (2/4) Epoch 13, batch 1900, datatang_loss[loss=0.1618, simple_loss=0.2419, pruned_loss=0.04087, over 4955.00 frames.], tot_loss[loss=0.1655, simple_loss=0.2419, pruned_loss=0.0446, over 985293.90 frames.], batch size: 91, aishell_tot_loss[loss=0.1654, simple_loss=0.2478, pruned_loss=0.04155, over 975618.38 frames.], datatang_tot_loss[loss=0.1663, simple_loss=0.237, pruned_loss=0.04776, over 977904.10 frames.], batch size: 91, lr: 6.24e-04
+2022-06-18 19:30:51,588 INFO [train.py:874] (2/4) Epoch 13, batch 1950, datatang_loss[loss=0.1645, simple_loss=0.244, pruned_loss=0.04247, over 4971.00 frames.], tot_loss[loss=0.1657, simple_loss=0.2421, pruned_loss=0.04463, over 985530.89 frames.], batch size: 34, aishell_tot_loss[loss=0.166, simple_loss=0.2481, pruned_loss=0.04189, over 976464.78 frames.], datatang_tot_loss[loss=0.1658, simple_loss=0.237, pruned_loss=0.04732, over 979252.78 frames.], batch size: 34, lr: 6.24e-04
+2022-06-18 19:31:23,219 INFO [train.py:874] (2/4) Epoch 13, batch 2000, aishell_loss[loss=0.1546, simple_loss=0.24, pruned_loss=0.0346, over 4932.00 frames.], tot_loss[loss=0.1669, simple_loss=0.2432, pruned_loss=0.04526, over 985911.20 frames.], batch size: 49, aishell_tot_loss[loss=0.1667, simple_loss=0.2485, pruned_loss=0.04247, over 978065.75 frames.], datatang_tot_loss[loss=0.1662, simple_loss=0.2374, pruned_loss=0.04755, over 979916.13 frames.], batch size: 49, lr: 6.24e-04
+2022-06-18 19:31:23,220 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 19:31:40,586 INFO [train.py:914] (2/4) Epoch 13, validation: loss=0.1662, simple_loss=0.2512, pruned_loss=0.04057, over 1622729.00 frames.
+2022-06-18 19:32:10,589 INFO [train.py:874] (2/4) Epoch 13, batch 2050, aishell_loss[loss=0.1561, simple_loss=0.2462, pruned_loss=0.03298, over 4878.00 frames.], tot_loss[loss=0.1655, simple_loss=0.2422, pruned_loss=0.04435, over 985610.19 frames.], batch size: 42, aishell_tot_loss[loss=0.167, simple_loss=0.2488, pruned_loss=0.04259, over 978999.87 frames.], datatang_tot_loss[loss=0.1645, simple_loss=0.2359, pruned_loss=0.04657, over 980294.87 frames.], batch size: 42, lr: 6.23e-04
+2022-06-18 19:32:41,259 INFO [train.py:874] (2/4) Epoch 13, batch 2100, aishell_loss[loss=0.1927, simple_loss=0.2798, pruned_loss=0.05277, over 4959.00 frames.], tot_loss[loss=0.1646, simple_loss=0.2417, pruned_loss=0.04373, over 986157.11 frames.], batch size: 64, aishell_tot_loss[loss=0.1666, simple_loss=0.2486, pruned_loss=0.0423, over 980034.33 frames.], datatang_tot_loss[loss=0.1639, simple_loss=0.2356, pruned_loss=0.04608, over 981225.04 frames.], batch size: 64, lr: 6.23e-04
+2022-06-18 19:33:09,876 INFO [train.py:874] (2/4) Epoch 13, batch 2150, datatang_loss[loss=0.1716, simple_loss=0.2338, pruned_loss=0.05475, over 4887.00 frames.], tot_loss[loss=0.1639, simple_loss=0.2411, pruned_loss=0.04331, over 986378.08 frames.], batch size: 42, aishell_tot_loss[loss=0.1657, simple_loss=0.2478, pruned_loss=0.04179, over 980713.05 frames.], datatang_tot_loss[loss=0.1639, simple_loss=0.2357, pruned_loss=0.04602, over 982074.19 frames.], batch size: 42, lr: 6.23e-04
+2022-06-18 19:33:41,240 INFO [train.py:874] (2/4) Epoch 13, batch 2200, aishell_loss[loss=0.1452, simple_loss=0.2334, pruned_loss=0.02851, over 4885.00 frames.], tot_loss[loss=0.1639, simple_loss=0.2413, pruned_loss=0.0433, over 985853.64 frames.], batch size: 28, aishell_tot_loss[loss=0.1657, simple_loss=0.2481, pruned_loss=0.0416, over 981215.84 frames.], datatang_tot_loss[loss=0.1638, simple_loss=0.2354, pruned_loss=0.04606, over 982165.23 frames.], batch size: 28, lr: 6.22e-04
+2022-06-18 19:34:10,779 INFO [train.py:874] (2/4) Epoch 13, batch 2250, aishell_loss[loss=0.1791, simple_loss=0.272, pruned_loss=0.04306, over 4909.00 frames.], tot_loss[loss=0.1647, simple_loss=0.2423, pruned_loss=0.04354, over 985682.19 frames.], batch size: 78, aishell_tot_loss[loss=0.1663, simple_loss=0.249, pruned_loss=0.04182, over 981435.34 frames.], datatang_tot_loss[loss=0.1637, simple_loss=0.2355, pruned_loss=0.04597, over 982740.18 frames.], batch size: 78, lr: 6.22e-04
+2022-06-18 19:34:40,236 INFO [train.py:874] (2/4) Epoch 13, batch 2300, datatang_loss[loss=0.169, simple_loss=0.2456, pruned_loss=0.04618, over 4934.00 frames.], tot_loss[loss=0.164, simple_loss=0.2415, pruned_loss=0.04327, over 985797.41 frames.], batch size: 88, aishell_tot_loss[loss=0.1659, simple_loss=0.2486, pruned_loss=0.04154, over 981863.33 frames.], datatang_tot_loss[loss=0.1634, simple_loss=0.2352, pruned_loss=0.04584, over 983266.93 frames.], batch size: 88, lr: 6.22e-04
+2022-06-18 19:35:11,446 INFO [train.py:874] (2/4) Epoch 13, batch 2350, aishell_loss[loss=0.1822, simple_loss=0.2746, pruned_loss=0.04496, over 4927.00 frames.], tot_loss[loss=0.1653, simple_loss=0.2426, pruned_loss=0.04395, over 985925.79 frames.], batch size: 68, aishell_tot_loss[loss=0.1664, simple_loss=0.2489, pruned_loss=0.04194, over 982443.23 frames.], datatang_tot_loss[loss=0.164, simple_loss=0.2356, pruned_loss=0.04619, over 983604.75 frames.], batch size: 68, lr: 6.21e-04
+2022-06-18 19:35:41,093 INFO [train.py:874] (2/4) Epoch 13, batch 2400, datatang_loss[loss=0.1649, simple_loss=0.2457, pruned_loss=0.04202, over 4957.00 frames.], tot_loss[loss=0.1654, simple_loss=0.2427, pruned_loss=0.04402, over 985791.42 frames.], batch size: 86, aishell_tot_loss[loss=0.1661, simple_loss=0.2488, pruned_loss=0.04171, over 982567.29 frames.], datatang_tot_loss[loss=0.1645, simple_loss=0.2361, pruned_loss=0.04642, over 983985.42 frames.], batch size: 86, lr: 6.21e-04
+2022-06-18 19:36:10,582 INFO [train.py:874] (2/4) Epoch 13, batch 2450, datatang_loss[loss=0.1722, simple_loss=0.2528, pruned_loss=0.04579, over 4935.00 frames.], tot_loss[loss=0.1654, simple_loss=0.243, pruned_loss=0.04387, over 985229.76 frames.], batch size: 94, aishell_tot_loss[loss=0.1664, simple_loss=0.2491, pruned_loss=0.04185, over 982597.38 frames.], datatang_tot_loss[loss=0.1641, simple_loss=0.236, pruned_loss=0.04612, over 983962.85 frames.], batch size: 94, lr: 6.21e-04
+2022-06-18 19:36:46,668 INFO [train.py:874] (2/4) Epoch 13, batch 2500, datatang_loss[loss=0.1712, simple_loss=0.252, pruned_loss=0.04516, over 4960.00 frames.], tot_loss[loss=0.1648, simple_loss=0.2427, pruned_loss=0.04352, over 985172.14 frames.], batch size: 37, aishell_tot_loss[loss=0.1657, simple_loss=0.2485, pruned_loss=0.0414, over 982726.08 frames.], datatang_tot_loss[loss=0.1643, simple_loss=0.2362, pruned_loss=0.04622, over 984242.47 frames.], batch size: 37, lr: 6.21e-04
+2022-06-18 19:37:16,483 INFO [train.py:874] (2/4) Epoch 13, batch 2550, datatang_loss[loss=0.1869, simple_loss=0.2517, pruned_loss=0.06107, over 4901.00 frames.], tot_loss[loss=0.1644, simple_loss=0.242, pruned_loss=0.04337, over 985190.05 frames.], batch size: 52, aishell_tot_loss[loss=0.1654, simple_loss=0.2487, pruned_loss=0.0411, over 983030.66 frames.], datatang_tot_loss[loss=0.164, simple_loss=0.2354, pruned_loss=0.04627, over 984381.86 frames.], batch size: 52, lr: 6.20e-04
+2022-06-18 19:37:46,578 INFO [train.py:874] (2/4) Epoch 13, batch 2600, datatang_loss[loss=0.1958, simple_loss=0.2638, pruned_loss=0.0639, over 4958.00 frames.], tot_loss[loss=0.1653, simple_loss=0.2429, pruned_loss=0.04391, over 985171.84 frames.], batch size: 99, aishell_tot_loss[loss=0.1651, simple_loss=0.2484, pruned_loss=0.04088, over 983250.71 frames.], datatang_tot_loss[loss=0.1652, simple_loss=0.2365, pruned_loss=0.04697, over 984493.10 frames.], batch size: 99, lr: 6.20e-04
+2022-06-18 19:38:17,672 INFO [train.py:874] (2/4) Epoch 13, batch 2650, aishell_loss[loss=0.1623, simple_loss=0.2555, pruned_loss=0.03452, over 4913.00 frames.], tot_loss[loss=0.1658, simple_loss=0.2437, pruned_loss=0.04395, over 985263.40 frames.], batch size: 46, aishell_tot_loss[loss=0.1653, simple_loss=0.2489, pruned_loss=0.04086, over 983657.99 frames.], datatang_tot_loss[loss=0.1656, simple_loss=0.237, pruned_loss=0.04708, over 984492.29 frames.], batch size: 46, lr: 6.20e-04
+2022-06-18 19:38:46,887 INFO [train.py:874] (2/4) Epoch 13, batch 2700, datatang_loss[loss=0.1755, simple_loss=0.2406, pruned_loss=0.05519, over 4965.00 frames.], tot_loss[loss=0.165, simple_loss=0.2427, pruned_loss=0.04363, over 985466.32 frames.], batch size: 45, aishell_tot_loss[loss=0.1655, simple_loss=0.2492, pruned_loss=0.04091, over 983779.25 frames.], datatang_tot_loss[loss=0.1646, simple_loss=0.2361, pruned_loss=0.04658, over 984836.56 frames.], batch size: 45, lr: 6.19e-04
+2022-06-18 19:39:16,957 INFO [train.py:874] (2/4) Epoch 13, batch 2750, datatang_loss[loss=0.171, simple_loss=0.2372, pruned_loss=0.05244, over 4911.00 frames.], tot_loss[loss=0.1654, simple_loss=0.2428, pruned_loss=0.04404, over 985481.16 frames.], batch size: 47, aishell_tot_loss[loss=0.1653, simple_loss=0.2491, pruned_loss=0.04073, over 983990.03 frames.], datatang_tot_loss[loss=0.1653, simple_loss=0.2364, pruned_loss=0.0471, over 984900.21 frames.], batch size: 47, lr: 6.19e-04
+2022-06-18 19:39:48,811 INFO [train.py:874] (2/4) Epoch 13, batch 2800, aishell_loss[loss=0.1544, simple_loss=0.2388, pruned_loss=0.035, over 4857.00 frames.], tot_loss[loss=0.1642, simple_loss=0.2412, pruned_loss=0.0436, over 985466.52 frames.], batch size: 37, aishell_tot_loss[loss=0.1643, simple_loss=0.2481, pruned_loss=0.04029, over 984006.36 frames.], datatang_tot_loss[loss=0.165, simple_loss=0.236, pruned_loss=0.04706, over 985091.65 frames.], batch size: 37, lr: 6.19e-04
+2022-06-18 19:40:18,682 INFO [train.py:874] (2/4) Epoch 13, batch 2850, aishell_loss[loss=0.1587, simple_loss=0.2556, pruned_loss=0.03093, over 4938.00 frames.], tot_loss[loss=0.1648, simple_loss=0.2418, pruned_loss=0.04389, over 985918.76 frames.], batch size: 58, aishell_tot_loss[loss=0.1647, simple_loss=0.2481, pruned_loss=0.04071, over 984309.52 frames.], datatang_tot_loss[loss=0.1652, simple_loss=0.2361, pruned_loss=0.04712, over 985522.87 frames.], batch size: 58, lr: 6.18e-04
+2022-06-18 19:40:48,089 INFO [train.py:874] (2/4) Epoch 13, batch 2900, datatang_loss[loss=0.1362, simple_loss=0.2053, pruned_loss=0.03359, over 4913.00 frames.], tot_loss[loss=0.1648, simple_loss=0.2419, pruned_loss=0.04388, over 985922.36 frames.], batch size: 52, aishell_tot_loss[loss=0.1649, simple_loss=0.2481, pruned_loss=0.04084, over 984611.23 frames.], datatang_tot_loss[loss=0.165, simple_loss=0.2359, pruned_loss=0.04705, over 985446.67 frames.], batch size: 52, lr: 6.18e-04
+2022-06-18 19:41:19,749 INFO [train.py:874] (2/4) Epoch 13, batch 2950, aishell_loss[loss=0.1344, simple_loss=0.2185, pruned_loss=0.02508, over 4907.00 frames.], tot_loss[loss=0.1649, simple_loss=0.2417, pruned_loss=0.04404, over 985761.47 frames.], batch size: 28, aishell_tot_loss[loss=0.1653, simple_loss=0.2483, pruned_loss=0.04113, over 984583.56 frames.], datatang_tot_loss[loss=0.1646, simple_loss=0.2356, pruned_loss=0.04684, over 985484.61 frames.], batch size: 28, lr: 6.18e-04
+2022-06-18 19:41:48,994 INFO [train.py:874] (2/4) Epoch 13, batch 3000, aishell_loss[loss=0.1854, simple_loss=0.26, pruned_loss=0.05544, over 4880.00 frames.], tot_loss[loss=0.1653, simple_loss=0.242, pruned_loss=0.0443, over 985418.28 frames.], batch size: 42, aishell_tot_loss[loss=0.1655, simple_loss=0.2484, pruned_loss=0.04126, over 984402.19 frames.], datatang_tot_loss[loss=0.1649, simple_loss=0.2359, pruned_loss=0.04694, over 985451.94 frames.], batch size: 42, lr: 6.18e-04
+2022-06-18 19:41:48,995 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 19:42:05,182 INFO [train.py:914] (2/4) Epoch 13, validation: loss=0.1653, simple_loss=0.2497, pruned_loss=0.04043, over 1622729.00 frames.
+2022-06-18 19:42:35,289 INFO [train.py:874] (2/4) Epoch 13, batch 3050, datatang_loss[loss=0.165, simple_loss=0.238, pruned_loss=0.04603, over 4937.00 frames.], tot_loss[loss=0.1649, simple_loss=0.242, pruned_loss=0.04389, over 985382.21 frames.], batch size: 79, aishell_tot_loss[loss=0.1657, simple_loss=0.2488, pruned_loss=0.04129, over 984646.87 frames.], datatang_tot_loss[loss=0.1643, simple_loss=0.2355, pruned_loss=0.04649, over 985278.93 frames.], batch size: 79, lr: 6.17e-04
+2022-06-18 19:43:05,348 INFO [train.py:874] (2/4) Epoch 13, batch 3100, datatang_loss[loss=0.1584, simple_loss=0.2313, pruned_loss=0.04271, over 4950.00 frames.], tot_loss[loss=0.1634, simple_loss=0.2405, pruned_loss=0.04321, over 985606.88 frames.], batch size: 88, aishell_tot_loss[loss=0.1652, simple_loss=0.2483, pruned_loss=0.04101, over 984755.72 frames.], datatang_tot_loss[loss=0.1633, simple_loss=0.2347, pruned_loss=0.04597, over 985483.10 frames.], batch size: 88, lr: 6.17e-04
+2022-06-18 19:43:36,263 INFO [train.py:874] (2/4) Epoch 13, batch 3150, aishell_loss[loss=0.1732, simple_loss=0.2584, pruned_loss=0.04399, over 4956.00 frames.], tot_loss[loss=0.1637, simple_loss=0.2407, pruned_loss=0.04336, over 985647.55 frames.], batch size: 56, aishell_tot_loss[loss=0.165, simple_loss=0.2482, pruned_loss=0.04093, over 984728.96 frames.], datatang_tot_loss[loss=0.1636, simple_loss=0.2348, pruned_loss=0.04619, over 985682.43 frames.], batch size: 56, lr: 6.17e-04
+2022-06-18 19:44:06,009 INFO [train.py:874] (2/4) Epoch 13, batch 3200, aishell_loss[loss=0.1786, simple_loss=0.264, pruned_loss=0.04661, over 4887.00 frames.], tot_loss[loss=0.1635, simple_loss=0.2407, pruned_loss=0.04313, over 985744.40 frames.], batch size: 34, aishell_tot_loss[loss=0.1651, simple_loss=0.2485, pruned_loss=0.04087, over 984845.97 frames.], datatang_tot_loss[loss=0.1632, simple_loss=0.2347, pruned_loss=0.04585, over 985756.08 frames.], batch size: 34, lr: 6.16e-04
+2022-06-18 19:44:35,533 INFO [train.py:874] (2/4) Epoch 13, batch 3250, datatang_loss[loss=0.2197, simple_loss=0.2785, pruned_loss=0.08041, over 4933.00 frames.], tot_loss[loss=0.1649, simple_loss=0.2419, pruned_loss=0.04392, over 985877.69 frames.], batch size: 108, aishell_tot_loss[loss=0.1659, simple_loss=0.2491, pruned_loss=0.04137, over 984957.22 frames.], datatang_tot_loss[loss=0.1636, simple_loss=0.2351, pruned_loss=0.04611, over 985888.20 frames.], batch size: 108, lr: 6.16e-04
+2022-06-18 19:45:06,292 INFO [train.py:874] (2/4) Epoch 13, batch 3300, datatang_loss[loss=0.1598, simple_loss=0.2393, pruned_loss=0.04014, over 4923.00 frames.], tot_loss[loss=0.1649, simple_loss=0.2422, pruned_loss=0.04385, over 985818.73 frames.], batch size: 42, aishell_tot_loss[loss=0.1653, simple_loss=0.2486, pruned_loss=0.04101, over 984942.85 frames.], datatang_tot_loss[loss=0.1644, simple_loss=0.2357, pruned_loss=0.04652, over 985962.16 frames.], batch size: 42, lr: 6.16e-04
+2022-06-18 19:45:35,734 INFO [train.py:874] (2/4) Epoch 13, batch 3350, datatang_loss[loss=0.2065, simple_loss=0.2716, pruned_loss=0.07066, over 4919.00 frames.], tot_loss[loss=0.1645, simple_loss=0.2413, pruned_loss=0.04379, over 985742.67 frames.], batch size: 98, aishell_tot_loss[loss=0.1647, simple_loss=0.2479, pruned_loss=0.0408, over 984735.10 frames.], datatang_tot_loss[loss=0.1644, simple_loss=0.2357, pruned_loss=0.04656, over 986151.72 frames.], batch size: 98, lr: 6.16e-04
+2022-06-18 19:46:06,203 INFO [train.py:874] (2/4) Epoch 13, batch 3400, datatang_loss[loss=0.1478, simple_loss=0.2076, pruned_loss=0.04401, over 4954.00 frames.], tot_loss[loss=0.1647, simple_loss=0.2414, pruned_loss=0.04397, over 985659.11 frames.], batch size: 50, aishell_tot_loss[loss=0.1649, simple_loss=0.248, pruned_loss=0.04085, over 984671.69 frames.], datatang_tot_loss[loss=0.1645, simple_loss=0.2357, pruned_loss=0.04662, over 986181.54 frames.], batch size: 50, lr: 6.15e-04
+2022-06-18 19:46:38,157 INFO [train.py:874] (2/4) Epoch 13, batch 3450, datatang_loss[loss=0.1542, simple_loss=0.2247, pruned_loss=0.0418, over 4919.00 frames.], tot_loss[loss=0.1659, simple_loss=0.2429, pruned_loss=0.04445, over 985683.45 frames.], batch size: 57, aishell_tot_loss[loss=0.1658, simple_loss=0.2491, pruned_loss=0.04123, over 984574.57 frames.], datatang_tot_loss[loss=0.1649, simple_loss=0.2362, pruned_loss=0.04677, over 986351.45 frames.], batch size: 57, lr: 6.15e-04
+2022-06-18 19:47:07,974 INFO [train.py:874] (2/4) Epoch 13, batch 3500, datatang_loss[loss=0.1726, simple_loss=0.2501, pruned_loss=0.04755, over 4819.00 frames.], tot_loss[loss=0.1657, simple_loss=0.2429, pruned_loss=0.04424, over 986045.97 frames.], batch size: 30, aishell_tot_loss[loss=0.1657, simple_loss=0.249, pruned_loss=0.04116, over 985195.37 frames.], datatang_tot_loss[loss=0.165, simple_loss=0.2364, pruned_loss=0.04675, over 986175.96 frames.], batch size: 30, lr: 6.15e-04
+2022-06-18 19:47:37,790 INFO [train.py:874] (2/4) Epoch 13, batch 3550, aishell_loss[loss=0.1729, simple_loss=0.2642, pruned_loss=0.0408, over 4905.00 frames.], tot_loss[loss=0.1645, simple_loss=0.2419, pruned_loss=0.04352, over 985437.63 frames.], batch size: 68, aishell_tot_loss[loss=0.1646, simple_loss=0.248, pruned_loss=0.04064, over 984776.35 frames.], datatang_tot_loss[loss=0.1647, simple_loss=0.2362, pruned_loss=0.04661, over 986042.00 frames.], batch size: 68, lr: 6.14e-04
+2022-06-18 19:48:09,654 INFO [train.py:874] (2/4) Epoch 13, batch 3600, datatang_loss[loss=0.155, simple_loss=0.2283, pruned_loss=0.04084, over 4905.00 frames.], tot_loss[loss=0.1644, simple_loss=0.2419, pruned_loss=0.04352, over 985181.32 frames.], batch size: 64, aishell_tot_loss[loss=0.1642, simple_loss=0.2473, pruned_loss=0.0405, over 984600.19 frames.], datatang_tot_loss[loss=0.1651, simple_loss=0.2366, pruned_loss=0.04681, over 985971.03 frames.], batch size: 64, lr: 6.14e-04
+2022-06-18 19:48:39,129 INFO [train.py:874] (2/4) Epoch 13, batch 3650, datatang_loss[loss=0.1414, simple_loss=0.2182, pruned_loss=0.03229, over 4945.00 frames.], tot_loss[loss=0.1629, simple_loss=0.2401, pruned_loss=0.04283, over 984961.55 frames.], batch size: 55, aishell_tot_loss[loss=0.1634, simple_loss=0.2465, pruned_loss=0.0402, over 984376.69 frames.], datatang_tot_loss[loss=0.1641, simple_loss=0.2355, pruned_loss=0.04638, over 985949.60 frames.], batch size: 55, lr: 6.14e-04
+2022-06-18 19:49:09,236 INFO [train.py:874] (2/4) Epoch 13, batch 3700, datatang_loss[loss=0.1677, simple_loss=0.2349, pruned_loss=0.0502, over 4925.00 frames.], tot_loss[loss=0.1624, simple_loss=0.2398, pruned_loss=0.04248, over 985040.46 frames.], batch size: 50, aishell_tot_loss[loss=0.1629, simple_loss=0.2462, pruned_loss=0.0398, over 984340.52 frames.], datatang_tot_loss[loss=0.1639, simple_loss=0.2353, pruned_loss=0.04624, over 986001.54 frames.], batch size: 50, lr: 6.14e-04
+2022-06-18 19:49:40,820 INFO [train.py:874] (2/4) Epoch 13, batch 3750, aishell_loss[loss=0.1746, simple_loss=0.2641, pruned_loss=0.04259, over 4879.00 frames.], tot_loss[loss=0.1632, simple_loss=0.2407, pruned_loss=0.04289, over 985153.32 frames.], batch size: 34, aishell_tot_loss[loss=0.1635, simple_loss=0.2467, pruned_loss=0.04008, over 984278.25 frames.], datatang_tot_loss[loss=0.164, simple_loss=0.2355, pruned_loss=0.0463, over 986151.83 frames.], batch size: 34, lr: 6.13e-04
+2022-06-18 19:50:08,757 INFO [train.py:874] (2/4) Epoch 13, batch 3800, aishell_loss[loss=0.1405, simple_loss=0.2274, pruned_loss=0.02677, over 4976.00 frames.], tot_loss[loss=0.1631, simple_loss=0.2407, pruned_loss=0.04274, over 984978.66 frames.], batch size: 30, aishell_tot_loss[loss=0.1637, simple_loss=0.2469, pruned_loss=0.04021, over 984280.72 frames.], datatang_tot_loss[loss=0.1635, simple_loss=0.2352, pruned_loss=0.04593, over 985950.94 frames.], batch size: 30, lr: 6.13e-04
+2022-06-18 19:50:39,578 INFO [train.py:874] (2/4) Epoch 13, batch 3850, datatang_loss[loss=0.1511, simple_loss=0.2205, pruned_loss=0.04083, over 4963.00 frames.], tot_loss[loss=0.1637, simple_loss=0.2413, pruned_loss=0.04306, over 985148.14 frames.], batch size: 45, aishell_tot_loss[loss=0.1639, simple_loss=0.2472, pruned_loss=0.04028, over 984445.29 frames.], datatang_tot_loss[loss=0.1638, simple_loss=0.2352, pruned_loss=0.04621, over 985954.17 frames.], batch size: 45, lr: 6.13e-04
+2022-06-18 19:51:08,419 INFO [train.py:874] (2/4) Epoch 13, batch 3900, aishell_loss[loss=0.166, simple_loss=0.2462, pruned_loss=0.04287, over 4879.00 frames.], tot_loss[loss=0.1649, simple_loss=0.2421, pruned_loss=0.04385, over 985587.25 frames.], batch size: 42, aishell_tot_loss[loss=0.1646, simple_loss=0.2479, pruned_loss=0.04066, over 984652.37 frames.], datatang_tot_loss[loss=0.1643, simple_loss=0.2354, pruned_loss=0.04664, over 986203.47 frames.], batch size: 42, lr: 6.12e-04
+2022-06-18 19:51:38,455 INFO [train.py:874] (2/4) Epoch 13, batch 3950, aishell_loss[loss=0.1306, simple_loss=0.2053, pruned_loss=0.02793, over 4802.00 frames.], tot_loss[loss=0.1632, simple_loss=0.2404, pruned_loss=0.043, over 985459.41 frames.], batch size: 24, aishell_tot_loss[loss=0.163, simple_loss=0.2459, pruned_loss=0.04004, over 984476.44 frames.], datatang_tot_loss[loss=0.1643, simple_loss=0.2356, pruned_loss=0.04646, over 986300.12 frames.], batch size: 24, lr: 6.12e-04
+2022-06-18 19:52:05,669 INFO [train.py:874] (2/4) Epoch 13, batch 4000, datatang_loss[loss=0.1434, simple_loss=0.2106, pruned_loss=0.03812, over 4798.00 frames.], tot_loss[loss=0.1625, simple_loss=0.2399, pruned_loss=0.04259, over 985231.83 frames.], batch size: 24, aishell_tot_loss[loss=0.1628, simple_loss=0.2456, pruned_loss=0.04003, over 984404.66 frames.], datatang_tot_loss[loss=0.1637, simple_loss=0.2353, pruned_loss=0.04603, over 986184.96 frames.], batch size: 24, lr: 6.12e-04
+2022-06-18 19:52:05,669 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 19:52:22,448 INFO [train.py:914] (2/4) Epoch 13, validation: loss=0.1652, simple_loss=0.2502, pruned_loss=0.04007, over 1622729.00 frames.
+2022-06-18 19:52:50,034 INFO [train.py:874] (2/4) Epoch 13, batch 4050, datatang_loss[loss=0.1662, simple_loss=0.2466, pruned_loss=0.04296, over 4943.00 frames.], tot_loss[loss=0.163, simple_loss=0.2405, pruned_loss=0.0427, over 985715.57 frames.], batch size: 88, aishell_tot_loss[loss=0.1631, simple_loss=0.246, pruned_loss=0.04005, over 984781.84 frames.], datatang_tot_loss[loss=0.1637, simple_loss=0.2354, pruned_loss=0.04602, over 986303.36 frames.], batch size: 88, lr: 6.12e-04
+2022-06-18 19:53:20,286 INFO [train.py:874] (2/4) Epoch 13, batch 4100, datatang_loss[loss=0.1614, simple_loss=0.2382, pruned_loss=0.0423, over 4908.00 frames.], tot_loss[loss=0.164, simple_loss=0.2416, pruned_loss=0.04321, over 985550.79 frames.], batch size: 47, aishell_tot_loss[loss=0.1643, simple_loss=0.247, pruned_loss=0.04076, over 984973.56 frames.], datatang_tot_loss[loss=0.1635, simple_loss=0.2353, pruned_loss=0.04586, over 986010.82 frames.], batch size: 47, lr: 6.11e-04
+2022-06-18 19:53:48,789 INFO [train.py:874] (2/4) Epoch 13, batch 4150, aishell_loss[loss=0.1645, simple_loss=0.2615, pruned_loss=0.03377, over 4910.00 frames.], tot_loss[loss=0.1645, simple_loss=0.2419, pruned_loss=0.04358, over 985310.33 frames.], batch size: 41, aishell_tot_loss[loss=0.1653, simple_loss=0.248, pruned_loss=0.04124, over 984643.79 frames.], datatang_tot_loss[loss=0.1631, simple_loss=0.2349, pruned_loss=0.04567, over 986078.43 frames.], batch size: 41, lr: 6.11e-04
+2022-06-18 19:55:12,920 INFO [train.py:874] (2/4) Epoch 14, batch 50, aishell_loss[loss=0.1522, simple_loss=0.2448, pruned_loss=0.0298, over 4925.00 frames.], tot_loss[loss=0.1591, simple_loss=0.2361, pruned_loss=0.04109, over 217997.21 frames.], batch size: 58, aishell_tot_loss[loss=0.1673, simple_loss=0.2509, pruned_loss=0.04183, over 93674.71 frames.], datatang_tot_loss[loss=0.1535, simple_loss=0.2258, pruned_loss=0.04055, over 137402.50 frames.], batch size: 58, lr: 5.91e-04
+2022-06-18 19:55:42,407 INFO [train.py:874] (2/4) Epoch 14, batch 100, datatang_loss[loss=0.1573, simple_loss=0.2222, pruned_loss=0.0462, over 4924.00 frames.], tot_loss[loss=0.1595, simple_loss=0.2363, pruned_loss=0.04136, over 388163.52 frames.], batch size: 81, aishell_tot_loss[loss=0.165, simple_loss=0.2482, pruned_loss=0.04087, over 194715.39 frames.], datatang_tot_loss[loss=0.1551, simple_loss=0.2267, pruned_loss=0.04177, over 241108.69 frames.], batch size: 81, lr: 5.91e-04
+2022-06-18 19:56:14,027 INFO [train.py:874] (2/4) Epoch 14, batch 150, datatang_loss[loss=0.1528, simple_loss=0.2347, pruned_loss=0.03543, over 4954.00 frames.], tot_loss[loss=0.1587, simple_loss=0.2358, pruned_loss=0.04076, over 520899.61 frames.], batch size: 91, aishell_tot_loss[loss=0.1648, simple_loss=0.2487, pruned_loss=0.04048, over 266656.10 frames.], datatang_tot_loss[loss=0.1543, simple_loss=0.2265, pruned_loss=0.04108, over 348474.41 frames.], batch size: 91, lr: 5.91e-04
+2022-06-18 19:56:43,538 INFO [train.py:874] (2/4) Epoch 14, batch 200, datatang_loss[loss=0.1481, simple_loss=0.2326, pruned_loss=0.03177, over 4905.00 frames.], tot_loss[loss=0.1583, simple_loss=0.2354, pruned_loss=0.04062, over 623537.50 frames.], batch size: 64, aishell_tot_loss[loss=0.1628, simple_loss=0.2463, pruned_loss=0.03965, over 354227.91 frames.], datatang_tot_loss[loss=0.1549, simple_loss=0.2266, pruned_loss=0.04161, over 420576.46 frames.], batch size: 64, lr: 5.90e-04
+2022-06-18 19:57:13,589 INFO [train.py:874] (2/4) Epoch 14, batch 250, aishell_loss[loss=0.1965, simple_loss=0.286, pruned_loss=0.05351, over 4910.00 frames.], tot_loss[loss=0.1583, simple_loss=0.2352, pruned_loss=0.04073, over 703631.66 frames.], batch size: 79, aishell_tot_loss[loss=0.1627, simple_loss=0.2453, pruned_loss=0.0401, over 439452.39 frames.], datatang_tot_loss[loss=0.1544, simple_loss=0.226, pruned_loss=0.04146, over 477041.01 frames.], batch size: 79, lr: 5.90e-04
+2022-06-18 19:57:44,737 INFO [train.py:874] (2/4) Epoch 14, batch 300, aishell_loss[loss=0.1667, simple_loss=0.2412, pruned_loss=0.04607, over 4929.00 frames.], tot_loss[loss=0.1582, simple_loss=0.2354, pruned_loss=0.04046, over 766074.11 frames.], batch size: 33, aishell_tot_loss[loss=0.1612, simple_loss=0.2435, pruned_loss=0.03945, over 513064.91 frames.], datatang_tot_loss[loss=0.1553, simple_loss=0.2273, pruned_loss=0.0417, over 528056.61 frames.], batch size: 33, lr: 5.90e-04
+2022-06-18 19:58:15,354 INFO [train.py:874] (2/4) Epoch 14, batch 350, datatang_loss[loss=0.1443, simple_loss=0.2204, pruned_loss=0.03407, over 4911.00 frames.], tot_loss[loss=0.1594, simple_loss=0.2369, pruned_loss=0.04096, over 814926.13 frames.], batch size: 75, aishell_tot_loss[loss=0.1627, simple_loss=0.2454, pruned_loss=0.03999, over 568729.45 frames.], datatang_tot_loss[loss=0.1558, simple_loss=0.2277, pruned_loss=0.04189, over 582129.18 frames.], batch size: 75, lr: 5.89e-04
+2022-06-18 19:58:44,788 INFO [train.py:874] (2/4) Epoch 14, batch 400, datatang_loss[loss=0.1839, simple_loss=0.2501, pruned_loss=0.05888, over 4914.00 frames.], tot_loss[loss=0.1602, simple_loss=0.2373, pruned_loss=0.04155, over 852691.11 frames.], batch size: 75, aishell_tot_loss[loss=0.1626, simple_loss=0.2457, pruned_loss=0.03972, over 604585.04 frames.], datatang_tot_loss[loss=0.1573, simple_loss=0.229, pruned_loss=0.04283, over 641956.61 frames.], batch size: 75, lr: 5.89e-04
+2022-06-18 19:59:15,070 INFO [train.py:874] (2/4) Epoch 14, batch 450, datatang_loss[loss=0.1494, simple_loss=0.2014, pruned_loss=0.04875, over 4939.00 frames.], tot_loss[loss=0.1598, simple_loss=0.2371, pruned_loss=0.04119, over 881825.03 frames.], batch size: 34, aishell_tot_loss[loss=0.1613, simple_loss=0.2449, pruned_loss=0.03886, over 648821.14 frames.], datatang_tot_loss[loss=0.1581, simple_loss=0.2296, pruned_loss=0.04324, over 682721.99 frames.], batch size: 34, lr: 5.89e-04
+2022-06-18 19:59:45,586 INFO [train.py:874] (2/4) Epoch 14, batch 500, datatang_loss[loss=0.1865, simple_loss=0.2423, pruned_loss=0.06533, over 4980.00 frames.], tot_loss[loss=0.1601, simple_loss=0.2377, pruned_loss=0.04123, over 904315.94 frames.], batch size: 40, aishell_tot_loss[loss=0.162, simple_loss=0.2455, pruned_loss=0.03926, over 690875.32 frames.], datatang_tot_loss[loss=0.1577, simple_loss=0.2295, pruned_loss=0.04297, over 715727.17 frames.], batch size: 40, lr: 5.89e-04
+2022-06-18 20:00:15,461 INFO [train.py:874] (2/4) Epoch 14, batch 550, datatang_loss[loss=0.1753, simple_loss=0.2417, pruned_loss=0.05442, over 4928.00 frames.], tot_loss[loss=0.161, simple_loss=0.2385, pruned_loss=0.04173, over 922259.52 frames.], batch size: 83, aishell_tot_loss[loss=0.1626, simple_loss=0.2462, pruned_loss=0.03946, over 724110.84 frames.], datatang_tot_loss[loss=0.1585, simple_loss=0.23, pruned_loss=0.04343, over 748792.05 frames.], batch size: 83, lr: 5.88e-04
+2022-06-18 20:00:45,856 INFO [train.py:874] (2/4) Epoch 14, batch 600, datatang_loss[loss=0.1626, simple_loss=0.2403, pruned_loss=0.04244, over 4918.00 frames.], tot_loss[loss=0.1613, simple_loss=0.239, pruned_loss=0.04181, over 936432.19 frames.], batch size: 64, aishell_tot_loss[loss=0.163, simple_loss=0.2464, pruned_loss=0.03981, over 757369.45 frames.], datatang_tot_loss[loss=0.1585, simple_loss=0.2304, pruned_loss=0.04332, over 774559.22 frames.], batch size: 64, lr: 5.88e-04
+2022-06-18 20:01:15,111 INFO [train.py:874] (2/4) Epoch 14, batch 650, aishell_loss[loss=0.1568, simple_loss=0.2423, pruned_loss=0.03565, over 4877.00 frames.], tot_loss[loss=0.1625, simple_loss=0.2402, pruned_loss=0.0424, over 947456.13 frames.], batch size: 35, aishell_tot_loss[loss=0.1639, simple_loss=0.2473, pruned_loss=0.04025, over 782154.94 frames.], datatang_tot_loss[loss=0.1594, simple_loss=0.2314, pruned_loss=0.04368, over 801481.55 frames.], batch size: 35, lr: 5.88e-04
+2022-06-18 20:01:44,219 INFO [train.py:874] (2/4) Epoch 14, batch 700, datatang_loss[loss=0.157, simple_loss=0.2295, pruned_loss=0.04223, over 4938.00 frames.], tot_loss[loss=0.1626, simple_loss=0.2405, pruned_loss=0.04231, over 955996.87 frames.], batch size: 79, aishell_tot_loss[loss=0.1632, simple_loss=0.2466, pruned_loss=0.03994, over 806267.73 frames.], datatang_tot_loss[loss=0.1604, simple_loss=0.2328, pruned_loss=0.04399, over 823108.44 frames.], batch size: 79, lr: 5.88e-04
+2022-06-18 20:02:13,951 INFO [train.py:874] (2/4) Epoch 14, batch 750, aishell_loss[loss=0.1755, simple_loss=0.2606, pruned_loss=0.04516, over 4961.00 frames.], tot_loss[loss=0.1628, simple_loss=0.2403, pruned_loss=0.04265, over 962423.97 frames.], batch size: 64, aishell_tot_loss[loss=0.1633, simple_loss=0.2465, pruned_loss=0.04001, over 825640.26 frames.], datatang_tot_loss[loss=0.1608, simple_loss=0.2329, pruned_loss=0.04438, over 843675.70 frames.], batch size: 64, lr: 5.87e-04
+2022-06-18 20:02:45,442 INFO [train.py:874] (2/4) Epoch 14, batch 800, datatang_loss[loss=0.1388, simple_loss=0.222, pruned_loss=0.02787, over 4975.00 frames.], tot_loss[loss=0.164, simple_loss=0.2415, pruned_loss=0.04323, over 967718.31 frames.], batch size: 60, aishell_tot_loss[loss=0.1643, simple_loss=0.2475, pruned_loss=0.04054, over 842264.75 frames.], datatang_tot_loss[loss=0.1615, simple_loss=0.2338, pruned_loss=0.0446, over 862496.90 frames.], batch size: 60, lr: 5.87e-04
+2022-06-18 20:03:15,188 INFO [train.py:874] (2/4) Epoch 14, batch 850, aishell_loss[loss=0.1524, simple_loss=0.2321, pruned_loss=0.03642, over 4938.00 frames.], tot_loss[loss=0.1644, simple_loss=0.2426, pruned_loss=0.04304, over 971783.36 frames.], batch size: 32, aishell_tot_loss[loss=0.1648, simple_loss=0.2484, pruned_loss=0.04061, over 862382.16 frames.], datatang_tot_loss[loss=0.1616, simple_loss=0.2341, pruned_loss=0.04457, over 874234.56 frames.], batch size: 32, lr: 5.87e-04
+2022-06-18 20:03:44,470 INFO [train.py:874] (2/4) Epoch 14, batch 900, aishell_loss[loss=0.1789, simple_loss=0.2673, pruned_loss=0.04526, over 4939.00 frames.], tot_loss[loss=0.1638, simple_loss=0.2419, pruned_loss=0.04282, over 974809.58 frames.], batch size: 49, aishell_tot_loss[loss=0.1645, simple_loss=0.2479, pruned_loss=0.04059, over 877460.86 frames.], datatang_tot_loss[loss=0.1615, simple_loss=0.234, pruned_loss=0.04445, over 886803.82 frames.], batch size: 49, lr: 5.86e-04
+2022-06-18 20:04:15,154 INFO [train.py:874] (2/4) Epoch 14, batch 950, datatang_loss[loss=0.1284, simple_loss=0.2073, pruned_loss=0.02474, over 4937.00 frames.], tot_loss[loss=0.1627, simple_loss=0.2404, pruned_loss=0.04254, over 977197.69 frames.], batch size: 71, aishell_tot_loss[loss=0.1635, simple_loss=0.2466, pruned_loss=0.04015, over 887800.76 frames.], datatang_tot_loss[loss=0.1616, simple_loss=0.2341, pruned_loss=0.04455, over 900557.61 frames.], batch size: 71, lr: 5.86e-04
+2022-06-18 20:04:45,283 INFO [train.py:874] (2/4) Epoch 14, batch 1000, datatang_loss[loss=0.1514, simple_loss=0.2325, pruned_loss=0.03521, over 4968.00 frames.], tot_loss[loss=0.1628, simple_loss=0.2405, pruned_loss=0.04252, over 979260.38 frames.], batch size: 60, aishell_tot_loss[loss=0.1638, simple_loss=0.247, pruned_loss=0.04031, over 898562.79 frames.], datatang_tot_loss[loss=0.1614, simple_loss=0.234, pruned_loss=0.04439, over 911424.12 frames.], batch size: 60, lr: 5.86e-04
+2022-06-18 20:04:45,284 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 20:05:01,198 INFO [train.py:914] (2/4) Epoch 14, validation: loss=0.1656, simple_loss=0.2496, pruned_loss=0.04079, over 1622729.00 frames.
+2022-06-18 20:05:31,525 INFO [train.py:874] (2/4) Epoch 14, batch 1050, aishell_loss[loss=0.1486, simple_loss=0.2402, pruned_loss=0.02847, over 4866.00 frames.], tot_loss[loss=0.1632, simple_loss=0.2412, pruned_loss=0.04265, over 980636.84 frames.], batch size: 37, aishell_tot_loss[loss=0.1637, simple_loss=0.2471, pruned_loss=0.04013, over 909087.50 frames.], datatang_tot_loss[loss=0.1621, simple_loss=0.2346, pruned_loss=0.04479, over 919926.53 frames.], batch size: 37, lr: 5.86e-04
+2022-06-18 20:06:01,967 INFO [train.py:874] (2/4) Epoch 14, batch 1100, aishell_loss[loss=0.1698, simple_loss=0.2657, pruned_loss=0.03697, over 4876.00 frames.], tot_loss[loss=0.1621, simple_loss=0.24, pruned_loss=0.04212, over 981533.35 frames.], batch size: 35, aishell_tot_loss[loss=0.1635, simple_loss=0.2469, pruned_loss=0.04001, over 916211.33 frames.], datatang_tot_loss[loss=0.1612, simple_loss=0.2339, pruned_loss=0.04426, over 929058.48 frames.], batch size: 35, lr: 5.85e-04
+2022-06-18 20:06:31,920 INFO [train.py:874] (2/4) Epoch 14, batch 1150, datatang_loss[loss=0.173, simple_loss=0.2421, pruned_loss=0.05198, over 4955.00 frames.], tot_loss[loss=0.1638, simple_loss=0.2417, pruned_loss=0.04294, over 982326.08 frames.], batch size: 91, aishell_tot_loss[loss=0.1635, simple_loss=0.2471, pruned_loss=0.04, over 924645.17 frames.], datatang_tot_loss[loss=0.1628, simple_loss=0.2353, pruned_loss=0.04519, over 935439.05 frames.], batch size: 91, lr: 5.85e-04
+2022-06-18 20:07:00,192 INFO [train.py:874] (2/4) Epoch 14, batch 1200, datatang_loss[loss=0.1888, simple_loss=0.2594, pruned_loss=0.05909, over 4918.00 frames.], tot_loss[loss=0.1643, simple_loss=0.2423, pruned_loss=0.04308, over 983192.21 frames.], batch size: 98, aishell_tot_loss[loss=0.1638, simple_loss=0.2473, pruned_loss=0.04017, over 933162.67 frames.], datatang_tot_loss[loss=0.1632, simple_loss=0.2357, pruned_loss=0.04536, over 940350.65 frames.], batch size: 98, lr: 5.85e-04
+2022-06-18 20:07:32,155 INFO [train.py:874] (2/4) Epoch 14, batch 1250, aishell_loss[loss=0.1666, simple_loss=0.2471, pruned_loss=0.04308, over 4978.00 frames.], tot_loss[loss=0.1636, simple_loss=0.2418, pruned_loss=0.04271, over 983350.95 frames.], batch size: 48, aishell_tot_loss[loss=0.1636, simple_loss=0.2472, pruned_loss=0.03999, over 938679.46 frames.], datatang_tot_loss[loss=0.1629, simple_loss=0.2353, pruned_loss=0.04518, over 945903.84 frames.], batch size: 48, lr: 5.85e-04
+2022-06-18 20:08:00,523 INFO [train.py:874] (2/4) Epoch 14, batch 1300, aishell_loss[loss=0.1571, simple_loss=0.238, pruned_loss=0.03807, over 4881.00 frames.], tot_loss[loss=0.1636, simple_loss=0.2422, pruned_loss=0.04252, over 984008.37 frames.], batch size: 35, aishell_tot_loss[loss=0.1635, simple_loss=0.2473, pruned_loss=0.03986, over 945615.26 frames.], datatang_tot_loss[loss=0.163, simple_loss=0.2354, pruned_loss=0.04532, over 949521.81 frames.], batch size: 35, lr: 5.84e-04
+2022-06-18 20:08:31,648 INFO [train.py:874] (2/4) Epoch 14, batch 1350, aishell_loss[loss=0.175, simple_loss=0.2727, pruned_loss=0.03864, over 4910.00 frames.], tot_loss[loss=0.1639, simple_loss=0.2421, pruned_loss=0.04286, over 983938.03 frames.], batch size: 68, aishell_tot_loss[loss=0.1628, simple_loss=0.2469, pruned_loss=0.03938, over 949177.48 frames.], datatang_tot_loss[loss=0.1641, simple_loss=0.2361, pruned_loss=0.04603, over 954388.75 frames.], batch size: 68, lr: 5.84e-04
+2022-06-18 20:09:02,305 INFO [train.py:874] (2/4) Epoch 14, batch 1400, aishell_loss[loss=0.1652, simple_loss=0.2489, pruned_loss=0.04074, over 4945.00 frames.], tot_loss[loss=0.1635, simple_loss=0.2419, pruned_loss=0.04252, over 984226.64 frames.], batch size: 32, aishell_tot_loss[loss=0.1623, simple_loss=0.2466, pruned_loss=0.03898, over 953876.70 frames.], datatang_tot_loss[loss=0.1643, simple_loss=0.2361, pruned_loss=0.04623, over 957662.44 frames.], batch size: 32, lr: 5.84e-04
+2022-06-18 20:09:32,256 INFO [train.py:874] (2/4) Epoch 14, batch 1450, aishell_loss[loss=0.1576, simple_loss=0.2492, pruned_loss=0.033, over 4889.00 frames.], tot_loss[loss=0.1631, simple_loss=0.2413, pruned_loss=0.04247, over 984049.98 frames.], batch size: 42, aishell_tot_loss[loss=0.1623, simple_loss=0.2465, pruned_loss=0.03905, over 957018.39 frames.], datatang_tot_loss[loss=0.1639, simple_loss=0.2357, pruned_loss=0.0461, over 961033.76 frames.], batch size: 42, lr: 5.84e-04
+2022-06-18 20:10:03,008 INFO [train.py:874] (2/4) Epoch 14, batch 1500, datatang_loss[loss=0.1633, simple_loss=0.2424, pruned_loss=0.04214, over 4957.00 frames.], tot_loss[loss=0.1628, simple_loss=0.2411, pruned_loss=0.04221, over 984099.53 frames.], batch size: 91, aishell_tot_loss[loss=0.1625, simple_loss=0.2468, pruned_loss=0.03907, over 959641.08 frames.], datatang_tot_loss[loss=0.1633, simple_loss=0.2353, pruned_loss=0.04568, over 964305.85 frames.], batch size: 91, lr: 5.83e-04
+2022-06-18 20:10:33,048 INFO [train.py:874] (2/4) Epoch 14, batch 1550, datatang_loss[loss=0.1469, simple_loss=0.2245, pruned_loss=0.03458, over 4923.00 frames.], tot_loss[loss=0.1626, simple_loss=0.2409, pruned_loss=0.04214, over 983967.66 frames.], batch size: 77, aishell_tot_loss[loss=0.1628, simple_loss=0.2467, pruned_loss=0.03941, over 962267.66 frames.], datatang_tot_loss[loss=0.1629, simple_loss=0.2353, pruned_loss=0.04525, over 966757.64 frames.], batch size: 77, lr: 5.83e-04
+2022-06-18 20:11:02,408 INFO [train.py:874] (2/4) Epoch 14, batch 1600, aishell_loss[loss=0.1559, simple_loss=0.2408, pruned_loss=0.03549, over 4957.00 frames.], tot_loss[loss=0.1625, simple_loss=0.2412, pruned_loss=0.04191, over 984049.41 frames.], batch size: 31, aishell_tot_loss[loss=0.1628, simple_loss=0.2465, pruned_loss=0.03953, over 965920.46 frames.], datatang_tot_loss[loss=0.1627, simple_loss=0.2351, pruned_loss=0.04517, over 967959.29 frames.], batch size: 31, lr: 5.83e-04
+2022-06-18 20:11:33,206 INFO [train.py:874] (2/4) Epoch 14, batch 1650, datatang_loss[loss=0.1554, simple_loss=0.2252, pruned_loss=0.04282, over 4940.00 frames.], tot_loss[loss=0.1623, simple_loss=0.2409, pruned_loss=0.04189, over 984603.34 frames.], batch size: 62, aishell_tot_loss[loss=0.1626, simple_loss=0.2464, pruned_loss=0.03942, over 968470.43 frames.], datatang_tot_loss[loss=0.1627, simple_loss=0.235, pruned_loss=0.04515, over 970024.31 frames.], batch size: 62, lr: 5.83e-04
+2022-06-18 20:12:02,943 INFO [train.py:874] (2/4) Epoch 14, batch 1700, aishell_loss[loss=0.1611, simple_loss=0.2472, pruned_loss=0.03754, over 4978.00 frames.], tot_loss[loss=0.1613, simple_loss=0.2398, pruned_loss=0.04142, over 984593.58 frames.], batch size: 51, aishell_tot_loss[loss=0.162, simple_loss=0.2457, pruned_loss=0.03911, over 970261.63 frames.], datatang_tot_loss[loss=0.1621, simple_loss=0.2344, pruned_loss=0.04489, over 971833.92 frames.], batch size: 51, lr: 5.82e-04
+2022-06-18 20:12:32,526 INFO [train.py:874] (2/4) Epoch 14, batch 1750, datatang_loss[loss=0.1393, simple_loss=0.2177, pruned_loss=0.0304, over 4922.00 frames.], tot_loss[loss=0.1626, simple_loss=0.2409, pruned_loss=0.04212, over 985112.28 frames.], batch size: 77, aishell_tot_loss[loss=0.1622, simple_loss=0.246, pruned_loss=0.03914, over 972243.24 frames.], datatang_tot_loss[loss=0.1631, simple_loss=0.2353, pruned_loss=0.04547, over 973591.39 frames.], batch size: 77, lr: 5.82e-04
+2022-06-18 20:13:04,405 INFO [train.py:874] (2/4) Epoch 14, batch 1800, datatang_loss[loss=0.1436, simple_loss=0.2261, pruned_loss=0.03062, over 4957.00 frames.], tot_loss[loss=0.1623, simple_loss=0.2406, pruned_loss=0.04205, over 985183.21 frames.], batch size: 67, aishell_tot_loss[loss=0.162, simple_loss=0.2461, pruned_loss=0.03897, over 973546.51 frames.], datatang_tot_loss[loss=0.163, simple_loss=0.235, pruned_loss=0.04545, over 975225.85 frames.], batch size: 67, lr: 5.82e-04
+2022-06-18 20:13:34,476 INFO [train.py:874] (2/4) Epoch 14, batch 1850, datatang_loss[loss=0.1786, simple_loss=0.2543, pruned_loss=0.0514, over 4916.00 frames.], tot_loss[loss=0.1624, simple_loss=0.2403, pruned_loss=0.04222, over 985656.49 frames.], batch size: 98, aishell_tot_loss[loss=0.1619, simple_loss=0.2458, pruned_loss=0.039, over 974987.30 frames.], datatang_tot_loss[loss=0.1631, simple_loss=0.2354, pruned_loss=0.0454, over 976793.32 frames.], batch size: 98, lr: 5.81e-04
+2022-06-18 20:14:03,928 INFO [train.py:874] (2/4) Epoch 14, batch 1900, aishell_loss[loss=0.1782, simple_loss=0.2606, pruned_loss=0.04787, over 4970.00 frames.], tot_loss[loss=0.163, simple_loss=0.2408, pruned_loss=0.0426, over 985743.75 frames.], batch size: 64, aishell_tot_loss[loss=0.1623, simple_loss=0.2461, pruned_loss=0.03925, over 976125.49 frames.], datatang_tot_loss[loss=0.1634, simple_loss=0.2358, pruned_loss=0.04549, over 978028.26 frames.], batch size: 64, lr: 5.81e-04
+2022-06-18 20:14:34,998 INFO [train.py:874] (2/4) Epoch 14, batch 1950, datatang_loss[loss=0.1605, simple_loss=0.228, pruned_loss=0.04648, over 4914.00 frames.], tot_loss[loss=0.1632, simple_loss=0.2412, pruned_loss=0.04267, over 985742.13 frames.], batch size: 64, aishell_tot_loss[loss=0.1628, simple_loss=0.2468, pruned_loss=0.03945, over 977392.94 frames.], datatang_tot_loss[loss=0.1631, simple_loss=0.2354, pruned_loss=0.04544, over 978815.28 frames.], batch size: 64, lr: 5.81e-04
+2022-06-18 20:15:05,413 INFO [train.py:874] (2/4) Epoch 14, batch 2000, aishell_loss[loss=0.173, simple_loss=0.2645, pruned_loss=0.04073, over 4957.00 frames.], tot_loss[loss=0.1631, simple_loss=0.2416, pruned_loss=0.04226, over 986224.03 frames.], batch size: 40, aishell_tot_loss[loss=0.163, simple_loss=0.2472, pruned_loss=0.03945, over 979008.42 frames.], datatang_tot_loss[loss=0.1628, simple_loss=0.2351, pruned_loss=0.04526, over 979542.99 frames.], batch size: 40, lr: 5.81e-04
+2022-06-18 20:15:05,414 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 20:15:22,469 INFO [train.py:914] (2/4) Epoch 14, validation: loss=0.1654, simple_loss=0.2497, pruned_loss=0.04057, over 1622729.00 frames.
+2022-06-18 20:15:52,261 INFO [train.py:874] (2/4) Epoch 14, batch 2050, datatang_loss[loss=0.1782, simple_loss=0.2482, pruned_loss=0.05413, over 4920.00 frames.], tot_loss[loss=0.1631, simple_loss=0.2418, pruned_loss=0.04222, over 985895.61 frames.], batch size: 73, aishell_tot_loss[loss=0.1633, simple_loss=0.2475, pruned_loss=0.03956, over 979558.70 frames.], datatang_tot_loss[loss=0.1626, simple_loss=0.235, pruned_loss=0.04508, over 980275.37 frames.], batch size: 73, lr: 5.80e-04
+2022-06-18 20:16:21,741 INFO [train.py:874] (2/4) Epoch 14, batch 2100, datatang_loss[loss=0.1438, simple_loss=0.2208, pruned_loss=0.03336, over 4922.00 frames.], tot_loss[loss=0.163, simple_loss=0.2417, pruned_loss=0.04217, over 985575.36 frames.], batch size: 81, aishell_tot_loss[loss=0.1637, simple_loss=0.2479, pruned_loss=0.03972, over 980035.12 frames.], datatang_tot_loss[loss=0.1622, simple_loss=0.2349, pruned_loss=0.04478, over 980862.57 frames.], batch size: 81, lr: 5.80e-04
+2022-06-18 20:16:52,433 INFO [train.py:874] (2/4) Epoch 14, batch 2150, datatang_loss[loss=0.1441, simple_loss=0.2123, pruned_loss=0.03798, over 4935.00 frames.], tot_loss[loss=0.1629, simple_loss=0.2419, pruned_loss=0.04194, over 985714.61 frames.], batch size: 79, aishell_tot_loss[loss=0.1632, simple_loss=0.2473, pruned_loss=0.03958, over 980864.14 frames.], datatang_tot_loss[loss=0.1625, simple_loss=0.2353, pruned_loss=0.04485, over 981409.83 frames.], batch size: 79, lr: 5.80e-04
+2022-06-18 20:17:23,084 INFO [train.py:874] (2/4) Epoch 14, batch 2200, datatang_loss[loss=0.1481, simple_loss=0.217, pruned_loss=0.03957, over 4977.00 frames.], tot_loss[loss=0.162, simple_loss=0.2404, pruned_loss=0.04175, over 985972.24 frames.], batch size: 60, aishell_tot_loss[loss=0.1624, simple_loss=0.2462, pruned_loss=0.03926, over 981631.74 frames.], datatang_tot_loss[loss=0.1624, simple_loss=0.235, pruned_loss=0.04485, over 981969.71 frames.], batch size: 60, lr: 5.80e-04
+2022-06-18 20:17:52,999 INFO [train.py:874] (2/4) Epoch 14, batch 2250, aishell_loss[loss=0.1596, simple_loss=0.2416, pruned_loss=0.03875, over 4920.00 frames.], tot_loss[loss=0.162, simple_loss=0.2404, pruned_loss=0.04175, over 986079.83 frames.], batch size: 46, aishell_tot_loss[loss=0.1621, simple_loss=0.2459, pruned_loss=0.0391, over 982493.55 frames.], datatang_tot_loss[loss=0.1626, simple_loss=0.2351, pruned_loss=0.04507, over 982200.38 frames.], batch size: 46, lr: 5.79e-04
+2022-06-18 20:18:23,985 INFO [train.py:874] (2/4) Epoch 14, batch 2300, datatang_loss[loss=0.1458, simple_loss=0.2163, pruned_loss=0.03764, over 4926.00 frames.], tot_loss[loss=0.162, simple_loss=0.2399, pruned_loss=0.04204, over 986224.21 frames.], batch size: 73, aishell_tot_loss[loss=0.162, simple_loss=0.246, pruned_loss=0.03901, over 982890.62 frames.], datatang_tot_loss[loss=0.1626, simple_loss=0.2348, pruned_loss=0.04519, over 982834.34 frames.], batch size: 73, lr: 5.79e-04
+2022-06-18 20:18:58,987 INFO [train.py:874] (2/4) Epoch 14, batch 2350, datatang_loss[loss=0.1467, simple_loss=0.2151, pruned_loss=0.03917, over 4886.00 frames.], tot_loss[loss=0.162, simple_loss=0.24, pruned_loss=0.04197, over 985883.88 frames.], batch size: 52, aishell_tot_loss[loss=0.1619, simple_loss=0.2456, pruned_loss=0.03915, over 983296.37 frames.], datatang_tot_loss[loss=0.1625, simple_loss=0.2349, pruned_loss=0.0451, over 982847.74 frames.], batch size: 52, lr: 5.79e-04
+2022-06-18 20:19:28,609 INFO [train.py:874] (2/4) Epoch 14, batch 2400, aishell_loss[loss=0.1448, simple_loss=0.2303, pruned_loss=0.02967, over 4873.00 frames.], tot_loss[loss=0.1616, simple_loss=0.2393, pruned_loss=0.04193, over 985457.24 frames.], batch size: 35, aishell_tot_loss[loss=0.1615, simple_loss=0.2449, pruned_loss=0.03911, over 983208.91 frames.], datatang_tot_loss[loss=0.1625, simple_loss=0.235, pruned_loss=0.04496, over 983141.71 frames.], batch size: 35, lr: 5.79e-04
+2022-06-18 20:19:59,655 INFO [train.py:874] (2/4) Epoch 14, batch 2450, datatang_loss[loss=0.2349, simple_loss=0.2863, pruned_loss=0.0918, over 4959.00 frames.], tot_loss[loss=0.1615, simple_loss=0.239, pruned_loss=0.04198, over 985909.25 frames.], batch size: 109, aishell_tot_loss[loss=0.1614, simple_loss=0.2447, pruned_loss=0.03909, over 983713.72 frames.], datatang_tot_loss[loss=0.1623, simple_loss=0.235, pruned_loss=0.04486, over 983656.86 frames.], batch size: 109, lr: 5.78e-04
+2022-06-18 20:20:30,392 INFO [train.py:874] (2/4) Epoch 14, batch 2500, datatang_loss[loss=0.1565, simple_loss=0.2366, pruned_loss=0.03822, over 4927.00 frames.], tot_loss[loss=0.1614, simple_loss=0.2391, pruned_loss=0.04181, over 985612.87 frames.], batch size: 73, aishell_tot_loss[loss=0.1614, simple_loss=0.2444, pruned_loss=0.03921, over 983864.16 frames.], datatang_tot_loss[loss=0.1622, simple_loss=0.235, pruned_loss=0.04465, over 983728.25 frames.], batch size: 73, lr: 5.78e-04
+2022-06-18 20:20:59,931 INFO [train.py:874] (2/4) Epoch 14, batch 2550, datatang_loss[loss=0.1964, simple_loss=0.2631, pruned_loss=0.06485, over 4900.00 frames.], tot_loss[loss=0.1609, simple_loss=0.2389, pruned_loss=0.04148, over 985357.09 frames.], batch size: 47, aishell_tot_loss[loss=0.1608, simple_loss=0.2438, pruned_loss=0.03887, over 983820.97 frames.], datatang_tot_loss[loss=0.1622, simple_loss=0.2352, pruned_loss=0.04465, over 983931.03 frames.], batch size: 47, lr: 5.78e-04
+2022-06-18 20:21:29,263 INFO [train.py:874] (2/4) Epoch 14, batch 2600, datatang_loss[loss=0.14, simple_loss=0.221, pruned_loss=0.02948, over 4931.00 frames.], tot_loss[loss=0.1614, simple_loss=0.2392, pruned_loss=0.0418, over 985335.33 frames.], batch size: 73, aishell_tot_loss[loss=0.1608, simple_loss=0.2438, pruned_loss=0.03886, over 983905.00 frames.], datatang_tot_loss[loss=0.1626, simple_loss=0.2354, pruned_loss=0.0449, over 984174.02 frames.], batch size: 73, lr: 5.78e-04
+2022-06-18 20:22:01,015 INFO [train.py:874] (2/4) Epoch 14, batch 2650, datatang_loss[loss=0.1592, simple_loss=0.2292, pruned_loss=0.04463, over 4956.00 frames.], tot_loss[loss=0.1612, simple_loss=0.239, pruned_loss=0.04167, over 985786.26 frames.], batch size: 60, aishell_tot_loss[loss=0.1607, simple_loss=0.2439, pruned_loss=0.03878, over 984265.39 frames.], datatang_tot_loss[loss=0.1623, simple_loss=0.2352, pruned_loss=0.04477, over 984585.61 frames.], batch size: 60, lr: 5.77e-04
+2022-06-18 20:22:30,526 INFO [train.py:874] (2/4) Epoch 14, batch 2700, datatang_loss[loss=0.173, simple_loss=0.2451, pruned_loss=0.05042, over 4957.00 frames.], tot_loss[loss=0.1626, simple_loss=0.2403, pruned_loss=0.04243, over 986117.52 frames.], batch size: 67, aishell_tot_loss[loss=0.1612, simple_loss=0.2444, pruned_loss=0.03897, over 984544.64 frames.], datatang_tot_loss[loss=0.1633, simple_loss=0.2361, pruned_loss=0.04526, over 984981.11 frames.], batch size: 67, lr: 5.77e-04
+2022-06-18 20:22:59,814 INFO [train.py:874] (2/4) Epoch 14, batch 2750, datatang_loss[loss=0.1711, simple_loss=0.244, pruned_loss=0.0491, over 4936.00 frames.], tot_loss[loss=0.1626, simple_loss=0.2402, pruned_loss=0.04249, over 986075.47 frames.], batch size: 88, aishell_tot_loss[loss=0.1617, simple_loss=0.2445, pruned_loss=0.03941, over 984798.81 frames.], datatang_tot_loss[loss=0.1629, simple_loss=0.2357, pruned_loss=0.04503, over 985005.60 frames.], batch size: 88, lr: 5.77e-04
+2022-06-18 20:23:30,611 INFO [train.py:874] (2/4) Epoch 14, batch 2800, aishell_loss[loss=0.1753, simple_loss=0.2593, pruned_loss=0.04568, over 4878.00 frames.], tot_loss[loss=0.1619, simple_loss=0.2401, pruned_loss=0.04184, over 985532.48 frames.], batch size: 42, aishell_tot_loss[loss=0.1618, simple_loss=0.2449, pruned_loss=0.03937, over 984582.13 frames.], datatang_tot_loss[loss=0.1621, simple_loss=0.2351, pruned_loss=0.04454, over 984937.77 frames.], batch size: 42, lr: 5.77e-04
+2022-06-18 20:24:01,487 INFO [train.py:874] (2/4) Epoch 14, batch 2850, aishell_loss[loss=0.1698, simple_loss=0.2581, pruned_loss=0.04074, over 4946.00 frames.], tot_loss[loss=0.1616, simple_loss=0.24, pruned_loss=0.04158, over 985690.44 frames.], batch size: 49, aishell_tot_loss[loss=0.1616, simple_loss=0.2451, pruned_loss=0.03908, over 984808.11 frames.], datatang_tot_loss[loss=0.162, simple_loss=0.2348, pruned_loss=0.04454, over 985066.14 frames.], batch size: 49, lr: 5.76e-04
+2022-06-18 20:24:30,564 INFO [train.py:874] (2/4) Epoch 14, batch 2900, aishell_loss[loss=0.1483, simple_loss=0.233, pruned_loss=0.03175, over 4940.00 frames.], tot_loss[loss=0.1617, simple_loss=0.2397, pruned_loss=0.04191, over 985866.83 frames.], batch size: 32, aishell_tot_loss[loss=0.1616, simple_loss=0.245, pruned_loss=0.03907, over 984749.78 frames.], datatang_tot_loss[loss=0.1621, simple_loss=0.2347, pruned_loss=0.04477, over 985469.33 frames.], batch size: 32, lr: 5.76e-04
+2022-06-18 20:25:01,167 INFO [train.py:874] (2/4) Epoch 14, batch 2950, aishell_loss[loss=0.1785, simple_loss=0.2564, pruned_loss=0.05027, over 4979.00 frames.], tot_loss[loss=0.1615, simple_loss=0.2397, pruned_loss=0.04168, over 985813.45 frames.], batch size: 48, aishell_tot_loss[loss=0.161, simple_loss=0.2447, pruned_loss=0.03867, over 984905.80 frames.], datatang_tot_loss[loss=0.1624, simple_loss=0.2351, pruned_loss=0.04487, over 985441.28 frames.], batch size: 48, lr: 5.76e-04
+2022-06-18 20:25:31,691 INFO [train.py:874] (2/4) Epoch 14, batch 3000, datatang_loss[loss=0.1575, simple_loss=0.2334, pruned_loss=0.04081, over 4925.00 frames.], tot_loss[loss=0.1611, simple_loss=0.2392, pruned_loss=0.04151, over 985676.11 frames.], batch size: 88, aishell_tot_loss[loss=0.1609, simple_loss=0.2445, pruned_loss=0.03861, over 984842.21 frames.], datatang_tot_loss[loss=0.162, simple_loss=0.2348, pruned_loss=0.04462, over 985495.97 frames.], batch size: 88, lr: 5.76e-04
+2022-06-18 20:25:31,692 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 20:25:47,764 INFO [train.py:914] (2/4) Epoch 14, validation: loss=0.1656, simple_loss=0.2493, pruned_loss=0.0409, over 1622729.00 frames.
+2022-06-18 20:26:17,835 INFO [train.py:874] (2/4) Epoch 14, batch 3050, aishell_loss[loss=0.1519, simple_loss=0.2336, pruned_loss=0.03508, over 4884.00 frames.], tot_loss[loss=0.1616, simple_loss=0.2397, pruned_loss=0.04171, over 985627.43 frames.], batch size: 42, aishell_tot_loss[loss=0.1614, simple_loss=0.2451, pruned_loss=0.03882, over 984871.40 frames.], datatang_tot_loss[loss=0.1619, simple_loss=0.2346, pruned_loss=0.04461, over 985537.18 frames.], batch size: 42, lr: 5.75e-04
+2022-06-18 20:26:48,725 INFO [train.py:874] (2/4) Epoch 14, batch 3100, datatang_loss[loss=0.1571, simple_loss=0.2314, pruned_loss=0.0414, over 4974.00 frames.], tot_loss[loss=0.1632, simple_loss=0.2409, pruned_loss=0.04274, over 986074.20 frames.], batch size: 45, aishell_tot_loss[loss=0.1614, simple_loss=0.2454, pruned_loss=0.03875, over 985142.96 frames.], datatang_tot_loss[loss=0.1635, simple_loss=0.2359, pruned_loss=0.04554, over 985821.91 frames.], batch size: 45, lr: 5.75e-04
+2022-06-18 20:27:19,599 INFO [train.py:874] (2/4) Epoch 14, batch 3150, aishell_loss[loss=0.1781, simple_loss=0.2552, pruned_loss=0.05045, over 4930.00 frames.], tot_loss[loss=0.1631, simple_loss=0.2408, pruned_loss=0.04271, over 985955.36 frames.], batch size: 49, aishell_tot_loss[loss=0.1617, simple_loss=0.2457, pruned_loss=0.03884, over 985187.62 frames.], datatang_tot_loss[loss=0.1633, simple_loss=0.2356, pruned_loss=0.04548, over 985773.47 frames.], batch size: 49, lr: 5.75e-04
+2022-06-18 20:27:49,660 INFO [train.py:874] (2/4) Epoch 14, batch 3200, aishell_loss[loss=0.1474, simple_loss=0.221, pruned_loss=0.0369, over 4967.00 frames.], tot_loss[loss=0.1625, simple_loss=0.2404, pruned_loss=0.04236, over 985813.68 frames.], batch size: 25, aishell_tot_loss[loss=0.1615, simple_loss=0.2455, pruned_loss=0.03873, over 985078.09 frames.], datatang_tot_loss[loss=0.1631, simple_loss=0.2354, pruned_loss=0.04536, over 985824.60 frames.], batch size: 25, lr: 5.75e-04
+2022-06-18 20:28:20,113 INFO [train.py:874] (2/4) Epoch 14, batch 3250, datatang_loss[loss=0.1331, simple_loss=0.199, pruned_loss=0.03366, over 4967.00 frames.], tot_loss[loss=0.1627, simple_loss=0.2407, pruned_loss=0.04233, over 985919.66 frames.], batch size: 37, aishell_tot_loss[loss=0.1622, simple_loss=0.2464, pruned_loss=0.03902, over 985347.96 frames.], datatang_tot_loss[loss=0.1626, simple_loss=0.2346, pruned_loss=0.04525, over 985753.44 frames.], batch size: 37, lr: 5.74e-04
+2022-06-18 20:28:49,465 INFO [train.py:874] (2/4) Epoch 14, batch 3300, aishell_loss[loss=0.1238, simple_loss=0.2129, pruned_loss=0.0173, over 4852.00 frames.], tot_loss[loss=0.1628, simple_loss=0.2409, pruned_loss=0.04236, over 985702.51 frames.], batch size: 28, aishell_tot_loss[loss=0.1616, simple_loss=0.2456, pruned_loss=0.0388, over 985275.33 frames.], datatang_tot_loss[loss=0.1634, simple_loss=0.2354, pruned_loss=0.04569, over 985686.25 frames.], batch size: 28, lr: 5.74e-04
+2022-06-18 20:29:19,700 INFO [train.py:874] (2/4) Epoch 14, batch 3350, datatang_loss[loss=0.2067, simple_loss=0.2629, pruned_loss=0.07525, over 4928.00 frames.], tot_loss[loss=0.1618, simple_loss=0.24, pruned_loss=0.04181, over 985168.71 frames.], batch size: 57, aishell_tot_loss[loss=0.1614, simple_loss=0.2454, pruned_loss=0.03874, over 984912.90 frames.], datatang_tot_loss[loss=0.1626, simple_loss=0.2347, pruned_loss=0.04523, over 985526.18 frames.], batch size: 57, lr: 5.74e-04
+2022-06-18 20:29:49,996 INFO [train.py:874] (2/4) Epoch 14, batch 3400, aishell_loss[loss=0.1532, simple_loss=0.2447, pruned_loss=0.03089, over 4858.00 frames.], tot_loss[loss=0.1604, simple_loss=0.2389, pruned_loss=0.04098, over 985073.76 frames.], batch size: 37, aishell_tot_loss[loss=0.1608, simple_loss=0.2448, pruned_loss=0.03837, over 984851.10 frames.], datatang_tot_loss[loss=0.1617, simple_loss=0.2338, pruned_loss=0.0448, over 985487.10 frames.], batch size: 37, lr: 5.74e-04
+2022-06-18 20:30:18,778 INFO [train.py:874] (2/4) Epoch 14, batch 3450, aishell_loss[loss=0.1896, simple_loss=0.2625, pruned_loss=0.05832, over 4937.00 frames.], tot_loss[loss=0.1602, simple_loss=0.2389, pruned_loss=0.04077, over 985162.15 frames.], batch size: 56, aishell_tot_loss[loss=0.1608, simple_loss=0.2449, pruned_loss=0.03836, over 984829.42 frames.], datatang_tot_loss[loss=0.1613, simple_loss=0.2335, pruned_loss=0.04453, over 985585.71 frames.], batch size: 56, lr: 5.73e-04
+2022-06-18 20:30:50,342 INFO [train.py:874] (2/4) Epoch 14, batch 3500, aishell_loss[loss=0.1474, simple_loss=0.2358, pruned_loss=0.0295, over 4949.00 frames.], tot_loss[loss=0.161, simple_loss=0.2394, pruned_loss=0.04132, over 985305.73 frames.], batch size: 56, aishell_tot_loss[loss=0.1616, simple_loss=0.2457, pruned_loss=0.03877, over 984708.87 frames.], datatang_tot_loss[loss=0.1612, simple_loss=0.2333, pruned_loss=0.04451, over 985846.28 frames.], batch size: 56, lr: 5.73e-04
+2022-06-18 20:31:19,803 INFO [train.py:874] (2/4) Epoch 14, batch 3550, datatang_loss[loss=0.2025, simple_loss=0.2687, pruned_loss=0.06813, over 4945.00 frames.], tot_loss[loss=0.1609, simple_loss=0.239, pruned_loss=0.04142, over 984940.93 frames.], batch size: 45, aishell_tot_loss[loss=0.1615, simple_loss=0.2455, pruned_loss=0.03877, over 984563.92 frames.], datatang_tot_loss[loss=0.1611, simple_loss=0.2332, pruned_loss=0.0445, over 985591.68 frames.], batch size: 45, lr: 5.73e-04
+2022-06-18 20:31:49,278 INFO [train.py:874] (2/4) Epoch 14, batch 3600, aishell_loss[loss=0.1629, simple_loss=0.2578, pruned_loss=0.03396, over 4874.00 frames.], tot_loss[loss=0.1603, simple_loss=0.2389, pruned_loss=0.04086, over 985139.41 frames.], batch size: 47, aishell_tot_loss[loss=0.1607, simple_loss=0.2449, pruned_loss=0.0383, over 984620.63 frames.], datatang_tot_loss[loss=0.1611, simple_loss=0.2333, pruned_loss=0.04446, over 985748.04 frames.], batch size: 47, lr: 5.73e-04
+2022-06-18 20:32:20,183 INFO [train.py:874] (2/4) Epoch 14, batch 3650, datatang_loss[loss=0.2195, simple_loss=0.2762, pruned_loss=0.08145, over 4939.00 frames.], tot_loss[loss=0.1602, simple_loss=0.2384, pruned_loss=0.04094, over 985201.21 frames.], batch size: 108, aishell_tot_loss[loss=0.1608, simple_loss=0.2449, pruned_loss=0.03833, over 984712.63 frames.], datatang_tot_loss[loss=0.1608, simple_loss=0.233, pruned_loss=0.04432, over 985712.19 frames.], batch size: 108, lr: 5.72e-04
+2022-06-18 20:32:50,537 INFO [train.py:874] (2/4) Epoch 14, batch 3700, aishell_loss[loss=0.1655, simple_loss=0.2625, pruned_loss=0.0343, over 4960.00 frames.], tot_loss[loss=0.1613, simple_loss=0.2393, pruned_loss=0.04159, over 985159.85 frames.], batch size: 44, aishell_tot_loss[loss=0.1614, simple_loss=0.2455, pruned_loss=0.03866, over 984613.46 frames.], datatang_tot_loss[loss=0.1612, simple_loss=0.2333, pruned_loss=0.04453, over 985748.85 frames.], batch size: 44, lr: 5.72e-04
+2022-06-18 20:33:20,206 INFO [train.py:874] (2/4) Epoch 14, batch 3750, datatang_loss[loss=0.1763, simple_loss=0.238, pruned_loss=0.05733, over 4987.00 frames.], tot_loss[loss=0.1608, simple_loss=0.2393, pruned_loss=0.04112, over 984948.82 frames.], batch size: 40, aishell_tot_loss[loss=0.1612, simple_loss=0.2453, pruned_loss=0.0385, over 984431.91 frames.], datatang_tot_loss[loss=0.1609, simple_loss=0.2332, pruned_loss=0.04427, over 985701.68 frames.], batch size: 40, lr: 5.72e-04
+2022-06-18 20:33:50,717 INFO [train.py:874] (2/4) Epoch 14, batch 3800, datatang_loss[loss=0.153, simple_loss=0.233, pruned_loss=0.03649, over 4955.00 frames.], tot_loss[loss=0.1617, simple_loss=0.2404, pruned_loss=0.04154, over 985308.11 frames.], batch size: 91, aishell_tot_loss[loss=0.1617, simple_loss=0.2461, pruned_loss=0.03862, over 984659.67 frames.], datatang_tot_loss[loss=0.1613, simple_loss=0.2338, pruned_loss=0.04443, over 985801.02 frames.], batch size: 91, lr: 5.72e-04
+2022-06-18 20:34:19,271 INFO [train.py:874] (2/4) Epoch 14, batch 3850, aishell_loss[loss=0.1372, simple_loss=0.2271, pruned_loss=0.0236, over 4981.00 frames.], tot_loss[loss=0.1605, simple_loss=0.2391, pruned_loss=0.041, over 985462.63 frames.], batch size: 30, aishell_tot_loss[loss=0.1608, simple_loss=0.2451, pruned_loss=0.03822, over 984700.76 frames.], datatang_tot_loss[loss=0.161, simple_loss=0.2335, pruned_loss=0.04425, over 985938.98 frames.], batch size: 30, lr: 5.71e-04
+2022-06-18 20:34:49,124 INFO [train.py:874] (2/4) Epoch 14, batch 3900, aishell_loss[loss=0.1736, simple_loss=0.2602, pruned_loss=0.04354, over 4946.00 frames.], tot_loss[loss=0.1604, simple_loss=0.2391, pruned_loss=0.04086, over 985377.51 frames.], batch size: 45, aishell_tot_loss[loss=0.1613, simple_loss=0.2455, pruned_loss=0.0386, over 984547.99 frames.], datatang_tot_loss[loss=0.1602, simple_loss=0.233, pruned_loss=0.04368, over 986049.86 frames.], batch size: 45, lr: 5.71e-04
+2022-06-18 20:35:17,121 INFO [train.py:874] (2/4) Epoch 14, batch 3950, datatang_loss[loss=0.1802, simple_loss=0.2543, pruned_loss=0.05306, over 4908.00 frames.], tot_loss[loss=0.16, simple_loss=0.2391, pruned_loss=0.04042, over 984621.63 frames.], batch size: 98, aishell_tot_loss[loss=0.1607, simple_loss=0.2451, pruned_loss=0.03813, over 984348.12 frames.], datatang_tot_loss[loss=0.1603, simple_loss=0.2332, pruned_loss=0.04373, over 985498.02 frames.], batch size: 98, lr: 5.71e-04
+2022-06-18 20:35:46,888 INFO [train.py:874] (2/4) Epoch 14, batch 4000, datatang_loss[loss=0.1622, simple_loss=0.242, pruned_loss=0.04123, over 4885.00 frames.], tot_loss[loss=0.1603, simple_loss=0.2391, pruned_loss=0.04076, over 985034.87 frames.], batch size: 47, aishell_tot_loss[loss=0.1614, simple_loss=0.2458, pruned_loss=0.03853, over 984580.53 frames.], datatang_tot_loss[loss=0.1598, simple_loss=0.2326, pruned_loss=0.0435, over 985616.71 frames.], batch size: 47, lr: 5.71e-04
+2022-06-18 20:35:46,889 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 20:36:02,937 INFO [train.py:914] (2/4) Epoch 14, validation: loss=0.1656, simple_loss=0.2513, pruned_loss=0.03999, over 1622729.00 frames.
+2022-06-18 20:36:31,812 INFO [train.py:874] (2/4) Epoch 14, batch 4050, datatang_loss[loss=0.1504, simple_loss=0.2161, pruned_loss=0.04231, over 4887.00 frames.], tot_loss[loss=0.1602, simple_loss=0.2391, pruned_loss=0.04065, over 984922.45 frames.], batch size: 39, aishell_tot_loss[loss=0.1618, simple_loss=0.2462, pruned_loss=0.03875, over 984385.98 frames.], datatang_tot_loss[loss=0.1593, simple_loss=0.2321, pruned_loss=0.0432, over 985706.68 frames.], batch size: 39, lr: 5.70e-04
+2022-06-18 20:36:59,613 INFO [train.py:874] (2/4) Epoch 14, batch 4100, datatang_loss[loss=0.1432, simple_loss=0.2022, pruned_loss=0.04206, over 4940.00 frames.], tot_loss[loss=0.1608, simple_loss=0.2395, pruned_loss=0.0411, over 984391.27 frames.], batch size: 34, aishell_tot_loss[loss=0.162, simple_loss=0.2465, pruned_loss=0.03871, over 983766.59 frames.], datatang_tot_loss[loss=0.1598, simple_loss=0.2324, pruned_loss=0.0436, over 985709.14 frames.], batch size: 34, lr: 5.70e-04
+2022-06-18 20:38:15,679 INFO [train.py:874] (2/4) Epoch 15, batch 50, datatang_loss[loss=0.1656, simple_loss=0.2335, pruned_loss=0.04879, over 4924.00 frames.], tot_loss[loss=0.1554, simple_loss=0.2327, pruned_loss=0.03905, over 218724.99 frames.], batch size: 81, aishell_tot_loss[loss=0.1582, simple_loss=0.2405, pruned_loss=0.03793, over 116148.49 frames.], datatang_tot_loss[loss=0.1523, simple_loss=0.2246, pruned_loss=0.04, over 116247.17 frames.], batch size: 81, lr: 5.52e-04
+2022-06-18 20:38:46,641 INFO [train.py:874] (2/4) Epoch 15, batch 100, aishell_loss[loss=0.1875, simple_loss=0.2691, pruned_loss=0.05298, over 4890.00 frames.], tot_loss[loss=0.1579, simple_loss=0.2372, pruned_loss=0.03928, over 388691.92 frames.], batch size: 34, aishell_tot_loss[loss=0.1636, simple_loss=0.2485, pruned_loss=0.0393, over 226226.72 frames.], datatang_tot_loss[loss=0.1511, simple_loss=0.2241, pruned_loss=0.03909, over 210860.86 frames.], batch size: 34, lr: 5.52e-04
+2022-06-18 20:39:16,994 INFO [train.py:874] (2/4) Epoch 15, batch 150, aishell_loss[loss=0.1261, simple_loss=0.206, pruned_loss=0.02311, over 4978.00 frames.], tot_loss[loss=0.1552, simple_loss=0.2344, pruned_loss=0.03803, over 521058.15 frames.], batch size: 25, aishell_tot_loss[loss=0.1607, simple_loss=0.2454, pruned_loss=0.03802, over 322467.14 frames.], datatang_tot_loss[loss=0.1494, simple_loss=0.2222, pruned_loss=0.03825, over 295134.33 frames.], batch size: 25, lr: 5.52e-04
+2022-06-18 20:39:45,837 INFO [train.py:874] (2/4) Epoch 15, batch 200, aishell_loss[loss=0.1557, simple_loss=0.2481, pruned_loss=0.03164, over 4905.00 frames.], tot_loss[loss=0.1553, simple_loss=0.2338, pruned_loss=0.03839, over 623681.73 frames.], batch size: 52, aishell_tot_loss[loss=0.1606, simple_loss=0.2451, pruned_loss=0.03809, over 394179.93 frames.], datatang_tot_loss[loss=0.15, simple_loss=0.2225, pruned_loss=0.03875, over 382634.97 frames.], batch size: 52, lr: 5.52e-04
+2022-06-18 20:40:16,148 INFO [train.py:874] (2/4) Epoch 15, batch 250, aishell_loss[loss=0.1728, simple_loss=0.2433, pruned_loss=0.0511, over 4956.00 frames.], tot_loss[loss=0.156, simple_loss=0.2349, pruned_loss=0.03861, over 704409.91 frames.], batch size: 31, aishell_tot_loss[loss=0.1604, simple_loss=0.2444, pruned_loss=0.03814, over 489612.68 frames.], datatang_tot_loss[loss=0.1505, simple_loss=0.2228, pruned_loss=0.03917, over 426561.31 frames.], batch size: 31, lr: 5.51e-04
+2022-06-18 20:40:47,473 INFO [train.py:874] (2/4) Epoch 15, batch 300, datatang_loss[loss=0.1444, simple_loss=0.2059, pruned_loss=0.04142, over 4958.00 frames.], tot_loss[loss=0.1558, simple_loss=0.2346, pruned_loss=0.03851, over 767005.11 frames.], batch size: 50, aishell_tot_loss[loss=0.1606, simple_loss=0.2444, pruned_loss=0.03836, over 554712.50 frames.], datatang_tot_loss[loss=0.1499, simple_loss=0.2223, pruned_loss=0.03873, over 485050.55 frames.], batch size: 50, lr: 5.51e-04
+2022-06-18 20:41:16,893 INFO [train.py:874] (2/4) Epoch 15, batch 350, datatang_loss[loss=0.1662, simple_loss=0.2367, pruned_loss=0.04782, over 4856.00 frames.], tot_loss[loss=0.157, simple_loss=0.2355, pruned_loss=0.03921, over 815397.33 frames.], batch size: 30, aishell_tot_loss[loss=0.1607, simple_loss=0.2444, pruned_loss=0.03851, over 603559.83 frames.], datatang_tot_loss[loss=0.1519, simple_loss=0.2244, pruned_loss=0.03968, over 546145.75 frames.], batch size: 30, lr: 5.51e-04
+2022-06-18 20:41:47,846 INFO [train.py:874] (2/4) Epoch 15, batch 400, aishell_loss[loss=0.1321, simple_loss=0.2149, pruned_loss=0.02468, over 4877.00 frames.], tot_loss[loss=0.158, simple_loss=0.2367, pruned_loss=0.0397, over 852924.89 frames.], batch size: 28, aishell_tot_loss[loss=0.1613, simple_loss=0.2452, pruned_loss=0.03866, over 648238.08 frames.], datatang_tot_loss[loss=0.1531, simple_loss=0.2256, pruned_loss=0.04028, over 598012.38 frames.], batch size: 28, lr: 5.51e-04
+2022-06-18 20:42:19,899 INFO [train.py:874] (2/4) Epoch 15, batch 450, aishell_loss[loss=0.1856, simple_loss=0.2628, pruned_loss=0.0542, over 4968.00 frames.], tot_loss[loss=0.1594, simple_loss=0.2379, pruned_loss=0.04046, over 882567.99 frames.], batch size: 79, aishell_tot_loss[loss=0.1621, simple_loss=0.2463, pruned_loss=0.03899, over 679037.64 frames.], datatang_tot_loss[loss=0.1548, simple_loss=0.2275, pruned_loss=0.04101, over 653883.48 frames.], batch size: 79, lr: 5.51e-04
+2022-06-18 20:42:49,001 INFO [train.py:874] (2/4) Epoch 15, batch 500, aishell_loss[loss=0.16, simple_loss=0.2476, pruned_loss=0.03615, over 4871.00 frames.], tot_loss[loss=0.1592, simple_loss=0.2376, pruned_loss=0.04038, over 905258.83 frames.], batch size: 28, aishell_tot_loss[loss=0.1619, simple_loss=0.246, pruned_loss=0.03886, over 706539.02 frames.], datatang_tot_loss[loss=0.1553, simple_loss=0.2284, pruned_loss=0.04113, over 701793.30 frames.], batch size: 28, lr: 5.50e-04
+2022-06-18 20:43:19,968 INFO [train.py:874] (2/4) Epoch 15, batch 550, aishell_loss[loss=0.1534, simple_loss=0.2344, pruned_loss=0.03625, over 4978.00 frames.], tot_loss[loss=0.16, simple_loss=0.2376, pruned_loss=0.04117, over 923060.87 frames.], batch size: 30, aishell_tot_loss[loss=0.1615, simple_loss=0.245, pruned_loss=0.03902, over 734448.45 frames.], datatang_tot_loss[loss=0.157, simple_loss=0.23, pruned_loss=0.04206, over 740149.09 frames.], batch size: 30, lr: 5.50e-04
+2022-06-18 20:43:50,626 INFO [train.py:874] (2/4) Epoch 15, batch 600, aishell_loss[loss=0.1682, simple_loss=0.2545, pruned_loss=0.04094, over 4972.00 frames.], tot_loss[loss=0.1603, simple_loss=0.2383, pruned_loss=0.04116, over 937023.06 frames.], batch size: 39, aishell_tot_loss[loss=0.1617, simple_loss=0.2452, pruned_loss=0.03911, over 766465.41 frames.], datatang_tot_loss[loss=0.1575, simple_loss=0.2307, pruned_loss=0.04219, over 766747.70 frames.], batch size: 39, lr: 5.50e-04
+2022-06-18 20:44:19,782 INFO [train.py:874] (2/4) Epoch 15, batch 650, aishell_loss[loss=0.1636, simple_loss=0.2532, pruned_loss=0.03694, over 4925.00 frames.], tot_loss[loss=0.1613, simple_loss=0.2395, pruned_loss=0.04153, over 947842.06 frames.], batch size: 68, aishell_tot_loss[loss=0.1618, simple_loss=0.2454, pruned_loss=0.03913, over 793115.14 frames.], datatang_tot_loss[loss=0.1588, simple_loss=0.2319, pruned_loss=0.04284, over 791748.42 frames.], batch size: 68, lr: 5.50e-04
+2022-06-18 20:44:51,125 INFO [train.py:874] (2/4) Epoch 15, batch 700, aishell_loss[loss=0.1518, simple_loss=0.2462, pruned_loss=0.02869, over 4937.00 frames.], tot_loss[loss=0.1608, simple_loss=0.2391, pruned_loss=0.04126, over 956393.38 frames.], batch size: 54, aishell_tot_loss[loss=0.1619, simple_loss=0.2456, pruned_loss=0.03911, over 811343.09 frames.], datatang_tot_loss[loss=0.1585, simple_loss=0.2319, pruned_loss=0.04258, over 819109.28 frames.], batch size: 54, lr: 5.49e-04
+2022-06-18 20:45:21,686 INFO [train.py:874] (2/4) Epoch 15, batch 750, datatang_loss[loss=0.1476, simple_loss=0.2145, pruned_loss=0.04032, over 4853.00 frames.], tot_loss[loss=0.1604, simple_loss=0.2389, pruned_loss=0.04097, over 962705.30 frames.], batch size: 33, aishell_tot_loss[loss=0.1617, simple_loss=0.2453, pruned_loss=0.03904, over 833376.48 frames.], datatang_tot_loss[loss=0.1584, simple_loss=0.2319, pruned_loss=0.04246, over 837101.03 frames.], batch size: 33, lr: 5.49e-04
+2022-06-18 20:45:51,522 INFO [train.py:874] (2/4) Epoch 15, batch 800, datatang_loss[loss=0.1495, simple_loss=0.2009, pruned_loss=0.04902, over 4983.00 frames.], tot_loss[loss=0.1609, simple_loss=0.239, pruned_loss=0.04142, over 967702.78 frames.], batch size: 40, aishell_tot_loss[loss=0.1619, simple_loss=0.2453, pruned_loss=0.03923, over 847810.87 frames.], datatang_tot_loss[loss=0.159, simple_loss=0.2324, pruned_loss=0.04277, over 857821.16 frames.], batch size: 40, lr: 5.49e-04
+2022-06-18 20:46:21,636 INFO [train.py:874] (2/4) Epoch 15, batch 850, datatang_loss[loss=0.163, simple_loss=0.2333, pruned_loss=0.04639, over 4887.00 frames.], tot_loss[loss=0.1606, simple_loss=0.2383, pruned_loss=0.04142, over 971662.81 frames.], batch size: 52, aishell_tot_loss[loss=0.161, simple_loss=0.2441, pruned_loss=0.03895, over 863514.44 frames.], datatang_tot_loss[loss=0.1596, simple_loss=0.2329, pruned_loss=0.04313, over 873343.72 frames.], batch size: 52, lr: 5.49e-04
+2022-06-18 20:46:52,651 INFO [train.py:874] (2/4) Epoch 15, batch 900, aishell_loss[loss=0.1793, simple_loss=0.2478, pruned_loss=0.05539, over 4965.00 frames.], tot_loss[loss=0.1601, simple_loss=0.2378, pruned_loss=0.04118, over 974747.17 frames.], batch size: 31, aishell_tot_loss[loss=0.1606, simple_loss=0.2437, pruned_loss=0.03876, over 876076.58 frames.], datatang_tot_loss[loss=0.1595, simple_loss=0.2329, pruned_loss=0.04305, over 888226.38 frames.], batch size: 31, lr: 5.48e-04
+2022-06-18 20:47:21,755 INFO [train.py:874] (2/4) Epoch 15, batch 950, datatang_loss[loss=0.1466, simple_loss=0.2192, pruned_loss=0.03698, over 4968.00 frames.], tot_loss[loss=0.161, simple_loss=0.2388, pruned_loss=0.04165, over 977391.27 frames.], batch size: 34, aishell_tot_loss[loss=0.1615, simple_loss=0.2447, pruned_loss=0.03917, over 887637.36 frames.], datatang_tot_loss[loss=0.1597, simple_loss=0.2331, pruned_loss=0.0432, over 901106.78 frames.], batch size: 34, lr: 5.48e-04
+2022-06-18 20:47:53,005 INFO [train.py:874] (2/4) Epoch 15, batch 1000, aishell_loss[loss=0.175, simple_loss=0.2555, pruned_loss=0.04725, over 4865.00 frames.], tot_loss[loss=0.1599, simple_loss=0.238, pruned_loss=0.04086, over 978998.76 frames.], batch size: 37, aishell_tot_loss[loss=0.1613, simple_loss=0.2446, pruned_loss=0.03905, over 899468.92 frames.], datatang_tot_loss[loss=0.1587, simple_loss=0.2322, pruned_loss=0.04256, over 910596.42 frames.], batch size: 37, lr: 5.48e-04
+2022-06-18 20:47:53,006 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 20:48:10,191 INFO [train.py:914] (2/4) Epoch 15, validation: loss=0.1651, simple_loss=0.2495, pruned_loss=0.0404, over 1622729.00 frames.
+2022-06-18 20:48:40,014 INFO [train.py:874] (2/4) Epoch 15, batch 1050, datatang_loss[loss=0.1278, simple_loss=0.2074, pruned_loss=0.02413, over 4925.00 frames.], tot_loss[loss=0.16, simple_loss=0.238, pruned_loss=0.04101, over 980448.97 frames.], batch size: 75, aishell_tot_loss[loss=0.1618, simple_loss=0.2451, pruned_loss=0.03931, over 908590.14 frames.], datatang_tot_loss[loss=0.1583, simple_loss=0.2318, pruned_loss=0.04244, over 920288.91 frames.], batch size: 75, lr: 5.48e-04
+2022-06-18 20:49:10,955 INFO [train.py:874] (2/4) Epoch 15, batch 1100, aishell_loss[loss=0.1952, simple_loss=0.2749, pruned_loss=0.05775, over 4954.00 frames.], tot_loss[loss=0.16, simple_loss=0.2385, pruned_loss=0.0408, over 981883.24 frames.], batch size: 61, aishell_tot_loss[loss=0.1622, simple_loss=0.2457, pruned_loss=0.03935, over 919224.88 frames.], datatang_tot_loss[loss=0.1579, simple_loss=0.2312, pruned_loss=0.04224, over 926901.63 frames.], batch size: 61, lr: 5.48e-04
+2022-06-18 20:49:39,093 INFO [train.py:874] (2/4) Epoch 15, batch 1150, datatang_loss[loss=0.1687, simple_loss=0.2366, pruned_loss=0.05037, over 4948.00 frames.], tot_loss[loss=0.1607, simple_loss=0.2394, pruned_loss=0.041, over 982583.04 frames.], batch size: 62, aishell_tot_loss[loss=0.1626, simple_loss=0.2462, pruned_loss=0.03955, over 927523.74 frames.], datatang_tot_loss[loss=0.1581, simple_loss=0.2315, pruned_loss=0.04231, over 933292.51 frames.], batch size: 62, lr: 5.47e-04
+2022-06-18 20:50:10,426 INFO [train.py:874] (2/4) Epoch 15, batch 1200, aishell_loss[loss=0.1391, simple_loss=0.2145, pruned_loss=0.03189, over 4794.00 frames.], tot_loss[loss=0.1607, simple_loss=0.2395, pruned_loss=0.04089, over 983209.51 frames.], batch size: 20, aishell_tot_loss[loss=0.1628, simple_loss=0.2467, pruned_loss=0.03949, over 934096.21 frames.], datatang_tot_loss[loss=0.158, simple_loss=0.2315, pruned_loss=0.04225, over 939656.90 frames.], batch size: 20, lr: 5.47e-04
+2022-06-18 20:50:40,894 INFO [train.py:874] (2/4) Epoch 15, batch 1250, aishell_loss[loss=0.1788, simple_loss=0.2663, pruned_loss=0.04567, over 4918.00 frames.], tot_loss[loss=0.1609, simple_loss=0.2399, pruned_loss=0.04094, over 984006.18 frames.], batch size: 46, aishell_tot_loss[loss=0.1636, simple_loss=0.2474, pruned_loss=0.0399, over 939986.15 frames.], datatang_tot_loss[loss=0.1576, simple_loss=0.2314, pruned_loss=0.04191, over 945520.27 frames.], batch size: 46, lr: 5.47e-04
+2022-06-18 20:51:09,557 INFO [train.py:874] (2/4) Epoch 15, batch 1300, aishell_loss[loss=0.1645, simple_loss=0.2481, pruned_loss=0.04049, over 4879.00 frames.], tot_loss[loss=0.1616, simple_loss=0.2406, pruned_loss=0.04134, over 984345.27 frames.], batch size: 42, aishell_tot_loss[loss=0.1637, simple_loss=0.2475, pruned_loss=0.04001, over 946130.06 frames.], datatang_tot_loss[loss=0.1582, simple_loss=0.2319, pruned_loss=0.04228, over 949544.08 frames.], batch size: 42, lr: 5.47e-04
+2022-06-18 20:51:39,854 INFO [train.py:874] (2/4) Epoch 15, batch 1350, datatang_loss[loss=0.1733, simple_loss=0.2503, pruned_loss=0.04813, over 4934.00 frames.], tot_loss[loss=0.1614, simple_loss=0.2405, pruned_loss=0.04121, over 984642.51 frames.], batch size: 94, aishell_tot_loss[loss=0.163, simple_loss=0.2471, pruned_loss=0.03946, over 949922.31 frames.], datatang_tot_loss[loss=0.159, simple_loss=0.2327, pruned_loss=0.04265, over 954547.79 frames.], batch size: 94, lr: 5.46e-04
+2022-06-18 20:52:11,215 INFO [train.py:874] (2/4) Epoch 15, batch 1400, datatang_loss[loss=0.1662, simple_loss=0.2481, pruned_loss=0.04208, over 4950.00 frames.], tot_loss[loss=0.1621, simple_loss=0.2408, pruned_loss=0.04175, over 984860.10 frames.], batch size: 67, aishell_tot_loss[loss=0.1637, simple_loss=0.2477, pruned_loss=0.03979, over 953560.62 frames.], datatang_tot_loss[loss=0.1593, simple_loss=0.2327, pruned_loss=0.0429, over 958709.14 frames.], batch size: 67, lr: 5.46e-04
+2022-06-18 20:52:39,833 INFO [train.py:874] (2/4) Epoch 15, batch 1450, aishell_loss[loss=0.1728, simple_loss=0.2572, pruned_loss=0.04426, over 4964.00 frames.], tot_loss[loss=0.1614, simple_loss=0.2399, pruned_loss=0.04148, over 985225.27 frames.], batch size: 44, aishell_tot_loss[loss=0.1631, simple_loss=0.2472, pruned_loss=0.03948, over 957207.35 frames.], datatang_tot_loss[loss=0.1593, simple_loss=0.2325, pruned_loss=0.04306, over 962222.81 frames.], batch size: 44, lr: 5.46e-04
+2022-06-18 20:53:10,614 INFO [train.py:874] (2/4) Epoch 15, batch 1500, aishell_loss[loss=0.1624, simple_loss=0.2517, pruned_loss=0.03659, over 4970.00 frames.], tot_loss[loss=0.1608, simple_loss=0.2394, pruned_loss=0.04108, over 985615.54 frames.], batch size: 61, aishell_tot_loss[loss=0.1629, simple_loss=0.2472, pruned_loss=0.03932, over 960198.31 frames.], datatang_tot_loss[loss=0.1589, simple_loss=0.2323, pruned_loss=0.04276, over 965551.93 frames.], batch size: 61, lr: 5.46e-04
+2022-06-18 20:53:41,103 INFO [train.py:874] (2/4) Epoch 15, batch 1550, aishell_loss[loss=0.1463, simple_loss=0.2311, pruned_loss=0.03073, over 4976.00 frames.], tot_loss[loss=0.1608, simple_loss=0.2394, pruned_loss=0.04109, over 985605.26 frames.], batch size: 51, aishell_tot_loss[loss=0.1619, simple_loss=0.2461, pruned_loss=0.0388, over 963028.79 frames.], datatang_tot_loss[loss=0.1599, simple_loss=0.2332, pruned_loss=0.04331, over 968087.16 frames.], batch size: 51, lr: 5.45e-04
+2022-06-18 20:54:09,708 INFO [train.py:874] (2/4) Epoch 15, batch 1600, datatang_loss[loss=0.1741, simple_loss=0.2429, pruned_loss=0.05266, over 4925.00 frames.], tot_loss[loss=0.1614, simple_loss=0.2401, pruned_loss=0.04131, over 985807.05 frames.], batch size: 57, aishell_tot_loss[loss=0.1627, simple_loss=0.2471, pruned_loss=0.03909, over 965403.34 frames.], datatang_tot_loss[loss=0.1597, simple_loss=0.233, pruned_loss=0.04321, over 970603.05 frames.], batch size: 57, lr: 5.45e-04
+2022-06-18 20:54:39,885 INFO [train.py:874] (2/4) Epoch 15, batch 1650, aishell_loss[loss=0.1429, simple_loss=0.2297, pruned_loss=0.02802, over 4859.00 frames.], tot_loss[loss=0.1607, simple_loss=0.2397, pruned_loss=0.04086, over 985849.13 frames.], batch size: 36, aishell_tot_loss[loss=0.1623, simple_loss=0.2469, pruned_loss=0.03882, over 967680.12 frames.], datatang_tot_loss[loss=0.1594, simple_loss=0.2328, pruned_loss=0.04305, over 972582.08 frames.], batch size: 36, lr: 5.45e-04
+2022-06-18 20:55:11,397 INFO [train.py:874] (2/4) Epoch 15, batch 1700, aishell_loss[loss=0.1479, simple_loss=0.2137, pruned_loss=0.04102, over 4975.00 frames.], tot_loss[loss=0.1608, simple_loss=0.2395, pruned_loss=0.0411, over 986099.39 frames.], batch size: 22, aishell_tot_loss[loss=0.1621, simple_loss=0.2467, pruned_loss=0.03877, over 969248.96 frames.], datatang_tot_loss[loss=0.1598, simple_loss=0.2333, pruned_loss=0.04317, over 974762.77 frames.], batch size: 22, lr: 5.45e-04
+2022-06-18 20:55:40,695 INFO [train.py:874] (2/4) Epoch 15, batch 1750, aishell_loss[loss=0.1642, simple_loss=0.2527, pruned_loss=0.03782, over 4958.00 frames.], tot_loss[loss=0.1598, simple_loss=0.2385, pruned_loss=0.04058, over 986492.78 frames.], batch size: 64, aishell_tot_loss[loss=0.1612, simple_loss=0.2456, pruned_loss=0.03839, over 971090.66 frames.], datatang_tot_loss[loss=0.1597, simple_loss=0.2334, pruned_loss=0.04294, over 976567.67 frames.], batch size: 64, lr: 5.45e-04
+2022-06-18 20:56:11,379 INFO [train.py:874] (2/4) Epoch 15, batch 1800, datatang_loss[loss=0.1487, simple_loss=0.2187, pruned_loss=0.03932, over 4928.00 frames.], tot_loss[loss=0.1597, simple_loss=0.2384, pruned_loss=0.04055, over 986350.06 frames.], batch size: 71, aishell_tot_loss[loss=0.161, simple_loss=0.2457, pruned_loss=0.03818, over 972535.86 frames.], datatang_tot_loss[loss=0.1596, simple_loss=0.2332, pruned_loss=0.04302, over 977889.54 frames.], batch size: 71, lr: 5.44e-04
+2022-06-18 20:56:41,336 INFO [train.py:874] (2/4) Epoch 15, batch 1850, datatang_loss[loss=0.1685, simple_loss=0.2445, pruned_loss=0.04621, over 4906.00 frames.], tot_loss[loss=0.1587, simple_loss=0.2375, pruned_loss=0.03996, over 986028.56 frames.], batch size: 52, aishell_tot_loss[loss=0.1604, simple_loss=0.2449, pruned_loss=0.03795, over 974096.44 frames.], datatang_tot_loss[loss=0.159, simple_loss=0.2326, pruned_loss=0.04265, over 978626.86 frames.], batch size: 52, lr: 5.44e-04
+2022-06-18 20:57:10,549 INFO [train.py:874] (2/4) Epoch 15, batch 1900, aishell_loss[loss=0.161, simple_loss=0.2406, pruned_loss=0.0407, over 4957.00 frames.], tot_loss[loss=0.1597, simple_loss=0.2381, pruned_loss=0.04063, over 986139.89 frames.], batch size: 64, aishell_tot_loss[loss=0.1603, simple_loss=0.2447, pruned_loss=0.03791, over 975691.93 frames.], datatang_tot_loss[loss=0.1599, simple_loss=0.2331, pruned_loss=0.04334, over 979453.56 frames.], batch size: 64, lr: 5.44e-04
+2022-06-18 20:57:39,924 INFO [train.py:874] (2/4) Epoch 15, batch 1950, aishell_loss[loss=0.1414, simple_loss=0.2289, pruned_loss=0.02695, over 4968.00 frames.], tot_loss[loss=0.159, simple_loss=0.2375, pruned_loss=0.04029, over 986005.82 frames.], batch size: 27, aishell_tot_loss[loss=0.1605, simple_loss=0.2447, pruned_loss=0.0381, over 976921.33 frames.], datatang_tot_loss[loss=0.1589, simple_loss=0.2322, pruned_loss=0.04279, over 980101.56 frames.], batch size: 27, lr: 5.44e-04
+2022-06-18 20:58:10,962 INFO [train.py:874] (2/4) Epoch 15, batch 2000, aishell_loss[loss=0.1597, simple_loss=0.2478, pruned_loss=0.03582, over 4949.00 frames.], tot_loss[loss=0.1591, simple_loss=0.2374, pruned_loss=0.04046, over 985867.24 frames.], batch size: 54, aishell_tot_loss[loss=0.1603, simple_loss=0.2444, pruned_loss=0.03811, over 977814.63 frames.], datatang_tot_loss[loss=0.159, simple_loss=0.2323, pruned_loss=0.04287, over 980790.01 frames.], batch size: 54, lr: 5.43e-04
+2022-06-18 20:58:10,962 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 20:58:27,550 INFO [train.py:914] (2/4) Epoch 15, validation: loss=0.166, simple_loss=0.2499, pruned_loss=0.04104, over 1622729.00 frames.
+2022-06-18 20:58:58,110 INFO [train.py:874] (2/4) Epoch 15, batch 2050, datatang_loss[loss=0.1431, simple_loss=0.2195, pruned_loss=0.03332, over 4959.00 frames.], tot_loss[loss=0.1603, simple_loss=0.2386, pruned_loss=0.04099, over 985439.59 frames.], batch size: 67, aishell_tot_loss[loss=0.1611, simple_loss=0.245, pruned_loss=0.03862, over 978380.99 frames.], datatang_tot_loss[loss=0.1594, simple_loss=0.2328, pruned_loss=0.04296, over 981322.13 frames.], batch size: 67, lr: 5.43e-04
+2022-06-18 20:59:28,181 INFO [train.py:874] (2/4) Epoch 15, batch 2100, datatang_loss[loss=0.166, simple_loss=0.2454, pruned_loss=0.04334, over 4830.00 frames.], tot_loss[loss=0.1594, simple_loss=0.2378, pruned_loss=0.04052, over 985601.61 frames.], batch size: 30, aishell_tot_loss[loss=0.1606, simple_loss=0.2445, pruned_loss=0.03832, over 979321.95 frames.], datatang_tot_loss[loss=0.159, simple_loss=0.2324, pruned_loss=0.0428, over 981867.05 frames.], batch size: 30, lr: 5.43e-04
+2022-06-18 20:59:58,263 INFO [train.py:874] (2/4) Epoch 15, batch 2150, aishell_loss[loss=0.1653, simple_loss=0.2489, pruned_loss=0.04084, over 4962.00 frames.], tot_loss[loss=0.1595, simple_loss=0.2379, pruned_loss=0.04058, over 985706.54 frames.], batch size: 40, aishell_tot_loss[loss=0.1608, simple_loss=0.2449, pruned_loss=0.03838, over 980055.78 frames.], datatang_tot_loss[loss=0.1588, simple_loss=0.2321, pruned_loss=0.04271, over 982370.88 frames.], batch size: 40, lr: 5.43e-04
+2022-06-18 21:00:33,783 INFO [train.py:874] (2/4) Epoch 15, batch 2200, aishell_loss[loss=0.1845, simple_loss=0.2653, pruned_loss=0.05183, over 4904.00 frames.], tot_loss[loss=0.1596, simple_loss=0.2379, pruned_loss=0.04066, over 985583.81 frames.], batch size: 41, aishell_tot_loss[loss=0.1607, simple_loss=0.2447, pruned_loss=0.03837, over 980707.30 frames.], datatang_tot_loss[loss=0.1589, simple_loss=0.2323, pruned_loss=0.04276, over 982637.44 frames.], batch size: 41, lr: 5.43e-04
+2022-06-18 21:01:02,746 INFO [train.py:874] (2/4) Epoch 15, batch 2250, aishell_loss[loss=0.1592, simple_loss=0.2478, pruned_loss=0.0353, over 4962.00 frames.], tot_loss[loss=0.159, simple_loss=0.2375, pruned_loss=0.04025, over 985473.23 frames.], batch size: 61, aishell_tot_loss[loss=0.1607, simple_loss=0.2446, pruned_loss=0.03834, over 981119.48 frames.], datatang_tot_loss[loss=0.1582, simple_loss=0.2315, pruned_loss=0.04248, over 983062.88 frames.], batch size: 61, lr: 5.42e-04
+2022-06-18 21:01:33,949 INFO [train.py:874] (2/4) Epoch 15, batch 2300, datatang_loss[loss=0.1978, simple_loss=0.2613, pruned_loss=0.06717, over 4956.00 frames.], tot_loss[loss=0.159, simple_loss=0.2371, pruned_loss=0.04039, over 985871.88 frames.], batch size: 86, aishell_tot_loss[loss=0.1599, simple_loss=0.244, pruned_loss=0.03795, over 981840.36 frames.], datatang_tot_loss[loss=0.1588, simple_loss=0.2317, pruned_loss=0.04294, over 983538.16 frames.], batch size: 86, lr: 5.42e-04
+2022-06-18 21:02:05,450 INFO [train.py:874] (2/4) Epoch 15, batch 2350, aishell_loss[loss=0.1801, simple_loss=0.2623, pruned_loss=0.04898, over 4974.00 frames.], tot_loss[loss=0.159, simple_loss=0.2371, pruned_loss=0.04041, over 985769.44 frames.], batch size: 44, aishell_tot_loss[loss=0.1602, simple_loss=0.244, pruned_loss=0.03817, over 982431.15 frames.], datatang_tot_loss[loss=0.1585, simple_loss=0.2315, pruned_loss=0.04271, over 983578.19 frames.], batch size: 44, lr: 5.42e-04
+2022-06-18 21:02:33,670 INFO [train.py:874] (2/4) Epoch 15, batch 2400, datatang_loss[loss=0.1681, simple_loss=0.2435, pruned_loss=0.04632, over 4919.00 frames.], tot_loss[loss=0.1595, simple_loss=0.2377, pruned_loss=0.04064, over 985867.97 frames.], batch size: 83, aishell_tot_loss[loss=0.1606, simple_loss=0.2444, pruned_loss=0.03836, over 982853.32 frames.], datatang_tot_loss[loss=0.1585, simple_loss=0.2315, pruned_loss=0.0428, over 983920.95 frames.], batch size: 83, lr: 5.42e-04
+2022-06-18 21:03:04,144 INFO [train.py:874] (2/4) Epoch 15, batch 2450, aishell_loss[loss=0.1639, simple_loss=0.2506, pruned_loss=0.03863, over 4974.00 frames.], tot_loss[loss=0.16, simple_loss=0.2383, pruned_loss=0.04088, over 985840.46 frames.], batch size: 39, aishell_tot_loss[loss=0.1606, simple_loss=0.2447, pruned_loss=0.03824, over 983123.91 frames.], datatang_tot_loss[loss=0.1591, simple_loss=0.232, pruned_loss=0.04305, over 984174.87 frames.], batch size: 39, lr: 5.41e-04
+2022-06-18 21:03:35,629 INFO [train.py:874] (2/4) Epoch 15, batch 2500, aishell_loss[loss=0.1125, simple_loss=0.1813, pruned_loss=0.02189, over 4933.00 frames.], tot_loss[loss=0.1607, simple_loss=0.2393, pruned_loss=0.04106, over 985964.92 frames.], batch size: 21, aishell_tot_loss[loss=0.1608, simple_loss=0.2449, pruned_loss=0.03832, over 983544.57 frames.], datatang_tot_loss[loss=0.1596, simple_loss=0.2326, pruned_loss=0.04333, over 984433.26 frames.], batch size: 21, lr: 5.41e-04
+2022-06-18 21:04:05,191 INFO [train.py:874] (2/4) Epoch 15, batch 2550, datatang_loss[loss=0.1674, simple_loss=0.2443, pruned_loss=0.04519, over 4956.00 frames.], tot_loss[loss=0.16, simple_loss=0.2385, pruned_loss=0.04071, over 986114.22 frames.], batch size: 86, aishell_tot_loss[loss=0.1604, simple_loss=0.2445, pruned_loss=0.03816, over 983910.23 frames.], datatang_tot_loss[loss=0.1593, simple_loss=0.2322, pruned_loss=0.04323, over 984700.78 frames.], batch size: 86, lr: 5.41e-04
+2022-06-18 21:04:36,451 INFO [train.py:874] (2/4) Epoch 15, batch 2600, datatang_loss[loss=0.1723, simple_loss=0.239, pruned_loss=0.05283, over 4921.00 frames.], tot_loss[loss=0.1594, simple_loss=0.2381, pruned_loss=0.04037, over 985942.14 frames.], batch size: 42, aishell_tot_loss[loss=0.1603, simple_loss=0.2445, pruned_loss=0.03804, over 984130.49 frames.], datatang_tot_loss[loss=0.1588, simple_loss=0.2317, pruned_loss=0.04301, over 984722.42 frames.], batch size: 42, lr: 5.41e-04
+2022-06-18 21:05:08,092 INFO [train.py:874] (2/4) Epoch 15, batch 2650, datatang_loss[loss=0.1383, simple_loss=0.2138, pruned_loss=0.03139, over 4930.00 frames.], tot_loss[loss=0.1592, simple_loss=0.2378, pruned_loss=0.04035, over 985682.25 frames.], batch size: 77, aishell_tot_loss[loss=0.1597, simple_loss=0.2439, pruned_loss=0.03773, over 984175.00 frames.], datatang_tot_loss[loss=0.1593, simple_loss=0.2321, pruned_loss=0.04326, over 984745.43 frames.], batch size: 77, lr: 5.41e-04
+2022-06-18 21:05:37,258 INFO [train.py:874] (2/4) Epoch 15, batch 2700, datatang_loss[loss=0.1568, simple_loss=0.2393, pruned_loss=0.03714, over 4956.00 frames.], tot_loss[loss=0.159, simple_loss=0.2376, pruned_loss=0.04022, over 985843.46 frames.], batch size: 86, aishell_tot_loss[loss=0.16, simple_loss=0.2443, pruned_loss=0.03782, over 984112.37 frames.], datatang_tot_loss[loss=0.1587, simple_loss=0.2317, pruned_loss=0.04286, over 985238.28 frames.], batch size: 86, lr: 5.40e-04
+2022-06-18 21:06:06,727 INFO [train.py:874] (2/4) Epoch 15, batch 2750, datatang_loss[loss=0.1358, simple_loss=0.2162, pruned_loss=0.02771, over 4921.00 frames.], tot_loss[loss=0.1583, simple_loss=0.2375, pruned_loss=0.03953, over 985737.46 frames.], batch size: 77, aishell_tot_loss[loss=0.1598, simple_loss=0.2445, pruned_loss=0.0376, over 984160.94 frames.], datatang_tot_loss[loss=0.158, simple_loss=0.2312, pruned_loss=0.04239, over 985358.02 frames.], batch size: 77, lr: 5.40e-04
+2022-06-18 21:06:38,005 INFO [train.py:874] (2/4) Epoch 15, batch 2800, datatang_loss[loss=0.1473, simple_loss=0.2163, pruned_loss=0.03917, over 4916.00 frames.], tot_loss[loss=0.1578, simple_loss=0.2369, pruned_loss=0.03931, over 985773.37 frames.], batch size: 42, aishell_tot_loss[loss=0.1595, simple_loss=0.2442, pruned_loss=0.03739, over 984093.24 frames.], datatang_tot_loss[loss=0.1577, simple_loss=0.2309, pruned_loss=0.04222, over 985670.45 frames.], batch size: 42, lr: 5.40e-04
+2022-06-18 21:07:05,638 INFO [train.py:874] (2/4) Epoch 15, batch 2850, aishell_loss[loss=0.1972, simple_loss=0.2718, pruned_loss=0.06129, over 4871.00 frames.], tot_loss[loss=0.1589, simple_loss=0.238, pruned_loss=0.03992, over 985661.22 frames.], batch size: 36, aishell_tot_loss[loss=0.1597, simple_loss=0.2442, pruned_loss=0.03761, over 984004.08 frames.], datatang_tot_loss[loss=0.1584, simple_loss=0.2318, pruned_loss=0.04254, over 985864.08 frames.], batch size: 36, lr: 5.40e-04
+2022-06-18 21:07:36,982 INFO [train.py:874] (2/4) Epoch 15, batch 2900, datatang_loss[loss=0.1632, simple_loss=0.2274, pruned_loss=0.04951, over 4951.00 frames.], tot_loss[loss=0.1597, simple_loss=0.2379, pruned_loss=0.04073, over 986024.29 frames.], batch size: 55, aishell_tot_loss[loss=0.1596, simple_loss=0.244, pruned_loss=0.03766, over 984290.22 frames.], datatang_tot_loss[loss=0.1593, simple_loss=0.2322, pruned_loss=0.04322, over 986095.24 frames.], batch size: 55, lr: 5.39e-04
+2022-06-18 21:08:07,678 INFO [train.py:874] (2/4) Epoch 15, batch 2950, aishell_loss[loss=0.1585, simple_loss=0.2423, pruned_loss=0.0374, over 4870.00 frames.], tot_loss[loss=0.1585, simple_loss=0.2372, pruned_loss=0.03983, over 985921.30 frames.], batch size: 35, aishell_tot_loss[loss=0.1589, simple_loss=0.2432, pruned_loss=0.0373, over 984341.89 frames.], datatang_tot_loss[loss=0.1588, simple_loss=0.232, pruned_loss=0.04281, over 986168.78 frames.], batch size: 35, lr: 5.39e-04
+2022-06-18 21:08:37,247 INFO [train.py:874] (2/4) Epoch 15, batch 3000, datatang_loss[loss=0.1655, simple_loss=0.2352, pruned_loss=0.0479, over 4890.00 frames.], tot_loss[loss=0.159, simple_loss=0.2368, pruned_loss=0.04057, over 985808.50 frames.], batch size: 52, aishell_tot_loss[loss=0.1598, simple_loss=0.2438, pruned_loss=0.03787, over 984494.57 frames.], datatang_tot_loss[loss=0.1584, simple_loss=0.2309, pruned_loss=0.0429, over 986035.86 frames.], batch size: 52, lr: 5.39e-04
+2022-06-18 21:08:37,247 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 21:08:54,378 INFO [train.py:914] (2/4) Epoch 15, validation: loss=0.1663, simple_loss=0.2504, pruned_loss=0.04115, over 1622729.00 frames.
+2022-06-18 21:09:23,247 INFO [train.py:874] (2/4) Epoch 15, batch 3050, aishell_loss[loss=0.1286, simple_loss=0.2163, pruned_loss=0.02044, over 4978.00 frames.], tot_loss[loss=0.1581, simple_loss=0.2367, pruned_loss=0.03972, over 985554.46 frames.], batch size: 30, aishell_tot_loss[loss=0.1588, simple_loss=0.2433, pruned_loss=0.03716, over 984581.09 frames.], datatang_tot_loss[loss=0.1584, simple_loss=0.231, pruned_loss=0.04291, over 985837.74 frames.], batch size: 30, lr: 5.39e-04
+2022-06-18 21:09:54,958 INFO [train.py:874] (2/4) Epoch 15, batch 3100, aishell_loss[loss=0.1639, simple_loss=0.2497, pruned_loss=0.03905, over 4943.00 frames.], tot_loss[loss=0.1584, simple_loss=0.2371, pruned_loss=0.03983, over 985240.95 frames.], batch size: 45, aishell_tot_loss[loss=0.1591, simple_loss=0.2436, pruned_loss=0.03729, over 984453.24 frames.], datatang_tot_loss[loss=0.1583, simple_loss=0.2311, pruned_loss=0.04277, over 985696.11 frames.], batch size: 45, lr: 5.39e-04
+2022-06-18 21:10:26,278 INFO [train.py:874] (2/4) Epoch 15, batch 3150, aishell_loss[loss=0.1629, simple_loss=0.2507, pruned_loss=0.03761, over 4938.00 frames.], tot_loss[loss=0.1588, simple_loss=0.2373, pruned_loss=0.04016, over 985144.13 frames.], batch size: 54, aishell_tot_loss[loss=0.1597, simple_loss=0.2441, pruned_loss=0.0376, over 984292.58 frames.], datatang_tot_loss[loss=0.1582, simple_loss=0.2308, pruned_loss=0.04279, over 985779.34 frames.], batch size: 54, lr: 5.38e-04
+2022-06-18 21:10:56,704 INFO [train.py:874] (2/4) Epoch 15, batch 3200, datatang_loss[loss=0.1451, simple_loss=0.2145, pruned_loss=0.03782, over 4920.00 frames.], tot_loss[loss=0.1594, simple_loss=0.238, pruned_loss=0.04039, over 985100.25 frames.], batch size: 77, aishell_tot_loss[loss=0.1602, simple_loss=0.2446, pruned_loss=0.0379, over 984308.61 frames.], datatang_tot_loss[loss=0.1582, simple_loss=0.2308, pruned_loss=0.04275, over 985749.70 frames.], batch size: 77, lr: 5.38e-04
+2022-06-18 21:11:27,555 INFO [train.py:874] (2/4) Epoch 15, batch 3250, aishell_loss[loss=0.1151, simple_loss=0.1935, pruned_loss=0.01832, over 4790.00 frames.], tot_loss[loss=0.1603, simple_loss=0.2384, pruned_loss=0.04114, over 984800.42 frames.], batch size: 24, aishell_tot_loss[loss=0.1601, simple_loss=0.2443, pruned_loss=0.03797, over 984052.68 frames.], datatang_tot_loss[loss=0.1593, simple_loss=0.2319, pruned_loss=0.0434, over 985677.71 frames.], batch size: 24, lr: 5.38e-04
+2022-06-18 21:11:59,149 INFO [train.py:874] (2/4) Epoch 15, batch 3300, datatang_loss[loss=0.1733, simple_loss=0.2406, pruned_loss=0.053, over 4943.00 frames.], tot_loss[loss=0.1603, simple_loss=0.2384, pruned_loss=0.0411, over 984824.36 frames.], batch size: 45, aishell_tot_loss[loss=0.1606, simple_loss=0.2451, pruned_loss=0.03808, over 984296.73 frames.], datatang_tot_loss[loss=0.1589, simple_loss=0.2314, pruned_loss=0.04326, over 985439.66 frames.], batch size: 45, lr: 5.38e-04
+2022-06-18 21:12:28,841 INFO [train.py:874] (2/4) Epoch 15, batch 3350, aishell_loss[loss=0.1643, simple_loss=0.2451, pruned_loss=0.04176, over 4941.00 frames.], tot_loss[loss=0.1603, simple_loss=0.2387, pruned_loss=0.04094, over 985153.70 frames.], batch size: 54, aishell_tot_loss[loss=0.1604, simple_loss=0.2449, pruned_loss=0.0379, over 984694.20 frames.], datatang_tot_loss[loss=0.1593, simple_loss=0.2319, pruned_loss=0.04337, over 985381.54 frames.], batch size: 54, lr: 5.37e-04
+2022-06-18 21:12:59,951 INFO [train.py:874] (2/4) Epoch 15, batch 3400, datatang_loss[loss=0.142, simple_loss=0.2217, pruned_loss=0.03114, over 4926.00 frames.], tot_loss[loss=0.1601, simple_loss=0.2389, pruned_loss=0.04069, over 985263.11 frames.], batch size: 73, aishell_tot_loss[loss=0.1606, simple_loss=0.245, pruned_loss=0.03805, over 984764.63 frames.], datatang_tot_loss[loss=0.159, simple_loss=0.2319, pruned_loss=0.0431, over 985456.84 frames.], batch size: 73, lr: 5.37e-04
+2022-06-18 21:13:28,564 INFO [train.py:874] (2/4) Epoch 15, batch 3450, aishell_loss[loss=0.1444, simple_loss=0.2369, pruned_loss=0.02601, over 4914.00 frames.], tot_loss[loss=0.1598, simple_loss=0.2387, pruned_loss=0.04047, over 985236.14 frames.], batch size: 58, aishell_tot_loss[loss=0.1614, simple_loss=0.2457, pruned_loss=0.03853, over 984879.33 frames.], datatang_tot_loss[loss=0.1579, simple_loss=0.2309, pruned_loss=0.04243, over 985361.34 frames.], batch size: 58, lr: 5.37e-04
+2022-06-18 21:13:59,925 INFO [train.py:874] (2/4) Epoch 15, batch 3500, aishell_loss[loss=0.1919, simple_loss=0.268, pruned_loss=0.05793, over 4946.00 frames.], tot_loss[loss=0.1599, simple_loss=0.2384, pruned_loss=0.04069, over 985192.10 frames.], batch size: 32, aishell_tot_loss[loss=0.1615, simple_loss=0.2457, pruned_loss=0.03866, over 984656.08 frames.], datatang_tot_loss[loss=0.1579, simple_loss=0.231, pruned_loss=0.04243, over 985556.70 frames.], batch size: 32, lr: 5.37e-04
+2022-06-18 21:14:32,149 INFO [train.py:874] (2/4) Epoch 15, batch 3550, aishell_loss[loss=0.1724, simple_loss=0.2594, pruned_loss=0.04268, over 4948.00 frames.], tot_loss[loss=0.1603, simple_loss=0.239, pruned_loss=0.0408, over 985679.04 frames.], batch size: 79, aishell_tot_loss[loss=0.1621, simple_loss=0.2461, pruned_loss=0.03908, over 984976.76 frames.], datatang_tot_loss[loss=0.1577, simple_loss=0.231, pruned_loss=0.04222, over 985790.73 frames.], batch size: 79, lr: 5.37e-04
+2022-06-18 21:15:01,932 INFO [train.py:874] (2/4) Epoch 15, batch 3600, aishell_loss[loss=0.2109, simple_loss=0.2784, pruned_loss=0.0717, over 4877.00 frames.], tot_loss[loss=0.1599, simple_loss=0.2388, pruned_loss=0.04051, over 985424.28 frames.], batch size: 47, aishell_tot_loss[loss=0.1622, simple_loss=0.2463, pruned_loss=0.03907, over 984657.69 frames.], datatang_tot_loss[loss=0.1573, simple_loss=0.2309, pruned_loss=0.04187, over 985885.39 frames.], batch size: 47, lr: 5.36e-04
+2022-06-18 21:15:34,134 INFO [train.py:874] (2/4) Epoch 15, batch 3650, datatang_loss[loss=0.1528, simple_loss=0.2276, pruned_loss=0.03901, over 4959.00 frames.], tot_loss[loss=0.1587, simple_loss=0.2376, pruned_loss=0.03989, over 985495.69 frames.], batch size: 67, aishell_tot_loss[loss=0.1618, simple_loss=0.2458, pruned_loss=0.0389, over 984669.17 frames.], datatang_tot_loss[loss=0.1566, simple_loss=0.2305, pruned_loss=0.04134, over 985950.95 frames.], batch size: 67, lr: 5.36e-04
+2022-06-18 21:16:05,733 INFO [train.py:874] (2/4) Epoch 15, batch 3700, aishell_loss[loss=0.1695, simple_loss=0.2595, pruned_loss=0.03972, over 4964.00 frames.], tot_loss[loss=0.1592, simple_loss=0.2386, pruned_loss=0.03985, over 985651.49 frames.], batch size: 69, aishell_tot_loss[loss=0.1621, simple_loss=0.2464, pruned_loss=0.03889, over 984912.95 frames.], datatang_tot_loss[loss=0.1566, simple_loss=0.2305, pruned_loss=0.0413, over 985938.88 frames.], batch size: 69, lr: 5.36e-04
+2022-06-18 21:16:34,879 INFO [train.py:874] (2/4) Epoch 15, batch 3750, datatang_loss[loss=0.1886, simple_loss=0.2608, pruned_loss=0.05821, over 4918.00 frames.], tot_loss[loss=0.1593, simple_loss=0.2384, pruned_loss=0.04007, over 985508.14 frames.], batch size: 98, aishell_tot_loss[loss=0.1618, simple_loss=0.2461, pruned_loss=0.03869, over 984831.56 frames.], datatang_tot_loss[loss=0.157, simple_loss=0.2308, pruned_loss=0.04162, over 985894.92 frames.], batch size: 98, lr: 5.36e-04
+2022-06-18 21:17:05,060 INFO [train.py:874] (2/4) Epoch 15, batch 3800, aishell_loss[loss=0.197, simple_loss=0.2808, pruned_loss=0.05658, over 4909.00 frames.], tot_loss[loss=0.1584, simple_loss=0.2379, pruned_loss=0.03942, over 985317.98 frames.], batch size: 79, aishell_tot_loss[loss=0.161, simple_loss=0.2455, pruned_loss=0.03825, over 984697.06 frames.], datatang_tot_loss[loss=0.1568, simple_loss=0.2307, pruned_loss=0.0414, over 985876.85 frames.], batch size: 79, lr: 5.35e-04
+2022-06-18 21:17:36,745 INFO [train.py:874] (2/4) Epoch 15, batch 3850, aishell_loss[loss=0.1556, simple_loss=0.2491, pruned_loss=0.03107, over 4976.00 frames.], tot_loss[loss=0.1584, simple_loss=0.2384, pruned_loss=0.03917, over 985458.26 frames.], batch size: 39, aishell_tot_loss[loss=0.161, simple_loss=0.2457, pruned_loss=0.03813, over 984715.70 frames.], datatang_tot_loss[loss=0.1566, simple_loss=0.2307, pruned_loss=0.04119, over 986026.59 frames.], batch size: 39, lr: 5.35e-04
+2022-06-18 21:18:05,687 INFO [train.py:874] (2/4) Epoch 15, batch 3900, datatang_loss[loss=0.1409, simple_loss=0.2178, pruned_loss=0.03201, over 4976.00 frames.], tot_loss[loss=0.157, simple_loss=0.2369, pruned_loss=0.03861, over 985252.01 frames.], batch size: 60, aishell_tot_loss[loss=0.1604, simple_loss=0.245, pruned_loss=0.03794, over 984434.71 frames.], datatang_tot_loss[loss=0.1557, simple_loss=0.2302, pruned_loss=0.0406, over 986083.30 frames.], batch size: 60, lr: 5.35e-04
+2022-06-18 21:18:34,802 INFO [train.py:874] (2/4) Epoch 15, batch 3950, datatang_loss[loss=0.1451, simple_loss=0.2279, pruned_loss=0.03117, over 4930.00 frames.], tot_loss[loss=0.1566, simple_loss=0.2364, pruned_loss=0.0384, over 985371.13 frames.], batch size: 57, aishell_tot_loss[loss=0.1599, simple_loss=0.2444, pruned_loss=0.03765, over 984674.75 frames.], datatang_tot_loss[loss=0.1554, simple_loss=0.2297, pruned_loss=0.0406, over 986015.02 frames.], batch size: 57, lr: 5.35e-04
+2022-06-18 21:19:03,470 INFO [train.py:874] (2/4) Epoch 15, batch 4000, datatang_loss[loss=0.1333, simple_loss=0.2115, pruned_loss=0.02755, over 4925.00 frames.], tot_loss[loss=0.1564, simple_loss=0.2363, pruned_loss=0.03828, over 985609.90 frames.], batch size: 57, aishell_tot_loss[loss=0.16, simple_loss=0.2445, pruned_loss=0.03776, over 984879.28 frames.], datatang_tot_loss[loss=0.1548, simple_loss=0.2291, pruned_loss=0.04023, over 986092.72 frames.], batch size: 57, lr: 5.35e-04
+2022-06-18 21:19:03,471 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 21:19:20,712 INFO [train.py:914] (2/4) Epoch 15, validation: loss=0.1641, simple_loss=0.2483, pruned_loss=0.0399, over 1622729.00 frames.
+2022-06-18 21:19:49,325 INFO [train.py:874] (2/4) Epoch 15, batch 4050, datatang_loss[loss=0.1587, simple_loss=0.2461, pruned_loss=0.03565, over 4861.00 frames.], tot_loss[loss=0.1563, simple_loss=0.2357, pruned_loss=0.03845, over 985737.54 frames.], batch size: 30, aishell_tot_loss[loss=0.1598, simple_loss=0.2443, pruned_loss=0.03761, over 985048.35 frames.], datatang_tot_loss[loss=0.1546, simple_loss=0.2284, pruned_loss=0.0404, over 986106.81 frames.], batch size: 30, lr: 5.34e-04
+2022-06-18 21:20:18,318 INFO [train.py:874] (2/4) Epoch 15, batch 4100, aishell_loss[loss=0.1326, simple_loss=0.2112, pruned_loss=0.02705, over 4981.00 frames.], tot_loss[loss=0.157, simple_loss=0.2365, pruned_loss=0.03876, over 985286.26 frames.], batch size: 27, aishell_tot_loss[loss=0.1599, simple_loss=0.2445, pruned_loss=0.03761, over 984694.05 frames.], datatang_tot_loss[loss=0.155, simple_loss=0.2286, pruned_loss=0.04064, over 986033.81 frames.], batch size: 27, lr: 5.34e-04
+2022-06-18 21:20:47,954 INFO [train.py:874] (2/4) Epoch 15, batch 4150, datatang_loss[loss=0.1541, simple_loss=0.2295, pruned_loss=0.03931, over 4896.00 frames.], tot_loss[loss=0.1569, simple_loss=0.2358, pruned_loss=0.03899, over 985412.28 frames.], batch size: 52, aishell_tot_loss[loss=0.1592, simple_loss=0.2438, pruned_loss=0.03733, over 984577.35 frames.], datatang_tot_loss[loss=0.1555, simple_loss=0.2289, pruned_loss=0.04103, over 986242.13 frames.], batch size: 52, lr: 5.34e-04
+2022-06-18 21:22:22,383 INFO [train.py:874] (2/4) Epoch 16, batch 50, aishell_loss[loss=0.1615, simple_loss=0.2433, pruned_loss=0.03983, over 4935.00 frames.], tot_loss[loss=0.153, simple_loss=0.2304, pruned_loss=0.0378, over 218396.88 frames.], batch size: 49, aishell_tot_loss[loss=0.1533, simple_loss=0.2372, pruned_loss=0.0347, over 111489.40 frames.], datatang_tot_loss[loss=0.1525, simple_loss=0.224, pruned_loss=0.04053, over 120528.26 frames.], batch size: 49, lr: 5.18e-04
+2022-06-18 21:22:53,185 INFO [train.py:874] (2/4) Epoch 16, batch 100, aishell_loss[loss=0.1571, simple_loss=0.2419, pruned_loss=0.03617, over 4925.00 frames.], tot_loss[loss=0.1528, simple_loss=0.2311, pruned_loss=0.03729, over 388549.22 frames.], batch size: 33, aishell_tot_loss[loss=0.1547, simple_loss=0.24, pruned_loss=0.03472, over 202849.61 frames.], datatang_tot_loss[loss=0.1512, simple_loss=0.2233, pruned_loss=0.03948, over 233793.95 frames.], batch size: 33, lr: 5.18e-04
+2022-06-18 21:23:22,588 INFO [train.py:874] (2/4) Epoch 16, batch 150, aishell_loss[loss=0.1614, simple_loss=0.2525, pruned_loss=0.03511, over 4913.00 frames.], tot_loss[loss=0.1537, simple_loss=0.2319, pruned_loss=0.03779, over 520566.98 frames.], batch size: 41, aishell_tot_loss[loss=0.1569, simple_loss=0.2414, pruned_loss=0.03621, over 280769.96 frames.], datatang_tot_loss[loss=0.1509, simple_loss=0.224, pruned_loss=0.03892, over 335458.78 frames.], batch size: 41, lr: 5.18e-04
+2022-06-18 21:23:53,695 INFO [train.py:874] (2/4) Epoch 16, batch 200, aishell_loss[loss=0.1514, simple_loss=0.245, pruned_loss=0.02896, over 4955.00 frames.], tot_loss[loss=0.1536, simple_loss=0.2319, pruned_loss=0.03761, over 623944.10 frames.], batch size: 56, aishell_tot_loss[loss=0.1574, simple_loss=0.2418, pruned_loss=0.03648, over 351423.73 frames.], datatang_tot_loss[loss=0.1504, simple_loss=0.224, pruned_loss=0.03844, over 423485.57 frames.], batch size: 56, lr: 5.17e-04
+2022-06-18 21:24:24,689 INFO [train.py:874] (2/4) Epoch 16, batch 250, aishell_loss[loss=0.1617, simple_loss=0.2478, pruned_loss=0.0378, over 4865.00 frames.], tot_loss[loss=0.1544, simple_loss=0.2337, pruned_loss=0.03758, over 703652.57 frames.], batch size: 35, aishell_tot_loss[loss=0.1587, simple_loss=0.2432, pruned_loss=0.0371, over 439501.37 frames.], datatang_tot_loss[loss=0.1499, simple_loss=0.2239, pruned_loss=0.03793, over 477066.14 frames.], batch size: 35, lr: 5.17e-04
+2022-06-18 21:24:55,264 INFO [train.py:874] (2/4) Epoch 16, batch 300, datatang_loss[loss=0.1571, simple_loss=0.2224, pruned_loss=0.0459, over 4933.00 frames.], tot_loss[loss=0.1553, simple_loss=0.2345, pruned_loss=0.03807, over 765798.84 frames.], batch size: 69, aishell_tot_loss[loss=0.1589, simple_loss=0.2436, pruned_loss=0.03713, over 508348.02 frames.], datatang_tot_loss[loss=0.151, simple_loss=0.2246, pruned_loss=0.03867, over 532347.75 frames.], batch size: 69, lr: 5.17e-04
+2022-06-18 21:25:25,761 INFO [train.py:874] (2/4) Epoch 16, batch 350, datatang_loss[loss=0.1558, simple_loss=0.236, pruned_loss=0.03783, over 4923.00 frames.], tot_loss[loss=0.1551, simple_loss=0.234, pruned_loss=0.03813, over 814575.16 frames.], batch size: 73, aishell_tot_loss[loss=0.1589, simple_loss=0.243, pruned_loss=0.03735, over 553901.28 frames.], datatang_tot_loss[loss=0.1512, simple_loss=0.2253, pruned_loss=0.03858, over 595661.28 frames.], batch size: 73, lr: 5.17e-04
+2022-06-18 21:25:56,577 INFO [train.py:874] (2/4) Epoch 16, batch 400, datatang_loss[loss=0.1579, simple_loss=0.226, pruned_loss=0.04492, over 4973.00 frames.], tot_loss[loss=0.1543, simple_loss=0.2333, pruned_loss=0.03766, over 852410.52 frames.], batch size: 60, aishell_tot_loss[loss=0.1583, simple_loss=0.2431, pruned_loss=0.03677, over 604547.98 frames.], datatang_tot_loss[loss=0.1507, simple_loss=0.2243, pruned_loss=0.03853, over 641695.54 frames.], batch size: 60, lr: 5.17e-04
+2022-06-18 21:26:25,941 INFO [train.py:874] (2/4) Epoch 16, batch 450, datatang_loss[loss=0.1565, simple_loss=0.219, pruned_loss=0.04698, over 4840.00 frames.], tot_loss[loss=0.1547, simple_loss=0.2341, pruned_loss=0.03769, over 881876.40 frames.], batch size: 25, aishell_tot_loss[loss=0.1581, simple_loss=0.2428, pruned_loss=0.03667, over 657656.26 frames.], datatang_tot_loss[loss=0.1512, simple_loss=0.225, pruned_loss=0.03875, over 674538.85 frames.], batch size: 25, lr: 5.16e-04
+2022-06-18 21:26:54,477 INFO [train.py:874] (2/4) Epoch 16, batch 500, datatang_loss[loss=0.1484, simple_loss=0.2309, pruned_loss=0.033, over 4932.00 frames.], tot_loss[loss=0.1554, simple_loss=0.2353, pruned_loss=0.03775, over 904951.38 frames.], batch size: 77, aishell_tot_loss[loss=0.1581, simple_loss=0.2429, pruned_loss=0.03667, over 710343.53 frames.], datatang_tot_loss[loss=0.1517, simple_loss=0.2255, pruned_loss=0.03899, over 697238.64 frames.], batch size: 77, lr: 5.16e-04
+2022-06-18 21:27:26,323 INFO [train.py:874] (2/4) Epoch 16, batch 550, aishell_loss[loss=0.1706, simple_loss=0.2505, pruned_loss=0.04536, over 4942.00 frames.], tot_loss[loss=0.1554, simple_loss=0.2354, pruned_loss=0.03766, over 922390.11 frames.], batch size: 31, aishell_tot_loss[loss=0.1584, simple_loss=0.2433, pruned_loss=0.03677, over 742289.26 frames.], datatang_tot_loss[loss=0.1515, simple_loss=0.2256, pruned_loss=0.03873, over 731203.76 frames.], batch size: 31, lr: 5.16e-04
+2022-06-18 21:27:56,249 INFO [train.py:874] (2/4) Epoch 16, batch 600, aishell_loss[loss=0.1622, simple_loss=0.2502, pruned_loss=0.0371, over 4947.00 frames.], tot_loss[loss=0.1567, simple_loss=0.2363, pruned_loss=0.03853, over 936376.52 frames.], batch size: 45, aishell_tot_loss[loss=0.1588, simple_loss=0.2434, pruned_loss=0.03708, over 770814.56 frames.], datatang_tot_loss[loss=0.153, simple_loss=0.2269, pruned_loss=0.03952, over 761273.63 frames.], batch size: 45, lr: 5.16e-04
+2022-06-18 21:28:26,886 INFO [train.py:874] (2/4) Epoch 16, batch 650, datatang_loss[loss=0.1528, simple_loss=0.2225, pruned_loss=0.0416, over 4848.00 frames.], tot_loss[loss=0.1574, simple_loss=0.2368, pruned_loss=0.03899, over 947126.21 frames.], batch size: 23, aishell_tot_loss[loss=0.1594, simple_loss=0.2439, pruned_loss=0.03741, over 793234.01 frames.], datatang_tot_loss[loss=0.1536, simple_loss=0.2276, pruned_loss=0.03979, over 790486.73 frames.], batch size: 23, lr: 5.16e-04
+2022-06-18 21:28:57,667 INFO [train.py:874] (2/4) Epoch 16, batch 700, aishell_loss[loss=0.1851, simple_loss=0.2734, pruned_loss=0.04835, over 4877.00 frames.], tot_loss[loss=0.1583, simple_loss=0.2376, pruned_loss=0.03953, over 955907.47 frames.], batch size: 35, aishell_tot_loss[loss=0.1604, simple_loss=0.2451, pruned_loss=0.03786, over 812638.75 frames.], datatang_tot_loss[loss=0.1541, simple_loss=0.2281, pruned_loss=0.04006, over 816958.97 frames.], batch size: 35, lr: 5.15e-04
+2022-06-18 21:29:27,943 INFO [train.py:874] (2/4) Epoch 16, batch 750, aishell_loss[loss=0.1686, simple_loss=0.2512, pruned_loss=0.04302, over 4959.00 frames.], tot_loss[loss=0.1589, simple_loss=0.2375, pruned_loss=0.0401, over 962666.68 frames.], batch size: 64, aishell_tot_loss[loss=0.1603, simple_loss=0.2448, pruned_loss=0.03794, over 829351.63 frames.], datatang_tot_loss[loss=0.1553, simple_loss=0.2291, pruned_loss=0.04077, over 840478.11 frames.], batch size: 64, lr: 5.15e-04
+2022-06-18 21:29:58,389 INFO [train.py:874] (2/4) Epoch 16, batch 800, aishell_loss[loss=0.1369, simple_loss=0.2282, pruned_loss=0.02278, over 4900.00 frames.], tot_loss[loss=0.159, simple_loss=0.2376, pruned_loss=0.04021, over 967787.79 frames.], batch size: 34, aishell_tot_loss[loss=0.1601, simple_loss=0.2446, pruned_loss=0.03777, over 847767.52 frames.], datatang_tot_loss[loss=0.156, simple_loss=0.2294, pruned_loss=0.04127, over 857596.45 frames.], batch size: 34, lr: 5.15e-04
+2022-06-18 21:30:28,993 INFO [train.py:874] (2/4) Epoch 16, batch 850, aishell_loss[loss=0.1495, simple_loss=0.2339, pruned_loss=0.03249, over 4916.00 frames.], tot_loss[loss=0.1575, simple_loss=0.2358, pruned_loss=0.0396, over 971421.05 frames.], batch size: 52, aishell_tot_loss[loss=0.1586, simple_loss=0.243, pruned_loss=0.03705, over 862671.11 frames.], datatang_tot_loss[loss=0.1561, simple_loss=0.2293, pruned_loss=0.04143, over 873577.28 frames.], batch size: 52, lr: 5.15e-04
+2022-06-18 21:30:58,972 INFO [train.py:874] (2/4) Epoch 16, batch 900, aishell_loss[loss=0.1289, simple_loss=0.2168, pruned_loss=0.02052, over 4985.00 frames.], tot_loss[loss=0.1565, simple_loss=0.2349, pruned_loss=0.03906, over 974869.16 frames.], batch size: 30, aishell_tot_loss[loss=0.1581, simple_loss=0.2427, pruned_loss=0.0368, over 874476.05 frames.], datatang_tot_loss[loss=0.1555, simple_loss=0.2289, pruned_loss=0.04103, over 889444.62 frames.], batch size: 30, lr: 5.15e-04
+2022-06-18 21:31:29,480 INFO [train.py:874] (2/4) Epoch 16, batch 950, datatang_loss[loss=0.1491, simple_loss=0.2304, pruned_loss=0.03389, over 4896.00 frames.], tot_loss[loss=0.1568, simple_loss=0.2352, pruned_loss=0.0392, over 977574.40 frames.], batch size: 59, aishell_tot_loss[loss=0.1578, simple_loss=0.2423, pruned_loss=0.03669, over 885643.24 frames.], datatang_tot_loss[loss=0.156, simple_loss=0.2296, pruned_loss=0.04122, over 902736.17 frames.], batch size: 59, lr: 5.14e-04
+2022-06-18 21:31:59,326 INFO [train.py:874] (2/4) Epoch 16, batch 1000, aishell_loss[loss=0.1478, simple_loss=0.2386, pruned_loss=0.02845, over 4938.00 frames.], tot_loss[loss=0.1569, simple_loss=0.236, pruned_loss=0.03896, over 979134.47 frames.], batch size: 45, aishell_tot_loss[loss=0.1581, simple_loss=0.2424, pruned_loss=0.03688, over 900677.55 frames.], datatang_tot_loss[loss=0.1559, simple_loss=0.2297, pruned_loss=0.04101, over 909477.00 frames.], batch size: 45, lr: 5.14e-04
+2022-06-18 21:31:59,327 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 21:32:15,094 INFO [train.py:914] (2/4) Epoch 16, validation: loss=0.1652, simple_loss=0.2489, pruned_loss=0.04072, over 1622729.00 frames.
+2022-06-18 21:32:45,850 INFO [train.py:874] (2/4) Epoch 16, batch 1050, aishell_loss[loss=0.1741, simple_loss=0.2601, pruned_loss=0.04402, over 4971.00 frames.], tot_loss[loss=0.1572, simple_loss=0.2366, pruned_loss=0.03894, over 980907.96 frames.], batch size: 80, aishell_tot_loss[loss=0.1579, simple_loss=0.2425, pruned_loss=0.0367, over 909701.79 frames.], datatang_tot_loss[loss=0.1564, simple_loss=0.2306, pruned_loss=0.0411, over 919637.84 frames.], batch size: 80, lr: 5.14e-04
+2022-06-18 21:33:17,604 INFO [train.py:874] (2/4) Epoch 16, batch 1100, aishell_loss[loss=0.1631, simple_loss=0.2468, pruned_loss=0.03971, over 4975.00 frames.], tot_loss[loss=0.1579, simple_loss=0.2366, pruned_loss=0.03956, over 982089.58 frames.], batch size: 51, aishell_tot_loss[loss=0.1582, simple_loss=0.2427, pruned_loss=0.03685, over 917819.85 frames.], datatang_tot_loss[loss=0.1568, simple_loss=0.2306, pruned_loss=0.04157, over 928230.86 frames.], batch size: 51, lr: 5.14e-04
+2022-06-18 21:33:46,499 INFO [train.py:874] (2/4) Epoch 16, batch 1150, aishell_loss[loss=0.141, simple_loss=0.2042, pruned_loss=0.0389, over 4927.00 frames.], tot_loss[loss=0.1571, simple_loss=0.2363, pruned_loss=0.03898, over 982828.03 frames.], batch size: 21, aishell_tot_loss[loss=0.1574, simple_loss=0.2421, pruned_loss=0.0363, over 926227.19 frames.], datatang_tot_loss[loss=0.1569, simple_loss=0.2307, pruned_loss=0.0416, over 934574.06 frames.], batch size: 21, lr: 5.14e-04
+2022-06-18 21:34:16,208 INFO [train.py:874] (2/4) Epoch 16, batch 1200, aishell_loss[loss=0.2134, simple_loss=0.2801, pruned_loss=0.07329, over 4858.00 frames.], tot_loss[loss=0.1582, simple_loss=0.2375, pruned_loss=0.03944, over 983544.21 frames.], batch size: 37, aishell_tot_loss[loss=0.1585, simple_loss=0.2433, pruned_loss=0.03689, over 933949.63 frames.], datatang_tot_loss[loss=0.1569, simple_loss=0.2306, pruned_loss=0.0416, over 940077.67 frames.], batch size: 37, lr: 5.13e-04
+2022-06-18 21:34:46,017 INFO [train.py:874] (2/4) Epoch 16, batch 1250, aishell_loss[loss=0.1895, simple_loss=0.2691, pruned_loss=0.05494, over 4915.00 frames.], tot_loss[loss=0.159, simple_loss=0.2383, pruned_loss=0.03982, over 983505.78 frames.], batch size: 41, aishell_tot_loss[loss=0.1592, simple_loss=0.2437, pruned_loss=0.03729, over 941076.06 frames.], datatang_tot_loss[loss=0.1572, simple_loss=0.2308, pruned_loss=0.04182, over 943990.59 frames.], batch size: 41, lr: 5.13e-04
+2022-06-18 21:35:16,551 INFO [train.py:874] (2/4) Epoch 16, batch 1300, datatang_loss[loss=0.175, simple_loss=0.2492, pruned_loss=0.05046, over 4964.00 frames.], tot_loss[loss=0.1583, simple_loss=0.2377, pruned_loss=0.03941, over 983834.90 frames.], batch size: 91, aishell_tot_loss[loss=0.1592, simple_loss=0.2437, pruned_loss=0.03733, over 945644.76 frames.], datatang_tot_loss[loss=0.1566, simple_loss=0.2306, pruned_loss=0.04133, over 949363.81 frames.], batch size: 91, lr: 5.13e-04
+2022-06-18 21:35:48,286 INFO [train.py:874] (2/4) Epoch 16, batch 1350, datatang_loss[loss=0.1689, simple_loss=0.2457, pruned_loss=0.04602, over 4940.00 frames.], tot_loss[loss=0.1583, simple_loss=0.2378, pruned_loss=0.03939, over 984431.05 frames.], batch size: 42, aishell_tot_loss[loss=0.1598, simple_loss=0.2444, pruned_loss=0.03761, over 950811.24 frames.], datatang_tot_loss[loss=0.1561, simple_loss=0.23, pruned_loss=0.0411, over 953409.39 frames.], batch size: 42, lr: 5.13e-04
+2022-06-18 21:36:18,257 INFO [train.py:874] (2/4) Epoch 16, batch 1400, aishell_loss[loss=0.1673, simple_loss=0.2538, pruned_loss=0.04045, over 4976.00 frames.], tot_loss[loss=0.1575, simple_loss=0.2369, pruned_loss=0.03906, over 984285.89 frames.], batch size: 39, aishell_tot_loss[loss=0.159, simple_loss=0.2438, pruned_loss=0.03713, over 953911.79 frames.], datatang_tot_loss[loss=0.1562, simple_loss=0.2301, pruned_loss=0.04114, over 957710.69 frames.], batch size: 39, lr: 5.13e-04
+2022-06-18 21:36:47,827 INFO [train.py:874] (2/4) Epoch 16, batch 1450, datatang_loss[loss=0.1638, simple_loss=0.2358, pruned_loss=0.04585, over 4942.00 frames.], tot_loss[loss=0.1577, simple_loss=0.2369, pruned_loss=0.03921, over 984012.43 frames.], batch size: 69, aishell_tot_loss[loss=0.1591, simple_loss=0.2437, pruned_loss=0.03729, over 957405.59 frames.], datatang_tot_loss[loss=0.1562, simple_loss=0.2302, pruned_loss=0.04115, over 960653.63 frames.], batch size: 69, lr: 5.12e-04
+2022-06-18 21:37:19,305 INFO [train.py:874] (2/4) Epoch 16, batch 1500, datatang_loss[loss=0.1436, simple_loss=0.2146, pruned_loss=0.03632, over 4891.00 frames.], tot_loss[loss=0.1571, simple_loss=0.2363, pruned_loss=0.03897, over 984476.65 frames.], batch size: 47, aishell_tot_loss[loss=0.1592, simple_loss=0.2437, pruned_loss=0.03733, over 960890.35 frames.], datatang_tot_loss[loss=0.1555, simple_loss=0.2293, pruned_loss=0.04086, over 963558.79 frames.], batch size: 47, lr: 5.12e-04
+2022-06-18 21:37:48,655 INFO [train.py:874] (2/4) Epoch 16, batch 1550, datatang_loss[loss=0.1877, simple_loss=0.2522, pruned_loss=0.06162, over 4863.00 frames.], tot_loss[loss=0.1568, simple_loss=0.236, pruned_loss=0.03876, over 984470.02 frames.], batch size: 39, aishell_tot_loss[loss=0.1584, simple_loss=0.243, pruned_loss=0.0369, over 963727.07 frames.], datatang_tot_loss[loss=0.1559, simple_loss=0.2297, pruned_loss=0.04107, over 965964.13 frames.], batch size: 39, lr: 5.12e-04
+2022-06-18 21:38:18,581 INFO [train.py:874] (2/4) Epoch 16, batch 1600, datatang_loss[loss=0.14, simple_loss=0.2146, pruned_loss=0.03276, over 4939.00 frames.], tot_loss[loss=0.156, simple_loss=0.2351, pruned_loss=0.03841, over 984667.13 frames.], batch size: 71, aishell_tot_loss[loss=0.1581, simple_loss=0.2427, pruned_loss=0.03669, over 966096.39 frames.], datatang_tot_loss[loss=0.1553, simple_loss=0.2289, pruned_loss=0.04085, over 968421.20 frames.], batch size: 71, lr: 5.12e-04
+2022-06-18 21:38:49,710 INFO [train.py:874] (2/4) Epoch 16, batch 1650, aishell_loss[loss=0.1464, simple_loss=0.2249, pruned_loss=0.03395, over 4984.00 frames.], tot_loss[loss=0.157, simple_loss=0.2358, pruned_loss=0.03907, over 984840.79 frames.], batch size: 25, aishell_tot_loss[loss=0.1587, simple_loss=0.2433, pruned_loss=0.03707, over 967878.11 frames.], datatang_tot_loss[loss=0.1556, simple_loss=0.2292, pruned_loss=0.04098, over 970857.67 frames.], batch size: 25, lr: 5.12e-04
+2022-06-18 21:39:20,244 INFO [train.py:874] (2/4) Epoch 16, batch 1700, aishell_loss[loss=0.1497, simple_loss=0.2375, pruned_loss=0.03098, over 4939.00 frames.], tot_loss[loss=0.1563, simple_loss=0.2353, pruned_loss=0.03866, over 985002.74 frames.], batch size: 58, aishell_tot_loss[loss=0.1588, simple_loss=0.2434, pruned_loss=0.03704, over 969830.80 frames.], datatang_tot_loss[loss=0.1549, simple_loss=0.2286, pruned_loss=0.04055, over 972684.58 frames.], batch size: 58, lr: 5.11e-04
+2022-06-18 21:39:48,453 INFO [train.py:874] (2/4) Epoch 16, batch 1750, datatang_loss[loss=0.1521, simple_loss=0.2245, pruned_loss=0.03985, over 4935.00 frames.], tot_loss[loss=0.1565, simple_loss=0.2357, pruned_loss=0.03866, over 985045.13 frames.], batch size: 73, aishell_tot_loss[loss=0.1579, simple_loss=0.2426, pruned_loss=0.03661, over 971513.74 frames.], datatang_tot_loss[loss=0.1558, simple_loss=0.2298, pruned_loss=0.04094, over 974296.18 frames.], batch size: 73, lr: 5.11e-04
+2022-06-18 21:40:19,017 INFO [train.py:874] (2/4) Epoch 16, batch 1800, aishell_loss[loss=0.1465, simple_loss=0.2387, pruned_loss=0.02716, over 4959.00 frames.], tot_loss[loss=0.1565, simple_loss=0.2363, pruned_loss=0.03833, over 984892.75 frames.], batch size: 64, aishell_tot_loss[loss=0.1571, simple_loss=0.2421, pruned_loss=0.03603, over 973118.77 frames.], datatang_tot_loss[loss=0.1565, simple_loss=0.2305, pruned_loss=0.04125, over 975440.61 frames.], batch size: 64, lr: 5.11e-04
+2022-06-18 21:40:47,575 INFO [train.py:874] (2/4) Epoch 16, batch 1850, aishell_loss[loss=0.1599, simple_loss=0.2512, pruned_loss=0.03426, over 4863.00 frames.], tot_loss[loss=0.1561, simple_loss=0.2358, pruned_loss=0.0382, over 984594.87 frames.], batch size: 35, aishell_tot_loss[loss=0.1571, simple_loss=0.2421, pruned_loss=0.03604, over 974138.65 frames.], datatang_tot_loss[loss=0.156, simple_loss=0.23, pruned_loss=0.04095, over 976592.31 frames.], batch size: 35, lr: 5.11e-04
+2022-06-18 21:41:17,887 INFO [train.py:874] (2/4) Epoch 16, batch 1900, datatang_loss[loss=0.1699, simple_loss=0.2484, pruned_loss=0.04572, over 4921.00 frames.], tot_loss[loss=0.1567, simple_loss=0.2363, pruned_loss=0.03855, over 984767.73 frames.], batch size: 81, aishell_tot_loss[loss=0.1575, simple_loss=0.2423, pruned_loss=0.0363, over 975657.68 frames.], datatang_tot_loss[loss=0.1561, simple_loss=0.23, pruned_loss=0.04108, over 977462.93 frames.], batch size: 81, lr: 5.11e-04
+2022-06-18 21:41:47,573 INFO [train.py:874] (2/4) Epoch 16, batch 1950, aishell_loss[loss=0.1228, simple_loss=0.2052, pruned_loss=0.02023, over 4813.00 frames.], tot_loss[loss=0.1564, simple_loss=0.2359, pruned_loss=0.03843, over 984763.69 frames.], batch size: 24, aishell_tot_loss[loss=0.1572, simple_loss=0.242, pruned_loss=0.03618, over 976610.81 frames.], datatang_tot_loss[loss=0.156, simple_loss=0.23, pruned_loss=0.04103, over 978436.87 frames.], batch size: 24, lr: 5.10e-04
+2022-06-18 21:42:22,161 INFO [train.py:874] (2/4) Epoch 16, batch 2000, datatang_loss[loss=0.1353, simple_loss=0.2107, pruned_loss=0.02998, over 4929.00 frames.], tot_loss[loss=0.1561, simple_loss=0.236, pruned_loss=0.03816, over 984889.86 frames.], batch size: 73, aishell_tot_loss[loss=0.1569, simple_loss=0.2419, pruned_loss=0.03595, over 977572.39 frames.], datatang_tot_loss[loss=0.156, simple_loss=0.23, pruned_loss=0.04094, over 979319.54 frames.], batch size: 73, lr: 5.10e-04
+2022-06-18 21:42:22,162 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 21:42:38,234 INFO [train.py:914] (2/4) Epoch 16, validation: loss=0.1651, simple_loss=0.2493, pruned_loss=0.04046, over 1622729.00 frames.
+2022-06-18 21:43:07,840 INFO [train.py:874] (2/4) Epoch 16, batch 2050, datatang_loss[loss=0.1207, simple_loss=0.2065, pruned_loss=0.01742, over 4934.00 frames.], tot_loss[loss=0.1558, simple_loss=0.2352, pruned_loss=0.03815, over 985092.96 frames.], batch size: 79, aishell_tot_loss[loss=0.1576, simple_loss=0.2424, pruned_loss=0.03642, over 978231.38 frames.], datatang_tot_loss[loss=0.1548, simple_loss=0.229, pruned_loss=0.04032, over 980353.24 frames.], batch size: 79, lr: 5.10e-04
+2022-06-18 21:43:37,797 INFO [train.py:874] (2/4) Epoch 16, batch 2100, aishell_loss[loss=0.1496, simple_loss=0.2384, pruned_loss=0.0304, over 4935.00 frames.], tot_loss[loss=0.1554, simple_loss=0.2351, pruned_loss=0.03788, over 985146.33 frames.], batch size: 64, aishell_tot_loss[loss=0.157, simple_loss=0.2419, pruned_loss=0.03602, over 979064.74 frames.], datatang_tot_loss[loss=0.155, simple_loss=0.2291, pruned_loss=0.04044, over 980961.90 frames.], batch size: 64, lr: 5.10e-04
+2022-06-18 21:44:09,095 INFO [train.py:874] (2/4) Epoch 16, batch 2150, aishell_loss[loss=0.1426, simple_loss=0.2297, pruned_loss=0.02777, over 4963.00 frames.], tot_loss[loss=0.156, simple_loss=0.2356, pruned_loss=0.0382, over 985293.53 frames.], batch size: 31, aishell_tot_loss[loss=0.1571, simple_loss=0.242, pruned_loss=0.03609, over 979704.35 frames.], datatang_tot_loss[loss=0.1553, simple_loss=0.2293, pruned_loss=0.04064, over 981693.75 frames.], batch size: 31, lr: 5.10e-04
+2022-06-18 21:44:39,262 INFO [train.py:874] (2/4) Epoch 16, batch 2200, datatang_loss[loss=0.2146, simple_loss=0.2753, pruned_loss=0.07698, over 4912.00 frames.], tot_loss[loss=0.1571, simple_loss=0.2362, pruned_loss=0.03897, over 985326.27 frames.], batch size: 109, aishell_tot_loss[loss=0.1574, simple_loss=0.2421, pruned_loss=0.0363, over 980341.95 frames.], datatang_tot_loss[loss=0.1561, simple_loss=0.2299, pruned_loss=0.04116, over 982179.18 frames.], batch size: 109, lr: 5.09e-04
+2022-06-18 21:45:09,176 INFO [train.py:874] (2/4) Epoch 16, batch 2250, datatang_loss[loss=0.145, simple_loss=0.2052, pruned_loss=0.04236, over 4956.00 frames.], tot_loss[loss=0.1573, simple_loss=0.236, pruned_loss=0.03926, over 985520.54 frames.], batch size: 25, aishell_tot_loss[loss=0.1574, simple_loss=0.242, pruned_loss=0.03638, over 980893.27 frames.], datatang_tot_loss[loss=0.1564, simple_loss=0.23, pruned_loss=0.0414, over 982757.12 frames.], batch size: 25, lr: 5.09e-04
+2022-06-18 21:45:40,317 INFO [train.py:874] (2/4) Epoch 16, batch 2300, aishell_loss[loss=0.1563, simple_loss=0.2416, pruned_loss=0.03548, over 4986.00 frames.], tot_loss[loss=0.157, simple_loss=0.2357, pruned_loss=0.03913, over 985493.46 frames.], batch size: 38, aishell_tot_loss[loss=0.1576, simple_loss=0.242, pruned_loss=0.03662, over 981221.43 frames.], datatang_tot_loss[loss=0.1559, simple_loss=0.2296, pruned_loss=0.04111, over 983257.20 frames.], batch size: 38, lr: 5.09e-04
+2022-06-18 21:46:10,852 INFO [train.py:874] (2/4) Epoch 16, batch 2350, aishell_loss[loss=0.1437, simple_loss=0.201, pruned_loss=0.04321, over 4946.00 frames.], tot_loss[loss=0.1573, simple_loss=0.2357, pruned_loss=0.03946, over 985158.99 frames.], batch size: 21, aishell_tot_loss[loss=0.1579, simple_loss=0.242, pruned_loss=0.0369, over 981263.74 frames.], datatang_tot_loss[loss=0.156, simple_loss=0.2296, pruned_loss=0.04126, over 983623.75 frames.], batch size: 21, lr: 5.09e-04
+2022-06-18 21:46:41,008 INFO [train.py:874] (2/4) Epoch 16, batch 2400, datatang_loss[loss=0.1398, simple_loss=0.2094, pruned_loss=0.03515, over 4979.00 frames.], tot_loss[loss=0.1571, simple_loss=0.2355, pruned_loss=0.03935, over 985700.86 frames.], batch size: 34, aishell_tot_loss[loss=0.1585, simple_loss=0.2428, pruned_loss=0.03715, over 982075.49 frames.], datatang_tot_loss[loss=0.1554, simple_loss=0.2289, pruned_loss=0.04093, over 983991.27 frames.], batch size: 34, lr: 5.09e-04
+2022-06-18 21:47:10,429 INFO [train.py:874] (2/4) Epoch 16, batch 2450, datatang_loss[loss=0.1739, simple_loss=0.2346, pruned_loss=0.05664, over 4924.00 frames.], tot_loss[loss=0.157, simple_loss=0.2359, pruned_loss=0.03905, over 985313.77 frames.], batch size: 73, aishell_tot_loss[loss=0.1579, simple_loss=0.2421, pruned_loss=0.03687, over 982265.08 frames.], datatang_tot_loss[loss=0.1558, simple_loss=0.2292, pruned_loss=0.04116, over 984126.52 frames.], batch size: 73, lr: 5.08e-04
+2022-06-18 21:47:41,056 INFO [train.py:874] (2/4) Epoch 16, batch 2500, datatang_loss[loss=0.1398, simple_loss=0.2125, pruned_loss=0.03351, over 4911.00 frames.], tot_loss[loss=0.1566, simple_loss=0.2355, pruned_loss=0.03887, over 985181.82 frames.], batch size: 42, aishell_tot_loss[loss=0.1579, simple_loss=0.242, pruned_loss=0.03688, over 982419.07 frames.], datatang_tot_loss[loss=0.1555, simple_loss=0.229, pruned_loss=0.04097, over 984352.26 frames.], batch size: 42, lr: 5.08e-04
+2022-06-18 21:48:11,422 INFO [train.py:874] (2/4) Epoch 16, batch 2550, datatang_loss[loss=0.1357, simple_loss=0.2192, pruned_loss=0.02613, over 4925.00 frames.], tot_loss[loss=0.157, simple_loss=0.2363, pruned_loss=0.03891, over 985362.58 frames.], batch size: 83, aishell_tot_loss[loss=0.1579, simple_loss=0.242, pruned_loss=0.03692, over 982773.08 frames.], datatang_tot_loss[loss=0.1558, simple_loss=0.2295, pruned_loss=0.04107, over 984644.60 frames.], batch size: 83, lr: 5.08e-04
+2022-06-18 21:48:41,675 INFO [train.py:874] (2/4) Epoch 16, batch 2600, datatang_loss[loss=0.1235, simple_loss=0.1969, pruned_loss=0.02502, over 4970.00 frames.], tot_loss[loss=0.1565, simple_loss=0.2352, pruned_loss=0.03887, over 985582.21 frames.], batch size: 45, aishell_tot_loss[loss=0.1576, simple_loss=0.2415, pruned_loss=0.03686, over 983121.35 frames.], datatang_tot_loss[loss=0.1555, simple_loss=0.2291, pruned_loss=0.041, over 984906.64 frames.], batch size: 45, lr: 5.08e-04
+2022-06-18 21:49:12,446 INFO [train.py:874] (2/4) Epoch 16, batch 2650, datatang_loss[loss=0.1435, simple_loss=0.2235, pruned_loss=0.03172, over 4895.00 frames.], tot_loss[loss=0.157, simple_loss=0.2354, pruned_loss=0.03933, over 985596.83 frames.], batch size: 42, aishell_tot_loss[loss=0.1575, simple_loss=0.2412, pruned_loss=0.03691, over 983544.95 frames.], datatang_tot_loss[loss=0.1562, simple_loss=0.2297, pruned_loss=0.04138, over 984853.52 frames.], batch size: 42, lr: 5.08e-04
+2022-06-18 21:49:42,475 INFO [train.py:874] (2/4) Epoch 16, batch 2700, aishell_loss[loss=0.1603, simple_loss=0.2583, pruned_loss=0.03119, over 4944.00 frames.], tot_loss[loss=0.1574, simple_loss=0.2364, pruned_loss=0.03924, over 986076.05 frames.], batch size: 58, aishell_tot_loss[loss=0.1581, simple_loss=0.242, pruned_loss=0.03711, over 984187.12 frames.], datatang_tot_loss[loss=0.156, simple_loss=0.2296, pruned_loss=0.04123, over 985078.47 frames.], batch size: 58, lr: 5.07e-04
+2022-06-18 21:50:12,463 INFO [train.py:874] (2/4) Epoch 16, batch 2750, datatang_loss[loss=0.1565, simple_loss=0.2375, pruned_loss=0.03775, over 4931.00 frames.], tot_loss[loss=0.1591, simple_loss=0.2378, pruned_loss=0.04019, over 986082.36 frames.], batch size: 94, aishell_tot_loss[loss=0.1592, simple_loss=0.2429, pruned_loss=0.0377, over 984292.92 frames.], datatang_tot_loss[loss=0.1568, simple_loss=0.2304, pruned_loss=0.04162, over 985295.86 frames.], batch size: 94, lr: 5.07e-04
+2022-06-18 21:50:43,000 INFO [train.py:874] (2/4) Epoch 16, batch 2800, datatang_loss[loss=0.1346, simple_loss=0.2047, pruned_loss=0.03227, over 4942.00 frames.], tot_loss[loss=0.1584, simple_loss=0.2374, pruned_loss=0.03975, over 986152.41 frames.], batch size: 50, aishell_tot_loss[loss=0.1595, simple_loss=0.2434, pruned_loss=0.03781, over 984494.69 frames.], datatang_tot_loss[loss=0.1561, simple_loss=0.23, pruned_loss=0.04112, over 985449.70 frames.], batch size: 50, lr: 5.07e-04
+2022-06-18 21:51:13,686 INFO [train.py:874] (2/4) Epoch 16, batch 2850, aishell_loss[loss=0.1416, simple_loss=0.2257, pruned_loss=0.02868, over 4823.00 frames.], tot_loss[loss=0.1579, simple_loss=0.2365, pruned_loss=0.03965, over 985766.30 frames.], batch size: 29, aishell_tot_loss[loss=0.1592, simple_loss=0.2428, pruned_loss=0.0378, over 984292.06 frames.], datatang_tot_loss[loss=0.156, simple_loss=0.2299, pruned_loss=0.04107, over 985518.47 frames.], batch size: 29, lr: 5.07e-04
+2022-06-18 21:51:43,105 INFO [train.py:874] (2/4) Epoch 16, batch 2900, datatang_loss[loss=0.1785, simple_loss=0.2435, pruned_loss=0.05674, over 4915.00 frames.], tot_loss[loss=0.1584, simple_loss=0.2374, pruned_loss=0.03969, over 985729.25 frames.], batch size: 42, aishell_tot_loss[loss=0.1595, simple_loss=0.2433, pruned_loss=0.03779, over 984351.82 frames.], datatang_tot_loss[loss=0.1563, simple_loss=0.2303, pruned_loss=0.04119, over 985631.88 frames.], batch size: 42, lr: 5.07e-04
+2022-06-18 21:52:12,469 INFO [train.py:874] (2/4) Epoch 16, batch 2950, datatang_loss[loss=0.1214, simple_loss=0.2015, pruned_loss=0.02066, over 4834.00 frames.], tot_loss[loss=0.1575, simple_loss=0.2363, pruned_loss=0.03934, over 985394.35 frames.], batch size: 30, aishell_tot_loss[loss=0.159, simple_loss=0.2431, pruned_loss=0.03745, over 984246.34 frames.], datatang_tot_loss[loss=0.156, simple_loss=0.2298, pruned_loss=0.04108, over 985513.67 frames.], batch size: 30, lr: 5.06e-04
+2022-06-18 21:52:44,018 INFO [train.py:874] (2/4) Epoch 16, batch 3000, aishell_loss[loss=0.1524, simple_loss=0.2372, pruned_loss=0.03383, over 4891.00 frames.], tot_loss[loss=0.157, simple_loss=0.2361, pruned_loss=0.03893, over 985435.53 frames.], batch size: 34, aishell_tot_loss[loss=0.1589, simple_loss=0.2433, pruned_loss=0.03726, over 984534.79 frames.], datatang_tot_loss[loss=0.1555, simple_loss=0.2293, pruned_loss=0.0409, over 985391.98 frames.], batch size: 34, lr: 5.06e-04
+2022-06-18 21:52:44,019 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 21:53:00,708 INFO [train.py:914] (2/4) Epoch 16, validation: loss=0.1647, simple_loss=0.249, pruned_loss=0.04025, over 1622729.00 frames.
+2022-06-18 21:53:29,644 INFO [train.py:874] (2/4) Epoch 16, batch 3050, datatang_loss[loss=0.1437, simple_loss=0.2186, pruned_loss=0.03438, over 4924.00 frames.], tot_loss[loss=0.1571, simple_loss=0.2369, pruned_loss=0.03865, over 985430.51 frames.], batch size: 71, aishell_tot_loss[loss=0.1589, simple_loss=0.2435, pruned_loss=0.03716, over 984656.42 frames.], datatang_tot_loss[loss=0.1554, simple_loss=0.2293, pruned_loss=0.04077, over 985405.32 frames.], batch size: 71, lr: 5.06e-04
+2022-06-18 21:53:59,446 INFO [train.py:874] (2/4) Epoch 16, batch 3100, datatang_loss[loss=0.1512, simple_loss=0.2288, pruned_loss=0.03677, over 4926.00 frames.], tot_loss[loss=0.1577, simple_loss=0.2376, pruned_loss=0.03894, over 985262.40 frames.], batch size: 79, aishell_tot_loss[loss=0.1587, simple_loss=0.2433, pruned_loss=0.03702, over 984652.67 frames.], datatang_tot_loss[loss=0.1563, simple_loss=0.2301, pruned_loss=0.04125, over 985341.01 frames.], batch size: 79, lr: 5.06e-04
+2022-06-18 21:54:29,143 INFO [train.py:874] (2/4) Epoch 16, batch 3150, aishell_loss[loss=0.119, simple_loss=0.2086, pruned_loss=0.01468, over 4979.00 frames.], tot_loss[loss=0.1585, simple_loss=0.2382, pruned_loss=0.03942, over 985302.08 frames.], batch size: 30, aishell_tot_loss[loss=0.1593, simple_loss=0.2437, pruned_loss=0.03745, over 984631.65 frames.], datatang_tot_loss[loss=0.1565, simple_loss=0.2305, pruned_loss=0.04129, over 985465.66 frames.], batch size: 30, lr: 5.06e-04
+2022-06-18 21:55:00,819 INFO [train.py:874] (2/4) Epoch 16, batch 3200, aishell_loss[loss=0.1314, simple_loss=0.2141, pruned_loss=0.02438, over 4952.00 frames.], tot_loss[loss=0.1579, simple_loss=0.2369, pruned_loss=0.03948, over 985367.10 frames.], batch size: 27, aishell_tot_loss[loss=0.1589, simple_loss=0.2428, pruned_loss=0.03744, over 984802.84 frames.], datatang_tot_loss[loss=0.1565, simple_loss=0.2301, pruned_loss=0.04143, over 985439.29 frames.], batch size: 27, lr: 5.05e-04
+2022-06-18 21:55:30,448 INFO [train.py:874] (2/4) Epoch 16, batch 3250, datatang_loss[loss=0.1426, simple_loss=0.2149, pruned_loss=0.03521, over 4891.00 frames.], tot_loss[loss=0.1583, simple_loss=0.2375, pruned_loss=0.03955, over 985603.44 frames.], batch size: 52, aishell_tot_loss[loss=0.1592, simple_loss=0.2436, pruned_loss=0.03746, over 984904.85 frames.], datatang_tot_loss[loss=0.1566, simple_loss=0.2304, pruned_loss=0.04141, over 985632.25 frames.], batch size: 52, lr: 5.05e-04
+2022-06-18 21:55:59,844 INFO [train.py:874] (2/4) Epoch 16, batch 3300, datatang_loss[loss=0.1335, simple_loss=0.213, pruned_loss=0.02704, over 4938.00 frames.], tot_loss[loss=0.1582, simple_loss=0.2372, pruned_loss=0.03956, over 985556.21 frames.], batch size: 69, aishell_tot_loss[loss=0.1595, simple_loss=0.244, pruned_loss=0.03754, over 984905.88 frames.], datatang_tot_loss[loss=0.1563, simple_loss=0.23, pruned_loss=0.04129, over 985638.61 frames.], batch size: 69, lr: 5.05e-04
+2022-06-18 21:56:30,787 INFO [train.py:874] (2/4) Epoch 16, batch 3350, datatang_loss[loss=0.1826, simple_loss=0.2558, pruned_loss=0.05468, over 4930.00 frames.], tot_loss[loss=0.1579, simple_loss=0.2376, pruned_loss=0.0391, over 985636.91 frames.], batch size: 94, aishell_tot_loss[loss=0.1592, simple_loss=0.2438, pruned_loss=0.03728, over 984919.01 frames.], datatang_tot_loss[loss=0.1563, simple_loss=0.2302, pruned_loss=0.04121, over 985803.05 frames.], batch size: 94, lr: 5.05e-04
+2022-06-18 21:57:00,628 INFO [train.py:874] (2/4) Epoch 16, batch 3400, aishell_loss[loss=0.1648, simple_loss=0.2488, pruned_loss=0.04038, over 4856.00 frames.], tot_loss[loss=0.1571, simple_loss=0.2366, pruned_loss=0.03883, over 985570.26 frames.], batch size: 35, aishell_tot_loss[loss=0.1591, simple_loss=0.2436, pruned_loss=0.03728, over 984895.74 frames.], datatang_tot_loss[loss=0.1557, simple_loss=0.2297, pruned_loss=0.04085, over 985811.12 frames.], batch size: 35, lr: 5.05e-04
+2022-06-18 21:57:30,359 INFO [train.py:874] (2/4) Epoch 16, batch 3450, aishell_loss[loss=0.1203, simple_loss=0.202, pruned_loss=0.01926, over 4878.00 frames.], tot_loss[loss=0.1561, simple_loss=0.2356, pruned_loss=0.03828, over 985706.98 frames.], batch size: 28, aishell_tot_loss[loss=0.1585, simple_loss=0.2428, pruned_loss=0.03708, over 984999.43 frames.], datatang_tot_loss[loss=0.155, simple_loss=0.229, pruned_loss=0.04049, over 985937.96 frames.], batch size: 28, lr: 5.05e-04
+2022-06-18 21:58:01,217 INFO [train.py:874] (2/4) Epoch 16, batch 3500, aishell_loss[loss=0.1715, simple_loss=0.2504, pruned_loss=0.04628, over 4966.00 frames.], tot_loss[loss=0.1558, simple_loss=0.2355, pruned_loss=0.03808, over 985373.08 frames.], batch size: 44, aishell_tot_loss[loss=0.1582, simple_loss=0.2427, pruned_loss=0.03685, over 984780.21 frames.], datatang_tot_loss[loss=0.155, simple_loss=0.2291, pruned_loss=0.0404, over 985860.36 frames.], batch size: 44, lr: 5.04e-04
+2022-06-18 21:58:30,715 INFO [train.py:874] (2/4) Epoch 16, batch 3550, aishell_loss[loss=0.1583, simple_loss=0.2446, pruned_loss=0.03605, over 4956.00 frames.], tot_loss[loss=0.1546, simple_loss=0.2343, pruned_loss=0.0374, over 985996.84 frames.], batch size: 40, aishell_tot_loss[loss=0.1573, simple_loss=0.2418, pruned_loss=0.03634, over 985262.42 frames.], datatang_tot_loss[loss=0.1543, simple_loss=0.2285, pruned_loss=0.04009, over 986055.84 frames.], batch size: 40, lr: 5.04e-04
+2022-06-18 21:59:01,870 INFO [train.py:874] (2/4) Epoch 16, batch 3600, datatang_loss[loss=0.145, simple_loss=0.2201, pruned_loss=0.03498, over 4945.00 frames.], tot_loss[loss=0.155, simple_loss=0.2347, pruned_loss=0.03765, over 986307.35 frames.], batch size: 37, aishell_tot_loss[loss=0.1578, simple_loss=0.2423, pruned_loss=0.03663, over 985615.12 frames.], datatang_tot_loss[loss=0.154, simple_loss=0.2282, pruned_loss=0.03992, over 986119.58 frames.], batch size: 37, lr: 5.04e-04
+2022-06-18 21:59:31,582 INFO [train.py:874] (2/4) Epoch 16, batch 3650, aishell_loss[loss=0.1514, simple_loss=0.2397, pruned_loss=0.03155, over 4977.00 frames.], tot_loss[loss=0.1549, simple_loss=0.2351, pruned_loss=0.03732, over 986197.79 frames.], batch size: 51, aishell_tot_loss[loss=0.1573, simple_loss=0.2421, pruned_loss=0.03626, over 985627.22 frames.], datatang_tot_loss[loss=0.1541, simple_loss=0.2285, pruned_loss=0.03983, over 986106.60 frames.], batch size: 51, lr: 5.04e-04
+2022-06-18 22:00:02,454 INFO [train.py:874] (2/4) Epoch 16, batch 3700, datatang_loss[loss=0.1655, simple_loss=0.2374, pruned_loss=0.04679, over 4965.00 frames.], tot_loss[loss=0.1558, simple_loss=0.2357, pruned_loss=0.03793, over 986046.26 frames.], batch size: 60, aishell_tot_loss[loss=0.1575, simple_loss=0.2422, pruned_loss=0.03641, over 985626.35 frames.], datatang_tot_loss[loss=0.1547, simple_loss=0.2292, pruned_loss=0.04009, over 986001.91 frames.], batch size: 60, lr: 5.04e-04
+2022-06-18 22:00:32,961 INFO [train.py:874] (2/4) Epoch 16, batch 3750, aishell_loss[loss=0.1588, simple_loss=0.2331, pruned_loss=0.04224, over 4936.00 frames.], tot_loss[loss=0.1549, simple_loss=0.2349, pruned_loss=0.03746, over 985762.30 frames.], batch size: 32, aishell_tot_loss[loss=0.1571, simple_loss=0.2418, pruned_loss=0.03618, over 985379.00 frames.], datatang_tot_loss[loss=0.1541, simple_loss=0.2288, pruned_loss=0.03976, over 986010.84 frames.], batch size: 32, lr: 5.03e-04
+2022-06-18 22:01:02,945 INFO [train.py:874] (2/4) Epoch 16, batch 3800, datatang_loss[loss=0.1458, simple_loss=0.2196, pruned_loss=0.036, over 4914.00 frames.], tot_loss[loss=0.1549, simple_loss=0.2346, pruned_loss=0.03761, over 986062.50 frames.], batch size: 57, aishell_tot_loss[loss=0.1569, simple_loss=0.2416, pruned_loss=0.03605, over 985703.75 frames.], datatang_tot_loss[loss=0.1542, simple_loss=0.2286, pruned_loss=0.0399, over 986039.09 frames.], batch size: 57, lr: 5.03e-04
+2022-06-18 22:01:31,720 INFO [train.py:874] (2/4) Epoch 16, batch 3850, datatang_loss[loss=0.1684, simple_loss=0.2384, pruned_loss=0.04922, over 4925.00 frames.], tot_loss[loss=0.1555, simple_loss=0.2353, pruned_loss=0.03779, over 985888.92 frames.], batch size: 57, aishell_tot_loss[loss=0.1568, simple_loss=0.2418, pruned_loss=0.03596, over 985521.41 frames.], datatang_tot_loss[loss=0.1546, simple_loss=0.2291, pruned_loss=0.04007, over 986071.50 frames.], batch size: 57, lr: 5.03e-04
+2022-06-18 22:01:59,925 INFO [train.py:874] (2/4) Epoch 16, batch 3900, datatang_loss[loss=0.1464, simple_loss=0.2195, pruned_loss=0.03665, over 4945.00 frames.], tot_loss[loss=0.1562, simple_loss=0.2364, pruned_loss=0.03806, over 986025.88 frames.], batch size: 69, aishell_tot_loss[loss=0.1566, simple_loss=0.2421, pruned_loss=0.03558, over 985566.13 frames.], datatang_tot_loss[loss=0.1556, simple_loss=0.2298, pruned_loss=0.04066, over 986189.01 frames.], batch size: 69, lr: 5.03e-04
+2022-06-18 22:02:29,032 INFO [train.py:874] (2/4) Epoch 16, batch 3950, aishell_loss[loss=0.1337, simple_loss=0.2076, pruned_loss=0.02991, over 4826.00 frames.], tot_loss[loss=0.1557, simple_loss=0.2358, pruned_loss=0.0378, over 985688.81 frames.], batch size: 29, aishell_tot_loss[loss=0.1563, simple_loss=0.2415, pruned_loss=0.03558, over 985318.88 frames.], datatang_tot_loss[loss=0.1553, simple_loss=0.2299, pruned_loss=0.0404, over 986112.12 frames.], batch size: 29, lr: 5.03e-04
+2022-06-18 22:02:58,082 INFO [train.py:874] (2/4) Epoch 16, batch 4000, aishell_loss[loss=0.1563, simple_loss=0.2494, pruned_loss=0.03161, over 4968.00 frames.], tot_loss[loss=0.1554, simple_loss=0.2355, pruned_loss=0.0377, over 985520.34 frames.], batch size: 51, aishell_tot_loss[loss=0.1565, simple_loss=0.2415, pruned_loss=0.03578, over 985330.14 frames.], datatang_tot_loss[loss=0.1548, simple_loss=0.2295, pruned_loss=0.04009, over 985913.22 frames.], batch size: 51, lr: 5.02e-04
+2022-06-18 22:02:58,083 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 22:03:14,113 INFO [train.py:914] (2/4) Epoch 16, validation: loss=0.164, simple_loss=0.2482, pruned_loss=0.03992, over 1622729.00 frames.
+2022-06-18 22:03:42,689 INFO [train.py:874] (2/4) Epoch 16, batch 4050, datatang_loss[loss=0.1958, simple_loss=0.2636, pruned_loss=0.06398, over 4941.00 frames.], tot_loss[loss=0.155, simple_loss=0.2348, pruned_loss=0.03754, over 985375.20 frames.], batch size: 94, aishell_tot_loss[loss=0.1565, simple_loss=0.2415, pruned_loss=0.03579, over 985120.49 frames.], datatang_tot_loss[loss=0.1543, simple_loss=0.2291, pruned_loss=0.03974, over 985916.06 frames.], batch size: 94, lr: 5.02e-04
+2022-06-18 22:04:12,004 INFO [train.py:874] (2/4) Epoch 16, batch 4100, aishell_loss[loss=0.1899, simple_loss=0.2695, pruned_loss=0.05518, over 4916.00 frames.], tot_loss[loss=0.1559, simple_loss=0.2355, pruned_loss=0.03808, over 985237.51 frames.], batch size: 41, aishell_tot_loss[loss=0.1569, simple_loss=0.2417, pruned_loss=0.03602, over 984805.58 frames.], datatang_tot_loss[loss=0.1548, simple_loss=0.2295, pruned_loss=0.04004, over 986035.08 frames.], batch size: 41, lr: 5.02e-04
+2022-06-18 22:05:35,075 INFO [train.py:874] (2/4) Epoch 17, batch 50, datatang_loss[loss=0.1249, simple_loss=0.2021, pruned_loss=0.0238, over 4929.00 frames.], tot_loss[loss=0.1528, simple_loss=0.2338, pruned_loss=0.03587, over 218573.13 frames.], batch size: 71, aishell_tot_loss[loss=0.1604, simple_loss=0.2463, pruned_loss=0.03724, over 129162.93 frames.], datatang_tot_loss[loss=0.1433, simple_loss=0.2182, pruned_loss=0.03415, over 102876.78 frames.], batch size: 71, lr: 4.88e-04
+2022-06-18 22:06:05,301 INFO [train.py:874] (2/4) Epoch 17, batch 100, aishell_loss[loss=0.1579, simple_loss=0.2391, pruned_loss=0.03835, over 4937.00 frames.], tot_loss[loss=0.152, simple_loss=0.2326, pruned_loss=0.03568, over 388592.92 frames.], batch size: 58, aishell_tot_loss[loss=0.1598, simple_loss=0.2459, pruned_loss=0.03682, over 222387.61 frames.], datatang_tot_loss[loss=0.1439, simple_loss=0.2189, pruned_loss=0.03448, over 214626.55 frames.], batch size: 58, lr: 4.88e-04
+2022-06-18 22:06:34,674 INFO [train.py:874] (2/4) Epoch 17, batch 150, datatang_loss[loss=0.1264, simple_loss=0.1966, pruned_loss=0.02814, over 4884.00 frames.], tot_loss[loss=0.15, simple_loss=0.2304, pruned_loss=0.03482, over 520678.21 frames.], batch size: 24, aishell_tot_loss[loss=0.1586, simple_loss=0.2447, pruned_loss=0.03625, over 301858.28 frames.], datatang_tot_loss[loss=0.1424, simple_loss=0.2176, pruned_loss=0.03363, over 315533.66 frames.], batch size: 24, lr: 4.87e-04
+2022-06-18 22:07:06,317 INFO [train.py:874] (2/4) Epoch 17, batch 200, datatang_loss[loss=0.1379, simple_loss=0.2144, pruned_loss=0.03069, over 4965.00 frames.], tot_loss[loss=0.1505, simple_loss=0.2305, pruned_loss=0.0353, over 623907.16 frames.], batch size: 37, aishell_tot_loss[loss=0.1581, simple_loss=0.2435, pruned_loss=0.03638, over 379487.01 frames.], datatang_tot_loss[loss=0.1437, simple_loss=0.2188, pruned_loss=0.03427, over 397435.21 frames.], batch size: 37, lr: 4.87e-04
+2022-06-18 22:07:35,620 INFO [train.py:874] (2/4) Epoch 17, batch 250, datatang_loss[loss=0.1619, simple_loss=0.2358, pruned_loss=0.04404, over 4894.00 frames.], tot_loss[loss=0.151, simple_loss=0.2306, pruned_loss=0.03574, over 703564.19 frames.], batch size: 47, aishell_tot_loss[loss=0.158, simple_loss=0.2429, pruned_loss=0.03653, over 447820.38 frames.], datatang_tot_loss[loss=0.1446, simple_loss=0.2195, pruned_loss=0.03483, over 469110.86 frames.], batch size: 47, lr: 4.87e-04
+2022-06-18 22:08:05,779 INFO [train.py:874] (2/4) Epoch 17, batch 300, aishell_loss[loss=0.1516, simple_loss=0.2311, pruned_loss=0.03608, over 4950.00 frames.], tot_loss[loss=0.1528, simple_loss=0.2322, pruned_loss=0.03665, over 766339.39 frames.], batch size: 31, aishell_tot_loss[loss=0.1582, simple_loss=0.243, pruned_loss=0.03671, over 509029.65 frames.], datatang_tot_loss[loss=0.1469, simple_loss=0.2218, pruned_loss=0.03605, over 532233.23 frames.], batch size: 31, lr: 4.87e-04
+2022-06-18 22:08:36,967 INFO [train.py:874] (2/4) Epoch 17, batch 350, aishell_loss[loss=0.1252, simple_loss=0.208, pruned_loss=0.02126, over 4976.00 frames.], tot_loss[loss=0.1529, simple_loss=0.2325, pruned_loss=0.03666, over 815129.30 frames.], batch size: 25, aishell_tot_loss[loss=0.1575, simple_loss=0.2423, pruned_loss=0.0363, over 563206.60 frames.], datatang_tot_loss[loss=0.148, simple_loss=0.2229, pruned_loss=0.03656, over 587673.27 frames.], batch size: 25, lr: 4.87e-04
+2022-06-18 22:09:07,067 INFO [train.py:874] (2/4) Epoch 17, batch 400, datatang_loss[loss=0.1613, simple_loss=0.2359, pruned_loss=0.0434, over 4955.00 frames.], tot_loss[loss=0.1541, simple_loss=0.2337, pruned_loss=0.03728, over 852913.72 frames.], batch size: 91, aishell_tot_loss[loss=0.1577, simple_loss=0.2424, pruned_loss=0.03647, over 618433.60 frames.], datatang_tot_loss[loss=0.1494, simple_loss=0.224, pruned_loss=0.03739, over 629303.33 frames.], batch size: 91, lr: 4.87e-04
+2022-06-18 22:09:36,431 INFO [train.py:874] (2/4) Epoch 17, batch 450, datatang_loss[loss=0.139, simple_loss=0.2135, pruned_loss=0.03222, over 4920.00 frames.], tot_loss[loss=0.1539, simple_loss=0.2333, pruned_loss=0.03727, over 882056.79 frames.], batch size: 34, aishell_tot_loss[loss=0.1572, simple_loss=0.2421, pruned_loss=0.03614, over 659546.38 frames.], datatang_tot_loss[loss=0.1499, simple_loss=0.2241, pruned_loss=0.03782, over 673049.09 frames.], batch size: 34, lr: 4.86e-04
+2022-06-18 22:10:07,605 INFO [train.py:874] (2/4) Epoch 17, batch 500, datatang_loss[loss=0.1397, simple_loss=0.2251, pruned_loss=0.02719, over 4926.00 frames.], tot_loss[loss=0.1537, simple_loss=0.2331, pruned_loss=0.03714, over 904761.75 frames.], batch size: 83, aishell_tot_loss[loss=0.1564, simple_loss=0.2413, pruned_loss=0.03578, over 701829.86 frames.], datatang_tot_loss[loss=0.1503, simple_loss=0.2244, pruned_loss=0.03811, over 705838.99 frames.], batch size: 83, lr: 4.86e-04
+2022-06-18 22:10:36,764 INFO [train.py:874] (2/4) Epoch 17, batch 550, aishell_loss[loss=0.1553, simple_loss=0.2415, pruned_loss=0.03453, over 4893.00 frames.], tot_loss[loss=0.1546, simple_loss=0.2339, pruned_loss=0.03764, over 922662.61 frames.], batch size: 50, aishell_tot_loss[loss=0.1561, simple_loss=0.2411, pruned_loss=0.03555, over 728522.92 frames.], datatang_tot_loss[loss=0.1521, simple_loss=0.2262, pruned_loss=0.03896, over 745241.99 frames.], batch size: 50, lr: 4.86e-04
+2022-06-18 22:11:06,714 INFO [train.py:874] (2/4) Epoch 17, batch 600, datatang_loss[loss=0.1733, simple_loss=0.2442, pruned_loss=0.05119, over 4950.00 frames.], tot_loss[loss=0.1557, simple_loss=0.2345, pruned_loss=0.0384, over 936531.85 frames.], batch size: 86, aishell_tot_loss[loss=0.1562, simple_loss=0.2415, pruned_loss=0.03546, over 750769.91 frames.], datatang_tot_loss[loss=0.1537, simple_loss=0.2274, pruned_loss=0.03998, over 780712.77 frames.], batch size: 86, lr: 4.86e-04
+2022-06-18 22:11:38,225 INFO [train.py:874] (2/4) Epoch 17, batch 650, aishell_loss[loss=0.1696, simple_loss=0.2591, pruned_loss=0.0401, over 4916.00 frames.], tot_loss[loss=0.1565, simple_loss=0.2359, pruned_loss=0.03855, over 947462.20 frames.], batch size: 52, aishell_tot_loss[loss=0.1566, simple_loss=0.2422, pruned_loss=0.03551, over 780545.79 frames.], datatang_tot_loss[loss=0.1544, simple_loss=0.2282, pruned_loss=0.04033, over 803036.17 frames.], batch size: 52, lr: 4.86e-04
+2022-06-18 22:12:08,227 INFO [train.py:874] (2/4) Epoch 17, batch 700, datatang_loss[loss=0.1541, simple_loss=0.226, pruned_loss=0.04107, over 4944.00 frames.], tot_loss[loss=0.1558, simple_loss=0.235, pruned_loss=0.03826, over 955726.05 frames.], batch size: 55, aishell_tot_loss[loss=0.1557, simple_loss=0.2408, pruned_loss=0.03527, over 805253.60 frames.], datatang_tot_loss[loss=0.1547, simple_loss=0.2286, pruned_loss=0.04042, over 823868.87 frames.], batch size: 55, lr: 4.86e-04
+2022-06-18 22:12:37,685 INFO [train.py:874] (2/4) Epoch 17, batch 750, aishell_loss[loss=0.1327, simple_loss=0.2, pruned_loss=0.03272, over 4805.00 frames.], tot_loss[loss=0.1558, simple_loss=0.2351, pruned_loss=0.03823, over 962452.63 frames.], batch size: 21, aishell_tot_loss[loss=0.1562, simple_loss=0.2414, pruned_loss=0.03552, over 827273.86 frames.], datatang_tot_loss[loss=0.1544, simple_loss=0.2281, pruned_loss=0.04031, over 842343.18 frames.], batch size: 21, lr: 4.85e-04
+2022-06-18 22:13:08,741 INFO [train.py:874] (2/4) Epoch 17, batch 800, aishell_loss[loss=0.1517, simple_loss=0.2374, pruned_loss=0.033, over 4934.00 frames.], tot_loss[loss=0.1547, simple_loss=0.2342, pruned_loss=0.03762, over 967539.24 frames.], batch size: 32, aishell_tot_loss[loss=0.1559, simple_loss=0.2411, pruned_loss=0.03535, over 846578.97 frames.], datatang_tot_loss[loss=0.1535, simple_loss=0.2273, pruned_loss=0.03987, over 858566.23 frames.], batch size: 32, lr: 4.85e-04
+2022-06-18 22:13:38,513 INFO [train.py:874] (2/4) Epoch 17, batch 850, aishell_loss[loss=0.1455, simple_loss=0.2389, pruned_loss=0.02609, over 4887.00 frames.], tot_loss[loss=0.1549, simple_loss=0.2347, pruned_loss=0.03753, over 971624.95 frames.], batch size: 42, aishell_tot_loss[loss=0.1562, simple_loss=0.2416, pruned_loss=0.03543, over 863664.05 frames.], datatang_tot_loss[loss=0.1534, simple_loss=0.2273, pruned_loss=0.0397, over 872963.95 frames.], batch size: 42, lr: 4.85e-04
+2022-06-18 22:14:08,882 INFO [train.py:874] (2/4) Epoch 17, batch 900, datatang_loss[loss=0.1531, simple_loss=0.2353, pruned_loss=0.03545, over 4910.00 frames.], tot_loss[loss=0.1553, simple_loss=0.2349, pruned_loss=0.03783, over 974796.22 frames.], batch size: 75, aishell_tot_loss[loss=0.1566, simple_loss=0.2418, pruned_loss=0.0357, over 878540.59 frames.], datatang_tot_loss[loss=0.1535, simple_loss=0.2274, pruned_loss=0.03981, over 885831.45 frames.], batch size: 75, lr: 4.85e-04
+2022-06-18 22:14:39,883 INFO [train.py:874] (2/4) Epoch 17, batch 950, datatang_loss[loss=0.166, simple_loss=0.2464, pruned_loss=0.04281, over 4923.00 frames.], tot_loss[loss=0.156, simple_loss=0.2354, pruned_loss=0.03833, over 977068.82 frames.], batch size: 64, aishell_tot_loss[loss=0.1571, simple_loss=0.2422, pruned_loss=0.03607, over 890446.95 frames.], datatang_tot_loss[loss=0.1539, simple_loss=0.2277, pruned_loss=0.04003, over 898143.09 frames.], batch size: 64, lr: 4.85e-04
+2022-06-18 22:15:10,131 INFO [train.py:874] (2/4) Epoch 17, batch 1000, datatang_loss[loss=0.1461, simple_loss=0.2251, pruned_loss=0.03352, over 4902.00 frames.], tot_loss[loss=0.1555, simple_loss=0.2351, pruned_loss=0.03796, over 978567.14 frames.], batch size: 34, aishell_tot_loss[loss=0.1569, simple_loss=0.2421, pruned_loss=0.03588, over 900829.35 frames.], datatang_tot_loss[loss=0.1537, simple_loss=0.2277, pruned_loss=0.03986, over 908816.39 frames.], batch size: 34, lr: 4.84e-04
+2022-06-18 22:15:10,132 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 22:15:27,676 INFO [train.py:914] (2/4) Epoch 17, validation: loss=0.1643, simple_loss=0.2484, pruned_loss=0.04015, over 1622729.00 frames.
+2022-06-18 22:15:57,376 INFO [train.py:874] (2/4) Epoch 17, batch 1050, aishell_loss[loss=0.1691, simple_loss=0.2471, pruned_loss=0.04557, over 4942.00 frames.], tot_loss[loss=0.1558, simple_loss=0.2359, pruned_loss=0.03782, over 980326.47 frames.], batch size: 45, aishell_tot_loss[loss=0.157, simple_loss=0.2424, pruned_loss=0.03583, over 913973.89 frames.], datatang_tot_loss[loss=0.1538, simple_loss=0.2277, pruned_loss=0.03994, over 915064.76 frames.], batch size: 45, lr: 4.84e-04
+2022-06-18 22:16:27,407 INFO [train.py:874] (2/4) Epoch 17, batch 1100, aishell_loss[loss=0.1442, simple_loss=0.2226, pruned_loss=0.03293, over 4867.00 frames.], tot_loss[loss=0.156, simple_loss=0.2362, pruned_loss=0.03792, over 981561.14 frames.], batch size: 36, aishell_tot_loss[loss=0.1572, simple_loss=0.2424, pruned_loss=0.03599, over 923608.74 frames.], datatang_tot_loss[loss=0.1539, simple_loss=0.2279, pruned_loss=0.03995, over 922251.20 frames.], batch size: 36, lr: 4.84e-04
+2022-06-18 22:16:58,178 INFO [train.py:874] (2/4) Epoch 17, batch 1150, datatang_loss[loss=0.1503, simple_loss=0.2266, pruned_loss=0.037, over 4865.00 frames.], tot_loss[loss=0.1558, simple_loss=0.2357, pruned_loss=0.038, over 982453.84 frames.], batch size: 39, aishell_tot_loss[loss=0.1569, simple_loss=0.2422, pruned_loss=0.03584, over 928780.75 frames.], datatang_tot_loss[loss=0.1542, simple_loss=0.2284, pruned_loss=0.04001, over 931805.79 frames.], batch size: 39, lr: 4.84e-04
+2022-06-18 22:17:27,460 INFO [train.py:874] (2/4) Epoch 17, batch 1200, aishell_loss[loss=0.1285, simple_loss=0.2031, pruned_loss=0.02697, over 4935.00 frames.], tot_loss[loss=0.155, simple_loss=0.2348, pruned_loss=0.03757, over 983190.16 frames.], batch size: 27, aishell_tot_loss[loss=0.1558, simple_loss=0.241, pruned_loss=0.03534, over 936596.28 frames.], datatang_tot_loss[loss=0.1544, simple_loss=0.2284, pruned_loss=0.04018, over 937079.85 frames.], batch size: 27, lr: 4.84e-04
+2022-06-18 22:17:58,100 INFO [train.py:874] (2/4) Epoch 17, batch 1250, datatang_loss[loss=0.1713, simple_loss=0.252, pruned_loss=0.04529, over 4955.00 frames.], tot_loss[loss=0.1558, simple_loss=0.2357, pruned_loss=0.03796, over 983856.93 frames.], batch size: 34, aishell_tot_loss[loss=0.1565, simple_loss=0.2417, pruned_loss=0.03563, over 941876.56 frames.], datatang_tot_loss[loss=0.1547, simple_loss=0.2289, pruned_loss=0.04019, over 943402.96 frames.], batch size: 34, lr: 4.84e-04
+2022-06-18 22:18:27,949 INFO [train.py:874] (2/4) Epoch 17, batch 1300, aishell_loss[loss=0.1683, simple_loss=0.2604, pruned_loss=0.03811, over 4936.00 frames.], tot_loss[loss=0.1556, simple_loss=0.2356, pruned_loss=0.03786, over 984339.35 frames.], batch size: 45, aishell_tot_loss[loss=0.1567, simple_loss=0.2419, pruned_loss=0.03574, over 947341.56 frames.], datatang_tot_loss[loss=0.1543, simple_loss=0.2286, pruned_loss=0.04, over 948151.36 frames.], batch size: 45, lr: 4.83e-04
+2022-06-18 22:18:58,724 INFO [train.py:874] (2/4) Epoch 17, batch 1350, aishell_loss[loss=0.1525, simple_loss=0.243, pruned_loss=0.03097, over 4898.00 frames.], tot_loss[loss=0.1552, simple_loss=0.235, pruned_loss=0.03769, over 984499.17 frames.], batch size: 60, aishell_tot_loss[loss=0.1567, simple_loss=0.2417, pruned_loss=0.03585, over 952094.96 frames.], datatang_tot_loss[loss=0.1539, simple_loss=0.2282, pruned_loss=0.03978, over 952215.92 frames.], batch size: 60, lr: 4.83e-04
+2022-06-18 22:19:29,795 INFO [train.py:874] (2/4) Epoch 17, batch 1400, aishell_loss[loss=0.1481, simple_loss=0.2385, pruned_loss=0.02887, over 4940.00 frames.], tot_loss[loss=0.1552, simple_loss=0.2349, pruned_loss=0.03773, over 984625.65 frames.], batch size: 54, aishell_tot_loss[loss=0.1562, simple_loss=0.2411, pruned_loss=0.03564, over 955570.03 frames.], datatang_tot_loss[loss=0.1543, simple_loss=0.2287, pruned_loss=0.04, over 956477.75 frames.], batch size: 54, lr: 4.83e-04
+2022-06-18 22:19:58,836 INFO [train.py:874] (2/4) Epoch 17, batch 1450, aishell_loss[loss=0.1938, simple_loss=0.2639, pruned_loss=0.06182, over 4901.00 frames.], tot_loss[loss=0.155, simple_loss=0.2344, pruned_loss=0.03775, over 984954.71 frames.], batch size: 34, aishell_tot_loss[loss=0.156, simple_loss=0.2408, pruned_loss=0.03561, over 959090.12 frames.], datatang_tot_loss[loss=0.1543, simple_loss=0.2285, pruned_loss=0.04003, over 960046.91 frames.], batch size: 34, lr: 4.83e-04
+2022-06-18 22:20:29,713 INFO [train.py:874] (2/4) Epoch 17, batch 1500, datatang_loss[loss=0.2344, simple_loss=0.2942, pruned_loss=0.08724, over 4954.00 frames.], tot_loss[loss=0.1553, simple_loss=0.2343, pruned_loss=0.03814, over 985147.63 frames.], batch size: 109, aishell_tot_loss[loss=0.1556, simple_loss=0.2403, pruned_loss=0.03549, over 961628.31 frames.], datatang_tot_loss[loss=0.155, simple_loss=0.2291, pruned_loss=0.04046, over 963681.13 frames.], batch size: 109, lr: 4.83e-04
+2022-06-18 22:20:59,674 INFO [train.py:874] (2/4) Epoch 17, batch 1550, datatang_loss[loss=0.1677, simple_loss=0.244, pruned_loss=0.04573, over 4956.00 frames.], tot_loss[loss=0.1552, simple_loss=0.2341, pruned_loss=0.0381, over 985173.74 frames.], batch size: 91, aishell_tot_loss[loss=0.1556, simple_loss=0.2401, pruned_loss=0.0355, over 964694.50 frames.], datatang_tot_loss[loss=0.1549, simple_loss=0.2289, pruned_loss=0.04051, over 965968.18 frames.], batch size: 91, lr: 4.82e-04
+2022-06-18 22:21:29,216 INFO [train.py:874] (2/4) Epoch 17, batch 1600, aishell_loss[loss=0.1525, simple_loss=0.2398, pruned_loss=0.03259, over 4862.00 frames.], tot_loss[loss=0.1546, simple_loss=0.2336, pruned_loss=0.03779, over 985305.95 frames.], batch size: 37, aishell_tot_loss[loss=0.1554, simple_loss=0.2401, pruned_loss=0.0354, over 966725.40 frames.], datatang_tot_loss[loss=0.1544, simple_loss=0.2285, pruned_loss=0.04016, over 968733.36 frames.], batch size: 37, lr: 4.82e-04
+2022-06-18 22:21:59,598 INFO [train.py:874] (2/4) Epoch 17, batch 1650, aishell_loss[loss=0.1607, simple_loss=0.2449, pruned_loss=0.03828, over 4949.00 frames.], tot_loss[loss=0.1544, simple_loss=0.2339, pruned_loss=0.03747, over 984872.71 frames.], batch size: 49, aishell_tot_loss[loss=0.1552, simple_loss=0.2399, pruned_loss=0.03525, over 968964.83 frames.], datatang_tot_loss[loss=0.1544, simple_loss=0.2288, pruned_loss=0.03999, over 970191.69 frames.], batch size: 49, lr: 4.82e-04
+2022-06-18 22:22:30,208 INFO [train.py:874] (2/4) Epoch 17, batch 1700, datatang_loss[loss=0.1589, simple_loss=0.2379, pruned_loss=0.03994, over 4951.00 frames.], tot_loss[loss=0.155, simple_loss=0.2348, pruned_loss=0.03764, over 984898.83 frames.], batch size: 67, aishell_tot_loss[loss=0.1558, simple_loss=0.2406, pruned_loss=0.03552, over 970937.69 frames.], datatang_tot_loss[loss=0.1544, simple_loss=0.2289, pruned_loss=0.03993, over 971851.45 frames.], batch size: 67, lr: 4.82e-04
+2022-06-18 22:22:59,503 INFO [train.py:874] (2/4) Epoch 17, batch 1750, datatang_loss[loss=0.1416, simple_loss=0.2213, pruned_loss=0.03092, over 4962.00 frames.], tot_loss[loss=0.1543, simple_loss=0.2339, pruned_loss=0.03738, over 985149.14 frames.], batch size: 67, aishell_tot_loss[loss=0.1553, simple_loss=0.2403, pruned_loss=0.03519, over 972037.54 frames.], datatang_tot_loss[loss=0.1542, simple_loss=0.2288, pruned_loss=0.03978, over 974144.87 frames.], batch size: 67, lr: 4.82e-04
+2022-06-18 22:23:31,054 INFO [train.py:874] (2/4) Epoch 17, batch 1800, datatang_loss[loss=0.1459, simple_loss=0.2266, pruned_loss=0.03264, over 4948.00 frames.], tot_loss[loss=0.1535, simple_loss=0.2328, pruned_loss=0.03708, over 984918.13 frames.], batch size: 67, aishell_tot_loss[loss=0.1552, simple_loss=0.2399, pruned_loss=0.03522, over 973124.05 frames.], datatang_tot_loss[loss=0.1533, simple_loss=0.228, pruned_loss=0.03932, over 975596.12 frames.], batch size: 67, lr: 4.82e-04
+2022-06-18 22:24:01,625 INFO [train.py:874] (2/4) Epoch 17, batch 1850, aishell_loss[loss=0.1534, simple_loss=0.2304, pruned_loss=0.03817, over 4974.00 frames.], tot_loss[loss=0.1549, simple_loss=0.2342, pruned_loss=0.03778, over 985282.99 frames.], batch size: 31, aishell_tot_loss[loss=0.1559, simple_loss=0.2408, pruned_loss=0.03549, over 974652.81 frames.], datatang_tot_loss[loss=0.1539, simple_loss=0.2283, pruned_loss=0.03975, over 976934.70 frames.], batch size: 31, lr: 4.81e-04
+2022-06-18 22:24:35,560 INFO [train.py:874] (2/4) Epoch 17, batch 1900, aishell_loss[loss=0.1683, simple_loss=0.2423, pruned_loss=0.04711, over 4869.00 frames.], tot_loss[loss=0.1544, simple_loss=0.2339, pruned_loss=0.03738, over 985217.33 frames.], batch size: 35, aishell_tot_loss[loss=0.1559, simple_loss=0.2406, pruned_loss=0.03555, over 975679.90 frames.], datatang_tot_loss[loss=0.1534, simple_loss=0.2283, pruned_loss=0.0392, over 978030.00 frames.], batch size: 35, lr: 4.81e-04
+2022-06-18 22:25:06,014 INFO [train.py:874] (2/4) Epoch 17, batch 1950, datatang_loss[loss=0.1534, simple_loss=0.2196, pruned_loss=0.04358, over 4834.00 frames.], tot_loss[loss=0.1548, simple_loss=0.2347, pruned_loss=0.03745, over 985530.29 frames.], batch size: 30, aishell_tot_loss[loss=0.1557, simple_loss=0.2406, pruned_loss=0.03539, over 976777.85 frames.], datatang_tot_loss[loss=0.154, simple_loss=0.2291, pruned_loss=0.0394, over 979205.33 frames.], batch size: 30, lr: 4.81e-04
+2022-06-18 22:25:36,666 INFO [train.py:874] (2/4) Epoch 17, batch 2000, datatang_loss[loss=0.1918, simple_loss=0.2584, pruned_loss=0.06257, over 4938.00 frames.], tot_loss[loss=0.1546, simple_loss=0.2344, pruned_loss=0.0374, over 985656.79 frames.], batch size: 108, aishell_tot_loss[loss=0.1559, simple_loss=0.2408, pruned_loss=0.03551, over 977695.48 frames.], datatang_tot_loss[loss=0.1536, simple_loss=0.2287, pruned_loss=0.03921, over 980178.71 frames.], batch size: 108, lr: 4.81e-04
+2022-06-18 22:25:36,667 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 22:25:53,689 INFO [train.py:914] (2/4) Epoch 17, validation: loss=0.1656, simple_loss=0.2494, pruned_loss=0.04087, over 1622729.00 frames.
+2022-06-18 22:26:23,333 INFO [train.py:874] (2/4) Epoch 17, batch 2050, aishell_loss[loss=0.147, simple_loss=0.2339, pruned_loss=0.0301, over 4968.00 frames.], tot_loss[loss=0.1546, simple_loss=0.2348, pruned_loss=0.0372, over 986071.88 frames.], batch size: 31, aishell_tot_loss[loss=0.1559, simple_loss=0.2411, pruned_loss=0.03541, over 978814.76 frames.], datatang_tot_loss[loss=0.1535, simple_loss=0.2286, pruned_loss=0.03917, over 981104.19 frames.], batch size: 31, lr: 4.81e-04
+2022-06-18 22:26:54,515 INFO [train.py:874] (2/4) Epoch 17, batch 2100, aishell_loss[loss=0.157, simple_loss=0.2405, pruned_loss=0.03678, over 4875.00 frames.], tot_loss[loss=0.1552, simple_loss=0.2354, pruned_loss=0.03746, over 986081.88 frames.], batch size: 28, aishell_tot_loss[loss=0.156, simple_loss=0.2413, pruned_loss=0.03538, over 979612.16 frames.], datatang_tot_loss[loss=0.1539, simple_loss=0.229, pruned_loss=0.03944, over 981767.37 frames.], batch size: 28, lr: 4.81e-04
+2022-06-18 22:27:23,934 INFO [train.py:874] (2/4) Epoch 17, batch 2150, datatang_loss[loss=0.1319, simple_loss=0.1957, pruned_loss=0.03402, over 4891.00 frames.], tot_loss[loss=0.1547, simple_loss=0.235, pruned_loss=0.03716, over 986040.58 frames.], batch size: 39, aishell_tot_loss[loss=0.1562, simple_loss=0.2416, pruned_loss=0.03538, over 980369.19 frames.], datatang_tot_loss[loss=0.1533, simple_loss=0.2283, pruned_loss=0.03913, over 982249.45 frames.], batch size: 39, lr: 4.80e-04
+2022-06-18 22:27:55,171 INFO [train.py:874] (2/4) Epoch 17, batch 2200, datatang_loss[loss=0.1662, simple_loss=0.2333, pruned_loss=0.04958, over 4924.00 frames.], tot_loss[loss=0.155, simple_loss=0.2355, pruned_loss=0.03723, over 985735.37 frames.], batch size: 42, aishell_tot_loss[loss=0.1565, simple_loss=0.242, pruned_loss=0.03553, over 980661.45 frames.], datatang_tot_loss[loss=0.1532, simple_loss=0.2283, pruned_loss=0.03907, over 982770.66 frames.], batch size: 42, lr: 4.80e-04
+2022-06-18 22:28:25,157 INFO [train.py:874] (2/4) Epoch 17, batch 2250, datatang_loss[loss=0.1504, simple_loss=0.2273, pruned_loss=0.03669, over 4939.00 frames.], tot_loss[loss=0.1549, simple_loss=0.2355, pruned_loss=0.03711, over 985601.19 frames.], batch size: 79, aishell_tot_loss[loss=0.1563, simple_loss=0.2418, pruned_loss=0.03546, over 981397.16 frames.], datatang_tot_loss[loss=0.1532, simple_loss=0.2283, pruned_loss=0.03907, over 982875.65 frames.], batch size: 79, lr: 4.80e-04
+2022-06-18 22:28:55,862 INFO [train.py:874] (2/4) Epoch 17, batch 2300, aishell_loss[loss=0.164, simple_loss=0.2603, pruned_loss=0.03388, over 4905.00 frames.], tot_loss[loss=0.1557, simple_loss=0.2361, pruned_loss=0.03767, over 985754.09 frames.], batch size: 41, aishell_tot_loss[loss=0.1564, simple_loss=0.2418, pruned_loss=0.03555, over 982147.25 frames.], datatang_tot_loss[loss=0.1541, simple_loss=0.2289, pruned_loss=0.03961, over 983110.90 frames.], batch size: 41, lr: 4.80e-04
+2022-06-18 22:29:26,647 INFO [train.py:874] (2/4) Epoch 17, batch 2350, aishell_loss[loss=0.1608, simple_loss=0.2477, pruned_loss=0.03698, over 4907.00 frames.], tot_loss[loss=0.155, simple_loss=0.2357, pruned_loss=0.03718, over 985823.84 frames.], batch size: 46, aishell_tot_loss[loss=0.1568, simple_loss=0.2422, pruned_loss=0.03569, over 982647.75 frames.], datatang_tot_loss[loss=0.1531, simple_loss=0.2281, pruned_loss=0.03906, over 983429.68 frames.], batch size: 46, lr: 4.80e-04
+2022-06-18 22:29:55,936 INFO [train.py:874] (2/4) Epoch 17, batch 2400, datatang_loss[loss=0.1379, simple_loss=0.2148, pruned_loss=0.03051, over 4930.00 frames.], tot_loss[loss=0.155, simple_loss=0.2354, pruned_loss=0.03734, over 985829.80 frames.], batch size: 71, aishell_tot_loss[loss=0.1573, simple_loss=0.2426, pruned_loss=0.03602, over 982901.90 frames.], datatang_tot_loss[loss=0.1526, simple_loss=0.2274, pruned_loss=0.03885, over 983852.58 frames.], batch size: 71, lr: 4.79e-04
+2022-06-18 22:30:25,169 INFO [train.py:874] (2/4) Epoch 17, batch 2450, aishell_loss[loss=0.15, simple_loss=0.2341, pruned_loss=0.03294, over 4929.00 frames.], tot_loss[loss=0.1545, simple_loss=0.2348, pruned_loss=0.03711, over 985708.28 frames.], batch size: 32, aishell_tot_loss[loss=0.1569, simple_loss=0.2421, pruned_loss=0.0358, over 983087.90 frames.], datatang_tot_loss[loss=0.1525, simple_loss=0.2272, pruned_loss=0.03886, over 984135.34 frames.], batch size: 32, lr: 4.79e-04
+2022-06-18 22:30:56,066 INFO [train.py:874] (2/4) Epoch 17, batch 2500, aishell_loss[loss=0.1413, simple_loss=0.227, pruned_loss=0.02783, over 4862.00 frames.], tot_loss[loss=0.1554, simple_loss=0.2355, pruned_loss=0.03765, over 985549.65 frames.], batch size: 36, aishell_tot_loss[loss=0.1573, simple_loss=0.2426, pruned_loss=0.03599, over 983106.82 frames.], datatang_tot_loss[loss=0.1529, simple_loss=0.2275, pruned_loss=0.03919, over 984458.49 frames.], batch size: 36, lr: 4.79e-04
+2022-06-18 22:31:26,715 INFO [train.py:874] (2/4) Epoch 17, batch 2550, aishell_loss[loss=0.1366, simple_loss=0.2241, pruned_loss=0.02453, over 4812.00 frames.], tot_loss[loss=0.1541, simple_loss=0.2342, pruned_loss=0.03698, over 985714.92 frames.], batch size: 26, aishell_tot_loss[loss=0.1565, simple_loss=0.2416, pruned_loss=0.03572, over 983423.97 frames.], datatang_tot_loss[loss=0.1523, simple_loss=0.227, pruned_loss=0.0388, over 984746.54 frames.], batch size: 26, lr: 4.79e-04
+2022-06-18 22:31:56,350 INFO [train.py:874] (2/4) Epoch 17, batch 2600, aishell_loss[loss=0.1179, simple_loss=0.1833, pruned_loss=0.02626, over 4924.00 frames.], tot_loss[loss=0.1539, simple_loss=0.234, pruned_loss=0.03688, over 985860.90 frames.], batch size: 20, aishell_tot_loss[loss=0.1557, simple_loss=0.2408, pruned_loss=0.03528, over 983773.67 frames.], datatang_tot_loss[loss=0.1529, simple_loss=0.2276, pruned_loss=0.0391, over 984942.50 frames.], batch size: 20, lr: 4.79e-04
+2022-06-18 22:32:27,294 INFO [train.py:874] (2/4) Epoch 17, batch 2650, aishell_loss[loss=0.1799, simple_loss=0.2598, pruned_loss=0.05001, over 4946.00 frames.], tot_loss[loss=0.1534, simple_loss=0.2339, pruned_loss=0.03644, over 985911.51 frames.], batch size: 56, aishell_tot_loss[loss=0.1557, simple_loss=0.2412, pruned_loss=0.03515, over 983830.44 frames.], datatang_tot_loss[loss=0.1523, simple_loss=0.2272, pruned_loss=0.03865, over 985287.80 frames.], batch size: 56, lr: 4.79e-04
+2022-06-18 22:32:58,304 INFO [train.py:874] (2/4) Epoch 17, batch 2700, datatang_loss[loss=0.1354, simple_loss=0.2149, pruned_loss=0.02794, over 4940.00 frames.], tot_loss[loss=0.1528, simple_loss=0.2328, pruned_loss=0.03635, over 985885.26 frames.], batch size: 37, aishell_tot_loss[loss=0.1552, simple_loss=0.2406, pruned_loss=0.03493, over 983827.93 frames.], datatang_tot_loss[loss=0.152, simple_loss=0.2269, pruned_loss=0.03858, over 985526.50 frames.], batch size: 37, lr: 4.78e-04
+2022-06-18 22:33:27,878 INFO [train.py:874] (2/4) Epoch 17, batch 2750, datatang_loss[loss=0.1416, simple_loss=0.2215, pruned_loss=0.03084, over 4925.00 frames.], tot_loss[loss=0.1531, simple_loss=0.2331, pruned_loss=0.03651, over 985550.37 frames.], batch size: 57, aishell_tot_loss[loss=0.1547, simple_loss=0.24, pruned_loss=0.03475, over 983755.33 frames.], datatang_tot_loss[loss=0.1526, simple_loss=0.2276, pruned_loss=0.03885, over 985531.55 frames.], batch size: 57, lr: 4.78e-04
+2022-06-18 22:33:57,729 INFO [train.py:874] (2/4) Epoch 17, batch 2800, aishell_loss[loss=0.1544, simple_loss=0.252, pruned_loss=0.02836, over 4976.00 frames.], tot_loss[loss=0.1521, simple_loss=0.2326, pruned_loss=0.03579, over 985582.05 frames.], batch size: 64, aishell_tot_loss[loss=0.1548, simple_loss=0.2402, pruned_loss=0.03467, over 983892.55 frames.], datatang_tot_loss[loss=0.1513, simple_loss=0.2264, pruned_loss=0.03814, over 985668.06 frames.], batch size: 64, lr: 4.78e-04
+2022-06-18 22:34:28,499 INFO [train.py:874] (2/4) Epoch 17, batch 2850, aishell_loss[loss=0.1525, simple_loss=0.2392, pruned_loss=0.03292, over 4951.00 frames.], tot_loss[loss=0.1528, simple_loss=0.2335, pruned_loss=0.03608, over 985823.67 frames.], batch size: 40, aishell_tot_loss[loss=0.1545, simple_loss=0.2402, pruned_loss=0.03439, over 984200.48 frames.], datatang_tot_loss[loss=0.1521, simple_loss=0.227, pruned_loss=0.03863, over 985838.34 frames.], batch size: 40, lr: 4.78e-04
+2022-06-18 22:34:58,998 INFO [train.py:874] (2/4) Epoch 17, batch 2900, datatang_loss[loss=0.158, simple_loss=0.2331, pruned_loss=0.04146, over 4919.00 frames.], tot_loss[loss=0.1531, simple_loss=0.2339, pruned_loss=0.03612, over 985971.17 frames.], batch size: 83, aishell_tot_loss[loss=0.1546, simple_loss=0.2404, pruned_loss=0.03443, over 984459.66 frames.], datatang_tot_loss[loss=0.1521, simple_loss=0.2269, pruned_loss=0.03862, over 985955.04 frames.], batch size: 83, lr: 4.78e-04
+2022-06-18 22:35:28,989 INFO [train.py:874] (2/4) Epoch 17, batch 2950, aishell_loss[loss=0.1659, simple_loss=0.2561, pruned_loss=0.03789, over 4882.00 frames.], tot_loss[loss=0.1541, simple_loss=0.2347, pruned_loss=0.03678, over 985855.09 frames.], batch size: 47, aishell_tot_loss[loss=0.155, simple_loss=0.2406, pruned_loss=0.03465, over 984621.32 frames.], datatang_tot_loss[loss=0.1528, simple_loss=0.2277, pruned_loss=0.03899, over 985847.99 frames.], batch size: 47, lr: 4.78e-04
+2022-06-18 22:35:58,591 INFO [train.py:874] (2/4) Epoch 17, batch 3000, datatang_loss[loss=0.1468, simple_loss=0.2162, pruned_loss=0.03873, over 4906.00 frames.], tot_loss[loss=0.1541, simple_loss=0.234, pruned_loss=0.03707, over 985697.93 frames.], batch size: 52, aishell_tot_loss[loss=0.1551, simple_loss=0.2408, pruned_loss=0.03471, over 984613.39 frames.], datatang_tot_loss[loss=0.1527, simple_loss=0.2271, pruned_loss=0.03915, over 985817.11 frames.], batch size: 52, lr: 4.77e-04
+2022-06-18 22:35:58,592 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 22:36:15,619 INFO [train.py:914] (2/4) Epoch 17, validation: loss=0.1675, simple_loss=0.2537, pruned_loss=0.04061, over 1622729.00 frames.
+2022-06-18 22:36:45,524 INFO [train.py:874] (2/4) Epoch 17, batch 3050, aishell_loss[loss=0.1546, simple_loss=0.2495, pruned_loss=0.02985, over 4921.00 frames.], tot_loss[loss=0.1547, simple_loss=0.2343, pruned_loss=0.03753, over 986070.47 frames.], batch size: 52, aishell_tot_loss[loss=0.1552, simple_loss=0.2409, pruned_loss=0.03477, over 984766.08 frames.], datatang_tot_loss[loss=0.1533, simple_loss=0.2276, pruned_loss=0.03953, over 986145.52 frames.], batch size: 52, lr: 4.77e-04
+2022-06-18 22:37:16,149 INFO [train.py:874] (2/4) Epoch 17, batch 3100, datatang_loss[loss=0.1492, simple_loss=0.2289, pruned_loss=0.03478, over 4896.00 frames.], tot_loss[loss=0.1543, simple_loss=0.2341, pruned_loss=0.03723, over 985761.87 frames.], batch size: 42, aishell_tot_loss[loss=0.1552, simple_loss=0.2409, pruned_loss=0.03472, over 984588.37 frames.], datatang_tot_loss[loss=0.1531, simple_loss=0.2278, pruned_loss=0.03925, over 986117.45 frames.], batch size: 42, lr: 4.77e-04
+2022-06-18 22:37:46,645 INFO [train.py:874] (2/4) Epoch 17, batch 3150, aishell_loss[loss=0.1579, simple_loss=0.2377, pruned_loss=0.03903, over 4924.00 frames.], tot_loss[loss=0.1541, simple_loss=0.2337, pruned_loss=0.03725, over 985860.37 frames.], batch size: 49, aishell_tot_loss[loss=0.1552, simple_loss=0.2406, pruned_loss=0.03492, over 984759.21 frames.], datatang_tot_loss[loss=0.1529, simple_loss=0.2276, pruned_loss=0.03911, over 986160.45 frames.], batch size: 49, lr: 4.77e-04
+2022-06-18 22:38:17,536 INFO [train.py:874] (2/4) Epoch 17, batch 3200, aishell_loss[loss=0.1596, simple_loss=0.2577, pruned_loss=0.03077, over 4924.00 frames.], tot_loss[loss=0.1541, simple_loss=0.2335, pruned_loss=0.03729, over 985737.59 frames.], batch size: 68, aishell_tot_loss[loss=0.1555, simple_loss=0.2407, pruned_loss=0.0351, over 984566.12 frames.], datatang_tot_loss[loss=0.1526, simple_loss=0.227, pruned_loss=0.03911, over 986354.35 frames.], batch size: 68, lr: 4.77e-04
+2022-06-18 22:38:47,532 INFO [train.py:874] (2/4) Epoch 17, batch 3250, aishell_loss[loss=0.1417, simple_loss=0.2355, pruned_loss=0.02398, over 4911.00 frames.], tot_loss[loss=0.154, simple_loss=0.2332, pruned_loss=0.03738, over 986103.44 frames.], batch size: 46, aishell_tot_loss[loss=0.1557, simple_loss=0.241, pruned_loss=0.03526, over 984878.78 frames.], datatang_tot_loss[loss=0.1523, simple_loss=0.2268, pruned_loss=0.03895, over 986442.55 frames.], batch size: 46, lr: 4.77e-04
+2022-06-18 22:39:17,042 INFO [train.py:874] (2/4) Epoch 17, batch 3300, aishell_loss[loss=0.1745, simple_loss=0.2614, pruned_loss=0.04376, over 4975.00 frames.], tot_loss[loss=0.1531, simple_loss=0.2324, pruned_loss=0.03692, over 985949.62 frames.], batch size: 44, aishell_tot_loss[loss=0.1553, simple_loss=0.2403, pruned_loss=0.03513, over 984898.34 frames.], datatang_tot_loss[loss=0.1519, simple_loss=0.2265, pruned_loss=0.03865, over 986341.75 frames.], batch size: 44, lr: 4.76e-04
+2022-06-18 22:39:47,733 INFO [train.py:874] (2/4) Epoch 17, batch 3350, aishell_loss[loss=0.1889, simple_loss=0.2686, pruned_loss=0.05466, over 4923.00 frames.], tot_loss[loss=0.1531, simple_loss=0.2326, pruned_loss=0.03678, over 985295.56 frames.], batch size: 68, aishell_tot_loss[loss=0.1553, simple_loss=0.2405, pruned_loss=0.03506, over 984768.81 frames.], datatang_tot_loss[loss=0.1517, simple_loss=0.2262, pruned_loss=0.03859, over 985844.16 frames.], batch size: 68, lr: 4.76e-04
+2022-06-18 22:40:18,435 INFO [train.py:874] (2/4) Epoch 17, batch 3400, datatang_loss[loss=0.1524, simple_loss=0.2257, pruned_loss=0.03955, over 4925.00 frames.], tot_loss[loss=0.1546, simple_loss=0.2342, pruned_loss=0.03745, over 985353.31 frames.], batch size: 81, aishell_tot_loss[loss=0.1555, simple_loss=0.2408, pruned_loss=0.03515, over 984816.03 frames.], datatang_tot_loss[loss=0.1529, simple_loss=0.2274, pruned_loss=0.03922, over 985847.29 frames.], batch size: 81, lr: 4.76e-04
+2022-06-18 22:40:47,877 INFO [train.py:874] (2/4) Epoch 17, batch 3450, aishell_loss[loss=0.1574, simple_loss=0.2453, pruned_loss=0.03475, over 4972.00 frames.], tot_loss[loss=0.155, simple_loss=0.2346, pruned_loss=0.03763, over 985419.33 frames.], batch size: 48, aishell_tot_loss[loss=0.1556, simple_loss=0.2408, pruned_loss=0.03523, over 984894.09 frames.], datatang_tot_loss[loss=0.1533, simple_loss=0.2279, pruned_loss=0.03937, over 985855.30 frames.], batch size: 48, lr: 4.76e-04
+2022-06-18 22:41:18,970 INFO [train.py:874] (2/4) Epoch 17, batch 3500, aishell_loss[loss=0.146, simple_loss=0.2341, pruned_loss=0.02901, over 4914.00 frames.], tot_loss[loss=0.1552, simple_loss=0.235, pruned_loss=0.03772, over 985223.27 frames.], batch size: 46, aishell_tot_loss[loss=0.1553, simple_loss=0.2405, pruned_loss=0.03508, over 984786.08 frames.], datatang_tot_loss[loss=0.1541, simple_loss=0.2287, pruned_loss=0.03971, over 985777.93 frames.], batch size: 46, lr: 4.76e-04
+2022-06-18 22:41:49,996 INFO [train.py:874] (2/4) Epoch 17, batch 3550, datatang_loss[loss=0.1375, simple_loss=0.213, pruned_loss=0.031, over 4927.00 frames.], tot_loss[loss=0.1548, simple_loss=0.2344, pruned_loss=0.03758, over 984785.64 frames.], batch size: 83, aishell_tot_loss[loss=0.1557, simple_loss=0.2408, pruned_loss=0.03528, over 984498.57 frames.], datatang_tot_loss[loss=0.1534, simple_loss=0.2281, pruned_loss=0.03936, over 985552.19 frames.], batch size: 83, lr: 4.76e-04
+2022-06-18 22:42:20,084 INFO [train.py:874] (2/4) Epoch 17, batch 3600, datatang_loss[loss=0.1329, simple_loss=0.2003, pruned_loss=0.03269, over 4872.00 frames.], tot_loss[loss=0.1544, simple_loss=0.2344, pruned_loss=0.03722, over 985112.22 frames.], batch size: 39, aishell_tot_loss[loss=0.1555, simple_loss=0.2406, pruned_loss=0.03522, over 984618.88 frames.], datatang_tot_loss[loss=0.1532, simple_loss=0.2279, pruned_loss=0.03922, over 985760.78 frames.], batch size: 39, lr: 4.75e-04
+2022-06-18 22:42:49,388 INFO [train.py:874] (2/4) Epoch 17, batch 3650, aishell_loss[loss=0.1804, simple_loss=0.2558, pruned_loss=0.05254, over 4969.00 frames.], tot_loss[loss=0.1546, simple_loss=0.2344, pruned_loss=0.03743, over 985573.98 frames.], batch size: 51, aishell_tot_loss[loss=0.156, simple_loss=0.241, pruned_loss=0.03548, over 984909.55 frames.], datatang_tot_loss[loss=0.1529, simple_loss=0.2275, pruned_loss=0.03917, over 985942.24 frames.], batch size: 51, lr: 4.75e-04
+2022-06-18 22:43:21,322 INFO [train.py:874] (2/4) Epoch 17, batch 3700, aishell_loss[loss=0.1516, simple_loss=0.2406, pruned_loss=0.03131, over 4974.00 frames.], tot_loss[loss=0.1544, simple_loss=0.2346, pruned_loss=0.03712, over 985896.65 frames.], batch size: 61, aishell_tot_loss[loss=0.1559, simple_loss=0.2412, pruned_loss=0.03529, over 985203.47 frames.], datatang_tot_loss[loss=0.1529, simple_loss=0.2276, pruned_loss=0.03908, over 986007.61 frames.], batch size: 61, lr: 4.75e-04
+2022-06-18 22:43:51,462 INFO [train.py:874] (2/4) Epoch 17, batch 3750, aishell_loss[loss=0.1656, simple_loss=0.259, pruned_loss=0.03608, over 4965.00 frames.], tot_loss[loss=0.1531, simple_loss=0.2334, pruned_loss=0.03645, over 986006.90 frames.], batch size: 56, aishell_tot_loss[loss=0.1551, simple_loss=0.2405, pruned_loss=0.0348, over 985517.26 frames.], datatang_tot_loss[loss=0.1524, simple_loss=0.2271, pruned_loss=0.03881, over 985875.80 frames.], batch size: 56, lr: 4.75e-04
+2022-06-18 22:44:21,504 INFO [train.py:874] (2/4) Epoch 17, batch 3800, aishell_loss[loss=0.1528, simple_loss=0.239, pruned_loss=0.03331, over 4952.00 frames.], tot_loss[loss=0.1538, simple_loss=0.234, pruned_loss=0.0368, over 985866.60 frames.], batch size: 56, aishell_tot_loss[loss=0.1554, simple_loss=0.2409, pruned_loss=0.03497, over 985543.84 frames.], datatang_tot_loss[loss=0.1526, simple_loss=0.2273, pruned_loss=0.0389, over 985778.30 frames.], batch size: 56, lr: 4.75e-04
+2022-06-18 22:44:50,826 INFO [train.py:874] (2/4) Epoch 17, batch 3850, aishell_loss[loss=0.1748, simple_loss=0.2546, pruned_loss=0.04756, over 4945.00 frames.], tot_loss[loss=0.1539, simple_loss=0.2346, pruned_loss=0.03663, over 985754.08 frames.], batch size: 33, aishell_tot_loss[loss=0.156, simple_loss=0.2416, pruned_loss=0.03518, over 985393.58 frames.], datatang_tot_loss[loss=0.152, simple_loss=0.227, pruned_loss=0.03853, over 985872.17 frames.], batch size: 33, lr: 4.75e-04
+2022-06-18 22:45:20,644 INFO [train.py:874] (2/4) Epoch 17, batch 3900, aishell_loss[loss=0.1493, simple_loss=0.2478, pruned_loss=0.02538, over 4939.00 frames.], tot_loss[loss=0.1534, simple_loss=0.2338, pruned_loss=0.03647, over 986026.07 frames.], batch size: 54, aishell_tot_loss[loss=0.1554, simple_loss=0.2408, pruned_loss=0.03493, over 985599.65 frames.], datatang_tot_loss[loss=0.152, simple_loss=0.2269, pruned_loss=0.03855, over 985978.95 frames.], batch size: 54, lr: 4.74e-04
+2022-06-18 22:45:49,951 INFO [train.py:874] (2/4) Epoch 17, batch 3950, aishell_loss[loss=0.1927, simple_loss=0.2698, pruned_loss=0.05777, over 4870.00 frames.], tot_loss[loss=0.1533, simple_loss=0.234, pruned_loss=0.03624, over 986105.07 frames.], batch size: 36, aishell_tot_loss[loss=0.1555, simple_loss=0.2413, pruned_loss=0.03487, over 985445.11 frames.], datatang_tot_loss[loss=0.1516, simple_loss=0.2267, pruned_loss=0.03831, over 986274.94 frames.], batch size: 36, lr: 4.74e-04
+2022-06-18 22:46:19,507 INFO [train.py:874] (2/4) Epoch 17, batch 4000, datatang_loss[loss=0.148, simple_loss=0.2158, pruned_loss=0.04006, over 4820.00 frames.], tot_loss[loss=0.1542, simple_loss=0.2346, pruned_loss=0.0369, over 986162.07 frames.], batch size: 24, aishell_tot_loss[loss=0.1563, simple_loss=0.2419, pruned_loss=0.03537, over 985638.22 frames.], datatang_tot_loss[loss=0.1518, simple_loss=0.2268, pruned_loss=0.03843, over 986205.57 frames.], batch size: 24, lr: 4.74e-04
+2022-06-18 22:46:19,508 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 22:46:35,708 INFO [train.py:914] (2/4) Epoch 17, validation: loss=0.1673, simple_loss=0.2516, pruned_loss=0.04154, over 1622729.00 frames.
+2022-06-18 22:47:04,906 INFO [train.py:874] (2/4) Epoch 17, batch 4050, datatang_loss[loss=0.1336, simple_loss=0.2079, pruned_loss=0.02969, over 4883.00 frames.], tot_loss[loss=0.155, simple_loss=0.2347, pruned_loss=0.03765, over 985879.26 frames.], batch size: 39, aishell_tot_loss[loss=0.1569, simple_loss=0.2422, pruned_loss=0.03581, over 985623.64 frames.], datatang_tot_loss[loss=0.1522, simple_loss=0.227, pruned_loss=0.03873, over 985954.52 frames.], batch size: 39, lr: 4.74e-04
+2022-06-18 22:47:35,447 INFO [train.py:874] (2/4) Epoch 17, batch 4100, aishell_loss[loss=0.1471, simple_loss=0.2378, pruned_loss=0.02821, over 4956.00 frames.], tot_loss[loss=0.1556, simple_loss=0.2353, pruned_loss=0.03801, over 985720.86 frames.], batch size: 61, aishell_tot_loss[loss=0.1567, simple_loss=0.2418, pruned_loss=0.03583, over 985410.67 frames.], datatang_tot_loss[loss=0.1533, simple_loss=0.2283, pruned_loss=0.0391, over 985998.02 frames.], batch size: 61, lr: 4.74e-04
+2022-06-18 22:48:03,147 INFO [train.py:874] (2/4) Epoch 17, batch 4150, aishell_loss[loss=0.1323, simple_loss=0.2183, pruned_loss=0.02318, over 4873.00 frames.], tot_loss[loss=0.156, simple_loss=0.2358, pruned_loss=0.03807, over 985355.76 frames.], batch size: 28, aishell_tot_loss[loss=0.1573, simple_loss=0.2423, pruned_loss=0.03613, over 985113.65 frames.], datatang_tot_loss[loss=0.1532, simple_loss=0.2284, pruned_loss=0.039, over 985927.51 frames.], batch size: 28, lr: 4.73e-04
+2022-06-18 22:49:24,001 INFO [train.py:874] (2/4) Epoch 18, batch 50, aishell_loss[loss=0.1758, simple_loss=0.2591, pruned_loss=0.04623, over 4934.00 frames.], tot_loss[loss=0.1453, simple_loss=0.2267, pruned_loss=0.032, over 218612.44 frames.], batch size: 49, aishell_tot_loss[loss=0.1491, simple_loss=0.2355, pruned_loss=0.03138, over 129037.03 frames.], datatang_tot_loss[loss=0.1404, simple_loss=0.2152, pruned_loss=0.03284, over 103079.99 frames.], batch size: 49, lr: 4.61e-04
+2022-06-18 22:49:54,848 INFO [train.py:874] (2/4) Epoch 18, batch 100, datatang_loss[loss=0.1658, simple_loss=0.2352, pruned_loss=0.04822, over 4903.00 frames.], tot_loss[loss=0.1474, simple_loss=0.2274, pruned_loss=0.03369, over 388782.79 frames.], batch size: 59, aishell_tot_loss[loss=0.1533, simple_loss=0.2404, pruned_loss=0.03313, over 222263.09 frames.], datatang_tot_loss[loss=0.1412, simple_loss=0.2143, pruned_loss=0.03405, over 214986.75 frames.], batch size: 59, lr: 4.61e-04
+2022-06-18 22:50:25,769 INFO [train.py:874] (2/4) Epoch 18, batch 150, datatang_loss[loss=0.1556, simple_loss=0.2205, pruned_loss=0.04533, over 4940.00 frames.], tot_loss[loss=0.1486, simple_loss=0.229, pruned_loss=0.0341, over 520430.93 frames.], batch size: 62, aishell_tot_loss[loss=0.1538, simple_loss=0.2406, pruned_loss=0.03349, over 324953.87 frames.], datatang_tot_loss[loss=0.1424, simple_loss=0.2157, pruned_loss=0.03457, over 291870.35 frames.], batch size: 62, lr: 4.60e-04
+2022-06-18 22:50:54,380 INFO [train.py:874] (2/4) Epoch 18, batch 200, aishell_loss[loss=0.1512, simple_loss=0.2421, pruned_loss=0.03016, over 4882.00 frames.], tot_loss[loss=0.149, simple_loss=0.2295, pruned_loss=0.03426, over 623547.97 frames.], batch size: 47, aishell_tot_loss[loss=0.1537, simple_loss=0.2403, pruned_loss=0.03356, over 414392.30 frames.], datatang_tot_loss[loss=0.143, simple_loss=0.2163, pruned_loss=0.03489, over 361100.36 frames.], batch size: 47, lr: 4.60e-04
+2022-06-18 22:51:24,568 INFO [train.py:874] (2/4) Epoch 18, batch 250, aishell_loss[loss=0.158, simple_loss=0.2327, pruned_loss=0.04172, over 4883.00 frames.], tot_loss[loss=0.1508, simple_loss=0.2306, pruned_loss=0.03556, over 703516.83 frames.], batch size: 42, aishell_tot_loss[loss=0.1546, simple_loss=0.2403, pruned_loss=0.03447, over 478833.93 frames.], datatang_tot_loss[loss=0.1455, simple_loss=0.2189, pruned_loss=0.03606, over 437393.57 frames.], batch size: 42, lr: 4.60e-04
+2022-06-18 22:51:56,057 INFO [train.py:874] (2/4) Epoch 18, batch 300, datatang_loss[loss=0.1512, simple_loss=0.2264, pruned_loss=0.03797, over 4918.00 frames.], tot_loss[loss=0.1498, simple_loss=0.2296, pruned_loss=0.03499, over 766181.65 frames.], batch size: 77, aishell_tot_loss[loss=0.1546, simple_loss=0.2405, pruned_loss=0.03438, over 527235.05 frames.], datatang_tot_loss[loss=0.1445, simple_loss=0.2185, pruned_loss=0.03529, over 514005.03 frames.], batch size: 77, lr: 4.60e-04
+2022-06-18 22:52:24,075 INFO [train.py:874] (2/4) Epoch 18, batch 350, datatang_loss[loss=0.1493, simple_loss=0.2283, pruned_loss=0.03512, over 4925.00 frames.], tot_loss[loss=0.1497, simple_loss=0.2303, pruned_loss=0.03458, over 814492.05 frames.], batch size: 42, aishell_tot_loss[loss=0.1542, simple_loss=0.2403, pruned_loss=0.03403, over 594823.13 frames.], datatang_tot_loss[loss=0.1445, simple_loss=0.2187, pruned_loss=0.03517, over 554675.57 frames.], batch size: 42, lr: 4.60e-04
+2022-06-18 22:52:55,581 INFO [train.py:874] (2/4) Epoch 18, batch 400, datatang_loss[loss=0.1388, simple_loss=0.2162, pruned_loss=0.03073, over 4920.00 frames.], tot_loss[loss=0.1512, simple_loss=0.2309, pruned_loss=0.0358, over 852233.10 frames.], batch size: 75, aishell_tot_loss[loss=0.1538, simple_loss=0.2394, pruned_loss=0.03403, over 632043.58 frames.], datatang_tot_loss[loss=0.1476, simple_loss=0.2214, pruned_loss=0.03689, over 614697.29 frames.], batch size: 75, lr: 4.60e-04
+2022-06-18 22:53:26,693 INFO [train.py:874] (2/4) Epoch 18, batch 450, datatang_loss[loss=0.1561, simple_loss=0.2343, pruned_loss=0.03888, over 4960.00 frames.], tot_loss[loss=0.151, simple_loss=0.231, pruned_loss=0.03545, over 881789.48 frames.], batch size: 91, aishell_tot_loss[loss=0.1537, simple_loss=0.2395, pruned_loss=0.03397, over 675131.13 frames.], datatang_tot_loss[loss=0.1474, simple_loss=0.2216, pruned_loss=0.03658, over 656833.93 frames.], batch size: 91, lr: 4.59e-04
+2022-06-18 22:53:54,537 INFO [train.py:874] (2/4) Epoch 18, batch 500, datatang_loss[loss=0.1402, simple_loss=0.2203, pruned_loss=0.03003, over 4925.00 frames.], tot_loss[loss=0.1507, simple_loss=0.2308, pruned_loss=0.03531, over 904593.99 frames.], batch size: 79, aishell_tot_loss[loss=0.1535, simple_loss=0.239, pruned_loss=0.03406, over 711330.47 frames.], datatang_tot_loss[loss=0.1473, simple_loss=0.222, pruned_loss=0.03637, over 695700.51 frames.], batch size: 79, lr: 4.59e-04
+2022-06-18 22:54:26,287 INFO [train.py:874] (2/4) Epoch 18, batch 550, datatang_loss[loss=0.1591, simple_loss=0.237, pruned_loss=0.0406, over 4923.00 frames.], tot_loss[loss=0.1512, simple_loss=0.2316, pruned_loss=0.03538, over 922629.53 frames.], batch size: 83, aishell_tot_loss[loss=0.1532, simple_loss=0.2388, pruned_loss=0.03376, over 746070.36 frames.], datatang_tot_loss[loss=0.1484, simple_loss=0.223, pruned_loss=0.03685, over 727319.11 frames.], batch size: 83, lr: 4.59e-04
+2022-06-18 22:54:57,082 INFO [train.py:874] (2/4) Epoch 18, batch 600, datatang_loss[loss=0.1597, simple_loss=0.2446, pruned_loss=0.03742, over 4949.00 frames.], tot_loss[loss=0.1527, simple_loss=0.2326, pruned_loss=0.0364, over 936565.63 frames.], batch size: 99, aishell_tot_loss[loss=0.1533, simple_loss=0.2386, pruned_loss=0.03401, over 768870.04 frames.], datatang_tot_loss[loss=0.1505, simple_loss=0.2252, pruned_loss=0.03787, over 763414.27 frames.], batch size: 99, lr: 4.59e-04
+2022-06-18 22:55:24,764 INFO [train.py:874] (2/4) Epoch 18, batch 650, datatang_loss[loss=0.1991, simple_loss=0.2714, pruned_loss=0.06347, over 4926.00 frames.], tot_loss[loss=0.1544, simple_loss=0.2342, pruned_loss=0.03726, over 947202.54 frames.], batch size: 108, aishell_tot_loss[loss=0.1542, simple_loss=0.2392, pruned_loss=0.03456, over 795166.14 frames.], datatang_tot_loss[loss=0.152, simple_loss=0.2268, pruned_loss=0.03854, over 788538.24 frames.], batch size: 108, lr: 4.59e-04
+2022-06-18 22:55:56,305 INFO [train.py:874] (2/4) Epoch 18, batch 700, datatang_loss[loss=0.1592, simple_loss=0.2311, pruned_loss=0.04367, over 4927.00 frames.], tot_loss[loss=0.1538, simple_loss=0.234, pruned_loss=0.0368, over 955963.43 frames.], batch size: 79, aishell_tot_loss[loss=0.1539, simple_loss=0.2389, pruned_loss=0.03445, over 821071.10 frames.], datatang_tot_loss[loss=0.1519, simple_loss=0.227, pruned_loss=0.03839, over 808364.14 frames.], batch size: 79, lr: 4.59e-04
+2022-06-18 22:56:25,875 INFO [train.py:874] (2/4) Epoch 18, batch 750, aishell_loss[loss=0.156, simple_loss=0.24, pruned_loss=0.03602, over 4935.00 frames.], tot_loss[loss=0.154, simple_loss=0.2349, pruned_loss=0.03657, over 962393.67 frames.], batch size: 49, aishell_tot_loss[loss=0.1546, simple_loss=0.2402, pruned_loss=0.03454, over 840825.61 frames.], datatang_tot_loss[loss=0.1516, simple_loss=0.227, pruned_loss=0.03813, over 828656.90 frames.], batch size: 49, lr: 4.58e-04
+2022-06-18 22:56:55,809 INFO [train.py:874] (2/4) Epoch 18, batch 800, aishell_loss[loss=0.1492, simple_loss=0.2368, pruned_loss=0.03084, over 4963.00 frames.], tot_loss[loss=0.1543, simple_loss=0.235, pruned_loss=0.03683, over 967669.98 frames.], batch size: 39, aishell_tot_loss[loss=0.1546, simple_loss=0.2401, pruned_loss=0.03452, over 856111.74 frames.], datatang_tot_loss[loss=0.1523, simple_loss=0.2277, pruned_loss=0.03844, over 849175.43 frames.], batch size: 39, lr: 4.58e-04
+2022-06-18 22:57:27,086 INFO [train.py:874] (2/4) Epoch 18, batch 850, aishell_loss[loss=0.1515, simple_loss=0.236, pruned_loss=0.03355, over 4979.00 frames.], tot_loss[loss=0.1537, simple_loss=0.2344, pruned_loss=0.03652, over 971541.11 frames.], batch size: 48, aishell_tot_loss[loss=0.1543, simple_loss=0.2397, pruned_loss=0.03448, over 871723.93 frames.], datatang_tot_loss[loss=0.152, simple_loss=0.2276, pruned_loss=0.03823, over 864749.26 frames.], batch size: 48, lr: 4.58e-04
+2022-06-18 22:57:56,138 INFO [train.py:874] (2/4) Epoch 18, batch 900, datatang_loss[loss=0.1356, simple_loss=0.2163, pruned_loss=0.02743, over 4925.00 frames.], tot_loss[loss=0.1547, simple_loss=0.2353, pruned_loss=0.03708, over 974667.94 frames.], batch size: 83, aishell_tot_loss[loss=0.1543, simple_loss=0.2398, pruned_loss=0.0344, over 884156.94 frames.], datatang_tot_loss[loss=0.1534, simple_loss=0.2289, pruned_loss=0.03895, over 880027.54 frames.], batch size: 83, lr: 4.58e-04
+2022-06-18 22:58:25,696 INFO [train.py:874] (2/4) Epoch 18, batch 950, datatang_loss[loss=0.1318, simple_loss=0.1992, pruned_loss=0.03218, over 4940.00 frames.], tot_loss[loss=0.1545, simple_loss=0.2348, pruned_loss=0.03706, over 977084.48 frames.], batch size: 50, aishell_tot_loss[loss=0.1547, simple_loss=0.2401, pruned_loss=0.03466, over 892811.88 frames.], datatang_tot_loss[loss=0.1529, simple_loss=0.2287, pruned_loss=0.03859, over 895764.19 frames.], batch size: 50, lr: 4.58e-04
+2022-06-18 22:58:57,121 INFO [train.py:874] (2/4) Epoch 18, batch 1000, aishell_loss[loss=0.1506, simple_loss=0.2396, pruned_loss=0.03085, over 4943.00 frames.], tot_loss[loss=0.1534, simple_loss=0.2343, pruned_loss=0.03627, over 978791.19 frames.], batch size: 54, aishell_tot_loss[loss=0.1537, simple_loss=0.2394, pruned_loss=0.03404, over 905914.68 frames.], datatang_tot_loss[loss=0.1529, simple_loss=0.2286, pruned_loss=0.03858, over 903986.04 frames.], batch size: 54, lr: 4.58e-04
+2022-06-18 22:58:57,122 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 22:59:14,039 INFO [train.py:914] (2/4) Epoch 18, validation: loss=0.1636, simple_loss=0.2479, pruned_loss=0.03969, over 1622729.00 frames.
+2022-06-18 22:59:44,457 INFO [train.py:874] (2/4) Epoch 18, batch 1050, datatang_loss[loss=0.1309, simple_loss=0.1988, pruned_loss=0.03152, over 4942.00 frames.], tot_loss[loss=0.1531, simple_loss=0.2339, pruned_loss=0.03621, over 980328.06 frames.], batch size: 50, aishell_tot_loss[loss=0.154, simple_loss=0.2397, pruned_loss=0.03408, over 914609.40 frames.], datatang_tot_loss[loss=0.1524, simple_loss=0.2279, pruned_loss=0.0384, over 914328.16 frames.], batch size: 50, lr: 4.58e-04
+2022-06-18 23:00:16,718 INFO [train.py:874] (2/4) Epoch 18, batch 1100, datatang_loss[loss=0.2323, simple_loss=0.2946, pruned_loss=0.08503, over 4943.00 frames.], tot_loss[loss=0.153, simple_loss=0.2331, pruned_loss=0.03647, over 981053.41 frames.], batch size: 109, aishell_tot_loss[loss=0.1538, simple_loss=0.2392, pruned_loss=0.03422, over 921859.13 frames.], datatang_tot_loss[loss=0.1523, simple_loss=0.2276, pruned_loss=0.03847, over 923364.74 frames.], batch size: 109, lr: 4.57e-04
+2022-06-18 23:00:44,250 INFO [train.py:874] (2/4) Epoch 18, batch 1150, datatang_loss[loss=0.1788, simple_loss=0.2557, pruned_loss=0.05095, over 4942.00 frames.], tot_loss[loss=0.1536, simple_loss=0.2335, pruned_loss=0.03686, over 982097.15 frames.], batch size: 37, aishell_tot_loss[loss=0.1537, simple_loss=0.2392, pruned_loss=0.03406, over 928580.91 frames.], datatang_tot_loss[loss=0.1531, simple_loss=0.2281, pruned_loss=0.03902, over 931481.92 frames.], batch size: 37, lr: 4.57e-04
+2022-06-18 23:01:15,668 INFO [train.py:874] (2/4) Epoch 18, batch 1200, aishell_loss[loss=0.1561, simple_loss=0.2223, pruned_loss=0.04496, over 4922.00 frames.], tot_loss[loss=0.1542, simple_loss=0.2343, pruned_loss=0.03699, over 983172.80 frames.], batch size: 25, aishell_tot_loss[loss=0.154, simple_loss=0.2397, pruned_loss=0.03421, over 936106.55 frames.], datatang_tot_loss[loss=0.1534, simple_loss=0.2285, pruned_loss=0.03915, over 937381.07 frames.], batch size: 25, lr: 4.57e-04
+2022-06-18 23:01:47,502 INFO [train.py:874] (2/4) Epoch 18, batch 1250, datatang_loss[loss=0.1357, simple_loss=0.2091, pruned_loss=0.03119, over 4922.00 frames.], tot_loss[loss=0.1538, simple_loss=0.2341, pruned_loss=0.03675, over 983527.18 frames.], batch size: 64, aishell_tot_loss[loss=0.1543, simple_loss=0.2397, pruned_loss=0.03443, over 941774.43 frames.], datatang_tot_loss[loss=0.1528, simple_loss=0.2281, pruned_loss=0.03877, over 943038.47 frames.], batch size: 64, lr: 4.57e-04
+2022-06-18 23:02:15,950 INFO [train.py:874] (2/4) Epoch 18, batch 1300, datatang_loss[loss=0.1484, simple_loss=0.2201, pruned_loss=0.03833, over 4926.00 frames.], tot_loss[loss=0.1539, simple_loss=0.234, pruned_loss=0.03696, over 983696.72 frames.], batch size: 83, aishell_tot_loss[loss=0.1539, simple_loss=0.2394, pruned_loss=0.03426, over 945743.02 frames.], datatang_tot_loss[loss=0.1535, simple_loss=0.2287, pruned_loss=0.03911, over 948881.62 frames.], batch size: 83, lr: 4.57e-04
+2022-06-18 23:02:45,770 INFO [train.py:874] (2/4) Epoch 18, batch 1350, aishell_loss[loss=0.1706, simple_loss=0.2516, pruned_loss=0.04486, over 4946.00 frames.], tot_loss[loss=0.1527, simple_loss=0.2331, pruned_loss=0.03615, over 984262.87 frames.], batch size: 64, aishell_tot_loss[loss=0.1536, simple_loss=0.2391, pruned_loss=0.03406, over 950901.48 frames.], datatang_tot_loss[loss=0.1525, simple_loss=0.2279, pruned_loss=0.03853, over 952923.29 frames.], batch size: 64, lr: 4.57e-04
+2022-06-18 23:03:17,853 INFO [train.py:874] (2/4) Epoch 18, batch 1400, datatang_loss[loss=0.1474, simple_loss=0.232, pruned_loss=0.03137, over 4924.00 frames.], tot_loss[loss=0.1519, simple_loss=0.2323, pruned_loss=0.03577, over 984952.25 frames.], batch size: 75, aishell_tot_loss[loss=0.1531, simple_loss=0.2386, pruned_loss=0.03386, over 955141.15 frames.], datatang_tot_loss[loss=0.152, simple_loss=0.2276, pruned_loss=0.03826, over 957036.38 frames.], batch size: 75, lr: 4.56e-04
+2022-06-18 23:03:45,939 INFO [train.py:874] (2/4) Epoch 18, batch 1450, aishell_loss[loss=0.1492, simple_loss=0.2461, pruned_loss=0.02613, over 4938.00 frames.], tot_loss[loss=0.153, simple_loss=0.2333, pruned_loss=0.03634, over 984938.96 frames.], batch size: 68, aishell_tot_loss[loss=0.1536, simple_loss=0.2391, pruned_loss=0.03405, over 958704.07 frames.], datatang_tot_loss[loss=0.1526, simple_loss=0.2279, pruned_loss=0.03861, over 960249.87 frames.], batch size: 68, lr: 4.56e-04
+2022-06-18 23:04:16,699 INFO [train.py:874] (2/4) Epoch 18, batch 1500, datatang_loss[loss=0.1417, simple_loss=0.2091, pruned_loss=0.03716, over 4929.00 frames.], tot_loss[loss=0.1533, simple_loss=0.2337, pruned_loss=0.03647, over 985384.75 frames.], batch size: 73, aishell_tot_loss[loss=0.1536, simple_loss=0.2391, pruned_loss=0.03405, over 962128.30 frames.], datatang_tot_loss[loss=0.1529, simple_loss=0.2282, pruned_loss=0.03878, over 963310.89 frames.], batch size: 73, lr: 4.56e-04
+2022-06-18 23:04:45,925 INFO [train.py:874] (2/4) Epoch 18, batch 1550, aishell_loss[loss=0.1538, simple_loss=0.2327, pruned_loss=0.03744, over 4936.00 frames.], tot_loss[loss=0.1534, simple_loss=0.2337, pruned_loss=0.03655, over 984971.54 frames.], batch size: 56, aishell_tot_loss[loss=0.1542, simple_loss=0.2395, pruned_loss=0.03444, over 964853.28 frames.], datatang_tot_loss[loss=0.1524, simple_loss=0.2276, pruned_loss=0.03859, over 965506.66 frames.], batch size: 56, lr: 4.56e-04
+2022-06-18 23:05:15,883 INFO [train.py:874] (2/4) Epoch 18, batch 1600, datatang_loss[loss=0.1509, simple_loss=0.2208, pruned_loss=0.04052, over 4879.00 frames.], tot_loss[loss=0.154, simple_loss=0.2344, pruned_loss=0.03681, over 985010.40 frames.], batch size: 39, aishell_tot_loss[loss=0.1542, simple_loss=0.2397, pruned_loss=0.0343, over 967363.49 frames.], datatang_tot_loss[loss=0.1531, simple_loss=0.2281, pruned_loss=0.03908, over 967716.90 frames.], batch size: 39, lr: 4.56e-04
+2022-06-18 23:05:47,596 INFO [train.py:874] (2/4) Epoch 18, batch 1650, datatang_loss[loss=0.1442, simple_loss=0.2222, pruned_loss=0.03306, over 4920.00 frames.], tot_loss[loss=0.1543, simple_loss=0.2344, pruned_loss=0.0371, over 984989.29 frames.], batch size: 75, aishell_tot_loss[loss=0.1544, simple_loss=0.2398, pruned_loss=0.03457, over 968952.39 frames.], datatang_tot_loss[loss=0.1533, simple_loss=0.2284, pruned_loss=0.03908, over 970213.18 frames.], batch size: 75, lr: 4.56e-04
+2022-06-18 23:06:22,682 INFO [train.py:874] (2/4) Epoch 18, batch 1700, datatang_loss[loss=0.1365, simple_loss=0.2084, pruned_loss=0.03233, over 4932.00 frames.], tot_loss[loss=0.1542, simple_loss=0.2347, pruned_loss=0.03679, over 985109.62 frames.], batch size: 71, aishell_tot_loss[loss=0.1551, simple_loss=0.2407, pruned_loss=0.03479, over 970724.02 frames.], datatang_tot_loss[loss=0.1526, simple_loss=0.2279, pruned_loss=0.03861, over 972206.31 frames.], batch size: 71, lr: 4.55e-04
+2022-06-18 23:06:51,253 INFO [train.py:874] (2/4) Epoch 18, batch 1750, aishell_loss[loss=0.1423, simple_loss=0.2274, pruned_loss=0.02863, over 4944.00 frames.], tot_loss[loss=0.1538, simple_loss=0.2344, pruned_loss=0.03665, over 985301.23 frames.], batch size: 45, aishell_tot_loss[loss=0.1551, simple_loss=0.2406, pruned_loss=0.03485, over 972527.10 frames.], datatang_tot_loss[loss=0.1523, simple_loss=0.2277, pruned_loss=0.03843, over 973837.49 frames.], batch size: 45, lr: 4.55e-04
+2022-06-18 23:07:22,827 INFO [train.py:874] (2/4) Epoch 18, batch 1800, datatang_loss[loss=0.1368, simple_loss=0.2058, pruned_loss=0.0339, over 4940.00 frames.], tot_loss[loss=0.1534, simple_loss=0.2337, pruned_loss=0.03659, over 985535.39 frames.], batch size: 45, aishell_tot_loss[loss=0.1555, simple_loss=0.2409, pruned_loss=0.03501, over 973961.90 frames.], datatang_tot_loss[loss=0.1516, simple_loss=0.2269, pruned_loss=0.03814, over 975482.12 frames.], batch size: 45, lr: 4.55e-04
+2022-06-18 23:07:53,158 INFO [train.py:874] (2/4) Epoch 18, batch 1850, aishell_loss[loss=0.1597, simple_loss=0.2352, pruned_loss=0.04208, over 4940.00 frames.], tot_loss[loss=0.1542, simple_loss=0.2343, pruned_loss=0.03705, over 985360.12 frames.], batch size: 32, aishell_tot_loss[loss=0.1556, simple_loss=0.2409, pruned_loss=0.03516, over 975096.67 frames.], datatang_tot_loss[loss=0.1523, simple_loss=0.2275, pruned_loss=0.03852, over 976727.39 frames.], batch size: 32, lr: 4.55e-04
+2022-06-18 23:08:22,176 INFO [train.py:874] (2/4) Epoch 18, batch 1900, aishell_loss[loss=0.16, simple_loss=0.252, pruned_loss=0.03394, over 4984.00 frames.], tot_loss[loss=0.1535, simple_loss=0.2336, pruned_loss=0.03676, over 985412.88 frames.], batch size: 43, aishell_tot_loss[loss=0.1554, simple_loss=0.2405, pruned_loss=0.03513, over 976315.02 frames.], datatang_tot_loss[loss=0.1518, simple_loss=0.227, pruned_loss=0.03834, over 977811.85 frames.], batch size: 43, lr: 4.55e-04
+2022-06-18 23:08:53,182 INFO [train.py:874] (2/4) Epoch 18, batch 1950, datatang_loss[loss=0.1216, simple_loss=0.193, pruned_loss=0.02513, over 4975.00 frames.], tot_loss[loss=0.1534, simple_loss=0.2329, pruned_loss=0.03701, over 985369.79 frames.], batch size: 45, aishell_tot_loss[loss=0.1548, simple_loss=0.2398, pruned_loss=0.03493, over 976641.93 frames.], datatang_tot_loss[loss=0.1524, simple_loss=0.2274, pruned_loss=0.03871, over 979323.36 frames.], batch size: 45, lr: 4.55e-04
+2022-06-18 23:09:24,291 INFO [train.py:874] (2/4) Epoch 18, batch 2000, datatang_loss[loss=0.1743, simple_loss=0.252, pruned_loss=0.04834, over 4963.00 frames.], tot_loss[loss=0.1539, simple_loss=0.2335, pruned_loss=0.03718, over 985487.08 frames.], batch size: 86, aishell_tot_loss[loss=0.155, simple_loss=0.24, pruned_loss=0.03501, over 977516.98 frames.], datatang_tot_loss[loss=0.1527, simple_loss=0.2278, pruned_loss=0.03883, over 980299.53 frames.], batch size: 86, lr: 4.55e-04
+2022-06-18 23:09:24,292 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 23:09:40,066 INFO [train.py:914] (2/4) Epoch 18, validation: loss=0.1648, simple_loss=0.2489, pruned_loss=0.0403, over 1622729.00 frames.
+2022-06-18 23:10:11,559 INFO [train.py:874] (2/4) Epoch 18, batch 2050, aishell_loss[loss=0.1712, simple_loss=0.2587, pruned_loss=0.04188, over 4915.00 frames.], tot_loss[loss=0.1541, simple_loss=0.2336, pruned_loss=0.03728, over 985298.15 frames.], batch size: 68, aishell_tot_loss[loss=0.1552, simple_loss=0.2402, pruned_loss=0.03512, over 978171.00 frames.], datatang_tot_loss[loss=0.1527, simple_loss=0.2276, pruned_loss=0.03888, over 981007.22 frames.], batch size: 68, lr: 4.54e-04
+2022-06-18 23:10:39,544 INFO [train.py:874] (2/4) Epoch 18, batch 2100, aishell_loss[loss=0.1615, simple_loss=0.2533, pruned_loss=0.03488, over 4962.00 frames.], tot_loss[loss=0.1539, simple_loss=0.2336, pruned_loss=0.03715, over 985204.47 frames.], batch size: 40, aishell_tot_loss[loss=0.1552, simple_loss=0.2401, pruned_loss=0.03513, over 978771.80 frames.], datatang_tot_loss[loss=0.1526, simple_loss=0.2275, pruned_loss=0.0388, over 981672.20 frames.], batch size: 40, lr: 4.54e-04
+2022-06-18 23:11:11,040 INFO [train.py:874] (2/4) Epoch 18, batch 2150, datatang_loss[loss=0.1488, simple_loss=0.2238, pruned_loss=0.03693, over 4943.00 frames.], tot_loss[loss=0.1546, simple_loss=0.2344, pruned_loss=0.03742, over 985640.58 frames.], batch size: 62, aishell_tot_loss[loss=0.1553, simple_loss=0.2404, pruned_loss=0.03509, over 979789.41 frames.], datatang_tot_loss[loss=0.1532, simple_loss=0.2279, pruned_loss=0.03918, over 982289.07 frames.], batch size: 62, lr: 4.54e-04
+2022-06-18 23:11:42,801 INFO [train.py:874] (2/4) Epoch 18, batch 2200, aishell_loss[loss=0.1433, simple_loss=0.2306, pruned_loss=0.02803, over 4977.00 frames.], tot_loss[loss=0.1546, simple_loss=0.2346, pruned_loss=0.03735, over 985805.43 frames.], batch size: 51, aishell_tot_loss[loss=0.1556, simple_loss=0.2409, pruned_loss=0.03514, over 980535.20 frames.], datatang_tot_loss[loss=0.153, simple_loss=0.2277, pruned_loss=0.03909, over 982785.86 frames.], batch size: 51, lr: 4.54e-04
+2022-06-18 23:12:10,475 INFO [train.py:874] (2/4) Epoch 18, batch 2250, datatang_loss[loss=0.1357, simple_loss=0.2141, pruned_loss=0.02864, over 4941.00 frames.], tot_loss[loss=0.1533, simple_loss=0.2337, pruned_loss=0.0364, over 985901.32 frames.], batch size: 62, aishell_tot_loss[loss=0.1549, simple_loss=0.2405, pruned_loss=0.03464, over 981013.30 frames.], datatang_tot_loss[loss=0.1523, simple_loss=0.2275, pruned_loss=0.0386, over 983341.27 frames.], batch size: 62, lr: 4.54e-04
+2022-06-18 23:12:42,756 INFO [train.py:874] (2/4) Epoch 18, batch 2300, datatang_loss[loss=0.1533, simple_loss=0.2298, pruned_loss=0.03837, over 4959.00 frames.], tot_loss[loss=0.1536, simple_loss=0.2338, pruned_loss=0.03672, over 986229.99 frames.], batch size: 67, aishell_tot_loss[loss=0.155, simple_loss=0.2405, pruned_loss=0.0347, over 981726.20 frames.], datatang_tot_loss[loss=0.1526, simple_loss=0.2276, pruned_loss=0.03878, over 983838.36 frames.], batch size: 67, lr: 4.54e-04
+2022-06-18 23:13:13,532 INFO [train.py:874] (2/4) Epoch 18, batch 2350, aishell_loss[loss=0.1696, simple_loss=0.2572, pruned_loss=0.04097, over 4878.00 frames.], tot_loss[loss=0.1539, simple_loss=0.2341, pruned_loss=0.03681, over 986154.15 frames.], batch size: 35, aishell_tot_loss[loss=0.1552, simple_loss=0.2405, pruned_loss=0.035, over 982282.56 frames.], datatang_tot_loss[loss=0.1525, simple_loss=0.2278, pruned_loss=0.03859, over 984038.71 frames.], batch size: 35, lr: 4.53e-04
+2022-06-18 23:13:42,678 INFO [train.py:874] (2/4) Epoch 18, batch 2400, datatang_loss[loss=0.2127, simple_loss=0.2718, pruned_loss=0.07678, over 4957.00 frames.], tot_loss[loss=0.1534, simple_loss=0.2333, pruned_loss=0.0368, over 985537.80 frames.], batch size: 110, aishell_tot_loss[loss=0.1544, simple_loss=0.2396, pruned_loss=0.03463, over 982276.74 frames.], datatang_tot_loss[loss=0.1528, simple_loss=0.2278, pruned_loss=0.03895, over 984102.54 frames.], batch size: 110, lr: 4.53e-04
+2022-06-18 23:14:13,859 INFO [train.py:874] (2/4) Epoch 18, batch 2450, aishell_loss[loss=0.1598, simple_loss=0.2483, pruned_loss=0.03564, over 4941.00 frames.], tot_loss[loss=0.1539, simple_loss=0.2344, pruned_loss=0.03669, over 985305.55 frames.], batch size: 45, aishell_tot_loss[loss=0.1548, simple_loss=0.2403, pruned_loss=0.03467, over 982462.33 frames.], datatang_tot_loss[loss=0.1528, simple_loss=0.2279, pruned_loss=0.0389, over 984296.26 frames.], batch size: 45, lr: 4.53e-04
+2022-06-18 23:14:43,646 INFO [train.py:874] (2/4) Epoch 18, batch 2500, aishell_loss[loss=0.1572, simple_loss=0.2429, pruned_loss=0.0357, over 4955.00 frames.], tot_loss[loss=0.1537, simple_loss=0.2344, pruned_loss=0.03654, over 985471.46 frames.], batch size: 40, aishell_tot_loss[loss=0.1545, simple_loss=0.24, pruned_loss=0.03454, over 982871.57 frames.], datatang_tot_loss[loss=0.1529, simple_loss=0.2278, pruned_loss=0.03899, over 984591.00 frames.], batch size: 40, lr: 4.53e-04
+2022-06-18 23:15:12,786 INFO [train.py:874] (2/4) Epoch 18, batch 2550, datatang_loss[loss=0.1613, simple_loss=0.2393, pruned_loss=0.0416, over 4913.00 frames.], tot_loss[loss=0.1539, simple_loss=0.2348, pruned_loss=0.03647, over 985443.68 frames.], batch size: 98, aishell_tot_loss[loss=0.1546, simple_loss=0.2402, pruned_loss=0.03455, over 982990.50 frames.], datatang_tot_loss[loss=0.1529, simple_loss=0.2279, pruned_loss=0.03895, over 984876.32 frames.], batch size: 98, lr: 4.53e-04
+2022-06-18 23:15:44,502 INFO [train.py:874] (2/4) Epoch 18, batch 2600, aishell_loss[loss=0.1566, simple_loss=0.2536, pruned_loss=0.02982, over 4915.00 frames.], tot_loss[loss=0.1525, simple_loss=0.2331, pruned_loss=0.03601, over 985061.49 frames.], batch size: 68, aishell_tot_loss[loss=0.1543, simple_loss=0.2396, pruned_loss=0.03451, over 983222.92 frames.], datatang_tot_loss[loss=0.1518, simple_loss=0.2268, pruned_loss=0.03844, over 984604.64 frames.], batch size: 68, lr: 4.53e-04
+2022-06-18 23:16:12,045 INFO [train.py:874] (2/4) Epoch 18, batch 2650, aishell_loss[loss=0.1573, simple_loss=0.244, pruned_loss=0.03527, over 4984.00 frames.], tot_loss[loss=0.1522, simple_loss=0.2329, pruned_loss=0.03576, over 985034.99 frames.], batch size: 39, aishell_tot_loss[loss=0.1543, simple_loss=0.2398, pruned_loss=0.03439, over 983324.99 frames.], datatang_tot_loss[loss=0.1514, simple_loss=0.2264, pruned_loss=0.03818, over 984744.37 frames.], batch size: 39, lr: 4.52e-04
+2022-06-18 23:16:42,926 INFO [train.py:874] (2/4) Epoch 18, batch 2700, aishell_loss[loss=0.1484, simple_loss=0.232, pruned_loss=0.0324, over 4861.00 frames.], tot_loss[loss=0.1529, simple_loss=0.2332, pruned_loss=0.03626, over 985072.28 frames.], batch size: 37, aishell_tot_loss[loss=0.1543, simple_loss=0.2397, pruned_loss=0.03446, over 983302.72 frames.], datatang_tot_loss[loss=0.1519, simple_loss=0.2268, pruned_loss=0.0385, over 985038.81 frames.], batch size: 37, lr: 4.52e-04
+2022-06-18 23:17:14,054 INFO [train.py:874] (2/4) Epoch 18, batch 2750, aishell_loss[loss=0.1427, simple_loss=0.2353, pruned_loss=0.02504, over 4964.00 frames.], tot_loss[loss=0.1521, simple_loss=0.2324, pruned_loss=0.03589, over 985098.98 frames.], batch size: 44, aishell_tot_loss[loss=0.1541, simple_loss=0.2395, pruned_loss=0.03439, over 983383.01 frames.], datatang_tot_loss[loss=0.1512, simple_loss=0.2264, pruned_loss=0.03805, over 985145.76 frames.], batch size: 44, lr: 4.52e-04
+2022-06-18 23:17:42,081 INFO [train.py:874] (2/4) Epoch 18, batch 2800, aishell_loss[loss=0.1537, simple_loss=0.2478, pruned_loss=0.02983, over 4908.00 frames.], tot_loss[loss=0.1514, simple_loss=0.2318, pruned_loss=0.03552, over 985267.10 frames.], batch size: 52, aishell_tot_loss[loss=0.1537, simple_loss=0.2391, pruned_loss=0.03414, over 983582.61 frames.], datatang_tot_loss[loss=0.1508, simple_loss=0.226, pruned_loss=0.03783, over 985340.09 frames.], batch size: 52, lr: 4.52e-04
+2022-06-18 23:18:13,839 INFO [train.py:874] (2/4) Epoch 18, batch 2850, aishell_loss[loss=0.1467, simple_loss=0.2382, pruned_loss=0.02761, over 4947.00 frames.], tot_loss[loss=0.1512, simple_loss=0.2314, pruned_loss=0.03548, over 985153.14 frames.], batch size: 64, aishell_tot_loss[loss=0.1534, simple_loss=0.2389, pruned_loss=0.03402, over 983806.33 frames.], datatang_tot_loss[loss=0.1506, simple_loss=0.2257, pruned_loss=0.0378, over 985190.54 frames.], batch size: 64, lr: 4.52e-04
+2022-06-18 23:18:44,553 INFO [train.py:874] (2/4) Epoch 18, batch 2900, aishell_loss[loss=0.1686, simple_loss=0.256, pruned_loss=0.04055, over 4949.00 frames.], tot_loss[loss=0.1518, simple_loss=0.2315, pruned_loss=0.03601, over 985275.00 frames.], batch size: 79, aishell_tot_loss[loss=0.1538, simple_loss=0.239, pruned_loss=0.0343, over 983780.36 frames.], datatang_tot_loss[loss=0.1508, simple_loss=0.2256, pruned_loss=0.03797, over 985492.47 frames.], batch size: 79, lr: 4.52e-04
+2022-06-18 23:19:14,110 INFO [train.py:874] (2/4) Epoch 18, batch 2950, datatang_loss[loss=0.1527, simple_loss=0.2334, pruned_loss=0.036, over 4883.00 frames.], tot_loss[loss=0.1518, simple_loss=0.2316, pruned_loss=0.03605, over 985668.88 frames.], batch size: 47, aishell_tot_loss[loss=0.1544, simple_loss=0.2397, pruned_loss=0.03452, over 984251.31 frames.], datatang_tot_loss[loss=0.1503, simple_loss=0.2253, pruned_loss=0.03766, over 985557.44 frames.], batch size: 47, lr: 4.52e-04
+2022-06-18 23:19:45,376 INFO [train.py:874] (2/4) Epoch 18, batch 3000, aishell_loss[loss=0.1536, simple_loss=0.2491, pruned_loss=0.02903, over 4915.00 frames.], tot_loss[loss=0.1516, simple_loss=0.2315, pruned_loss=0.03583, over 985398.78 frames.], batch size: 46, aishell_tot_loss[loss=0.1541, simple_loss=0.2396, pruned_loss=0.03424, over 984288.73 frames.], datatang_tot_loss[loss=0.1504, simple_loss=0.2255, pruned_loss=0.03763, over 985389.29 frames.], batch size: 46, lr: 4.51e-04
+2022-06-18 23:19:45,377 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 23:20:02,473 INFO [train.py:914] (2/4) Epoch 18, validation: loss=0.165, simple_loss=0.2484, pruned_loss=0.04079, over 1622729.00 frames.
+2022-06-18 23:20:31,594 INFO [train.py:874] (2/4) Epoch 18, batch 3050, datatang_loss[loss=0.1633, simple_loss=0.2216, pruned_loss=0.05248, over 4980.00 frames.], tot_loss[loss=0.1516, simple_loss=0.2317, pruned_loss=0.0358, over 985448.84 frames.], batch size: 34, aishell_tot_loss[loss=0.1544, simple_loss=0.24, pruned_loss=0.03438, over 984321.56 frames.], datatang_tot_loss[loss=0.1499, simple_loss=0.225, pruned_loss=0.03744, over 985537.50 frames.], batch size: 34, lr: 4.51e-04
+2022-06-18 23:21:04,152 INFO [train.py:874] (2/4) Epoch 18, batch 3100, datatang_loss[loss=0.1404, simple_loss=0.2236, pruned_loss=0.02862, over 4955.00 frames.], tot_loss[loss=0.1519, simple_loss=0.2321, pruned_loss=0.03582, over 985402.96 frames.], batch size: 86, aishell_tot_loss[loss=0.1542, simple_loss=0.2398, pruned_loss=0.03429, over 984509.65 frames.], datatang_tot_loss[loss=0.1503, simple_loss=0.2255, pruned_loss=0.03754, over 985430.27 frames.], batch size: 86, lr: 4.51e-04
+2022-06-18 23:21:34,427 INFO [train.py:874] (2/4) Epoch 18, batch 3150, datatang_loss[loss=0.1304, simple_loss=0.2099, pruned_loss=0.0254, over 4916.00 frames.], tot_loss[loss=0.1524, simple_loss=0.2323, pruned_loss=0.0363, over 985776.61 frames.], batch size: 77, aishell_tot_loss[loss=0.1541, simple_loss=0.2395, pruned_loss=0.03431, over 985015.32 frames.], datatang_tot_loss[loss=0.1509, simple_loss=0.2257, pruned_loss=0.03805, over 985429.44 frames.], batch size: 77, lr: 4.51e-04
+2022-06-18 23:22:05,202 INFO [train.py:874] (2/4) Epoch 18, batch 3200, datatang_loss[loss=0.1505, simple_loss=0.2222, pruned_loss=0.03944, over 4925.00 frames.], tot_loss[loss=0.1528, simple_loss=0.2322, pruned_loss=0.03665, over 985443.53 frames.], batch size: 73, aishell_tot_loss[loss=0.1545, simple_loss=0.2399, pruned_loss=0.0345, over 984651.95 frames.], datatang_tot_loss[loss=0.151, simple_loss=0.2256, pruned_loss=0.03818, over 985559.31 frames.], batch size: 73, lr: 4.51e-04
+2022-06-18 23:22:37,282 INFO [train.py:874] (2/4) Epoch 18, batch 3250, datatang_loss[loss=0.1419, simple_loss=0.2182, pruned_loss=0.03282, over 4933.00 frames.], tot_loss[loss=0.1528, simple_loss=0.2323, pruned_loss=0.03662, over 985597.87 frames.], batch size: 73, aishell_tot_loss[loss=0.1545, simple_loss=0.2402, pruned_loss=0.03444, over 984900.63 frames.], datatang_tot_loss[loss=0.151, simple_loss=0.2254, pruned_loss=0.03826, over 985554.32 frames.], batch size: 73, lr: 4.51e-04
+2022-06-18 23:23:06,593 INFO [train.py:874] (2/4) Epoch 18, batch 3300, aishell_loss[loss=0.149, simple_loss=0.2376, pruned_loss=0.03017, over 4945.00 frames.], tot_loss[loss=0.1527, simple_loss=0.2327, pruned_loss=0.03636, over 985588.05 frames.], batch size: 45, aishell_tot_loss[loss=0.1546, simple_loss=0.2404, pruned_loss=0.03444, over 984882.75 frames.], datatang_tot_loss[loss=0.1508, simple_loss=0.2255, pruned_loss=0.03805, over 985649.35 frames.], batch size: 45, lr: 4.50e-04
+2022-06-18 23:23:37,937 INFO [train.py:874] (2/4) Epoch 18, batch 3350, datatang_loss[loss=0.1271, simple_loss=0.2075, pruned_loss=0.0233, over 4940.00 frames.], tot_loss[loss=0.1521, simple_loss=0.2318, pruned_loss=0.03616, over 985713.30 frames.], batch size: 79, aishell_tot_loss[loss=0.1543, simple_loss=0.2398, pruned_loss=0.03441, over 985023.66 frames.], datatang_tot_loss[loss=0.1505, simple_loss=0.2252, pruned_loss=0.03786, over 985707.99 frames.], batch size: 79, lr: 4.50e-04
+2022-06-18 23:24:09,440 INFO [train.py:874] (2/4) Epoch 18, batch 3400, datatang_loss[loss=0.1663, simple_loss=0.247, pruned_loss=0.0428, over 4926.00 frames.], tot_loss[loss=0.1522, simple_loss=0.2318, pruned_loss=0.03631, over 985796.66 frames.], batch size: 94, aishell_tot_loss[loss=0.1543, simple_loss=0.2398, pruned_loss=0.03439, over 985020.03 frames.], datatang_tot_loss[loss=0.1507, simple_loss=0.2254, pruned_loss=0.03795, over 985862.17 frames.], batch size: 94, lr: 4.50e-04
+2022-06-18 23:24:39,078 INFO [train.py:874] (2/4) Epoch 18, batch 3450, datatang_loss[loss=0.1319, simple_loss=0.2039, pruned_loss=0.02997, over 4933.00 frames.], tot_loss[loss=0.1527, simple_loss=0.2323, pruned_loss=0.03659, over 986090.46 frames.], batch size: 71, aishell_tot_loss[loss=0.1547, simple_loss=0.2402, pruned_loss=0.03456, over 985307.34 frames.], datatang_tot_loss[loss=0.1508, simple_loss=0.2256, pruned_loss=0.03803, over 985951.51 frames.], batch size: 71, lr: 4.50e-04
+2022-06-18 23:25:10,231 INFO [train.py:874] (2/4) Epoch 18, batch 3500, aishell_loss[loss=0.1465, simple_loss=0.233, pruned_loss=0.03003, over 4873.00 frames.], tot_loss[loss=0.1528, simple_loss=0.2326, pruned_loss=0.03656, over 985538.10 frames.], batch size: 35, aishell_tot_loss[loss=0.1547, simple_loss=0.2402, pruned_loss=0.03457, over 984771.33 frames.], datatang_tot_loss[loss=0.151, simple_loss=0.226, pruned_loss=0.038, over 986000.36 frames.], batch size: 35, lr: 4.50e-04
+2022-06-18 23:25:41,313 INFO [train.py:874] (2/4) Epoch 18, batch 3550, aishell_loss[loss=0.1389, simple_loss=0.2322, pruned_loss=0.02282, over 4930.00 frames.], tot_loss[loss=0.1528, simple_loss=0.2326, pruned_loss=0.03654, over 985564.55 frames.], batch size: 58, aishell_tot_loss[loss=0.155, simple_loss=0.2405, pruned_loss=0.03476, over 985071.35 frames.], datatang_tot_loss[loss=0.1507, simple_loss=0.2256, pruned_loss=0.03784, over 985767.96 frames.], batch size: 58, lr: 4.50e-04
+2022-06-18 23:26:10,926 INFO [train.py:874] (2/4) Epoch 18, batch 3600, aishell_loss[loss=0.1299, simple_loss=0.2151, pruned_loss=0.02238, over 4975.00 frames.], tot_loss[loss=0.1526, simple_loss=0.2323, pruned_loss=0.03645, over 985553.37 frames.], batch size: 30, aishell_tot_loss[loss=0.1546, simple_loss=0.2401, pruned_loss=0.03455, over 985141.02 frames.], datatang_tot_loss[loss=0.1508, simple_loss=0.2255, pruned_loss=0.03804, over 985724.62 frames.], batch size: 30, lr: 4.50e-04
+2022-06-18 23:26:42,441 INFO [train.py:874] (2/4) Epoch 18, batch 3650, aishell_loss[loss=0.1643, simple_loss=0.2519, pruned_loss=0.03834, over 4927.00 frames.], tot_loss[loss=0.1526, simple_loss=0.2325, pruned_loss=0.03636, over 985368.53 frames.], batch size: 52, aishell_tot_loss[loss=0.1553, simple_loss=0.2407, pruned_loss=0.03489, over 984971.21 frames.], datatang_tot_loss[loss=0.1501, simple_loss=0.225, pruned_loss=0.03763, over 985719.13 frames.], batch size: 52, lr: 4.49e-04
+2022-06-18 23:27:14,774 INFO [train.py:874] (2/4) Epoch 18, batch 3700, aishell_loss[loss=0.1536, simple_loss=0.2475, pruned_loss=0.02979, over 4986.00 frames.], tot_loss[loss=0.1525, simple_loss=0.2328, pruned_loss=0.03613, over 985321.28 frames.], batch size: 51, aishell_tot_loss[loss=0.1552, simple_loss=0.2409, pruned_loss=0.03474, over 984889.75 frames.], datatang_tot_loss[loss=0.1501, simple_loss=0.2251, pruned_loss=0.03756, over 985757.14 frames.], batch size: 51, lr: 4.49e-04
+2022-06-18 23:27:43,237 INFO [train.py:874] (2/4) Epoch 18, batch 3750, aishell_loss[loss=0.1431, simple_loss=0.1975, pruned_loss=0.04434, over 4969.00 frames.], tot_loss[loss=0.1526, simple_loss=0.2333, pruned_loss=0.03595, over 985808.12 frames.], batch size: 21, aishell_tot_loss[loss=0.1549, simple_loss=0.2409, pruned_loss=0.03447, over 985216.45 frames.], datatang_tot_loss[loss=0.1503, simple_loss=0.2253, pruned_loss=0.03769, over 985956.35 frames.], batch size: 21, lr: 4.49e-04
+2022-06-18 23:28:15,699 INFO [train.py:874] (2/4) Epoch 18, batch 3800, datatang_loss[loss=0.1352, simple_loss=0.2103, pruned_loss=0.03002, over 4952.00 frames.], tot_loss[loss=0.1524, simple_loss=0.2333, pruned_loss=0.03573, over 985710.50 frames.], batch size: 25, aishell_tot_loss[loss=0.1543, simple_loss=0.2401, pruned_loss=0.03424, over 985148.57 frames.], datatang_tot_loss[loss=0.1506, simple_loss=0.2256, pruned_loss=0.03775, over 985990.33 frames.], batch size: 25, lr: 4.49e-04
+2022-06-18 23:28:45,427 INFO [train.py:874] (2/4) Epoch 18, batch 3850, datatang_loss[loss=0.1489, simple_loss=0.227, pruned_loss=0.03535, over 4919.00 frames.], tot_loss[loss=0.1538, simple_loss=0.2347, pruned_loss=0.03647, over 985713.85 frames.], batch size: 64, aishell_tot_loss[loss=0.155, simple_loss=0.2406, pruned_loss=0.03465, over 985234.75 frames.], datatang_tot_loss[loss=0.1513, simple_loss=0.2264, pruned_loss=0.03814, over 985961.33 frames.], batch size: 64, lr: 4.49e-04
+2022-06-18 23:29:15,480 INFO [train.py:874] (2/4) Epoch 18, batch 3900, datatang_loss[loss=0.2048, simple_loss=0.275, pruned_loss=0.0673, over 4933.00 frames.], tot_loss[loss=0.1533, simple_loss=0.2339, pruned_loss=0.03632, over 985476.92 frames.], batch size: 109, aishell_tot_loss[loss=0.1549, simple_loss=0.2403, pruned_loss=0.03477, over 985125.26 frames.], datatang_tot_loss[loss=0.1509, simple_loss=0.2261, pruned_loss=0.03789, over 985848.01 frames.], batch size: 109, lr: 4.49e-04
+2022-06-18 23:29:45,384 INFO [train.py:874] (2/4) Epoch 18, batch 3950, datatang_loss[loss=0.1675, simple_loss=0.242, pruned_loss=0.0465, over 4961.00 frames.], tot_loss[loss=0.1539, simple_loss=0.2343, pruned_loss=0.03669, over 985700.99 frames.], batch size: 34, aishell_tot_loss[loss=0.1553, simple_loss=0.2407, pruned_loss=0.03494, over 985381.82 frames.], datatang_tot_loss[loss=0.1514, simple_loss=0.2266, pruned_loss=0.03809, over 985821.71 frames.], batch size: 34, lr: 4.49e-04
+2022-06-18 23:30:15,531 INFO [train.py:874] (2/4) Epoch 18, batch 4000, aishell_loss[loss=0.1871, simple_loss=0.2815, pruned_loss=0.04633, over 4969.00 frames.], tot_loss[loss=0.1535, simple_loss=0.2339, pruned_loss=0.03651, over 985920.32 frames.], batch size: 80, aishell_tot_loss[loss=0.1555, simple_loss=0.2408, pruned_loss=0.03506, over 985574.82 frames.], datatang_tot_loss[loss=0.1509, simple_loss=0.2264, pruned_loss=0.03776, over 985880.17 frames.], batch size: 80, lr: 4.48e-04
+2022-06-18 23:30:15,532 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 23:30:33,176 INFO [train.py:914] (2/4) Epoch 18, validation: loss=0.1643, simple_loss=0.249, pruned_loss=0.03975, over 1622729.00 frames.
+2022-06-18 23:31:01,710 INFO [train.py:874] (2/4) Epoch 18, batch 4050, aishell_loss[loss=0.1372, simple_loss=0.2246, pruned_loss=0.02486, over 4913.00 frames.], tot_loss[loss=0.1533, simple_loss=0.2338, pruned_loss=0.03645, over 985918.91 frames.], batch size: 52, aishell_tot_loss[loss=0.1555, simple_loss=0.2408, pruned_loss=0.03513, over 985644.23 frames.], datatang_tot_loss[loss=0.1508, simple_loss=0.2263, pruned_loss=0.03765, over 985854.88 frames.], batch size: 52, lr: 4.48e-04
+2022-06-18 23:31:31,893 INFO [train.py:874] (2/4) Epoch 18, batch 4100, datatang_loss[loss=0.1633, simple_loss=0.2362, pruned_loss=0.04527, over 4917.00 frames.], tot_loss[loss=0.1532, simple_loss=0.2335, pruned_loss=0.03644, over 985984.60 frames.], batch size: 64, aishell_tot_loss[loss=0.1553, simple_loss=0.2407, pruned_loss=0.03496, over 985805.36 frames.], datatang_tot_loss[loss=0.1509, simple_loss=0.2263, pruned_loss=0.03777, over 985800.96 frames.], batch size: 64, lr: 4.48e-04
+2022-06-18 23:32:01,290 INFO [train.py:874] (2/4) Epoch 18, batch 4150, aishell_loss[loss=0.1945, simple_loss=0.2743, pruned_loss=0.05732, over 4939.00 frames.], tot_loss[loss=0.1534, simple_loss=0.2336, pruned_loss=0.03658, over 985474.02 frames.], batch size: 68, aishell_tot_loss[loss=0.156, simple_loss=0.2415, pruned_loss=0.03526, over 985619.21 frames.], datatang_tot_loss[loss=0.1505, simple_loss=0.2258, pruned_loss=0.03762, over 985493.51 frames.], batch size: 68, lr: 4.48e-04
+2022-06-18 23:32:31,802 INFO [train.py:874] (2/4) Epoch 18, batch 4200, datatang_loss[loss=0.1145, simple_loss=0.1804, pruned_loss=0.0243, over 4957.00 frames.], tot_loss[loss=0.1534, simple_loss=0.2335, pruned_loss=0.03665, over 985128.17 frames.], batch size: 31, aishell_tot_loss[loss=0.1563, simple_loss=0.2417, pruned_loss=0.03542, over 985175.53 frames.], datatang_tot_loss[loss=0.1503, simple_loss=0.2254, pruned_loss=0.03756, over 985547.46 frames.], batch size: 31, lr: 4.48e-04
+2022-06-18 23:33:39,356 INFO [train.py:874] (2/4) Epoch 19, batch 50, aishell_loss[loss=0.1442, simple_loss=0.2289, pruned_loss=0.02971, over 4937.00 frames.], tot_loss[loss=0.1455, simple_loss=0.2245, pruned_loss=0.03329, over 218296.24 frames.], batch size: 33, aishell_tot_loss[loss=0.1505, simple_loss=0.2364, pruned_loss=0.03233, over 111585.56 frames.], datatang_tot_loss[loss=0.1408, simple_loss=0.2136, pruned_loss=0.034, over 120341.46 frames.], batch size: 33, lr: 4.36e-04
+2022-06-18 23:34:10,845 INFO [train.py:874] (2/4) Epoch 19, batch 100, aishell_loss[loss=0.1421, simple_loss=0.2235, pruned_loss=0.03031, over 4924.00 frames.], tot_loss[loss=0.1452, simple_loss=0.2248, pruned_loss=0.03278, over 388291.00 frames.], batch size: 32, aishell_tot_loss[loss=0.15, simple_loss=0.2348, pruned_loss=0.03265, over 214317.76 frames.], datatang_tot_loss[loss=0.1404, simple_loss=0.2151, pruned_loss=0.03286, over 222346.44 frames.], batch size: 32, lr: 4.36e-04
+2022-06-18 23:34:43,807 INFO [train.py:874] (2/4) Epoch 19, batch 150, datatang_loss[loss=0.1421, simple_loss=0.2247, pruned_loss=0.0298, over 4974.00 frames.], tot_loss[loss=0.1439, simple_loss=0.2242, pruned_loss=0.0318, over 520803.78 frames.], batch size: 45, aishell_tot_loss[loss=0.15, simple_loss=0.235, pruned_loss=0.03249, over 298332.16 frames.], datatang_tot_loss[loss=0.1384, simple_loss=0.2142, pruned_loss=0.03127, over 319037.96 frames.], batch size: 45, lr: 4.36e-04
+2022-06-18 23:35:13,701 INFO [train.py:874] (2/4) Epoch 19, batch 200, aishell_loss[loss=0.1666, simple_loss=0.2477, pruned_loss=0.04275, over 4951.00 frames.], tot_loss[loss=0.1469, simple_loss=0.2267, pruned_loss=0.03353, over 624187.44 frames.], batch size: 56, aishell_tot_loss[loss=0.1516, simple_loss=0.2366, pruned_loss=0.03324, over 370275.73 frames.], datatang_tot_loss[loss=0.1419, simple_loss=0.2171, pruned_loss=0.03336, over 406512.31 frames.], batch size: 56, lr: 4.36e-04
+2022-06-18 23:35:44,911 INFO [train.py:874] (2/4) Epoch 19, batch 250, datatang_loss[loss=0.1334, simple_loss=0.2076, pruned_loss=0.02963, over 4921.00 frames.], tot_loss[loss=0.1472, simple_loss=0.2273, pruned_loss=0.03362, over 704279.90 frames.], batch size: 81, aishell_tot_loss[loss=0.1517, simple_loss=0.2368, pruned_loss=0.03335, over 445273.47 frames.], datatang_tot_loss[loss=0.1423, simple_loss=0.2175, pruned_loss=0.03351, over 472316.56 frames.], batch size: 81, lr: 4.36e-04
+2022-06-18 23:36:17,276 INFO [train.py:874] (2/4) Epoch 19, batch 300, aishell_loss[loss=0.1477, simple_loss=0.2337, pruned_loss=0.03091, over 4878.00 frames.], tot_loss[loss=0.1477, simple_loss=0.2279, pruned_loss=0.0337, over 766693.71 frames.], batch size: 28, aishell_tot_loss[loss=0.1516, simple_loss=0.2368, pruned_loss=0.03322, over 506478.00 frames.], datatang_tot_loss[loss=0.1432, simple_loss=0.2187, pruned_loss=0.03383, over 535131.90 frames.], batch size: 28, lr: 4.36e-04
+2022-06-18 23:36:46,750 INFO [train.py:874] (2/4) Epoch 19, batch 350, aishell_loss[loss=0.1211, simple_loss=0.2101, pruned_loss=0.01603, over 4881.00 frames.], tot_loss[loss=0.1475, simple_loss=0.2277, pruned_loss=0.03364, over 814459.07 frames.], batch size: 28, aishell_tot_loss[loss=0.1508, simple_loss=0.236, pruned_loss=0.03283, over 564283.07 frames.], datatang_tot_loss[loss=0.1437, simple_loss=0.2191, pruned_loss=0.03417, over 586105.71 frames.], batch size: 28, lr: 4.35e-04
+2022-06-18 23:37:18,102 INFO [train.py:874] (2/4) Epoch 19, batch 400, aishell_loss[loss=0.1779, simple_loss=0.2652, pruned_loss=0.0453, over 4968.00 frames.], tot_loss[loss=0.1473, simple_loss=0.2277, pruned_loss=0.03346, over 852389.39 frames.], batch size: 64, aishell_tot_loss[loss=0.1516, simple_loss=0.2368, pruned_loss=0.03323, over 610107.71 frames.], datatang_tot_loss[loss=0.1429, simple_loss=0.2186, pruned_loss=0.03358, over 636698.83 frames.], batch size: 64, lr: 4.35e-04
+2022-06-18 23:37:49,207 INFO [train.py:874] (2/4) Epoch 19, batch 450, datatang_loss[loss=0.1504, simple_loss=0.2249, pruned_loss=0.03794, over 4954.00 frames.], tot_loss[loss=0.1482, simple_loss=0.2284, pruned_loss=0.03405, over 881775.04 frames.], batch size: 34, aishell_tot_loss[loss=0.1523, simple_loss=0.2375, pruned_loss=0.03359, over 655739.36 frames.], datatang_tot_loss[loss=0.1434, simple_loss=0.2186, pruned_loss=0.03406, over 676341.56 frames.], batch size: 34, lr: 4.35e-04
+2022-06-18 23:38:17,523 INFO [train.py:874] (2/4) Epoch 19, batch 500, aishell_loss[loss=0.148, simple_loss=0.2279, pruned_loss=0.03401, over 4855.00 frames.], tot_loss[loss=0.1487, simple_loss=0.2289, pruned_loss=0.03424, over 904859.49 frames.], batch size: 28, aishell_tot_loss[loss=0.1521, simple_loss=0.2373, pruned_loss=0.03341, over 694502.36 frames.], datatang_tot_loss[loss=0.1443, simple_loss=0.2196, pruned_loss=0.03455, over 712942.82 frames.], batch size: 28, lr: 4.35e-04
+2022-06-18 23:38:49,413 INFO [train.py:874] (2/4) Epoch 19, batch 550, aishell_loss[loss=0.1398, simple_loss=0.2279, pruned_loss=0.02583, over 4933.00 frames.], tot_loss[loss=0.1488, simple_loss=0.229, pruned_loss=0.03433, over 923072.09 frames.], batch size: 32, aishell_tot_loss[loss=0.1517, simple_loss=0.2369, pruned_loss=0.0333, over 729074.03 frames.], datatang_tot_loss[loss=0.1451, simple_loss=0.2204, pruned_loss=0.03488, over 745111.73 frames.], batch size: 32, lr: 4.35e-04
+2022-06-18 23:39:21,330 INFO [train.py:874] (2/4) Epoch 19, batch 600, aishell_loss[loss=0.1745, simple_loss=0.262, pruned_loss=0.04349, over 4964.00 frames.], tot_loss[loss=0.1498, simple_loss=0.2303, pruned_loss=0.03462, over 937327.37 frames.], batch size: 39, aishell_tot_loss[loss=0.1523, simple_loss=0.2378, pruned_loss=0.03343, over 760798.12 frames.], datatang_tot_loss[loss=0.1458, simple_loss=0.2212, pruned_loss=0.03522, over 772405.63 frames.], batch size: 39, lr: 4.35e-04
+2022-06-18 23:39:49,966 INFO [train.py:874] (2/4) Epoch 19, batch 650, aishell_loss[loss=0.1247, simple_loss=0.2139, pruned_loss=0.01775, over 4980.00 frames.], tot_loss[loss=0.1496, simple_loss=0.23, pruned_loss=0.03453, over 947823.40 frames.], batch size: 30, aishell_tot_loss[loss=0.1521, simple_loss=0.2376, pruned_loss=0.03334, over 786116.62 frames.], datatang_tot_loss[loss=0.1459, simple_loss=0.2213, pruned_loss=0.03528, over 798390.20 frames.], batch size: 30, lr: 4.35e-04
+2022-06-18 23:40:22,420 INFO [train.py:874] (2/4) Epoch 19, batch 700, datatang_loss[loss=0.154, simple_loss=0.2283, pruned_loss=0.03988, over 4884.00 frames.], tot_loss[loss=0.1492, simple_loss=0.2294, pruned_loss=0.03445, over 956125.45 frames.], batch size: 47, aishell_tot_loss[loss=0.1516, simple_loss=0.2372, pruned_loss=0.03303, over 808606.55 frames.], datatang_tot_loss[loss=0.1462, simple_loss=0.2213, pruned_loss=0.03554, over 821317.16 frames.], batch size: 47, lr: 4.34e-04
+2022-06-18 23:40:54,862 INFO [train.py:874] (2/4) Epoch 19, batch 750, aishell_loss[loss=0.1509, simple_loss=0.2412, pruned_loss=0.03036, over 4949.00 frames.], tot_loss[loss=0.1496, simple_loss=0.2302, pruned_loss=0.03452, over 962479.63 frames.], batch size: 56, aishell_tot_loss[loss=0.1517, simple_loss=0.2373, pruned_loss=0.03308, over 828410.35 frames.], datatang_tot_loss[loss=0.1467, simple_loss=0.2222, pruned_loss=0.0356, over 841448.92 frames.], batch size: 56, lr: 4.34e-04
+2022-06-18 23:41:23,178 INFO [train.py:874] (2/4) Epoch 19, batch 800, aishell_loss[loss=0.1351, simple_loss=0.2298, pruned_loss=0.02018, over 4889.00 frames.], tot_loss[loss=0.1513, simple_loss=0.232, pruned_loss=0.03533, over 967494.33 frames.], batch size: 50, aishell_tot_loss[loss=0.1528, simple_loss=0.2384, pruned_loss=0.03359, over 849353.82 frames.], datatang_tot_loss[loss=0.1478, simple_loss=0.2232, pruned_loss=0.03618, over 856064.80 frames.], batch size: 50, lr: 4.34e-04
+2022-06-18 23:41:55,285 INFO [train.py:874] (2/4) Epoch 19, batch 850, aishell_loss[loss=0.1796, simple_loss=0.2557, pruned_loss=0.05178, over 4928.00 frames.], tot_loss[loss=0.1517, simple_loss=0.2325, pruned_loss=0.03551, over 971678.87 frames.], batch size: 79, aishell_tot_loss[loss=0.153, simple_loss=0.2385, pruned_loss=0.03374, over 865240.10 frames.], datatang_tot_loss[loss=0.1483, simple_loss=0.2239, pruned_loss=0.03635, over 871654.10 frames.], batch size: 79, lr: 4.34e-04
+2022-06-18 23:42:25,136 INFO [train.py:874] (2/4) Epoch 19, batch 900, aishell_loss[loss=0.1678, simple_loss=0.2575, pruned_loss=0.03907, over 4946.00 frames.], tot_loss[loss=0.1522, simple_loss=0.2333, pruned_loss=0.03553, over 974845.98 frames.], batch size: 79, aishell_tot_loss[loss=0.1533, simple_loss=0.2391, pruned_loss=0.03378, over 879887.61 frames.], datatang_tot_loss[loss=0.1488, simple_loss=0.2246, pruned_loss=0.03648, over 884708.28 frames.], batch size: 79, lr: 4.34e-04
+2022-06-18 23:42:56,451 INFO [train.py:874] (2/4) Epoch 19, batch 950, aishell_loss[loss=0.1539, simple_loss=0.2376, pruned_loss=0.03514, over 4965.00 frames.], tot_loss[loss=0.1519, simple_loss=0.2329, pruned_loss=0.03552, over 977090.30 frames.], batch size: 25, aishell_tot_loss[loss=0.1531, simple_loss=0.2387, pruned_loss=0.03369, over 891319.63 frames.], datatang_tot_loss[loss=0.1491, simple_loss=0.225, pruned_loss=0.03662, over 897414.59 frames.], batch size: 25, lr: 4.34e-04
+2022-06-18 23:43:29,243 INFO [train.py:874] (2/4) Epoch 19, batch 1000, aishell_loss[loss=0.1188, simple_loss=0.1913, pruned_loss=0.02315, over 4933.00 frames.], tot_loss[loss=0.1522, simple_loss=0.2326, pruned_loss=0.03587, over 978969.41 frames.], batch size: 22, aishell_tot_loss[loss=0.1532, simple_loss=0.2386, pruned_loss=0.03387, over 899719.81 frames.], datatang_tot_loss[loss=0.1495, simple_loss=0.2254, pruned_loss=0.03683, over 910269.07 frames.], batch size: 22, lr: 4.34e-04
+2022-06-18 23:43:29,244 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 23:43:46,288 INFO [train.py:914] (2/4) Epoch 19, validation: loss=0.1644, simple_loss=0.2486, pruned_loss=0.04009, over 1622729.00 frames.
+2022-06-18 23:44:16,846 INFO [train.py:874] (2/4) Epoch 19, batch 1050, datatang_loss[loss=0.153, simple_loss=0.2329, pruned_loss=0.03652, over 4931.00 frames.], tot_loss[loss=0.1519, simple_loss=0.2326, pruned_loss=0.03564, over 980040.33 frames.], batch size: 94, aishell_tot_loss[loss=0.1528, simple_loss=0.2383, pruned_loss=0.03365, over 908657.83 frames.], datatang_tot_loss[loss=0.1499, simple_loss=0.226, pruned_loss=0.03688, over 919777.40 frames.], batch size: 94, lr: 4.33e-04
+2022-06-18 23:44:49,111 INFO [train.py:874] (2/4) Epoch 19, batch 1100, aishell_loss[loss=0.1417, simple_loss=0.2371, pruned_loss=0.02309, over 4925.00 frames.], tot_loss[loss=0.1514, simple_loss=0.2319, pruned_loss=0.03541, over 981243.49 frames.], batch size: 58, aishell_tot_loss[loss=0.153, simple_loss=0.2384, pruned_loss=0.03379, over 916628.84 frames.], datatang_tot_loss[loss=0.1493, simple_loss=0.2254, pruned_loss=0.03655, over 928424.07 frames.], batch size: 58, lr: 4.33e-04
+2022-06-18 23:45:17,819 INFO [train.py:874] (2/4) Epoch 19, batch 1150, datatang_loss[loss=0.1327, simple_loss=0.2181, pruned_loss=0.02368, over 4920.00 frames.], tot_loss[loss=0.1506, simple_loss=0.2311, pruned_loss=0.03501, over 982215.15 frames.], batch size: 83, aishell_tot_loss[loss=0.1525, simple_loss=0.2382, pruned_loss=0.03344, over 923792.29 frames.], datatang_tot_loss[loss=0.149, simple_loss=0.225, pruned_loss=0.03649, over 935952.44 frames.], batch size: 83, lr: 4.33e-04
+2022-06-18 23:45:50,864 INFO [train.py:874] (2/4) Epoch 19, batch 1200, aishell_loss[loss=0.1497, simple_loss=0.2485, pruned_loss=0.02547, over 4951.00 frames.], tot_loss[loss=0.1507, simple_loss=0.2314, pruned_loss=0.03503, over 982820.17 frames.], batch size: 40, aishell_tot_loss[loss=0.1527, simple_loss=0.2383, pruned_loss=0.03349, over 931353.59 frames.], datatang_tot_loss[loss=0.149, simple_loss=0.225, pruned_loss=0.03648, over 941410.27 frames.], batch size: 40, lr: 4.33e-04
+2022-06-18 23:46:23,203 INFO [train.py:874] (2/4) Epoch 19, batch 1250, aishell_loss[loss=0.1815, simple_loss=0.2689, pruned_loss=0.04708, over 4931.00 frames.], tot_loss[loss=0.1509, simple_loss=0.2319, pruned_loss=0.03498, over 983452.25 frames.], batch size: 33, aishell_tot_loss[loss=0.1532, simple_loss=0.2389, pruned_loss=0.03375, over 938166.10 frames.], datatang_tot_loss[loss=0.1486, simple_loss=0.2248, pruned_loss=0.0362, over 946289.86 frames.], batch size: 33, lr: 4.33e-04
+2022-06-18 23:46:52,175 INFO [train.py:874] (2/4) Epoch 19, batch 1300, aishell_loss[loss=0.1494, simple_loss=0.2421, pruned_loss=0.02832, over 4951.00 frames.], tot_loss[loss=0.1521, simple_loss=0.2323, pruned_loss=0.03595, over 984132.97 frames.], batch size: 44, aishell_tot_loss[loss=0.1537, simple_loss=0.2393, pruned_loss=0.03405, over 942544.81 frames.], datatang_tot_loss[loss=0.1495, simple_loss=0.2252, pruned_loss=0.03692, over 952094.23 frames.], batch size: 44, lr: 4.33e-04
+2022-06-18 23:47:23,373 INFO [train.py:874] (2/4) Epoch 19, batch 1350, aishell_loss[loss=0.1568, simple_loss=0.2412, pruned_loss=0.03625, over 4889.00 frames.], tot_loss[loss=0.1516, simple_loss=0.232, pruned_loss=0.03561, over 984162.96 frames.], batch size: 42, aishell_tot_loss[loss=0.1531, simple_loss=0.2388, pruned_loss=0.03375, over 947985.55 frames.], datatang_tot_loss[loss=0.1497, simple_loss=0.2253, pruned_loss=0.03701, over 955485.94 frames.], batch size: 42, lr: 4.33e-04
+2022-06-18 23:47:56,177 INFO [train.py:874] (2/4) Epoch 19, batch 1400, aishell_loss[loss=0.1595, simple_loss=0.2469, pruned_loss=0.03602, over 4956.00 frames.], tot_loss[loss=0.1513, simple_loss=0.2315, pruned_loss=0.03554, over 984818.56 frames.], batch size: 78, aishell_tot_loss[loss=0.1533, simple_loss=0.2388, pruned_loss=0.03386, over 951363.38 frames.], datatang_tot_loss[loss=0.1493, simple_loss=0.2251, pruned_loss=0.03678, over 960159.00 frames.], batch size: 78, lr: 4.32e-04
+2022-06-18 23:48:24,798 INFO [train.py:874] (2/4) Epoch 19, batch 1450, datatang_loss[loss=0.1495, simple_loss=0.234, pruned_loss=0.03248, over 4925.00 frames.], tot_loss[loss=0.1507, simple_loss=0.2309, pruned_loss=0.03523, over 984797.31 frames.], batch size: 81, aishell_tot_loss[loss=0.153, simple_loss=0.2385, pruned_loss=0.03373, over 955494.64 frames.], datatang_tot_loss[loss=0.1489, simple_loss=0.2245, pruned_loss=0.03665, over 962925.63 frames.], batch size: 81, lr: 4.32e-04
+2022-06-18 23:49:02,514 INFO [train.py:874] (2/4) Epoch 19, batch 1500, datatang_loss[loss=0.1517, simple_loss=0.2293, pruned_loss=0.03708, over 4943.00 frames.], tot_loss[loss=0.1512, simple_loss=0.2314, pruned_loss=0.03548, over 985164.40 frames.], batch size: 69, aishell_tot_loss[loss=0.1529, simple_loss=0.2384, pruned_loss=0.03373, over 959450.87 frames.], datatang_tot_loss[loss=0.1494, simple_loss=0.225, pruned_loss=0.03696, over 965460.82 frames.], batch size: 69, lr: 4.32e-04
+2022-06-18 23:49:33,538 INFO [train.py:874] (2/4) Epoch 19, batch 1550, aishell_loss[loss=0.1826, simple_loss=0.2648, pruned_loss=0.05018, over 4926.00 frames.], tot_loss[loss=0.1522, simple_loss=0.2325, pruned_loss=0.03593, over 985524.79 frames.], batch size: 79, aishell_tot_loss[loss=0.1533, simple_loss=0.2387, pruned_loss=0.03392, over 962932.66 frames.], datatang_tot_loss[loss=0.1501, simple_loss=0.2255, pruned_loss=0.03737, over 967815.61 frames.], batch size: 79, lr: 4.32e-04
+2022-06-18 23:50:04,151 INFO [train.py:874] (2/4) Epoch 19, batch 1600, aishell_loss[loss=0.155, simple_loss=0.2413, pruned_loss=0.03433, over 4955.00 frames.], tot_loss[loss=0.1516, simple_loss=0.2317, pruned_loss=0.03577, over 985748.00 frames.], batch size: 44, aishell_tot_loss[loss=0.1529, simple_loss=0.2382, pruned_loss=0.03384, over 965641.07 frames.], datatang_tot_loss[loss=0.1499, simple_loss=0.2252, pruned_loss=0.03732, over 970081.74 frames.], batch size: 44, lr: 4.32e-04
+2022-06-18 23:50:37,457 INFO [train.py:874] (2/4) Epoch 19, batch 1650, aishell_loss[loss=0.1424, simple_loss=0.2267, pruned_loss=0.02899, over 4908.00 frames.], tot_loss[loss=0.1516, simple_loss=0.232, pruned_loss=0.03562, over 985343.80 frames.], batch size: 28, aishell_tot_loss[loss=0.1529, simple_loss=0.2382, pruned_loss=0.0338, over 968039.79 frames.], datatang_tot_loss[loss=0.15, simple_loss=0.2254, pruned_loss=0.03734, over 971559.47 frames.], batch size: 28, lr: 4.32e-04
+2022-06-18 23:51:09,470 INFO [train.py:874] (2/4) Epoch 19, batch 1700, datatang_loss[loss=0.1573, simple_loss=0.235, pruned_loss=0.03976, over 4951.00 frames.], tot_loss[loss=0.1514, simple_loss=0.2311, pruned_loss=0.03586, over 985258.12 frames.], batch size: 86, aishell_tot_loss[loss=0.1523, simple_loss=0.2372, pruned_loss=0.03369, over 969501.32 frames.], datatang_tot_loss[loss=0.1505, simple_loss=0.2257, pruned_loss=0.03765, over 973576.30 frames.], batch size: 86, lr: 4.32e-04
+2022-06-18 23:51:39,494 INFO [train.py:874] (2/4) Epoch 19, batch 1750, datatang_loss[loss=0.156, simple_loss=0.2254, pruned_loss=0.04328, over 4980.00 frames.], tot_loss[loss=0.1521, simple_loss=0.2318, pruned_loss=0.03626, over 985701.21 frames.], batch size: 40, aishell_tot_loss[loss=0.1526, simple_loss=0.2375, pruned_loss=0.03379, over 971215.19 frames.], datatang_tot_loss[loss=0.1511, simple_loss=0.2262, pruned_loss=0.03796, over 975494.90 frames.], batch size: 40, lr: 4.31e-04
+2022-06-18 23:52:12,623 INFO [train.py:874] (2/4) Epoch 19, batch 1800, aishell_loss[loss=0.1673, simple_loss=0.2506, pruned_loss=0.04199, over 4936.00 frames.], tot_loss[loss=0.1519, simple_loss=0.2323, pruned_loss=0.0358, over 985689.33 frames.], batch size: 49, aishell_tot_loss[loss=0.1526, simple_loss=0.2382, pruned_loss=0.03349, over 972969.63 frames.], datatang_tot_loss[loss=0.1509, simple_loss=0.2261, pruned_loss=0.03789, over 976645.53 frames.], batch size: 49, lr: 4.31e-04
+2022-06-18 23:52:41,166 INFO [train.py:874] (2/4) Epoch 19, batch 1850, datatang_loss[loss=0.1581, simple_loss=0.2296, pruned_loss=0.0433, over 4834.00 frames.], tot_loss[loss=0.1515, simple_loss=0.2324, pruned_loss=0.03534, over 985334.33 frames.], batch size: 25, aishell_tot_loss[loss=0.1524, simple_loss=0.2381, pruned_loss=0.0334, over 974605.73 frames.], datatang_tot_loss[loss=0.1506, simple_loss=0.226, pruned_loss=0.03765, over 977311.03 frames.], batch size: 25, lr: 4.31e-04
+2022-06-18 23:53:12,992 INFO [train.py:874] (2/4) Epoch 19, batch 1900, aishell_loss[loss=0.1551, simple_loss=0.243, pruned_loss=0.03363, over 4985.00 frames.], tot_loss[loss=0.1517, simple_loss=0.2325, pruned_loss=0.03549, over 985084.66 frames.], batch size: 43, aishell_tot_loss[loss=0.1526, simple_loss=0.2382, pruned_loss=0.03354, over 975686.81 frames.], datatang_tot_loss[loss=0.1506, simple_loss=0.2259, pruned_loss=0.03766, over 978179.44 frames.], batch size: 43, lr: 4.31e-04
+2022-06-18 23:53:45,784 INFO [train.py:874] (2/4) Epoch 19, batch 1950, datatang_loss[loss=0.1123, simple_loss=0.1884, pruned_loss=0.01807, over 4975.00 frames.], tot_loss[loss=0.1517, simple_loss=0.2325, pruned_loss=0.03547, over 985237.33 frames.], batch size: 31, aishell_tot_loss[loss=0.1528, simple_loss=0.2384, pruned_loss=0.03361, over 976900.70 frames.], datatang_tot_loss[loss=0.1504, simple_loss=0.2258, pruned_loss=0.03754, over 979050.96 frames.], batch size: 31, lr: 4.31e-04
+2022-06-18 23:54:14,556 INFO [train.py:874] (2/4) Epoch 19, batch 2000, datatang_loss[loss=0.1318, simple_loss=0.2056, pruned_loss=0.02905, over 4976.00 frames.], tot_loss[loss=0.1514, simple_loss=0.2324, pruned_loss=0.03519, over 985154.78 frames.], batch size: 26, aishell_tot_loss[loss=0.1528, simple_loss=0.2386, pruned_loss=0.03352, over 977882.54 frames.], datatang_tot_loss[loss=0.1501, simple_loss=0.2257, pruned_loss=0.0373, over 979691.84 frames.], batch size: 26, lr: 4.31e-04
+2022-06-18 23:54:14,557 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-18 23:54:31,135 INFO [train.py:914] (2/4) Epoch 19, validation: loss=0.1648, simple_loss=0.2482, pruned_loss=0.04066, over 1622729.00 frames.
+2022-06-18 23:55:03,609 INFO [train.py:874] (2/4) Epoch 19, batch 2050, datatang_loss[loss=0.1601, simple_loss=0.2332, pruned_loss=0.04346, over 4894.00 frames.], tot_loss[loss=0.1517, simple_loss=0.2328, pruned_loss=0.03532, over 985322.29 frames.], batch size: 47, aishell_tot_loss[loss=0.1536, simple_loss=0.2395, pruned_loss=0.03382, over 978919.74 frames.], datatang_tot_loss[loss=0.1497, simple_loss=0.2251, pruned_loss=0.03715, over 980346.65 frames.], batch size: 47, lr: 4.31e-04
+2022-06-18 23:55:34,125 INFO [train.py:874] (2/4) Epoch 19, batch 2100, datatang_loss[loss=0.1497, simple_loss=0.2243, pruned_loss=0.03761, over 4941.00 frames.], tot_loss[loss=0.1514, simple_loss=0.2322, pruned_loss=0.03524, over 985069.40 frames.], batch size: 37, aishell_tot_loss[loss=0.1532, simple_loss=0.2393, pruned_loss=0.03355, over 979377.52 frames.], datatang_tot_loss[loss=0.1497, simple_loss=0.225, pruned_loss=0.03722, over 980934.29 frames.], batch size: 37, lr: 4.30e-04
+2022-06-18 23:56:07,334 INFO [train.py:874] (2/4) Epoch 19, batch 2150, aishell_loss[loss=0.2133, simple_loss=0.2815, pruned_loss=0.07252, over 4940.00 frames.], tot_loss[loss=0.1506, simple_loss=0.231, pruned_loss=0.03511, over 985161.44 frames.], batch size: 49, aishell_tot_loss[loss=0.1533, simple_loss=0.239, pruned_loss=0.03383, over 979909.57 frames.], datatang_tot_loss[loss=0.1489, simple_loss=0.2243, pruned_loss=0.03669, over 981629.87 frames.], batch size: 49, lr: 4.30e-04
+2022-06-18 23:56:40,091 INFO [train.py:874] (2/4) Epoch 19, batch 2200, datatang_loss[loss=0.1337, simple_loss=0.2044, pruned_loss=0.03147, over 4955.00 frames.], tot_loss[loss=0.1503, simple_loss=0.2306, pruned_loss=0.03502, over 984738.58 frames.], batch size: 34, aishell_tot_loss[loss=0.1534, simple_loss=0.2388, pruned_loss=0.03399, over 979951.64 frames.], datatang_tot_loss[loss=0.1483, simple_loss=0.2239, pruned_loss=0.03637, over 982156.61 frames.], batch size: 34, lr: 4.30e-04
+2022-06-18 23:57:09,674 INFO [train.py:874] (2/4) Epoch 19, batch 2250, aishell_loss[loss=0.1902, simple_loss=0.2684, pruned_loss=0.05605, over 4946.00 frames.], tot_loss[loss=0.15, simple_loss=0.2304, pruned_loss=0.03477, over 984861.60 frames.], batch size: 80, aishell_tot_loss[loss=0.1536, simple_loss=0.2392, pruned_loss=0.03402, over 980405.84 frames.], datatang_tot_loss[loss=0.1477, simple_loss=0.2234, pruned_loss=0.036, over 982668.73 frames.], batch size: 80, lr: 4.30e-04
+2022-06-18 23:57:42,508 INFO [train.py:874] (2/4) Epoch 19, batch 2300, datatang_loss[loss=0.1615, simple_loss=0.2358, pruned_loss=0.04364, over 4945.00 frames.], tot_loss[loss=0.1497, simple_loss=0.2301, pruned_loss=0.03467, over 984815.48 frames.], batch size: 69, aishell_tot_loss[loss=0.1529, simple_loss=0.2383, pruned_loss=0.03377, over 980924.12 frames.], datatang_tot_loss[loss=0.1479, simple_loss=0.2236, pruned_loss=0.03613, over 982911.32 frames.], batch size: 69, lr: 4.30e-04
+2022-06-18 23:58:14,308 INFO [train.py:874] (2/4) Epoch 19, batch 2350, aishell_loss[loss=0.1474, simple_loss=0.2386, pruned_loss=0.02812, over 4962.00 frames.], tot_loss[loss=0.151, simple_loss=0.2319, pruned_loss=0.03504, over 984858.85 frames.], batch size: 64, aishell_tot_loss[loss=0.1537, simple_loss=0.2394, pruned_loss=0.03402, over 981512.17 frames.], datatang_tot_loss[loss=0.1482, simple_loss=0.2238, pruned_loss=0.03627, over 983091.92 frames.], batch size: 64, lr: 4.30e-04
+2022-06-18 23:58:44,223 INFO [train.py:874] (2/4) Epoch 19, batch 2400, datatang_loss[loss=0.1142, simple_loss=0.183, pruned_loss=0.02273, over 4828.00 frames.], tot_loss[loss=0.1513, simple_loss=0.2316, pruned_loss=0.03551, over 985338.59 frames.], batch size: 25, aishell_tot_loss[loss=0.1537, simple_loss=0.2393, pruned_loss=0.03407, over 982134.85 frames.], datatang_tot_loss[loss=0.1486, simple_loss=0.2238, pruned_loss=0.03668, over 983582.63 frames.], batch size: 25, lr: 4.30e-04
+2022-06-18 23:59:16,943 INFO [train.py:874] (2/4) Epoch 19, batch 2450, datatang_loss[loss=0.1155, simple_loss=0.1973, pruned_loss=0.01685, over 4926.00 frames.], tot_loss[loss=0.152, simple_loss=0.2324, pruned_loss=0.03578, over 985309.15 frames.], batch size: 71, aishell_tot_loss[loss=0.1542, simple_loss=0.2398, pruned_loss=0.03431, over 982443.36 frames.], datatang_tot_loss[loss=0.1489, simple_loss=0.2244, pruned_loss=0.03674, over 983799.01 frames.], batch size: 71, lr: 4.30e-04
+2022-06-18 23:59:48,704 INFO [train.py:874] (2/4) Epoch 19, batch 2500, datatang_loss[loss=0.1677, simple_loss=0.247, pruned_loss=0.04423, over 4931.00 frames.], tot_loss[loss=0.1517, simple_loss=0.2327, pruned_loss=0.03532, over 985607.00 frames.], batch size: 45, aishell_tot_loss[loss=0.1539, simple_loss=0.2396, pruned_loss=0.03405, over 983036.23 frames.], datatang_tot_loss[loss=0.1489, simple_loss=0.2246, pruned_loss=0.03662, over 984057.20 frames.], batch size: 45, lr: 4.29e-04
+2022-06-19 00:00:18,830 INFO [train.py:874] (2/4) Epoch 19, batch 2550, aishell_loss[loss=0.1405, simple_loss=0.2276, pruned_loss=0.02667, over 4883.00 frames.], tot_loss[loss=0.1525, simple_loss=0.2334, pruned_loss=0.03574, over 985967.45 frames.], batch size: 28, aishell_tot_loss[loss=0.1541, simple_loss=0.2399, pruned_loss=0.03413, over 983512.93 frames.], datatang_tot_loss[loss=0.1496, simple_loss=0.2252, pruned_loss=0.03701, over 984454.60 frames.], batch size: 28, lr: 4.29e-04
+2022-06-19 00:00:52,788 INFO [train.py:874] (2/4) Epoch 19, batch 2600, datatang_loss[loss=0.1512, simple_loss=0.2341, pruned_loss=0.03416, over 4971.00 frames.], tot_loss[loss=0.1519, simple_loss=0.2325, pruned_loss=0.0357, over 985744.79 frames.], batch size: 60, aishell_tot_loss[loss=0.154, simple_loss=0.2395, pruned_loss=0.03425, over 983600.01 frames.], datatang_tot_loss[loss=0.1494, simple_loss=0.225, pruned_loss=0.03686, over 984582.03 frames.], batch size: 60, lr: 4.29e-04
+2022-06-19 00:01:22,176 INFO [train.py:874] (2/4) Epoch 19, batch 2650, aishell_loss[loss=0.1539, simple_loss=0.246, pruned_loss=0.03095, over 4906.00 frames.], tot_loss[loss=0.1521, simple_loss=0.2327, pruned_loss=0.03577, over 985786.79 frames.], batch size: 79, aishell_tot_loss[loss=0.1542, simple_loss=0.2398, pruned_loss=0.03435, over 983783.43 frames.], datatang_tot_loss[loss=0.1494, simple_loss=0.225, pruned_loss=0.03684, over 984831.80 frames.], batch size: 79, lr: 4.29e-04
+2022-06-19 00:01:54,086 INFO [train.py:874] (2/4) Epoch 19, batch 2700, datatang_loss[loss=0.1592, simple_loss=0.223, pruned_loss=0.04775, over 4904.00 frames.], tot_loss[loss=0.152, simple_loss=0.2321, pruned_loss=0.0359, over 985241.48 frames.], batch size: 47, aishell_tot_loss[loss=0.154, simple_loss=0.2396, pruned_loss=0.0342, over 983610.83 frames.], datatang_tot_loss[loss=0.1495, simple_loss=0.2248, pruned_loss=0.03715, over 984775.76 frames.], batch size: 47, lr: 4.29e-04
+2022-06-19 00:02:27,628 INFO [train.py:874] (2/4) Epoch 19, batch 2750, aishell_loss[loss=0.1672, simple_loss=0.2613, pruned_loss=0.03654, over 4910.00 frames.], tot_loss[loss=0.152, simple_loss=0.2318, pruned_loss=0.03612, over 985314.30 frames.], batch size: 78, aishell_tot_loss[loss=0.1548, simple_loss=0.2402, pruned_loss=0.03468, over 983786.67 frames.], datatang_tot_loss[loss=0.149, simple_loss=0.2243, pruned_loss=0.03686, over 984883.49 frames.], batch size: 78, lr: 4.29e-04
+2022-06-19 00:03:00,484 INFO [train.py:874] (2/4) Epoch 19, batch 2800, aishell_loss[loss=0.1192, simple_loss=0.2098, pruned_loss=0.01426, over 4869.00 frames.], tot_loss[loss=0.1524, simple_loss=0.2322, pruned_loss=0.03631, over 985111.38 frames.], batch size: 28, aishell_tot_loss[loss=0.1547, simple_loss=0.2403, pruned_loss=0.03459, over 983934.64 frames.], datatang_tot_loss[loss=0.1497, simple_loss=0.225, pruned_loss=0.03716, over 984712.93 frames.], batch size: 28, lr: 4.29e-04
+2022-06-19 00:03:29,848 INFO [train.py:874] (2/4) Epoch 19, batch 2850, aishell_loss[loss=0.1625, simple_loss=0.2521, pruned_loss=0.0365, over 4948.00 frames.], tot_loss[loss=0.1524, simple_loss=0.2326, pruned_loss=0.0361, over 985118.95 frames.], batch size: 56, aishell_tot_loss[loss=0.1543, simple_loss=0.2398, pruned_loss=0.03437, over 984113.99 frames.], datatang_tot_loss[loss=0.15, simple_loss=0.2253, pruned_loss=0.03734, over 984737.33 frames.], batch size: 56, lr: 4.28e-04
+2022-06-19 00:04:03,075 INFO [train.py:874] (2/4) Epoch 19, batch 2900, datatang_loss[loss=0.1298, simple_loss=0.206, pruned_loss=0.02683, over 4935.00 frames.], tot_loss[loss=0.1524, simple_loss=0.233, pruned_loss=0.03592, over 985358.33 frames.], batch size: 50, aishell_tot_loss[loss=0.154, simple_loss=0.2398, pruned_loss=0.03412, over 984363.51 frames.], datatang_tot_loss[loss=0.1503, simple_loss=0.2255, pruned_loss=0.03752, over 984917.89 frames.], batch size: 50, lr: 4.28e-04
+2022-06-19 00:04:31,754 INFO [train.py:874] (2/4) Epoch 19, batch 2950, datatang_loss[loss=0.1391, simple_loss=0.2152, pruned_loss=0.03151, over 4919.00 frames.], tot_loss[loss=0.1526, simple_loss=0.2331, pruned_loss=0.03606, over 985544.78 frames.], batch size: 75, aishell_tot_loss[loss=0.1539, simple_loss=0.2398, pruned_loss=0.03407, over 984528.98 frames.], datatang_tot_loss[loss=0.1506, simple_loss=0.2257, pruned_loss=0.03772, over 985108.11 frames.], batch size: 75, lr: 4.28e-04
+2022-06-19 00:05:05,079 INFO [train.py:874] (2/4) Epoch 19, batch 3000, aishell_loss[loss=0.154, simple_loss=0.248, pruned_loss=0.03, over 4906.00 frames.], tot_loss[loss=0.1521, simple_loss=0.2329, pruned_loss=0.03565, over 985604.19 frames.], batch size: 52, aishell_tot_loss[loss=0.1537, simple_loss=0.2398, pruned_loss=0.03386, over 984568.09 frames.], datatang_tot_loss[loss=0.1503, simple_loss=0.2254, pruned_loss=0.03758, over 985312.52 frames.], batch size: 52, lr: 4.28e-04
+2022-06-19 00:05:05,080 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 00:05:22,957 INFO [train.py:914] (2/4) Epoch 19, validation: loss=0.1643, simple_loss=0.2485, pruned_loss=0.04007, over 1622729.00 frames.
+2022-06-19 00:05:51,965 INFO [train.py:874] (2/4) Epoch 19, batch 3050, aishell_loss[loss=0.1459, simple_loss=0.2392, pruned_loss=0.02633, over 4936.00 frames.], tot_loss[loss=0.1522, simple_loss=0.2333, pruned_loss=0.03559, over 985409.29 frames.], batch size: 54, aishell_tot_loss[loss=0.1532, simple_loss=0.2391, pruned_loss=0.03362, over 984537.17 frames.], datatang_tot_loss[loss=0.1509, simple_loss=0.2262, pruned_loss=0.03786, over 985327.70 frames.], batch size: 54, lr: 4.28e-04
+2022-06-19 00:06:24,274 INFO [train.py:874] (2/4) Epoch 19, batch 3100, datatang_loss[loss=0.1565, simple_loss=0.2342, pruned_loss=0.03941, over 4919.00 frames.], tot_loss[loss=0.1523, simple_loss=0.2332, pruned_loss=0.0357, over 985545.25 frames.], batch size: 77, aishell_tot_loss[loss=0.1533, simple_loss=0.2391, pruned_loss=0.03373, over 984818.13 frames.], datatang_tot_loss[loss=0.1509, simple_loss=0.2261, pruned_loss=0.03787, over 985305.55 frames.], batch size: 77, lr: 4.28e-04
+2022-06-19 00:06:52,802 INFO [train.py:874] (2/4) Epoch 19, batch 3150, aishell_loss[loss=0.1494, simple_loss=0.2436, pruned_loss=0.02754, over 4877.00 frames.], tot_loss[loss=0.1523, simple_loss=0.2332, pruned_loss=0.03567, over 985542.06 frames.], batch size: 47, aishell_tot_loss[loss=0.1534, simple_loss=0.2394, pruned_loss=0.0337, over 984800.22 frames.], datatang_tot_loss[loss=0.1509, simple_loss=0.2262, pruned_loss=0.03781, over 985454.51 frames.], batch size: 47, lr: 4.28e-04
+2022-06-19 00:07:26,191 INFO [train.py:874] (2/4) Epoch 19, batch 3200, aishell_loss[loss=0.1459, simple_loss=0.2279, pruned_loss=0.032, over 4915.00 frames.], tot_loss[loss=0.1523, simple_loss=0.2326, pruned_loss=0.03599, over 985132.83 frames.], batch size: 46, aishell_tot_loss[loss=0.1536, simple_loss=0.2394, pruned_loss=0.03395, over 984363.76 frames.], datatang_tot_loss[loss=0.1507, simple_loss=0.2258, pruned_loss=0.03783, over 985560.86 frames.], batch size: 46, lr: 4.27e-04
+2022-06-19 00:07:58,158 INFO [train.py:874] (2/4) Epoch 19, batch 3250, aishell_loss[loss=0.1424, simple_loss=0.2281, pruned_loss=0.02836, over 4867.00 frames.], tot_loss[loss=0.1527, simple_loss=0.2334, pruned_loss=0.03602, over 985318.95 frames.], batch size: 35, aishell_tot_loss[loss=0.154, simple_loss=0.2398, pruned_loss=0.03405, over 984471.19 frames.], datatang_tot_loss[loss=0.1509, simple_loss=0.226, pruned_loss=0.03784, over 985703.22 frames.], batch size: 35, lr: 4.27e-04
+2022-06-19 00:08:26,501 INFO [train.py:874] (2/4) Epoch 19, batch 3300, aishell_loss[loss=0.1365, simple_loss=0.2138, pruned_loss=0.02961, over 4812.00 frames.], tot_loss[loss=0.1519, simple_loss=0.2327, pruned_loss=0.03551, over 985036.50 frames.], batch size: 26, aishell_tot_loss[loss=0.1538, simple_loss=0.2397, pruned_loss=0.03396, over 984062.38 frames.], datatang_tot_loss[loss=0.1501, simple_loss=0.2255, pruned_loss=0.0374, over 985861.14 frames.], batch size: 26, lr: 4.27e-04
+2022-06-19 00:08:58,834 INFO [train.py:874] (2/4) Epoch 19, batch 3350, aishell_loss[loss=0.156, simple_loss=0.2519, pruned_loss=0.03004, over 4934.00 frames.], tot_loss[loss=0.151, simple_loss=0.2315, pruned_loss=0.03528, over 985143.36 frames.], batch size: 78, aishell_tot_loss[loss=0.1533, simple_loss=0.239, pruned_loss=0.0338, over 984227.10 frames.], datatang_tot_loss[loss=0.1497, simple_loss=0.2249, pruned_loss=0.03728, over 985825.18 frames.], batch size: 78, lr: 4.27e-04
+2022-06-19 00:09:31,436 INFO [train.py:874] (2/4) Epoch 19, batch 3400, aishell_loss[loss=0.1754, simple_loss=0.2636, pruned_loss=0.04363, over 4877.00 frames.], tot_loss[loss=0.1508, simple_loss=0.231, pruned_loss=0.03531, over 985089.39 frames.], batch size: 36, aishell_tot_loss[loss=0.1531, simple_loss=0.2388, pruned_loss=0.03369, over 984243.94 frames.], datatang_tot_loss[loss=0.1496, simple_loss=0.2246, pruned_loss=0.0373, over 985759.51 frames.], batch size: 36, lr: 4.27e-04
+2022-06-19 00:09:59,985 INFO [train.py:874] (2/4) Epoch 19, batch 3450, datatang_loss[loss=0.1781, simple_loss=0.2506, pruned_loss=0.05282, over 4953.00 frames.], tot_loss[loss=0.1507, simple_loss=0.2307, pruned_loss=0.03535, over 985068.72 frames.], batch size: 99, aishell_tot_loss[loss=0.1528, simple_loss=0.2383, pruned_loss=0.03369, over 984190.02 frames.], datatang_tot_loss[loss=0.1496, simple_loss=0.2246, pruned_loss=0.0373, over 985807.04 frames.], batch size: 99, lr: 4.27e-04
+2022-06-19 00:10:33,118 INFO [train.py:874] (2/4) Epoch 19, batch 3500, aishell_loss[loss=0.1195, simple_loss=0.1958, pruned_loss=0.02159, over 4972.00 frames.], tot_loss[loss=0.1501, simple_loss=0.2302, pruned_loss=0.03497, over 985050.55 frames.], batch size: 25, aishell_tot_loss[loss=0.1524, simple_loss=0.2377, pruned_loss=0.03358, over 984125.77 frames.], datatang_tot_loss[loss=0.1492, simple_loss=0.2245, pruned_loss=0.03702, over 985878.93 frames.], batch size: 25, lr: 4.27e-04
+2022-06-19 00:11:03,925 INFO [train.py:874] (2/4) Epoch 19, batch 3550, aishell_loss[loss=0.1483, simple_loss=0.236, pruned_loss=0.03033, over 4952.00 frames.], tot_loss[loss=0.1508, simple_loss=0.231, pruned_loss=0.03527, over 985254.11 frames.], batch size: 45, aishell_tot_loss[loss=0.1527, simple_loss=0.2381, pruned_loss=0.03362, over 984472.82 frames.], datatang_tot_loss[loss=0.1496, simple_loss=0.2248, pruned_loss=0.03722, over 985742.18 frames.], batch size: 45, lr: 4.26e-04
+2022-06-19 00:11:33,771 INFO [train.py:874] (2/4) Epoch 19, batch 3600, aishell_loss[loss=0.1476, simple_loss=0.2348, pruned_loss=0.03024, over 4920.00 frames.], tot_loss[loss=0.15, simple_loss=0.2299, pruned_loss=0.03505, over 985445.80 frames.], batch size: 46, aishell_tot_loss[loss=0.153, simple_loss=0.2383, pruned_loss=0.03383, over 984574.33 frames.], datatang_tot_loss[loss=0.1485, simple_loss=0.2236, pruned_loss=0.03668, over 985843.80 frames.], batch size: 46, lr: 4.26e-04
+2022-06-19 00:12:05,687 INFO [train.py:874] (2/4) Epoch 19, batch 3650, aishell_loss[loss=0.1649, simple_loss=0.2511, pruned_loss=0.03938, over 4877.00 frames.], tot_loss[loss=0.151, simple_loss=0.2313, pruned_loss=0.03532, over 985396.72 frames.], batch size: 47, aishell_tot_loss[loss=0.1533, simple_loss=0.2386, pruned_loss=0.03399, over 984546.22 frames.], datatang_tot_loss[loss=0.149, simple_loss=0.2244, pruned_loss=0.0368, over 985891.42 frames.], batch size: 47, lr: 4.26e-04
+2022-06-19 00:12:38,523 INFO [train.py:874] (2/4) Epoch 19, batch 3700, datatang_loss[loss=0.154, simple_loss=0.2357, pruned_loss=0.03616, over 4923.00 frames.], tot_loss[loss=0.1498, simple_loss=0.2302, pruned_loss=0.0347, over 985310.96 frames.], batch size: 98, aishell_tot_loss[loss=0.1533, simple_loss=0.2387, pruned_loss=0.03399, over 984683.27 frames.], datatang_tot_loss[loss=0.1478, simple_loss=0.2234, pruned_loss=0.03607, over 985694.31 frames.], batch size: 98, lr: 4.26e-04
+2022-06-19 00:13:06,424 INFO [train.py:874] (2/4) Epoch 19, batch 3750, datatang_loss[loss=0.1675, simple_loss=0.2295, pruned_loss=0.05272, over 4918.00 frames.], tot_loss[loss=0.1501, simple_loss=0.2306, pruned_loss=0.03483, over 985747.34 frames.], batch size: 64, aishell_tot_loss[loss=0.1536, simple_loss=0.2389, pruned_loss=0.03411, over 985012.27 frames.], datatang_tot_loss[loss=0.1477, simple_loss=0.2233, pruned_loss=0.03602, over 985856.62 frames.], batch size: 64, lr: 4.26e-04
+2022-06-19 00:13:39,436 INFO [train.py:874] (2/4) Epoch 19, batch 3800, aishell_loss[loss=0.1459, simple_loss=0.2396, pruned_loss=0.0261, over 4938.00 frames.], tot_loss[loss=0.1497, simple_loss=0.2302, pruned_loss=0.03462, over 985309.10 frames.], batch size: 58, aishell_tot_loss[loss=0.153, simple_loss=0.2383, pruned_loss=0.0338, over 984729.98 frames.], datatang_tot_loss[loss=0.1477, simple_loss=0.2233, pruned_loss=0.03611, over 985775.06 frames.], batch size: 58, lr: 4.26e-04
+2022-06-19 00:14:10,391 INFO [train.py:874] (2/4) Epoch 19, batch 3850, aishell_loss[loss=0.1655, simple_loss=0.2538, pruned_loss=0.03859, over 4919.00 frames.], tot_loss[loss=0.1503, simple_loss=0.2308, pruned_loss=0.03487, over 985378.52 frames.], batch size: 41, aishell_tot_loss[loss=0.1534, simple_loss=0.2388, pruned_loss=0.03402, over 984771.45 frames.], datatang_tot_loss[loss=0.1478, simple_loss=0.2236, pruned_loss=0.03603, over 985796.69 frames.], batch size: 41, lr: 4.26e-04
+2022-06-19 00:14:40,618 INFO [train.py:874] (2/4) Epoch 19, batch 3900, aishell_loss[loss=0.1713, simple_loss=0.2535, pruned_loss=0.04453, over 4976.00 frames.], tot_loss[loss=0.1508, simple_loss=0.2311, pruned_loss=0.03522, over 985527.38 frames.], batch size: 51, aishell_tot_loss[loss=0.1539, simple_loss=0.2391, pruned_loss=0.03435, over 984868.41 frames.], datatang_tot_loss[loss=0.1478, simple_loss=0.2236, pruned_loss=0.036, over 985881.42 frames.], batch size: 51, lr: 4.26e-04
+2022-06-19 00:15:09,675 INFO [train.py:874] (2/4) Epoch 19, batch 3950, aishell_loss[loss=0.1608, simple_loss=0.2489, pruned_loss=0.03629, over 4947.00 frames.], tot_loss[loss=0.1501, simple_loss=0.231, pruned_loss=0.03464, over 985731.86 frames.], batch size: 32, aishell_tot_loss[loss=0.154, simple_loss=0.2394, pruned_loss=0.03433, over 984874.50 frames.], datatang_tot_loss[loss=0.1469, simple_loss=0.2229, pruned_loss=0.03543, over 986132.88 frames.], batch size: 32, lr: 4.25e-04
+2022-06-19 00:15:40,629 INFO [train.py:874] (2/4) Epoch 19, batch 4000, aishell_loss[loss=0.1357, simple_loss=0.2243, pruned_loss=0.02356, over 4973.00 frames.], tot_loss[loss=0.1507, simple_loss=0.2316, pruned_loss=0.03494, over 985739.72 frames.], batch size: 27, aishell_tot_loss[loss=0.1547, simple_loss=0.2401, pruned_loss=0.03467, over 984903.72 frames.], datatang_tot_loss[loss=0.1467, simple_loss=0.2227, pruned_loss=0.03536, over 986160.94 frames.], batch size: 27, lr: 4.25e-04
+2022-06-19 00:15:40,629 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 00:15:58,331 INFO [train.py:914] (2/4) Epoch 19, validation: loss=0.165, simple_loss=0.2492, pruned_loss=0.0404, over 1622729.00 frames.
+2022-06-19 00:16:27,765 INFO [train.py:874] (2/4) Epoch 19, batch 4050, aishell_loss[loss=0.1573, simple_loss=0.2474, pruned_loss=0.03361, over 4916.00 frames.], tot_loss[loss=0.1496, simple_loss=0.2309, pruned_loss=0.03413, over 985433.06 frames.], batch size: 52, aishell_tot_loss[loss=0.1547, simple_loss=0.2404, pruned_loss=0.03448, over 984760.68 frames.], datatang_tot_loss[loss=0.1455, simple_loss=0.2218, pruned_loss=0.03464, over 986025.33 frames.], batch size: 52, lr: 4.25e-04
+2022-06-19 00:16:58,278 INFO [train.py:874] (2/4) Epoch 19, batch 4100, datatang_loss[loss=0.1401, simple_loss=0.2176, pruned_loss=0.03129, over 4943.00 frames.], tot_loss[loss=0.1499, simple_loss=0.2313, pruned_loss=0.03426, over 985566.49 frames.], batch size: 50, aishell_tot_loss[loss=0.1543, simple_loss=0.2403, pruned_loss=0.03417, over 984812.21 frames.], datatang_tot_loss[loss=0.1462, simple_loss=0.2224, pruned_loss=0.03501, over 986110.30 frames.], batch size: 50, lr: 4.25e-04
+2022-06-19 00:17:27,287 INFO [train.py:874] (2/4) Epoch 19, batch 4150, aishell_loss[loss=0.1379, simple_loss=0.2265, pruned_loss=0.02468, over 4892.00 frames.], tot_loss[loss=0.1496, simple_loss=0.2312, pruned_loss=0.03403, over 985719.26 frames.], batch size: 34, aishell_tot_loss[loss=0.1535, simple_loss=0.2397, pruned_loss=0.03369, over 984998.64 frames.], datatang_tot_loss[loss=0.1465, simple_loss=0.2227, pruned_loss=0.03517, over 986134.34 frames.], batch size: 34, lr: 4.25e-04
+2022-06-19 00:19:01,262 INFO [train.py:874] (2/4) Epoch 20, batch 50, datatang_loss[loss=0.1578, simple_loss=0.23, pruned_loss=0.04281, over 4968.00 frames.], tot_loss[loss=0.1482, simple_loss=0.2282, pruned_loss=0.03411, over 218502.12 frames.], batch size: 91, aishell_tot_loss[loss=0.1555, simple_loss=0.2422, pruned_loss=0.03444, over 89493.63 frames.], datatang_tot_loss[loss=0.1436, simple_loss=0.2196, pruned_loss=0.0338, over 141899.53 frames.], batch size: 91, lr: 4.14e-04
+2022-06-19 00:19:31,411 INFO [train.py:874] (2/4) Epoch 20, batch 100, aishell_loss[loss=0.2057, simple_loss=0.2759, pruned_loss=0.06777, over 4928.00 frames.], tot_loss[loss=0.1507, simple_loss=0.2314, pruned_loss=0.03501, over 388973.29 frames.], batch size: 49, aishell_tot_loss[loss=0.1581, simple_loss=0.2436, pruned_loss=0.03631, over 214693.48 frames.], datatang_tot_loss[loss=0.143, simple_loss=0.2189, pruned_loss=0.03356, over 222728.45 frames.], batch size: 49, lr: 4.14e-04
+2022-06-19 00:20:03,663 INFO [train.py:874] (2/4) Epoch 20, batch 150, datatang_loss[loss=0.1209, simple_loss=0.1953, pruned_loss=0.02327, over 4945.00 frames.], tot_loss[loss=0.1499, simple_loss=0.2311, pruned_loss=0.03433, over 521293.09 frames.], batch size: 55, aishell_tot_loss[loss=0.1559, simple_loss=0.2424, pruned_loss=0.03471, over 312232.22 frames.], datatang_tot_loss[loss=0.1436, simple_loss=0.2192, pruned_loss=0.034, over 305909.09 frames.], batch size: 55, lr: 4.14e-04
+2022-06-19 00:20:32,882 INFO [train.py:874] (2/4) Epoch 20, batch 200, datatang_loss[loss=0.1299, simple_loss=0.2091, pruned_loss=0.02537, over 4905.00 frames.], tot_loss[loss=0.1485, simple_loss=0.23, pruned_loss=0.03352, over 624267.84 frames.], batch size: 59, aishell_tot_loss[loss=0.1556, simple_loss=0.2422, pruned_loss=0.0345, over 391413.16 frames.], datatang_tot_loss[loss=0.1418, simple_loss=0.2178, pruned_loss=0.03286, over 386133.76 frames.], batch size: 59, lr: 4.14e-04
+2022-06-19 00:21:05,313 INFO [train.py:874] (2/4) Epoch 20, batch 250, aishell_loss[loss=0.1517, simple_loss=0.2374, pruned_loss=0.03303, over 4949.00 frames.], tot_loss[loss=0.1488, simple_loss=0.2291, pruned_loss=0.03418, over 704217.08 frames.], batch size: 54, aishell_tot_loss[loss=0.1559, simple_loss=0.2416, pruned_loss=0.03507, over 464003.51 frames.], datatang_tot_loss[loss=0.1415, simple_loss=0.2165, pruned_loss=0.03324, over 453924.53 frames.], batch size: 54, lr: 4.14e-04
+2022-06-19 00:21:36,972 INFO [train.py:874] (2/4) Epoch 20, batch 300, datatang_loss[loss=0.1266, simple_loss=0.1944, pruned_loss=0.02936, over 4951.00 frames.], tot_loss[loss=0.1481, simple_loss=0.228, pruned_loss=0.03414, over 766873.26 frames.], batch size: 34, aishell_tot_loss[loss=0.1562, simple_loss=0.241, pruned_loss=0.03567, over 521007.29 frames.], datatang_tot_loss[loss=0.1406, simple_loss=0.2158, pruned_loss=0.03266, over 521307.40 frames.], batch size: 34, lr: 4.14e-04
+2022-06-19 00:22:05,466 INFO [train.py:874] (2/4) Epoch 20, batch 350, aishell_loss[loss=0.1688, simple_loss=0.2542, pruned_loss=0.04169, over 4915.00 frames.], tot_loss[loss=0.1489, simple_loss=0.2292, pruned_loss=0.03431, over 815418.93 frames.], batch size: 33, aishell_tot_loss[loss=0.1562, simple_loss=0.2415, pruned_loss=0.03547, over 581867.20 frames.], datatang_tot_loss[loss=0.1413, simple_loss=0.2165, pruned_loss=0.03304, over 569826.21 frames.], batch size: 33, lr: 4.14e-04
+2022-06-19 00:22:38,813 INFO [train.py:874] (2/4) Epoch 20, batch 400, datatang_loss[loss=0.1431, simple_loss=0.2121, pruned_loss=0.03707, over 4901.00 frames.], tot_loss[loss=0.1496, simple_loss=0.2298, pruned_loss=0.03473, over 853141.60 frames.], batch size: 52, aishell_tot_loss[loss=0.1564, simple_loss=0.2419, pruned_loss=0.03546, over 616560.70 frames.], datatang_tot_loss[loss=0.1429, simple_loss=0.2183, pruned_loss=0.03378, over 631559.04 frames.], batch size: 52, lr: 4.13e-04
+2022-06-19 00:23:11,536 INFO [train.py:874] (2/4) Epoch 20, batch 450, aishell_loss[loss=0.167, simple_loss=0.2499, pruned_loss=0.04209, over 4910.00 frames.], tot_loss[loss=0.1491, simple_loss=0.2295, pruned_loss=0.03437, over 882540.94 frames.], batch size: 52, aishell_tot_loss[loss=0.155, simple_loss=0.2406, pruned_loss=0.03469, over 661640.57 frames.], datatang_tot_loss[loss=0.1436, simple_loss=0.219, pruned_loss=0.03407, over 671738.73 frames.], batch size: 52, lr: 4.13e-04
+2022-06-19 00:23:40,896 INFO [train.py:874] (2/4) Epoch 20, batch 500, datatang_loss[loss=0.1266, simple_loss=0.2014, pruned_loss=0.02586, over 4927.00 frames.], tot_loss[loss=0.1489, simple_loss=0.2291, pruned_loss=0.03431, over 905226.86 frames.], batch size: 77, aishell_tot_loss[loss=0.1544, simple_loss=0.2398, pruned_loss=0.0345, over 696659.35 frames.], datatang_tot_loss[loss=0.144, simple_loss=0.2196, pruned_loss=0.03419, over 711562.25 frames.], batch size: 77, lr: 4.13e-04
+2022-06-19 00:24:13,537 INFO [train.py:874] (2/4) Epoch 20, batch 550, aishell_loss[loss=0.1487, simple_loss=0.2372, pruned_loss=0.03008, over 4927.00 frames.], tot_loss[loss=0.1489, simple_loss=0.2293, pruned_loss=0.03429, over 923692.37 frames.], batch size: 41, aishell_tot_loss[loss=0.1541, simple_loss=0.2395, pruned_loss=0.03439, over 729862.76 frames.], datatang_tot_loss[loss=0.1443, simple_loss=0.22, pruned_loss=0.03429, over 745265.36 frames.], batch size: 41, lr: 4.13e-04
+2022-06-19 00:24:45,794 INFO [train.py:874] (2/4) Epoch 20, batch 600, aishell_loss[loss=0.1312, simple_loss=0.1916, pruned_loss=0.03537, over 4802.00 frames.], tot_loss[loss=0.1492, simple_loss=0.2294, pruned_loss=0.03447, over 937042.47 frames.], batch size: 20, aishell_tot_loss[loss=0.1535, simple_loss=0.2386, pruned_loss=0.03415, over 761834.05 frames.], datatang_tot_loss[loss=0.1451, simple_loss=0.2207, pruned_loss=0.03475, over 771459.07 frames.], batch size: 20, lr: 4.13e-04
+2022-06-19 00:25:14,716 INFO [train.py:874] (2/4) Epoch 20, batch 650, datatang_loss[loss=0.1521, simple_loss=0.2253, pruned_loss=0.03946, over 4929.00 frames.], tot_loss[loss=0.1497, simple_loss=0.2302, pruned_loss=0.03462, over 947842.42 frames.], batch size: 71, aishell_tot_loss[loss=0.154, simple_loss=0.2393, pruned_loss=0.03432, over 790457.96 frames.], datatang_tot_loss[loss=0.1452, simple_loss=0.2208, pruned_loss=0.03477, over 794531.77 frames.], batch size: 71, lr: 4.13e-04
+2022-06-19 00:25:47,580 INFO [train.py:874] (2/4) Epoch 20, batch 700, datatang_loss[loss=0.1551, simple_loss=0.2258, pruned_loss=0.04221, over 4964.00 frames.], tot_loss[loss=0.1491, simple_loss=0.2296, pruned_loss=0.03428, over 956530.26 frames.], batch size: 45, aishell_tot_loss[loss=0.1535, simple_loss=0.239, pruned_loss=0.03403, over 813620.20 frames.], datatang_tot_loss[loss=0.1448, simple_loss=0.2204, pruned_loss=0.03465, over 817197.20 frames.], batch size: 45, lr: 4.13e-04
+2022-06-19 00:26:18,478 INFO [train.py:874] (2/4) Epoch 20, batch 750, datatang_loss[loss=0.171, simple_loss=0.2329, pruned_loss=0.05457, over 4882.00 frames.], tot_loss[loss=0.1498, simple_loss=0.2304, pruned_loss=0.03458, over 963007.94 frames.], batch size: 25, aishell_tot_loss[loss=0.1537, simple_loss=0.2392, pruned_loss=0.03414, over 839867.54 frames.], datatang_tot_loss[loss=0.1451, simple_loss=0.2203, pruned_loss=0.03491, over 830973.07 frames.], batch size: 25, lr: 4.13e-04
+2022-06-19 00:26:48,879 INFO [train.py:874] (2/4) Epoch 20, batch 800, aishell_loss[loss=0.1535, simple_loss=0.2433, pruned_loss=0.03187, over 4933.00 frames.], tot_loss[loss=0.15, simple_loss=0.2309, pruned_loss=0.03455, over 967865.39 frames.], batch size: 49, aishell_tot_loss[loss=0.1533, simple_loss=0.2391, pruned_loss=0.0338, over 856181.23 frames.], datatang_tot_loss[loss=0.1459, simple_loss=0.2213, pruned_loss=0.03522, over 849908.53 frames.], batch size: 49, lr: 4.12e-04
+2022-06-19 00:27:20,460 INFO [train.py:874] (2/4) Epoch 20, batch 850, aishell_loss[loss=0.1432, simple_loss=0.2264, pruned_loss=0.03, over 4924.00 frames.], tot_loss[loss=0.1498, simple_loss=0.231, pruned_loss=0.03433, over 971354.63 frames.], batch size: 33, aishell_tot_loss[loss=0.1527, simple_loss=0.2382, pruned_loss=0.03364, over 873454.63 frames.], datatang_tot_loss[loss=0.1463, simple_loss=0.2222, pruned_loss=0.03516, over 863217.63 frames.], batch size: 33, lr: 4.12e-04
+2022-06-19 00:27:50,551 INFO [train.py:874] (2/4) Epoch 20, batch 900, aishell_loss[loss=0.146, simple_loss=0.2257, pruned_loss=0.03313, over 4931.00 frames.], tot_loss[loss=0.1496, simple_loss=0.2306, pruned_loss=0.03426, over 974747.21 frames.], batch size: 33, aishell_tot_loss[loss=0.1525, simple_loss=0.238, pruned_loss=0.03347, over 884241.66 frames.], datatang_tot_loss[loss=0.1464, simple_loss=0.2225, pruned_loss=0.03518, over 880462.31 frames.], batch size: 33, lr: 4.12e-04
+2022-06-19 00:28:20,648 INFO [train.py:874] (2/4) Epoch 20, batch 950, aishell_loss[loss=0.1688, simple_loss=0.2532, pruned_loss=0.04221, over 4861.00 frames.], tot_loss[loss=0.1492, simple_loss=0.2303, pruned_loss=0.03404, over 977063.14 frames.], batch size: 35, aishell_tot_loss[loss=0.1517, simple_loss=0.2374, pruned_loss=0.03305, over 896451.65 frames.], datatang_tot_loss[loss=0.1467, simple_loss=0.2228, pruned_loss=0.03533, over 892479.61 frames.], batch size: 35, lr: 4.12e-04
+2022-06-19 00:28:53,578 INFO [train.py:874] (2/4) Epoch 20, batch 1000, aishell_loss[loss=0.1499, simple_loss=0.2327, pruned_loss=0.03354, over 4912.00 frames.], tot_loss[loss=0.1486, simple_loss=0.2294, pruned_loss=0.03387, over 978728.80 frames.], batch size: 41, aishell_tot_loss[loss=0.1512, simple_loss=0.2369, pruned_loss=0.03279, over 905583.90 frames.], datatang_tot_loss[loss=0.1466, simple_loss=0.2225, pruned_loss=0.03533, over 904613.19 frames.], batch size: 41, lr: 4.12e-04
+2022-06-19 00:28:53,579 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 00:29:10,059 INFO [train.py:914] (2/4) Epoch 20, validation: loss=0.1639, simple_loss=0.248, pruned_loss=0.0399, over 1622729.00 frames.
+2022-06-19 00:29:43,678 INFO [train.py:874] (2/4) Epoch 20, batch 1050, aishell_loss[loss=0.1529, simple_loss=0.242, pruned_loss=0.03192, over 4937.00 frames.], tot_loss[loss=0.1492, simple_loss=0.2298, pruned_loss=0.03435, over 980197.00 frames.], batch size: 32, aishell_tot_loss[loss=0.1517, simple_loss=0.2372, pruned_loss=0.03312, over 914572.54 frames.], datatang_tot_loss[loss=0.1468, simple_loss=0.2227, pruned_loss=0.03549, over 914518.74 frames.], batch size: 32, lr: 4.12e-04
+2022-06-19 00:30:16,591 INFO [train.py:874] (2/4) Epoch 20, batch 1100, aishell_loss[loss=0.1704, simple_loss=0.2529, pruned_loss=0.04392, over 4879.00 frames.], tot_loss[loss=0.15, simple_loss=0.2306, pruned_loss=0.03476, over 981128.27 frames.], batch size: 36, aishell_tot_loss[loss=0.1521, simple_loss=0.2374, pruned_loss=0.03342, over 922707.21 frames.], datatang_tot_loss[loss=0.1473, simple_loss=0.2233, pruned_loss=0.03566, over 922841.25 frames.], batch size: 36, lr: 4.12e-04
+2022-06-19 00:30:45,837 INFO [train.py:874] (2/4) Epoch 20, batch 1150, aishell_loss[loss=0.1688, simple_loss=0.2493, pruned_loss=0.04412, over 4879.00 frames.], tot_loss[loss=0.15, simple_loss=0.2307, pruned_loss=0.03465, over 982146.78 frames.], batch size: 34, aishell_tot_loss[loss=0.1525, simple_loss=0.238, pruned_loss=0.03347, over 929565.73 frames.], datatang_tot_loss[loss=0.147, simple_loss=0.2231, pruned_loss=0.0355, over 930808.47 frames.], batch size: 34, lr: 4.11e-04
+2022-06-19 00:31:19,120 INFO [train.py:874] (2/4) Epoch 20, batch 1200, aishell_loss[loss=0.1405, simple_loss=0.2269, pruned_loss=0.02708, over 4910.00 frames.], tot_loss[loss=0.1503, simple_loss=0.2309, pruned_loss=0.03485, over 983105.83 frames.], batch size: 52, aishell_tot_loss[loss=0.1524, simple_loss=0.2381, pruned_loss=0.03334, over 935520.71 frames.], datatang_tot_loss[loss=0.1475, simple_loss=0.2234, pruned_loss=0.03585, over 938084.02 frames.], batch size: 52, lr: 4.11e-04
+2022-06-19 00:31:51,399 INFO [train.py:874] (2/4) Epoch 20, batch 1250, datatang_loss[loss=0.1731, simple_loss=0.2499, pruned_loss=0.0481, over 4954.00 frames.], tot_loss[loss=0.15, simple_loss=0.2306, pruned_loss=0.03471, over 983475.17 frames.], batch size: 99, aishell_tot_loss[loss=0.1522, simple_loss=0.238, pruned_loss=0.03319, over 940999.80 frames.], datatang_tot_loss[loss=0.1476, simple_loss=0.2233, pruned_loss=0.03589, over 943888.33 frames.], batch size: 99, lr: 4.11e-04
+2022-06-19 00:32:25,047 INFO [train.py:874] (2/4) Epoch 20, batch 1300, aishell_loss[loss=0.1473, simple_loss=0.2337, pruned_loss=0.0305, over 4933.00 frames.], tot_loss[loss=0.1498, simple_loss=0.2303, pruned_loss=0.03467, over 983925.75 frames.], batch size: 49, aishell_tot_loss[loss=0.1521, simple_loss=0.2379, pruned_loss=0.03319, over 946078.93 frames.], datatang_tot_loss[loss=0.1474, simple_loss=0.2231, pruned_loss=0.03587, over 948931.09 frames.], batch size: 49, lr: 4.11e-04
+2022-06-19 00:32:57,716 INFO [train.py:874] (2/4) Epoch 20, batch 1350, aishell_loss[loss=0.1929, simple_loss=0.2866, pruned_loss=0.04959, over 4915.00 frames.], tot_loss[loss=0.1501, simple_loss=0.2306, pruned_loss=0.03475, over 984600.26 frames.], batch size: 78, aishell_tot_loss[loss=0.1526, simple_loss=0.2382, pruned_loss=0.03345, over 951330.93 frames.], datatang_tot_loss[loss=0.1472, simple_loss=0.2229, pruned_loss=0.03574, over 953007.49 frames.], batch size: 78, lr: 4.11e-04
+2022-06-19 00:33:29,214 INFO [train.py:874] (2/4) Epoch 20, batch 1400, datatang_loss[loss=0.1405, simple_loss=0.2225, pruned_loss=0.02928, over 4913.00 frames.], tot_loss[loss=0.1501, simple_loss=0.231, pruned_loss=0.0346, over 985142.36 frames.], batch size: 75, aishell_tot_loss[loss=0.1527, simple_loss=0.2384, pruned_loss=0.0335, over 955902.50 frames.], datatang_tot_loss[loss=0.1471, simple_loss=0.2231, pruned_loss=0.03559, over 956658.92 frames.], batch size: 75, lr: 4.11e-04
+2022-06-19 00:33:59,640 INFO [train.py:874] (2/4) Epoch 20, batch 1450, datatang_loss[loss=0.1527, simple_loss=0.2307, pruned_loss=0.03736, over 4929.00 frames.], tot_loss[loss=0.1503, simple_loss=0.2314, pruned_loss=0.03458, over 985493.29 frames.], batch size: 83, aishell_tot_loss[loss=0.1528, simple_loss=0.2384, pruned_loss=0.03357, over 960647.14 frames.], datatang_tot_loss[loss=0.1471, simple_loss=0.223, pruned_loss=0.0356, over 959065.83 frames.], batch size: 83, lr: 4.11e-04
+2022-06-19 00:34:33,118 INFO [train.py:874] (2/4) Epoch 20, batch 1500, aishell_loss[loss=0.1183, simple_loss=0.1992, pruned_loss=0.01864, over 4989.00 frames.], tot_loss[loss=0.1495, simple_loss=0.2304, pruned_loss=0.03428, over 985578.14 frames.], batch size: 25, aishell_tot_loss[loss=0.1517, simple_loss=0.2374, pruned_loss=0.03299, over 963439.67 frames.], datatang_tot_loss[loss=0.1474, simple_loss=0.223, pruned_loss=0.03587, over 962410.23 frames.], batch size: 25, lr: 4.11e-04
+2022-06-19 00:35:03,389 INFO [train.py:874] (2/4) Epoch 20, batch 1550, datatang_loss[loss=0.1339, simple_loss=0.2151, pruned_loss=0.02635, over 4874.00 frames.], tot_loss[loss=0.1494, simple_loss=0.2299, pruned_loss=0.0344, over 985754.05 frames.], batch size: 36, aishell_tot_loss[loss=0.1518, simple_loss=0.2376, pruned_loss=0.03297, over 965082.26 frames.], datatang_tot_loss[loss=0.1474, simple_loss=0.2231, pruned_loss=0.03584, over 966282.46 frames.], batch size: 36, lr: 4.10e-04
+2022-06-19 00:35:34,668 INFO [train.py:874] (2/4) Epoch 20, batch 1600, datatang_loss[loss=0.1232, simple_loss=0.206, pruned_loss=0.02022, over 4941.00 frames.], tot_loss[loss=0.1497, simple_loss=0.2302, pruned_loss=0.0346, over 985439.81 frames.], batch size: 69, aishell_tot_loss[loss=0.1517, simple_loss=0.2373, pruned_loss=0.03303, over 967705.38 frames.], datatang_tot_loss[loss=0.1478, simple_loss=0.2234, pruned_loss=0.03607, over 968083.86 frames.], batch size: 69, lr: 4.10e-04
+2022-06-19 00:36:07,047 INFO [train.py:874] (2/4) Epoch 20, batch 1650, datatang_loss[loss=0.142, simple_loss=0.2134, pruned_loss=0.03528, over 4938.00 frames.], tot_loss[loss=0.1503, simple_loss=0.2306, pruned_loss=0.03496, over 985504.69 frames.], batch size: 79, aishell_tot_loss[loss=0.1513, simple_loss=0.2368, pruned_loss=0.03289, over 969797.73 frames.], datatang_tot_loss[loss=0.1487, simple_loss=0.2243, pruned_loss=0.03658, over 970197.11 frames.], batch size: 79, lr: 4.10e-04
+2022-06-19 00:36:36,975 INFO [train.py:874] (2/4) Epoch 20, batch 1700, aishell_loss[loss=0.1555, simple_loss=0.2471, pruned_loss=0.03196, over 4876.00 frames.], tot_loss[loss=0.1499, simple_loss=0.2303, pruned_loss=0.03475, over 985855.61 frames.], batch size: 35, aishell_tot_loss[loss=0.1512, simple_loss=0.2366, pruned_loss=0.03285, over 971498.07 frames.], datatang_tot_loss[loss=0.1485, simple_loss=0.2242, pruned_loss=0.03643, over 972517.79 frames.], batch size: 35, lr: 4.10e-04
+2022-06-19 00:37:08,875 INFO [train.py:874] (2/4) Epoch 20, batch 1750, aishell_loss[loss=0.1553, simple_loss=0.2366, pruned_loss=0.03694, over 4873.00 frames.], tot_loss[loss=0.1499, simple_loss=0.2306, pruned_loss=0.03465, over 985452.05 frames.], batch size: 42, aishell_tot_loss[loss=0.1512, simple_loss=0.2366, pruned_loss=0.0329, over 972743.00 frames.], datatang_tot_loss[loss=0.1486, simple_loss=0.2246, pruned_loss=0.03626, over 974102.94 frames.], batch size: 42, lr: 4.10e-04
+2022-06-19 00:37:42,115 INFO [train.py:874] (2/4) Epoch 20, batch 1800, aishell_loss[loss=0.1463, simple_loss=0.2258, pruned_loss=0.03335, over 4877.00 frames.], tot_loss[loss=0.1488, simple_loss=0.2297, pruned_loss=0.03397, over 985177.37 frames.], batch size: 42, aishell_tot_loss[loss=0.1505, simple_loss=0.2357, pruned_loss=0.03262, over 974209.10 frames.], datatang_tot_loss[loss=0.148, simple_loss=0.2243, pruned_loss=0.03592, over 975197.22 frames.], batch size: 42, lr: 4.10e-04
+2022-06-19 00:38:12,141 INFO [train.py:874] (2/4) Epoch 20, batch 1850, aishell_loss[loss=0.1399, simple_loss=0.2393, pruned_loss=0.02021, over 4972.00 frames.], tot_loss[loss=0.1493, simple_loss=0.2302, pruned_loss=0.03424, over 985353.61 frames.], batch size: 44, aishell_tot_loss[loss=0.1512, simple_loss=0.2367, pruned_loss=0.03285, over 975514.72 frames.], datatang_tot_loss[loss=0.1478, simple_loss=0.2236, pruned_loss=0.03597, over 976567.50 frames.], batch size: 44, lr: 4.10e-04
+2022-06-19 00:38:43,846 INFO [train.py:874] (2/4) Epoch 20, batch 1900, aishell_loss[loss=0.1462, simple_loss=0.2338, pruned_loss=0.02933, over 4882.00 frames.], tot_loss[loss=0.1493, simple_loss=0.2306, pruned_loss=0.03406, over 985369.29 frames.], batch size: 47, aishell_tot_loss[loss=0.1517, simple_loss=0.2374, pruned_loss=0.033, over 976310.94 frames.], datatang_tot_loss[loss=0.1473, simple_loss=0.2236, pruned_loss=0.03552, over 977942.76 frames.], batch size: 47, lr: 4.10e-04
+2022-06-19 00:39:16,001 INFO [train.py:874] (2/4) Epoch 20, batch 1950, datatang_loss[loss=0.1271, simple_loss=0.2099, pruned_loss=0.0222, over 4930.00 frames.], tot_loss[loss=0.1493, simple_loss=0.23, pruned_loss=0.03428, over 985476.81 frames.], batch size: 83, aishell_tot_loss[loss=0.1514, simple_loss=0.2368, pruned_loss=0.03301, over 977631.74 frames.], datatang_tot_loss[loss=0.1475, simple_loss=0.2235, pruned_loss=0.03574, over 978695.63 frames.], batch size: 83, lr: 4.09e-04
+2022-06-19 00:39:45,777 INFO [train.py:874] (2/4) Epoch 20, batch 2000, aishell_loss[loss=0.1575, simple_loss=0.2392, pruned_loss=0.0379, over 4864.00 frames.], tot_loss[loss=0.1492, simple_loss=0.2302, pruned_loss=0.03413, over 985871.50 frames.], batch size: 36, aishell_tot_loss[loss=0.1514, simple_loss=0.2371, pruned_loss=0.03288, over 978855.58 frames.], datatang_tot_loss[loss=0.1473, simple_loss=0.2233, pruned_loss=0.03568, over 979613.11 frames.], batch size: 36, lr: 4.09e-04
+2022-06-19 00:39:45,778 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 00:40:02,385 INFO [train.py:914] (2/4) Epoch 20, validation: loss=0.1645, simple_loss=0.2482, pruned_loss=0.0404, over 1622729.00 frames.
+2022-06-19 00:40:32,677 INFO [train.py:874] (2/4) Epoch 20, batch 2050, datatang_loss[loss=0.1713, simple_loss=0.2385, pruned_loss=0.05205, over 4912.00 frames.], tot_loss[loss=0.1487, simple_loss=0.2296, pruned_loss=0.03392, over 985243.07 frames.], batch size: 64, aishell_tot_loss[loss=0.1506, simple_loss=0.2362, pruned_loss=0.03253, over 979091.93 frames.], datatang_tot_loss[loss=0.1475, simple_loss=0.2234, pruned_loss=0.03582, over 980288.64 frames.], batch size: 64, lr: 4.09e-04
+2022-06-19 00:41:04,529 INFO [train.py:874] (2/4) Epoch 20, batch 2100, datatang_loss[loss=0.1387, simple_loss=0.2176, pruned_loss=0.02984, over 4940.00 frames.], tot_loss[loss=0.1492, simple_loss=0.2301, pruned_loss=0.03415, over 985344.55 frames.], batch size: 69, aishell_tot_loss[loss=0.151, simple_loss=0.2367, pruned_loss=0.03264, over 979600.26 frames.], datatang_tot_loss[loss=0.1477, simple_loss=0.2237, pruned_loss=0.03583, over 981166.62 frames.], batch size: 69, lr: 4.09e-04
+2022-06-19 00:41:37,685 INFO [train.py:874] (2/4) Epoch 20, batch 2150, aishell_loss[loss=0.1185, simple_loss=0.1814, pruned_loss=0.02783, over 4952.00 frames.], tot_loss[loss=0.1487, simple_loss=0.2294, pruned_loss=0.03399, over 985555.06 frames.], batch size: 21, aishell_tot_loss[loss=0.1503, simple_loss=0.2361, pruned_loss=0.03224, over 980305.75 frames.], datatang_tot_loss[loss=0.1479, simple_loss=0.2239, pruned_loss=0.03596, over 981816.18 frames.], batch size: 21, lr: 4.09e-04
+2022-06-19 00:42:09,207 INFO [train.py:874] (2/4) Epoch 20, batch 2200, aishell_loss[loss=0.1546, simple_loss=0.2296, pruned_loss=0.03985, over 4945.00 frames.], tot_loss[loss=0.1499, simple_loss=0.2302, pruned_loss=0.03477, over 985827.81 frames.], batch size: 25, aishell_tot_loss[loss=0.1507, simple_loss=0.2363, pruned_loss=0.03254, over 980735.56 frames.], datatang_tot_loss[loss=0.1487, simple_loss=0.2246, pruned_loss=0.03639, over 982700.66 frames.], batch size: 25, lr: 4.09e-04
+2022-06-19 00:42:40,347 INFO [train.py:874] (2/4) Epoch 20, batch 2250, datatang_loss[loss=0.1487, simple_loss=0.2182, pruned_loss=0.03955, over 4925.00 frames.], tot_loss[loss=0.15, simple_loss=0.2304, pruned_loss=0.03482, over 985773.26 frames.], batch size: 77, aishell_tot_loss[loss=0.1511, simple_loss=0.2368, pruned_loss=0.03271, over 981021.33 frames.], datatang_tot_loss[loss=0.1485, simple_loss=0.2246, pruned_loss=0.03623, over 983280.45 frames.], batch size: 77, lr: 4.09e-04
+2022-06-19 00:43:13,255 INFO [train.py:874] (2/4) Epoch 20, batch 2300, aishell_loss[loss=0.1564, simple_loss=0.2388, pruned_loss=0.03698, over 4971.00 frames.], tot_loss[loss=0.1509, simple_loss=0.2318, pruned_loss=0.03497, over 986132.68 frames.], batch size: 39, aishell_tot_loss[loss=0.1515, simple_loss=0.2374, pruned_loss=0.03278, over 981829.64 frames.], datatang_tot_loss[loss=0.149, simple_loss=0.2251, pruned_loss=0.0365, over 983754.10 frames.], batch size: 39, lr: 4.09e-04
+2022-06-19 00:43:44,359 INFO [train.py:874] (2/4) Epoch 20, batch 2350, datatang_loss[loss=0.1665, simple_loss=0.2335, pruned_loss=0.04976, over 4924.00 frames.], tot_loss[loss=0.1507, simple_loss=0.2314, pruned_loss=0.03499, over 985499.83 frames.], batch size: 42, aishell_tot_loss[loss=0.1511, simple_loss=0.2369, pruned_loss=0.03262, over 981925.15 frames.], datatang_tot_loss[loss=0.1494, simple_loss=0.2254, pruned_loss=0.0367, over 983764.15 frames.], batch size: 42, lr: 4.08e-04
+2022-06-19 00:44:15,710 INFO [train.py:874] (2/4) Epoch 20, batch 2400, datatang_loss[loss=0.1791, simple_loss=0.2452, pruned_loss=0.05646, over 4949.00 frames.], tot_loss[loss=0.1506, simple_loss=0.2317, pruned_loss=0.03478, over 985104.01 frames.], batch size: 37, aishell_tot_loss[loss=0.1516, simple_loss=0.2376, pruned_loss=0.0328, over 982008.34 frames.], datatang_tot_loss[loss=0.149, simple_loss=0.2252, pruned_loss=0.03636, over 983862.46 frames.], batch size: 37, lr: 4.08e-04
+2022-06-19 00:44:49,180 INFO [train.py:874] (2/4) Epoch 20, batch 2450, aishell_loss[loss=0.1143, simple_loss=0.1814, pruned_loss=0.02366, over 4814.00 frames.], tot_loss[loss=0.1503, simple_loss=0.2313, pruned_loss=0.03466, over 985012.18 frames.], batch size: 21, aishell_tot_loss[loss=0.1515, simple_loss=0.2374, pruned_loss=0.03277, over 982188.05 frames.], datatang_tot_loss[loss=0.1489, simple_loss=0.2252, pruned_loss=0.03625, over 984078.07 frames.], batch size: 21, lr: 4.08e-04
+2022-06-19 00:45:21,529 INFO [train.py:874] (2/4) Epoch 20, batch 2500, datatang_loss[loss=0.1782, simple_loss=0.2522, pruned_loss=0.0521, over 4888.00 frames.], tot_loss[loss=0.1498, simple_loss=0.2305, pruned_loss=0.03459, over 985336.40 frames.], batch size: 47, aishell_tot_loss[loss=0.151, simple_loss=0.2368, pruned_loss=0.03259, over 982677.56 frames.], datatang_tot_loss[loss=0.1489, simple_loss=0.2252, pruned_loss=0.03634, over 984344.79 frames.], batch size: 47, lr: 4.08e-04
+2022-06-19 00:45:51,665 INFO [train.py:874] (2/4) Epoch 20, batch 2550, datatang_loss[loss=0.1541, simple_loss=0.2356, pruned_loss=0.03627, over 4959.00 frames.], tot_loss[loss=0.15, simple_loss=0.2303, pruned_loss=0.03485, over 985288.46 frames.], batch size: 91, aishell_tot_loss[loss=0.1513, simple_loss=0.237, pruned_loss=0.03276, over 982956.53 frames.], datatang_tot_loss[loss=0.1488, simple_loss=0.225, pruned_loss=0.03637, over 984415.34 frames.], batch size: 91, lr: 4.08e-04
+2022-06-19 00:46:25,316 INFO [train.py:874] (2/4) Epoch 20, batch 2600, datatang_loss[loss=0.1341, simple_loss=0.1999, pruned_loss=0.03418, over 4983.00 frames.], tot_loss[loss=0.1491, simple_loss=0.2296, pruned_loss=0.03431, over 985526.33 frames.], batch size: 40, aishell_tot_loss[loss=0.1511, simple_loss=0.2371, pruned_loss=0.03258, over 983194.18 frames.], datatang_tot_loss[loss=0.1481, simple_loss=0.2242, pruned_loss=0.03598, over 984802.38 frames.], batch size: 40, lr: 4.08e-04
+2022-06-19 00:46:57,788 INFO [train.py:874] (2/4) Epoch 20, batch 2650, aishell_loss[loss=0.1493, simple_loss=0.2353, pruned_loss=0.0317, over 4972.00 frames.], tot_loss[loss=0.1486, simple_loss=0.2298, pruned_loss=0.03376, over 985436.73 frames.], batch size: 61, aishell_tot_loss[loss=0.151, simple_loss=0.2373, pruned_loss=0.03239, over 983235.21 frames.], datatang_tot_loss[loss=0.1475, simple_loss=0.2239, pruned_loss=0.0356, over 985028.28 frames.], batch size: 61, lr: 4.08e-04
+2022-06-19 00:47:28,191 INFO [train.py:874] (2/4) Epoch 20, batch 2700, aishell_loss[loss=0.1601, simple_loss=0.2509, pruned_loss=0.03462, over 4942.00 frames.], tot_loss[loss=0.1482, simple_loss=0.2298, pruned_loss=0.03327, over 985790.68 frames.], batch size: 49, aishell_tot_loss[loss=0.1507, simple_loss=0.2373, pruned_loss=0.03203, over 983503.02 frames.], datatang_tot_loss[loss=0.1473, simple_loss=0.2238, pruned_loss=0.03539, over 985426.35 frames.], batch size: 49, lr: 4.08e-04
+2022-06-19 00:48:02,031 INFO [train.py:874] (2/4) Epoch 20, batch 2750, datatang_loss[loss=0.1378, simple_loss=0.2218, pruned_loss=0.02692, over 4920.00 frames.], tot_loss[loss=0.1479, simple_loss=0.2295, pruned_loss=0.03313, over 985736.51 frames.], batch size: 81, aishell_tot_loss[loss=0.1501, simple_loss=0.2367, pruned_loss=0.03176, over 983576.88 frames.], datatang_tot_loss[loss=0.1474, simple_loss=0.224, pruned_loss=0.03538, over 985595.56 frames.], batch size: 81, lr: 4.07e-04
+2022-06-19 00:48:34,546 INFO [train.py:874] (2/4) Epoch 20, batch 2800, aishell_loss[loss=0.1683, simple_loss=0.2619, pruned_loss=0.03738, over 4967.00 frames.], tot_loss[loss=0.1476, simple_loss=0.2294, pruned_loss=0.03292, over 985697.37 frames.], batch size: 80, aishell_tot_loss[loss=0.1506, simple_loss=0.2375, pruned_loss=0.03179, over 983627.57 frames.], datatang_tot_loss[loss=0.1465, simple_loss=0.2229, pruned_loss=0.035, over 985766.94 frames.], batch size: 80, lr: 4.07e-04
+2022-06-19 00:49:04,721 INFO [train.py:874] (2/4) Epoch 20, batch 2850, aishell_loss[loss=0.1342, simple_loss=0.221, pruned_loss=0.02367, over 4951.00 frames.], tot_loss[loss=0.1476, simple_loss=0.2291, pruned_loss=0.03305, over 985700.60 frames.], batch size: 31, aishell_tot_loss[loss=0.1508, simple_loss=0.2378, pruned_loss=0.03187, over 983752.00 frames.], datatang_tot_loss[loss=0.1461, simple_loss=0.2223, pruned_loss=0.03493, over 985850.56 frames.], batch size: 31, lr: 4.07e-04
+2022-06-19 00:49:38,230 INFO [train.py:874] (2/4) Epoch 20, batch 2900, aishell_loss[loss=0.1691, simple_loss=0.2572, pruned_loss=0.04047, over 4976.00 frames.], tot_loss[loss=0.1487, simple_loss=0.2303, pruned_loss=0.03356, over 985744.77 frames.], batch size: 51, aishell_tot_loss[loss=0.1511, simple_loss=0.2381, pruned_loss=0.032, over 984043.91 frames.], datatang_tot_loss[loss=0.1468, simple_loss=0.2229, pruned_loss=0.0353, over 985835.13 frames.], batch size: 51, lr: 4.07e-04
+2022-06-19 00:50:09,757 INFO [train.py:874] (2/4) Epoch 20, batch 2950, datatang_loss[loss=0.1497, simple_loss=0.2343, pruned_loss=0.03261, over 4915.00 frames.], tot_loss[loss=0.1488, simple_loss=0.23, pruned_loss=0.03383, over 985901.85 frames.], batch size: 81, aishell_tot_loss[loss=0.151, simple_loss=0.2377, pruned_loss=0.03213, over 984145.42 frames.], datatang_tot_loss[loss=0.147, simple_loss=0.2232, pruned_loss=0.03539, over 986062.78 frames.], batch size: 81, lr: 4.07e-04
+2022-06-19 00:50:39,733 INFO [train.py:874] (2/4) Epoch 20, batch 3000, datatang_loss[loss=0.1298, simple_loss=0.207, pruned_loss=0.02633, over 4929.00 frames.], tot_loss[loss=0.1484, simple_loss=0.2297, pruned_loss=0.0335, over 985724.68 frames.], batch size: 73, aishell_tot_loss[loss=0.151, simple_loss=0.2378, pruned_loss=0.03206, over 984323.72 frames.], datatang_tot_loss[loss=0.1465, simple_loss=0.2227, pruned_loss=0.03515, over 985875.12 frames.], batch size: 73, lr: 4.07e-04
+2022-06-19 00:50:39,734 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 00:50:56,984 INFO [train.py:914] (2/4) Epoch 20, validation: loss=0.164, simple_loss=0.2487, pruned_loss=0.03966, over 1622729.00 frames.
+2022-06-19 00:51:26,460 INFO [train.py:874] (2/4) Epoch 20, batch 3050, datatang_loss[loss=0.1694, simple_loss=0.2424, pruned_loss=0.04821, over 4930.00 frames.], tot_loss[loss=0.1477, simple_loss=0.2292, pruned_loss=0.03315, over 985884.13 frames.], batch size: 83, aishell_tot_loss[loss=0.1509, simple_loss=0.2379, pruned_loss=0.03192, over 984754.03 frames.], datatang_tot_loss[loss=0.1459, simple_loss=0.2223, pruned_loss=0.03482, over 985742.77 frames.], batch size: 83, lr: 4.07e-04
+2022-06-19 00:51:58,721 INFO [train.py:874] (2/4) Epoch 20, batch 3100, datatang_loss[loss=0.1262, simple_loss=0.1968, pruned_loss=0.02778, over 4933.00 frames.], tot_loss[loss=0.1473, simple_loss=0.2287, pruned_loss=0.03291, over 985636.65 frames.], batch size: 57, aishell_tot_loss[loss=0.1507, simple_loss=0.2379, pruned_loss=0.03179, over 984816.44 frames.], datatang_tot_loss[loss=0.1455, simple_loss=0.2217, pruned_loss=0.03462, over 985550.97 frames.], batch size: 57, lr: 4.07e-04
+2022-06-19 00:52:31,201 INFO [train.py:874] (2/4) Epoch 20, batch 3150, datatang_loss[loss=0.1349, simple_loss=0.2147, pruned_loss=0.02758, over 4972.00 frames.], tot_loss[loss=0.1489, simple_loss=0.2299, pruned_loss=0.03393, over 985387.80 frames.], batch size: 60, aishell_tot_loss[loss=0.151, simple_loss=0.2381, pruned_loss=0.03197, over 984504.70 frames.], datatang_tot_loss[loss=0.1468, simple_loss=0.2228, pruned_loss=0.03539, over 985689.18 frames.], batch size: 60, lr: 4.06e-04
+2022-06-19 00:53:02,565 INFO [train.py:874] (2/4) Epoch 20, batch 3200, datatang_loss[loss=0.1428, simple_loss=0.2182, pruned_loss=0.03372, over 4946.00 frames.], tot_loss[loss=0.1491, simple_loss=0.2304, pruned_loss=0.03396, over 985251.20 frames.], batch size: 50, aishell_tot_loss[loss=0.1513, simple_loss=0.2383, pruned_loss=0.03217, over 984466.25 frames.], datatang_tot_loss[loss=0.1466, simple_loss=0.2225, pruned_loss=0.0354, over 985688.29 frames.], batch size: 50, lr: 4.06e-04
+2022-06-19 00:53:34,156 INFO [train.py:874] (2/4) Epoch 20, batch 3250, datatang_loss[loss=0.1634, simple_loss=0.2316, pruned_loss=0.04764, over 4942.00 frames.], tot_loss[loss=0.1484, simple_loss=0.2298, pruned_loss=0.03346, over 985205.80 frames.], batch size: 37, aishell_tot_loss[loss=0.1509, simple_loss=0.238, pruned_loss=0.03195, over 984540.96 frames.], datatang_tot_loss[loss=0.1462, simple_loss=0.2222, pruned_loss=0.03513, over 985619.09 frames.], batch size: 37, lr: 4.06e-04
+2022-06-19 00:54:06,357 INFO [train.py:874] (2/4) Epoch 20, batch 3300, datatang_loss[loss=0.1461, simple_loss=0.2186, pruned_loss=0.03679, over 4946.00 frames.], tot_loss[loss=0.1484, simple_loss=0.2297, pruned_loss=0.03357, over 985439.94 frames.], batch size: 69, aishell_tot_loss[loss=0.1503, simple_loss=0.2373, pruned_loss=0.03166, over 984573.83 frames.], datatang_tot_loss[loss=0.1469, simple_loss=0.2229, pruned_loss=0.03546, over 985843.87 frames.], batch size: 69, lr: 4.06e-04
+2022-06-19 00:54:36,925 INFO [train.py:874] (2/4) Epoch 20, batch 3350, datatang_loss[loss=0.146, simple_loss=0.2174, pruned_loss=0.03725, over 4920.00 frames.], tot_loss[loss=0.1486, simple_loss=0.2297, pruned_loss=0.03376, over 985273.87 frames.], batch size: 64, aishell_tot_loss[loss=0.1508, simple_loss=0.2376, pruned_loss=0.03198, over 984506.40 frames.], datatang_tot_loss[loss=0.1466, simple_loss=0.2225, pruned_loss=0.03536, over 985810.39 frames.], batch size: 64, lr: 4.06e-04
+2022-06-19 00:55:10,684 INFO [train.py:874] (2/4) Epoch 20, batch 3400, datatang_loss[loss=0.1856, simple_loss=0.2611, pruned_loss=0.055, over 4927.00 frames.], tot_loss[loss=0.1491, simple_loss=0.2304, pruned_loss=0.03385, over 985296.20 frames.], batch size: 108, aishell_tot_loss[loss=0.1509, simple_loss=0.2378, pruned_loss=0.03202, over 984774.12 frames.], datatang_tot_loss[loss=0.1469, simple_loss=0.2228, pruned_loss=0.03544, over 985609.59 frames.], batch size: 108, lr: 4.06e-04
+2022-06-19 00:55:42,665 INFO [train.py:874] (2/4) Epoch 20, batch 3450, datatang_loss[loss=0.163, simple_loss=0.2369, pruned_loss=0.04451, over 4953.00 frames.], tot_loss[loss=0.1498, simple_loss=0.2307, pruned_loss=0.03449, over 985009.21 frames.], batch size: 91, aishell_tot_loss[loss=0.1513, simple_loss=0.238, pruned_loss=0.03232, over 984660.33 frames.], datatang_tot_loss[loss=0.1474, simple_loss=0.2233, pruned_loss=0.03576, over 985431.84 frames.], batch size: 91, lr: 4.06e-04
+2022-06-19 00:56:12,548 INFO [train.py:874] (2/4) Epoch 20, batch 3500, datatang_loss[loss=0.1694, simple_loss=0.2381, pruned_loss=0.05036, over 4906.00 frames.], tot_loss[loss=0.1495, simple_loss=0.2303, pruned_loss=0.03438, over 984886.69 frames.], batch size: 47, aishell_tot_loss[loss=0.1508, simple_loss=0.2374, pruned_loss=0.03213, over 984487.55 frames.], datatang_tot_loss[loss=0.1477, simple_loss=0.2235, pruned_loss=0.03595, over 985471.98 frames.], batch size: 47, lr: 4.06e-04
+2022-06-19 00:56:46,160 INFO [train.py:874] (2/4) Epoch 20, batch 3550, aishell_loss[loss=0.1443, simple_loss=0.2368, pruned_loss=0.02593, over 4948.00 frames.], tot_loss[loss=0.1488, simple_loss=0.2293, pruned_loss=0.03418, over 985170.29 frames.], batch size: 64, aishell_tot_loss[loss=0.1503, simple_loss=0.2368, pruned_loss=0.03192, over 984626.50 frames.], datatang_tot_loss[loss=0.1476, simple_loss=0.2233, pruned_loss=0.03593, over 985578.45 frames.], batch size: 64, lr: 4.05e-04
+2022-06-19 00:57:18,976 INFO [train.py:874] (2/4) Epoch 20, batch 3600, aishell_loss[loss=0.1324, simple_loss=0.2199, pruned_loss=0.02248, over 4946.00 frames.], tot_loss[loss=0.1492, simple_loss=0.2297, pruned_loss=0.03436, over 984815.38 frames.], batch size: 32, aishell_tot_loss[loss=0.1507, simple_loss=0.2372, pruned_loss=0.03209, over 984352.12 frames.], datatang_tot_loss[loss=0.1477, simple_loss=0.2235, pruned_loss=0.03592, over 985461.30 frames.], batch size: 32, lr: 4.05e-04
+2022-06-19 00:57:49,530 INFO [train.py:874] (2/4) Epoch 20, batch 3650, aishell_loss[loss=0.1598, simple_loss=0.2477, pruned_loss=0.03596, over 4941.00 frames.], tot_loss[loss=0.1491, simple_loss=0.2297, pruned_loss=0.03423, over 985022.71 frames.], batch size: 45, aishell_tot_loss[loss=0.1512, simple_loss=0.2377, pruned_loss=0.03231, over 984216.37 frames.], datatang_tot_loss[loss=0.1471, simple_loss=0.223, pruned_loss=0.03561, over 985775.65 frames.], batch size: 45, lr: 4.05e-04
+2022-06-19 00:58:21,860 INFO [train.py:874] (2/4) Epoch 20, batch 3700, aishell_loss[loss=0.1448, simple_loss=0.2322, pruned_loss=0.02869, over 4979.00 frames.], tot_loss[loss=0.1484, simple_loss=0.2291, pruned_loss=0.03387, over 984844.41 frames.], batch size: 51, aishell_tot_loss[loss=0.1512, simple_loss=0.2377, pruned_loss=0.03234, over 984237.95 frames.], datatang_tot_loss[loss=0.1464, simple_loss=0.2223, pruned_loss=0.03524, over 985550.38 frames.], batch size: 51, lr: 4.05e-04
+2022-06-19 00:58:52,955 INFO [train.py:874] (2/4) Epoch 20, batch 3750, aishell_loss[loss=0.1415, simple_loss=0.2293, pruned_loss=0.02679, over 4941.00 frames.], tot_loss[loss=0.1478, simple_loss=0.2284, pruned_loss=0.03359, over 984861.32 frames.], batch size: 58, aishell_tot_loss[loss=0.151, simple_loss=0.2373, pruned_loss=0.03233, over 984357.69 frames.], datatang_tot_loss[loss=0.1459, simple_loss=0.2219, pruned_loss=0.03494, over 985418.44 frames.], batch size: 58, lr: 4.05e-04
+2022-06-19 00:59:23,337 INFO [train.py:874] (2/4) Epoch 20, batch 3800, aishell_loss[loss=0.1477, simple_loss=0.2263, pruned_loss=0.03454, over 4926.00 frames.], tot_loss[loss=0.147, simple_loss=0.2275, pruned_loss=0.03327, over 984667.55 frames.], batch size: 33, aishell_tot_loss[loss=0.1507, simple_loss=0.2367, pruned_loss=0.03241, over 984037.11 frames.], datatang_tot_loss[loss=0.1451, simple_loss=0.2211, pruned_loss=0.03457, over 985553.25 frames.], batch size: 33, lr: 4.05e-04
+2022-06-19 00:59:56,334 INFO [train.py:874] (2/4) Epoch 20, batch 3850, aishell_loss[loss=0.1489, simple_loss=0.2453, pruned_loss=0.02623, over 4947.00 frames.], tot_loss[loss=0.1477, simple_loss=0.2279, pruned_loss=0.03374, over 984602.62 frames.], batch size: 54, aishell_tot_loss[loss=0.1512, simple_loss=0.237, pruned_loss=0.0327, over 984046.29 frames.], datatang_tot_loss[loss=0.1454, simple_loss=0.2215, pruned_loss=0.03466, over 985382.27 frames.], batch size: 54, lr: 4.05e-04
+2022-06-19 01:00:26,602 INFO [train.py:874] (2/4) Epoch 20, batch 3900, aishell_loss[loss=0.1621, simple_loss=0.2466, pruned_loss=0.03878, over 4925.00 frames.], tot_loss[loss=0.1475, simple_loss=0.2278, pruned_loss=0.03359, over 984872.98 frames.], batch size: 41, aishell_tot_loss[loss=0.1513, simple_loss=0.2371, pruned_loss=0.03277, over 984117.85 frames.], datatang_tot_loss[loss=0.145, simple_loss=0.2211, pruned_loss=0.03444, over 985581.06 frames.], batch size: 41, lr: 4.05e-04
+2022-06-19 01:00:57,345 INFO [train.py:874] (2/4) Epoch 20, batch 3950, aishell_loss[loss=0.141, simple_loss=0.2332, pruned_loss=0.02438, over 4940.00 frames.], tot_loss[loss=0.1478, simple_loss=0.2285, pruned_loss=0.03351, over 984897.34 frames.], batch size: 54, aishell_tot_loss[loss=0.1514, simple_loss=0.2374, pruned_loss=0.03268, over 984156.58 frames.], datatang_tot_loss[loss=0.145, simple_loss=0.2212, pruned_loss=0.03443, over 985560.51 frames.], batch size: 54, lr: 4.04e-04
+2022-06-19 01:01:28,305 INFO [train.py:874] (2/4) Epoch 20, batch 4000, datatang_loss[loss=0.1081, simple_loss=0.1845, pruned_loss=0.01588, over 4921.00 frames.], tot_loss[loss=0.1474, simple_loss=0.2282, pruned_loss=0.03329, over 984566.28 frames.], batch size: 42, aishell_tot_loss[loss=0.1516, simple_loss=0.2376, pruned_loss=0.03276, over 983788.54 frames.], datatang_tot_loss[loss=0.1445, simple_loss=0.2208, pruned_loss=0.0341, over 985569.62 frames.], batch size: 42, lr: 4.04e-04
+2022-06-19 01:01:28,306 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 01:01:44,966 INFO [train.py:914] (2/4) Epoch 20, validation: loss=0.1658, simple_loss=0.2493, pruned_loss=0.04118, over 1622729.00 frames.
+2022-06-19 01:02:16,037 INFO [train.py:874] (2/4) Epoch 20, batch 4050, aishell_loss[loss=0.1349, simple_loss=0.2285, pruned_loss=0.02063, over 4879.00 frames.], tot_loss[loss=0.1477, simple_loss=0.2285, pruned_loss=0.0335, over 984838.23 frames.], batch size: 42, aishell_tot_loss[loss=0.1515, simple_loss=0.2375, pruned_loss=0.03272, over 983947.38 frames.], datatang_tot_loss[loss=0.1449, simple_loss=0.2212, pruned_loss=0.0343, over 985638.97 frames.], batch size: 42, lr: 4.04e-04
+2022-06-19 01:02:45,507 INFO [train.py:874] (2/4) Epoch 20, batch 4100, aishell_loss[loss=0.1433, simple_loss=0.2284, pruned_loss=0.02911, over 4961.00 frames.], tot_loss[loss=0.1482, simple_loss=0.2289, pruned_loss=0.03381, over 985003.01 frames.], batch size: 30, aishell_tot_loss[loss=0.1519, simple_loss=0.2377, pruned_loss=0.03307, over 984296.63 frames.], datatang_tot_loss[loss=0.1448, simple_loss=0.221, pruned_loss=0.03431, over 985490.49 frames.], batch size: 30, lr: 4.04e-04
+2022-06-19 01:03:16,321 INFO [train.py:874] (2/4) Epoch 20, batch 4150, datatang_loss[loss=0.1236, simple_loss=0.2056, pruned_loss=0.02079, over 4933.00 frames.], tot_loss[loss=0.1484, simple_loss=0.2293, pruned_loss=0.03379, over 985341.95 frames.], batch size: 94, aishell_tot_loss[loss=0.1518, simple_loss=0.2378, pruned_loss=0.03292, over 984529.57 frames.], datatang_tot_loss[loss=0.1451, simple_loss=0.2212, pruned_loss=0.03448, over 985631.59 frames.], batch size: 94, lr: 4.04e-04
+2022-06-19 01:03:46,107 INFO [train.py:874] (2/4) Epoch 20, batch 4200, aishell_loss[loss=0.1391, simple_loss=0.2293, pruned_loss=0.02445, over 4870.00 frames.], tot_loss[loss=0.149, simple_loss=0.2301, pruned_loss=0.03397, over 985059.66 frames.], batch size: 35, aishell_tot_loss[loss=0.1519, simple_loss=0.2379, pruned_loss=0.03294, over 984136.76 frames.], datatang_tot_loss[loss=0.1457, simple_loss=0.222, pruned_loss=0.03464, over 985769.58 frames.], batch size: 35, lr: 4.04e-04
+2022-06-19 01:04:16,241 INFO [train.py:874] (2/4) Epoch 20, batch 4250, aishell_loss[loss=0.1759, simple_loss=0.26, pruned_loss=0.04594, over 4917.00 frames.], tot_loss[loss=0.1491, simple_loss=0.2303, pruned_loss=0.03392, over 984742.66 frames.], batch size: 41, aishell_tot_loss[loss=0.1522, simple_loss=0.2383, pruned_loss=0.03304, over 983934.61 frames.], datatang_tot_loss[loss=0.1455, simple_loss=0.2219, pruned_loss=0.03452, over 985665.22 frames.], batch size: 41, lr: 4.04e-04
+2022-06-19 01:05:47,342 INFO [train.py:874] (2/4) Epoch 21, batch 50, datatang_loss[loss=0.1434, simple_loss=0.2278, pruned_loss=0.02954, over 4904.00 frames.], tot_loss[loss=0.1401, simple_loss=0.2209, pruned_loss=0.02967, over 218149.34 frames.], batch size: 59, aishell_tot_loss[loss=0.1469, simple_loss=0.233, pruned_loss=0.03046, over 93821.40 frames.], datatang_tot_loss[loss=0.1357, simple_loss=0.2129, pruned_loss=0.02924, over 137426.57 frames.], batch size: 59, lr: 3.94e-04
+2022-06-19 01:06:17,242 INFO [train.py:874] (2/4) Epoch 21, batch 100, aishell_loss[loss=0.1238, simple_loss=0.1887, pruned_loss=0.02951, over 4819.00 frames.], tot_loss[loss=0.1433, simple_loss=0.2246, pruned_loss=0.03098, over 388256.02 frames.], batch size: 24, aishell_tot_loss[loss=0.1497, simple_loss=0.2363, pruned_loss=0.03155, over 198978.16 frames.], datatang_tot_loss[loss=0.1377, simple_loss=0.2145, pruned_loss=0.03041, over 237173.30 frames.], batch size: 24, lr: 3.94e-04
+2022-06-19 01:06:49,606 INFO [train.py:874] (2/4) Epoch 21, batch 150, aishell_loss[loss=0.1462, simple_loss=0.2327, pruned_loss=0.02981, over 4948.00 frames.], tot_loss[loss=0.1438, simple_loss=0.225, pruned_loss=0.0313, over 520595.55 frames.], batch size: 45, aishell_tot_loss[loss=0.1503, simple_loss=0.2364, pruned_loss=0.03213, over 298527.60 frames.], datatang_tot_loss[loss=0.1372, simple_loss=0.2137, pruned_loss=0.03035, over 318608.31 frames.], batch size: 45, lr: 3.94e-04
+2022-06-19 01:07:21,604 INFO [train.py:874] (2/4) Epoch 21, batch 200, datatang_loss[loss=0.1273, simple_loss=0.2054, pruned_loss=0.02463, over 4954.00 frames.], tot_loss[loss=0.1446, simple_loss=0.2256, pruned_loss=0.03181, over 623719.59 frames.], batch size: 42, aishell_tot_loss[loss=0.1513, simple_loss=0.237, pruned_loss=0.03277, over 391438.29 frames.], datatang_tot_loss[loss=0.1371, simple_loss=0.213, pruned_loss=0.03057, over 385344.82 frames.], batch size: 42, lr: 3.94e-04
+2022-06-19 01:07:50,963 INFO [train.py:874] (2/4) Epoch 21, batch 250, datatang_loss[loss=0.1352, simple_loss=0.2049, pruned_loss=0.03273, over 4981.00 frames.], tot_loss[loss=0.1442, simple_loss=0.2253, pruned_loss=0.03153, over 703768.80 frames.], batch size: 48, aishell_tot_loss[loss=0.1505, simple_loss=0.236, pruned_loss=0.03252, over 476709.93 frames.], datatang_tot_loss[loss=0.1365, simple_loss=0.2125, pruned_loss=0.03024, over 439921.21 frames.], batch size: 48, lr: 3.94e-04
+2022-06-19 01:08:22,372 INFO [train.py:874] (2/4) Epoch 21, batch 300, datatang_loss[loss=0.1415, simple_loss=0.2193, pruned_loss=0.03186, over 4927.00 frames.], tot_loss[loss=0.1456, simple_loss=0.2267, pruned_loss=0.03227, over 766299.10 frames.], batch size: 79, aishell_tot_loss[loss=0.1505, simple_loss=0.2362, pruned_loss=0.03238, over 547628.39 frames.], datatang_tot_loss[loss=0.1386, simple_loss=0.214, pruned_loss=0.03156, over 492167.00 frames.], batch size: 79, lr: 3.94e-04
+2022-06-19 01:08:53,927 INFO [train.py:874] (2/4) Epoch 21, batch 350, aishell_loss[loss=0.1282, simple_loss=0.1839, pruned_loss=0.03621, over 4895.00 frames.], tot_loss[loss=0.1465, simple_loss=0.2276, pruned_loss=0.03271, over 815124.19 frames.], batch size: 21, aishell_tot_loss[loss=0.1507, simple_loss=0.2366, pruned_loss=0.03239, over 601203.97 frames.], datatang_tot_loss[loss=0.1401, simple_loss=0.2154, pruned_loss=0.03238, over 548253.12 frames.], batch size: 21, lr: 3.93e-04
+2022-06-19 01:09:24,472 INFO [train.py:874] (2/4) Epoch 21, batch 400, datatang_loss[loss=0.1549, simple_loss=0.2242, pruned_loss=0.0428, over 4964.00 frames.], tot_loss[loss=0.1466, simple_loss=0.2272, pruned_loss=0.03301, over 852597.83 frames.], batch size: 34, aishell_tot_loss[loss=0.1509, simple_loss=0.2365, pruned_loss=0.03259, over 649694.81 frames.], datatang_tot_loss[loss=0.1404, simple_loss=0.2153, pruned_loss=0.03277, over 595707.93 frames.], batch size: 34, lr: 3.93e-04
+2022-06-19 01:09:56,120 INFO [train.py:874] (2/4) Epoch 21, batch 450, datatang_loss[loss=0.2028, simple_loss=0.2691, pruned_loss=0.06821, over 4947.00 frames.], tot_loss[loss=0.147, simple_loss=0.2273, pruned_loss=0.03332, over 882133.59 frames.], batch size: 99, aishell_tot_loss[loss=0.1511, simple_loss=0.2368, pruned_loss=0.03276, over 686375.25 frames.], datatang_tot_loss[loss=0.1411, simple_loss=0.216, pruned_loss=0.03313, over 645007.91 frames.], batch size: 99, lr: 3.93e-04
+2022-06-19 01:10:28,215 INFO [train.py:874] (2/4) Epoch 21, batch 500, aishell_loss[loss=0.1781, simple_loss=0.2563, pruned_loss=0.04992, over 4947.00 frames.], tot_loss[loss=0.1479, simple_loss=0.2278, pruned_loss=0.03402, over 905095.55 frames.], batch size: 40, aishell_tot_loss[loss=0.1515, simple_loss=0.2369, pruned_loss=0.03306, over 717536.12 frames.], datatang_tot_loss[loss=0.1425, simple_loss=0.2172, pruned_loss=0.03389, over 689703.87 frames.], batch size: 40, lr: 3.93e-04
+2022-06-19 01:10:58,422 INFO [train.py:874] (2/4) Epoch 21, batch 550, aishell_loss[loss=0.1492, simple_loss=0.2439, pruned_loss=0.02725, over 4942.00 frames.], tot_loss[loss=0.1472, simple_loss=0.2276, pruned_loss=0.03342, over 923201.58 frames.], batch size: 64, aishell_tot_loss[loss=0.151, simple_loss=0.2369, pruned_loss=0.03257, over 747979.48 frames.], datatang_tot_loss[loss=0.1425, simple_loss=0.2174, pruned_loss=0.03375, over 726077.07 frames.], batch size: 64, lr: 3.93e-04
+2022-06-19 01:11:30,246 INFO [train.py:874] (2/4) Epoch 21, batch 600, datatang_loss[loss=0.1241, simple_loss=0.2034, pruned_loss=0.02244, over 4943.00 frames.], tot_loss[loss=0.1477, simple_loss=0.2279, pruned_loss=0.03377, over 937002.91 frames.], batch size: 62, aishell_tot_loss[loss=0.1522, simple_loss=0.2376, pruned_loss=0.03339, over 775885.49 frames.], datatang_tot_loss[loss=0.142, simple_loss=0.2171, pruned_loss=0.03346, over 756688.88 frames.], batch size: 62, lr: 3.93e-04
+2022-06-19 01:12:03,148 INFO [train.py:874] (2/4) Epoch 21, batch 650, datatang_loss[loss=0.1251, simple_loss=0.1969, pruned_loss=0.02669, over 4945.00 frames.], tot_loss[loss=0.1469, simple_loss=0.2269, pruned_loss=0.03341, over 947789.78 frames.], batch size: 62, aishell_tot_loss[loss=0.152, simple_loss=0.2377, pruned_loss=0.03319, over 794119.20 frames.], datatang_tot_loss[loss=0.1417, simple_loss=0.2168, pruned_loss=0.0333, over 790494.76 frames.], batch size: 62, lr: 3.93e-04
+2022-06-19 01:12:34,557 INFO [train.py:874] (2/4) Epoch 21, batch 700, datatang_loss[loss=0.1549, simple_loss=0.2373, pruned_loss=0.03629, over 4926.00 frames.], tot_loss[loss=0.1467, simple_loss=0.2271, pruned_loss=0.0332, over 955898.50 frames.], batch size: 79, aishell_tot_loss[loss=0.1512, simple_loss=0.2371, pruned_loss=0.03265, over 814600.93 frames.], datatang_tot_loss[loss=0.1424, simple_loss=0.2178, pruned_loss=0.03357, over 815272.24 frames.], batch size: 79, lr: 3.93e-04
+2022-06-19 01:13:05,866 INFO [train.py:874] (2/4) Epoch 21, batch 750, aishell_loss[loss=0.1456, simple_loss=0.2404, pruned_loss=0.02546, over 4945.00 frames.], tot_loss[loss=0.1466, simple_loss=0.227, pruned_loss=0.03305, over 962764.50 frames.], batch size: 45, aishell_tot_loss[loss=0.1504, simple_loss=0.2362, pruned_loss=0.03226, over 834919.38 frames.], datatang_tot_loss[loss=0.1431, simple_loss=0.2186, pruned_loss=0.03382, over 835458.15 frames.], batch size: 45, lr: 3.93e-04
+2022-06-19 01:13:36,442 INFO [train.py:874] (2/4) Epoch 21, batch 800, datatang_loss[loss=0.1451, simple_loss=0.2201, pruned_loss=0.03507, over 4969.00 frames.], tot_loss[loss=0.1465, simple_loss=0.2273, pruned_loss=0.03283, over 967764.08 frames.], batch size: 60, aishell_tot_loss[loss=0.1508, simple_loss=0.2366, pruned_loss=0.0325, over 853148.29 frames.], datatang_tot_loss[loss=0.1425, simple_loss=0.2183, pruned_loss=0.03332, over 852599.57 frames.], batch size: 60, lr: 3.92e-04
+2022-06-19 01:14:08,622 INFO [train.py:874] (2/4) Epoch 21, batch 850, aishell_loss[loss=0.1436, simple_loss=0.2278, pruned_loss=0.02974, over 4864.00 frames.], tot_loss[loss=0.1466, simple_loss=0.2277, pruned_loss=0.03272, over 972038.73 frames.], batch size: 28, aishell_tot_loss[loss=0.1505, simple_loss=0.2364, pruned_loss=0.03231, over 868970.06 frames.], datatang_tot_loss[loss=0.1428, simple_loss=0.2189, pruned_loss=0.03336, over 868382.82 frames.], batch size: 28, lr: 3.92e-04
+2022-06-19 01:14:40,458 INFO [train.py:874] (2/4) Epoch 21, batch 900, aishell_loss[loss=0.1339, simple_loss=0.2233, pruned_loss=0.02225, over 4886.00 frames.], tot_loss[loss=0.1473, simple_loss=0.2283, pruned_loss=0.03317, over 975221.80 frames.], batch size: 28, aishell_tot_loss[loss=0.1507, simple_loss=0.2366, pruned_loss=0.03242, over 883125.22 frames.], datatang_tot_loss[loss=0.1434, simple_loss=0.2194, pruned_loss=0.03376, over 881938.86 frames.], batch size: 28, lr: 3.92e-04
+2022-06-19 01:15:10,904 INFO [train.py:874] (2/4) Epoch 21, batch 950, datatang_loss[loss=0.151, simple_loss=0.2217, pruned_loss=0.04016, over 4892.00 frames.], tot_loss[loss=0.1479, simple_loss=0.2291, pruned_loss=0.03335, over 977504.29 frames.], batch size: 47, aishell_tot_loss[loss=0.1511, simple_loss=0.2368, pruned_loss=0.03276, over 898730.60 frames.], datatang_tot_loss[loss=0.1435, simple_loss=0.2195, pruned_loss=0.03371, over 890397.95 frames.], batch size: 47, lr: 3.92e-04
+2022-06-19 01:15:49,478 INFO [train.py:874] (2/4) Epoch 21, batch 1000, aishell_loss[loss=0.134, simple_loss=0.2266, pruned_loss=0.02076, over 4936.00 frames.], tot_loss[loss=0.1485, simple_loss=0.2296, pruned_loss=0.03373, over 979195.81 frames.], batch size: 45, aishell_tot_loss[loss=0.1517, simple_loss=0.2376, pruned_loss=0.03285, over 907226.26 frames.], datatang_tot_loss[loss=0.144, simple_loss=0.2199, pruned_loss=0.03406, over 903350.23 frames.], batch size: 45, lr: 3.92e-04
+2022-06-19 01:15:49,479 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 01:16:06,684 INFO [train.py:914] (2/4) Epoch 21, validation: loss=0.1651, simple_loss=0.2487, pruned_loss=0.04071, over 1622729.00 frames.
+2022-06-19 01:16:39,091 INFO [train.py:874] (2/4) Epoch 21, batch 1050, datatang_loss[loss=0.1456, simple_loss=0.2241, pruned_loss=0.0335, over 4912.00 frames.], tot_loss[loss=0.1485, simple_loss=0.2295, pruned_loss=0.03377, over 979948.37 frames.], batch size: 77, aishell_tot_loss[loss=0.1508, simple_loss=0.2368, pruned_loss=0.03246, over 917201.06 frames.], datatang_tot_loss[loss=0.1448, simple_loss=0.2205, pruned_loss=0.03459, over 911507.89 frames.], batch size: 77, lr: 3.92e-04
+2022-06-19 01:17:11,986 INFO [train.py:874] (2/4) Epoch 21, batch 1100, datatang_loss[loss=0.1411, simple_loss=0.2235, pruned_loss=0.0293, over 4926.00 frames.], tot_loss[loss=0.1485, simple_loss=0.2297, pruned_loss=0.03363, over 980911.57 frames.], batch size: 57, aishell_tot_loss[loss=0.1505, simple_loss=0.2364, pruned_loss=0.03228, over 926881.61 frames.], datatang_tot_loss[loss=0.1452, simple_loss=0.2209, pruned_loss=0.03475, over 918106.16 frames.], batch size: 57, lr: 3.92e-04
+2022-06-19 01:17:43,373 INFO [train.py:874] (2/4) Epoch 21, batch 1150, datatang_loss[loss=0.1752, simple_loss=0.2496, pruned_loss=0.05037, over 4927.00 frames.], tot_loss[loss=0.1489, simple_loss=0.2299, pruned_loss=0.0339, over 981900.04 frames.], batch size: 94, aishell_tot_loss[loss=0.1503, simple_loss=0.2361, pruned_loss=0.03222, over 933949.60 frames.], datatang_tot_loss[loss=0.1459, simple_loss=0.2216, pruned_loss=0.03514, over 925798.28 frames.], batch size: 94, lr: 3.92e-04
+2022-06-19 01:18:16,585 INFO [train.py:874] (2/4) Epoch 21, batch 1200, datatang_loss[loss=0.1346, simple_loss=0.218, pruned_loss=0.02556, over 4978.00 frames.], tot_loss[loss=0.1488, simple_loss=0.2302, pruned_loss=0.0337, over 982304.19 frames.], batch size: 53, aishell_tot_loss[loss=0.1506, simple_loss=0.2364, pruned_loss=0.03241, over 940424.01 frames.], datatang_tot_loss[loss=0.1456, simple_loss=0.2216, pruned_loss=0.03482, over 931884.45 frames.], batch size: 53, lr: 3.91e-04
+2022-06-19 01:18:47,829 INFO [train.py:874] (2/4) Epoch 21, batch 1250, aishell_loss[loss=0.1509, simple_loss=0.246, pruned_loss=0.02794, over 4912.00 frames.], tot_loss[loss=0.1485, simple_loss=0.2303, pruned_loss=0.0334, over 983151.29 frames.], batch size: 68, aishell_tot_loss[loss=0.1503, simple_loss=0.2362, pruned_loss=0.03219, over 947034.00 frames.], datatang_tot_loss[loss=0.1457, simple_loss=0.2218, pruned_loss=0.03482, over 936776.32 frames.], batch size: 68, lr: 3.91e-04
+2022-06-19 01:19:18,448 INFO [train.py:874] (2/4) Epoch 21, batch 1300, aishell_loss[loss=0.1265, simple_loss=0.2175, pruned_loss=0.01771, over 4856.00 frames.], tot_loss[loss=0.1478, simple_loss=0.2298, pruned_loss=0.03295, over 983485.23 frames.], batch size: 37, aishell_tot_loss[loss=0.15, simple_loss=0.236, pruned_loss=0.03199, over 951794.32 frames.], datatang_tot_loss[loss=0.1453, simple_loss=0.2216, pruned_loss=0.03452, over 941972.70 frames.], batch size: 37, lr: 3.91e-04
+2022-06-19 01:19:51,536 INFO [train.py:874] (2/4) Epoch 21, batch 1350, aishell_loss[loss=0.1451, simple_loss=0.2384, pruned_loss=0.02594, over 4914.00 frames.], tot_loss[loss=0.1492, simple_loss=0.231, pruned_loss=0.03369, over 983915.97 frames.], batch size: 46, aishell_tot_loss[loss=0.1506, simple_loss=0.2369, pruned_loss=0.03217, over 955679.23 frames.], datatang_tot_loss[loss=0.1462, simple_loss=0.2224, pruned_loss=0.03506, over 947207.86 frames.], batch size: 46, lr: 3.91e-04
+2022-06-19 01:20:23,984 INFO [train.py:874] (2/4) Epoch 21, batch 1400, datatang_loss[loss=0.1349, simple_loss=0.2172, pruned_loss=0.02631, over 4962.00 frames.], tot_loss[loss=0.1491, simple_loss=0.2308, pruned_loss=0.03365, over 984228.56 frames.], batch size: 86, aishell_tot_loss[loss=0.151, simple_loss=0.2373, pruned_loss=0.03238, over 959003.28 frames.], datatang_tot_loss[loss=0.1459, simple_loss=0.2222, pruned_loss=0.03481, over 951907.87 frames.], batch size: 86, lr: 3.91e-04
+2022-06-19 01:20:54,654 INFO [train.py:874] (2/4) Epoch 21, batch 1450, datatang_loss[loss=0.1349, simple_loss=0.2163, pruned_loss=0.02678, over 4950.00 frames.], tot_loss[loss=0.148, simple_loss=0.2299, pruned_loss=0.03308, over 984485.98 frames.], batch size: 91, aishell_tot_loss[loss=0.1503, simple_loss=0.2366, pruned_loss=0.03203, over 961928.99 frames.], datatang_tot_loss[loss=0.1456, simple_loss=0.2221, pruned_loss=0.03455, over 956021.76 frames.], batch size: 91, lr: 3.91e-04
+2022-06-19 01:21:28,252 INFO [train.py:874] (2/4) Epoch 21, batch 1500, datatang_loss[loss=0.1489, simple_loss=0.2182, pruned_loss=0.03983, over 4981.00 frames.], tot_loss[loss=0.1479, simple_loss=0.2294, pruned_loss=0.03319, over 984769.52 frames.], batch size: 37, aishell_tot_loss[loss=0.1501, simple_loss=0.2365, pruned_loss=0.03179, over 964450.29 frames.], datatang_tot_loss[loss=0.1458, simple_loss=0.222, pruned_loss=0.03482, over 959889.63 frames.], batch size: 37, lr: 3.91e-04
+2022-06-19 01:21:57,780 INFO [train.py:874] (2/4) Epoch 21, batch 1550, aishell_loss[loss=0.1885, simple_loss=0.2592, pruned_loss=0.05886, over 4944.00 frames.], tot_loss[loss=0.1478, simple_loss=0.2294, pruned_loss=0.03313, over 984928.32 frames.], batch size: 45, aishell_tot_loss[loss=0.1497, simple_loss=0.236, pruned_loss=0.03169, over 966907.52 frames.], datatang_tot_loss[loss=0.1461, simple_loss=0.2224, pruned_loss=0.03484, over 962885.27 frames.], batch size: 45, lr: 3.91e-04
+2022-06-19 01:22:31,080 INFO [train.py:874] (2/4) Epoch 21, batch 1600, datatang_loss[loss=0.1342, simple_loss=0.2118, pruned_loss=0.02832, over 4927.00 frames.], tot_loss[loss=0.1476, simple_loss=0.2292, pruned_loss=0.03303, over 985180.22 frames.], batch size: 79, aishell_tot_loss[loss=0.1494, simple_loss=0.2356, pruned_loss=0.03163, over 969447.88 frames.], datatang_tot_loss[loss=0.146, simple_loss=0.2223, pruned_loss=0.03482, over 965240.79 frames.], batch size: 79, lr: 3.91e-04
+2022-06-19 01:23:02,969 INFO [train.py:874] (2/4) Epoch 21, batch 1650, datatang_loss[loss=0.1698, simple_loss=0.2513, pruned_loss=0.04414, over 4923.00 frames.], tot_loss[loss=0.148, simple_loss=0.2293, pruned_loss=0.03328, over 985188.74 frames.], batch size: 94, aishell_tot_loss[loss=0.15, simple_loss=0.2361, pruned_loss=0.03195, over 971108.20 frames.], datatang_tot_loss[loss=0.1458, simple_loss=0.2221, pruned_loss=0.0347, over 967816.79 frames.], batch size: 94, lr: 3.90e-04
+2022-06-19 01:23:32,744 INFO [train.py:874] (2/4) Epoch 21, batch 1700, aishell_loss[loss=0.1524, simple_loss=0.2414, pruned_loss=0.03172, over 4944.00 frames.], tot_loss[loss=0.1479, simple_loss=0.2291, pruned_loss=0.03341, over 985100.59 frames.], batch size: 45, aishell_tot_loss[loss=0.1505, simple_loss=0.2366, pruned_loss=0.03217, over 972613.79 frames.], datatang_tot_loss[loss=0.1454, simple_loss=0.2216, pruned_loss=0.03458, over 969975.99 frames.], batch size: 45, lr: 3.90e-04
+2022-06-19 01:24:06,674 INFO [train.py:874] (2/4) Epoch 21, batch 1750, aishell_loss[loss=0.1372, simple_loss=0.2274, pruned_loss=0.02348, over 4878.00 frames.], tot_loss[loss=0.1475, simple_loss=0.2284, pruned_loss=0.03333, over 985086.78 frames.], batch size: 28, aishell_tot_loss[loss=0.1503, simple_loss=0.2364, pruned_loss=0.03205, over 973542.31 frames.], datatang_tot_loss[loss=0.1453, simple_loss=0.2214, pruned_loss=0.03456, over 972333.13 frames.], batch size: 28, lr: 3.90e-04
+2022-06-19 01:24:40,279 INFO [train.py:874] (2/4) Epoch 21, batch 1800, datatang_loss[loss=0.1243, simple_loss=0.205, pruned_loss=0.02182, over 4919.00 frames.], tot_loss[loss=0.1476, simple_loss=0.2285, pruned_loss=0.03328, over 984547.73 frames.], batch size: 73, aishell_tot_loss[loss=0.1504, simple_loss=0.2363, pruned_loss=0.0322, over 974573.91 frames.], datatang_tot_loss[loss=0.1451, simple_loss=0.2213, pruned_loss=0.0344, over 973564.87 frames.], batch size: 73, lr: 3.90e-04
+2022-06-19 01:25:09,443 INFO [train.py:874] (2/4) Epoch 21, batch 1850, datatang_loss[loss=0.1419, simple_loss=0.2168, pruned_loss=0.03348, over 4931.00 frames.], tot_loss[loss=0.1469, simple_loss=0.228, pruned_loss=0.03294, over 984866.61 frames.], batch size: 79, aishell_tot_loss[loss=0.15, simple_loss=0.2358, pruned_loss=0.03205, over 976007.65 frames.], datatang_tot_loss[loss=0.1447, simple_loss=0.221, pruned_loss=0.03419, over 974907.53 frames.], batch size: 79, lr: 3.90e-04
+2022-06-19 01:25:41,903 INFO [train.py:874] (2/4) Epoch 21, batch 1900, aishell_loss[loss=0.1489, simple_loss=0.2387, pruned_loss=0.02952, over 4919.00 frames.], tot_loss[loss=0.1476, simple_loss=0.2289, pruned_loss=0.03319, over 984703.96 frames.], batch size: 46, aishell_tot_loss[loss=0.1505, simple_loss=0.2363, pruned_loss=0.03239, over 977078.13 frames.], datatang_tot_loss[loss=0.1447, simple_loss=0.2211, pruned_loss=0.03413, over 975855.59 frames.], batch size: 46, lr: 3.90e-04
+2022-06-19 01:26:13,749 INFO [train.py:874] (2/4) Epoch 21, batch 1950, datatang_loss[loss=0.1446, simple_loss=0.2291, pruned_loss=0.03008, over 4939.00 frames.], tot_loss[loss=0.1482, simple_loss=0.2296, pruned_loss=0.03344, over 985041.98 frames.], batch size: 79, aishell_tot_loss[loss=0.1501, simple_loss=0.236, pruned_loss=0.0321, over 978067.19 frames.], datatang_tot_loss[loss=0.1458, simple_loss=0.2222, pruned_loss=0.03467, over 977173.58 frames.], batch size: 79, lr: 3.90e-04
+2022-06-19 01:26:45,060 INFO [train.py:874] (2/4) Epoch 21, batch 2000, aishell_loss[loss=0.1493, simple_loss=0.2357, pruned_loss=0.03148, over 4867.00 frames.], tot_loss[loss=0.1483, simple_loss=0.2297, pruned_loss=0.03344, over 985492.80 frames.], batch size: 38, aishell_tot_loss[loss=0.1504, simple_loss=0.2365, pruned_loss=0.03215, over 979240.09 frames.], datatang_tot_loss[loss=0.1455, simple_loss=0.2216, pruned_loss=0.03467, over 978221.94 frames.], batch size: 38, lr: 3.90e-04
+2022-06-19 01:26:45,061 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 01:27:02,716 INFO [train.py:914] (2/4) Epoch 21, validation: loss=0.164, simple_loss=0.2481, pruned_loss=0.03994, over 1622729.00 frames.
+2022-06-19 01:27:32,171 INFO [train.py:874] (2/4) Epoch 21, batch 2050, aishell_loss[loss=0.1364, simple_loss=0.2349, pruned_loss=0.01899, over 4916.00 frames.], tot_loss[loss=0.148, simple_loss=0.2296, pruned_loss=0.03321, over 985017.48 frames.], batch size: 52, aishell_tot_loss[loss=0.1497, simple_loss=0.2361, pruned_loss=0.03167, over 979614.31 frames.], datatang_tot_loss[loss=0.1459, simple_loss=0.2222, pruned_loss=0.03484, over 978947.53 frames.], batch size: 52, lr: 3.90e-04
+2022-06-19 01:28:06,084 INFO [train.py:874] (2/4) Epoch 21, batch 2100, aishell_loss[loss=0.1757, simple_loss=0.2567, pruned_loss=0.04733, over 4911.00 frames.], tot_loss[loss=0.1477, simple_loss=0.2288, pruned_loss=0.0333, over 984651.84 frames.], batch size: 41, aishell_tot_loss[loss=0.1496, simple_loss=0.2356, pruned_loss=0.03177, over 979863.04 frames.], datatang_tot_loss[loss=0.1458, simple_loss=0.2221, pruned_loss=0.03482, over 979659.44 frames.], batch size: 41, lr: 3.89e-04
+2022-06-19 01:28:39,721 INFO [train.py:874] (2/4) Epoch 21, batch 2150, aishell_loss[loss=0.1829, simple_loss=0.2662, pruned_loss=0.04982, over 4962.00 frames.], tot_loss[loss=0.148, simple_loss=0.2291, pruned_loss=0.03341, over 984902.40 frames.], batch size: 79, aishell_tot_loss[loss=0.1499, simple_loss=0.2359, pruned_loss=0.03194, over 980402.88 frames.], datatang_tot_loss[loss=0.1458, simple_loss=0.2222, pruned_loss=0.03473, over 980525.37 frames.], batch size: 79, lr: 3.89e-04
+2022-06-19 01:29:09,140 INFO [train.py:874] (2/4) Epoch 21, batch 2200, aishell_loss[loss=0.1708, simple_loss=0.2528, pruned_loss=0.04442, over 4932.00 frames.], tot_loss[loss=0.1481, simple_loss=0.2292, pruned_loss=0.03351, over 985120.38 frames.], batch size: 33, aishell_tot_loss[loss=0.1501, simple_loss=0.236, pruned_loss=0.0321, over 981229.08 frames.], datatang_tot_loss[loss=0.1458, simple_loss=0.2222, pruned_loss=0.03469, over 980970.83 frames.], batch size: 33, lr: 3.89e-04
+2022-06-19 01:29:41,129 INFO [train.py:874] (2/4) Epoch 21, batch 2250, datatang_loss[loss=0.1359, simple_loss=0.2001, pruned_loss=0.03589, over 4902.00 frames.], tot_loss[loss=0.1487, simple_loss=0.2299, pruned_loss=0.03372, over 985330.56 frames.], batch size: 47, aishell_tot_loss[loss=0.1502, simple_loss=0.2363, pruned_loss=0.03206, over 981875.08 frames.], datatang_tot_loss[loss=0.1463, simple_loss=0.2227, pruned_loss=0.03498, over 981499.24 frames.], batch size: 47, lr: 3.89e-04
+2022-06-19 01:30:15,114 INFO [train.py:874] (2/4) Epoch 21, batch 2300, aishell_loss[loss=0.1489, simple_loss=0.2367, pruned_loss=0.03056, over 4979.00 frames.], tot_loss[loss=0.1491, simple_loss=0.2303, pruned_loss=0.03401, over 985348.88 frames.], batch size: 51, aishell_tot_loss[loss=0.1503, simple_loss=0.2365, pruned_loss=0.03206, over 982090.25 frames.], datatang_tot_loss[loss=0.1469, simple_loss=0.2231, pruned_loss=0.03532, over 982159.75 frames.], batch size: 51, lr: 3.89e-04
+2022-06-19 01:30:45,555 INFO [train.py:874] (2/4) Epoch 21, batch 2350, aishell_loss[loss=0.1791, simple_loss=0.2649, pruned_loss=0.0466, over 4862.00 frames.], tot_loss[loss=0.1486, simple_loss=0.2295, pruned_loss=0.03385, over 985175.86 frames.], batch size: 36, aishell_tot_loss[loss=0.1499, simple_loss=0.2358, pruned_loss=0.03197, over 982347.59 frames.], datatang_tot_loss[loss=0.1469, simple_loss=0.2231, pruned_loss=0.03533, over 982475.60 frames.], batch size: 36, lr: 3.89e-04
+2022-06-19 01:31:19,319 INFO [train.py:874] (2/4) Epoch 21, batch 2400, datatang_loss[loss=0.1558, simple_loss=0.238, pruned_loss=0.03685, over 4963.00 frames.], tot_loss[loss=0.1489, simple_loss=0.2301, pruned_loss=0.03389, over 985317.24 frames.], batch size: 99, aishell_tot_loss[loss=0.1503, simple_loss=0.2362, pruned_loss=0.03217, over 982745.15 frames.], datatang_tot_loss[loss=0.1468, simple_loss=0.2232, pruned_loss=0.03524, over 982880.68 frames.], batch size: 99, lr: 3.89e-04
+2022-06-19 01:31:50,758 INFO [train.py:874] (2/4) Epoch 21, batch 2450, datatang_loss[loss=0.1688, simple_loss=0.2489, pruned_loss=0.04432, over 4969.00 frames.], tot_loss[loss=0.1488, simple_loss=0.2302, pruned_loss=0.03368, over 985457.03 frames.], batch size: 55, aishell_tot_loss[loss=0.1503, simple_loss=0.2362, pruned_loss=0.03217, over 983029.35 frames.], datatang_tot_loss[loss=0.1468, simple_loss=0.2235, pruned_loss=0.03503, over 983322.77 frames.], batch size: 55, lr: 3.89e-04
+2022-06-19 01:32:21,755 INFO [train.py:874] (2/4) Epoch 21, batch 2500, aishell_loss[loss=0.1547, simple_loss=0.2367, pruned_loss=0.03637, over 4963.00 frames.], tot_loss[loss=0.1491, simple_loss=0.2308, pruned_loss=0.03367, over 985565.39 frames.], batch size: 64, aishell_tot_loss[loss=0.1508, simple_loss=0.2368, pruned_loss=0.03243, over 983614.55 frames.], datatang_tot_loss[loss=0.1466, simple_loss=0.2234, pruned_loss=0.03487, over 983402.52 frames.], batch size: 64, lr: 3.89e-04
+2022-06-19 01:32:56,166 INFO [train.py:874] (2/4) Epoch 21, batch 2550, aishell_loss[loss=0.1549, simple_loss=0.2455, pruned_loss=0.03211, over 4947.00 frames.], tot_loss[loss=0.1485, simple_loss=0.2298, pruned_loss=0.03362, over 985284.03 frames.], batch size: 45, aishell_tot_loss[loss=0.1507, simple_loss=0.2368, pruned_loss=0.0323, over 983669.03 frames.], datatang_tot_loss[loss=0.1462, simple_loss=0.2226, pruned_loss=0.03492, over 983541.47 frames.], batch size: 45, lr: 3.88e-04
+2022-06-19 01:33:29,089 INFO [train.py:874] (2/4) Epoch 21, batch 2600, datatang_loss[loss=0.1216, simple_loss=0.2039, pruned_loss=0.01968, over 4895.00 frames.], tot_loss[loss=0.1477, simple_loss=0.2288, pruned_loss=0.03327, over 985592.13 frames.], batch size: 52, aishell_tot_loss[loss=0.1504, simple_loss=0.2364, pruned_loss=0.0322, over 984158.98 frames.], datatang_tot_loss[loss=0.1456, simple_loss=0.222, pruned_loss=0.03465, over 983772.39 frames.], batch size: 52, lr: 3.88e-04
+2022-06-19 01:33:59,963 INFO [train.py:874] (2/4) Epoch 21, batch 2650, aishell_loss[loss=0.1285, simple_loss=0.2082, pruned_loss=0.0244, over 4967.00 frames.], tot_loss[loss=0.1465, simple_loss=0.2275, pruned_loss=0.03273, over 985970.18 frames.], batch size: 30, aishell_tot_loss[loss=0.1499, simple_loss=0.236, pruned_loss=0.03186, over 984803.41 frames.], datatang_tot_loss[loss=0.1449, simple_loss=0.221, pruned_loss=0.03438, over 983908.76 frames.], batch size: 30, lr: 3.88e-04
+2022-06-19 01:34:33,243 INFO [train.py:874] (2/4) Epoch 21, batch 2700, aishell_loss[loss=0.1242, simple_loss=0.2131, pruned_loss=0.01764, over 4977.00 frames.], tot_loss[loss=0.146, simple_loss=0.2271, pruned_loss=0.03245, over 986123.74 frames.], batch size: 30, aishell_tot_loss[loss=0.1492, simple_loss=0.2356, pruned_loss=0.03144, over 985030.26 frames.], datatang_tot_loss[loss=0.1449, simple_loss=0.2211, pruned_loss=0.03434, over 984230.22 frames.], batch size: 30, lr: 3.88e-04
+2022-06-19 01:35:05,888 INFO [train.py:874] (2/4) Epoch 21, batch 2750, aishell_loss[loss=0.1366, simple_loss=0.2271, pruned_loss=0.0231, over 4883.00 frames.], tot_loss[loss=0.1466, simple_loss=0.2272, pruned_loss=0.03303, over 985822.58 frames.], batch size: 47, aishell_tot_loss[loss=0.1494, simple_loss=0.2355, pruned_loss=0.03166, over 984841.77 frames.], datatang_tot_loss[loss=0.1452, simple_loss=0.2212, pruned_loss=0.03461, over 984474.56 frames.], batch size: 47, lr: 3.88e-04
+2022-06-19 01:35:36,806 INFO [train.py:874] (2/4) Epoch 21, batch 2800, aishell_loss[loss=0.1431, simple_loss=0.2193, pruned_loss=0.03343, over 4966.00 frames.], tot_loss[loss=0.1473, simple_loss=0.2277, pruned_loss=0.0334, over 986229.04 frames.], batch size: 27, aishell_tot_loss[loss=0.1497, simple_loss=0.2357, pruned_loss=0.03186, over 985436.03 frames.], datatang_tot_loss[loss=0.1453, simple_loss=0.2212, pruned_loss=0.03474, over 984581.96 frames.], batch size: 27, lr: 3.88e-04
+2022-06-19 01:36:09,374 INFO [train.py:874] (2/4) Epoch 21, batch 2850, datatang_loss[loss=0.1552, simple_loss=0.238, pruned_loss=0.03618, over 4918.00 frames.], tot_loss[loss=0.1479, simple_loss=0.229, pruned_loss=0.03343, over 986043.49 frames.], batch size: 98, aishell_tot_loss[loss=0.15, simple_loss=0.2361, pruned_loss=0.03197, over 985473.53 frames.], datatang_tot_loss[loss=0.1455, simple_loss=0.2216, pruned_loss=0.03477, over 984616.02 frames.], batch size: 98, lr: 3.88e-04
+2022-06-19 01:36:39,714 INFO [train.py:874] (2/4) Epoch 21, batch 2900, datatang_loss[loss=0.1208, simple_loss=0.1972, pruned_loss=0.02214, over 4838.00 frames.], tot_loss[loss=0.1474, simple_loss=0.2279, pruned_loss=0.03338, over 985969.86 frames.], batch size: 30, aishell_tot_loss[loss=0.15, simple_loss=0.2359, pruned_loss=0.03201, over 985344.60 frames.], datatang_tot_loss[loss=0.1451, simple_loss=0.2209, pruned_loss=0.03468, over 984932.33 frames.], batch size: 30, lr: 3.88e-04
+2022-06-19 01:37:10,801 INFO [train.py:874] (2/4) Epoch 21, batch 2950, datatang_loss[loss=0.1515, simple_loss=0.2207, pruned_loss=0.04114, over 4897.00 frames.], tot_loss[loss=0.147, simple_loss=0.2279, pruned_loss=0.03307, over 985854.83 frames.], batch size: 52, aishell_tot_loss[loss=0.1498, simple_loss=0.2358, pruned_loss=0.03189, over 985329.55 frames.], datatang_tot_loss[loss=0.1448, simple_loss=0.2205, pruned_loss=0.03451, over 984998.96 frames.], batch size: 52, lr: 3.87e-04
+2022-06-19 01:37:43,718 INFO [train.py:874] (2/4) Epoch 21, batch 3000, aishell_loss[loss=0.168, simple_loss=0.2476, pruned_loss=0.04421, over 4895.00 frames.], tot_loss[loss=0.1467, simple_loss=0.2274, pruned_loss=0.03296, over 985493.34 frames.], batch size: 34, aishell_tot_loss[loss=0.1492, simple_loss=0.2351, pruned_loss=0.03161, over 985311.69 frames.], datatang_tot_loss[loss=0.1449, simple_loss=0.2205, pruned_loss=0.03465, over 984810.13 frames.], batch size: 34, lr: 3.87e-04
+2022-06-19 01:37:43,719 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 01:38:00,373 INFO [train.py:914] (2/4) Epoch 21, validation: loss=0.1648, simple_loss=0.2487, pruned_loss=0.04045, over 1622729.00 frames.
+2022-06-19 01:38:34,160 INFO [train.py:874] (2/4) Epoch 21, batch 3050, aishell_loss[loss=0.1315, simple_loss=0.2121, pruned_loss=0.02549, over 4989.00 frames.], tot_loss[loss=0.1471, simple_loss=0.2279, pruned_loss=0.03309, over 985664.76 frames.], batch size: 27, aishell_tot_loss[loss=0.1494, simple_loss=0.2354, pruned_loss=0.03174, over 985498.88 frames.], datatang_tot_loss[loss=0.1449, simple_loss=0.2205, pruned_loss=0.03469, over 984893.16 frames.], batch size: 27, lr: 3.87e-04
+2022-06-19 01:39:07,204 INFO [train.py:874] (2/4) Epoch 21, batch 3100, aishell_loss[loss=0.1564, simple_loss=0.239, pruned_loss=0.03684, over 4935.00 frames.], tot_loss[loss=0.1472, simple_loss=0.2283, pruned_loss=0.03303, over 985740.74 frames.], batch size: 49, aishell_tot_loss[loss=0.1496, simple_loss=0.2357, pruned_loss=0.03173, over 985607.02 frames.], datatang_tot_loss[loss=0.1449, simple_loss=0.2206, pruned_loss=0.0346, over 984968.67 frames.], batch size: 49, lr: 3.87e-04
+2022-06-19 01:39:37,361 INFO [train.py:874] (2/4) Epoch 21, batch 3150, aishell_loss[loss=0.1646, simple_loss=0.2517, pruned_loss=0.03874, over 4978.00 frames.], tot_loss[loss=0.1478, simple_loss=0.2286, pruned_loss=0.03352, over 986043.62 frames.], batch size: 44, aishell_tot_loss[loss=0.1502, simple_loss=0.236, pruned_loss=0.03215, over 985684.57 frames.], datatang_tot_loss[loss=0.145, simple_loss=0.2208, pruned_loss=0.03462, over 985325.08 frames.], batch size: 44, lr: 3.87e-04
+2022-06-19 01:40:10,716 INFO [train.py:874] (2/4) Epoch 21, batch 3200, aishell_loss[loss=0.1499, simple_loss=0.2309, pruned_loss=0.03449, over 4940.00 frames.], tot_loss[loss=0.1487, simple_loss=0.2298, pruned_loss=0.03378, over 985746.22 frames.], batch size: 31, aishell_tot_loss[loss=0.1505, simple_loss=0.2365, pruned_loss=0.03228, over 985319.47 frames.], datatang_tot_loss[loss=0.1455, simple_loss=0.2212, pruned_loss=0.0349, over 985500.93 frames.], batch size: 31, lr: 3.87e-04
+2022-06-19 01:40:43,324 INFO [train.py:874] (2/4) Epoch 21, batch 3250, datatang_loss[loss=0.1482, simple_loss=0.2142, pruned_loss=0.04103, over 4937.00 frames.], tot_loss[loss=0.1474, simple_loss=0.228, pruned_loss=0.03344, over 985607.62 frames.], batch size: 50, aishell_tot_loss[loss=0.1495, simple_loss=0.2353, pruned_loss=0.0318, over 985153.18 frames.], datatang_tot_loss[loss=0.1455, simple_loss=0.221, pruned_loss=0.03494, over 985588.14 frames.], batch size: 50, lr: 3.87e-04
+2022-06-19 01:41:14,432 INFO [train.py:874] (2/4) Epoch 21, batch 3300, datatang_loss[loss=0.1453, simple_loss=0.2163, pruned_loss=0.03716, over 4973.00 frames.], tot_loss[loss=0.1475, simple_loss=0.2279, pruned_loss=0.03354, over 985859.31 frames.], batch size: 60, aishell_tot_loss[loss=0.1494, simple_loss=0.2351, pruned_loss=0.03185, over 985094.34 frames.], datatang_tot_loss[loss=0.1455, simple_loss=0.221, pruned_loss=0.03503, over 985986.35 frames.], batch size: 60, lr: 3.87e-04
+2022-06-19 01:41:47,795 INFO [train.py:874] (2/4) Epoch 21, batch 3350, datatang_loss[loss=0.1202, simple_loss=0.2009, pruned_loss=0.01974, over 4918.00 frames.], tot_loss[loss=0.1468, simple_loss=0.2277, pruned_loss=0.03292, over 985584.11 frames.], batch size: 77, aishell_tot_loss[loss=0.1492, simple_loss=0.2351, pruned_loss=0.03166, over 984787.39 frames.], datatang_tot_loss[loss=0.1449, simple_loss=0.2205, pruned_loss=0.03466, over 986084.75 frames.], batch size: 77, lr: 3.87e-04
+2022-06-19 01:42:19,505 INFO [train.py:874] (2/4) Epoch 21, batch 3400, datatang_loss[loss=0.1379, simple_loss=0.2143, pruned_loss=0.03073, over 4920.00 frames.], tot_loss[loss=0.1474, simple_loss=0.2284, pruned_loss=0.03319, over 985956.51 frames.], batch size: 81, aishell_tot_loss[loss=0.149, simple_loss=0.2349, pruned_loss=0.03155, over 985042.30 frames.], datatang_tot_loss[loss=0.1457, simple_loss=0.2212, pruned_loss=0.03511, over 986294.00 frames.], batch size: 81, lr: 3.86e-04
+2022-06-19 01:42:51,684 INFO [train.py:874] (2/4) Epoch 21, batch 3450, datatang_loss[loss=0.1546, simple_loss=0.2322, pruned_loss=0.0385, over 4934.00 frames.], tot_loss[loss=0.1467, simple_loss=0.2278, pruned_loss=0.03286, over 985998.18 frames.], batch size: 88, aishell_tot_loss[loss=0.1489, simple_loss=0.2348, pruned_loss=0.03153, over 985024.24 frames.], datatang_tot_loss[loss=0.1451, simple_loss=0.2208, pruned_loss=0.03467, over 986400.20 frames.], batch size: 88, lr: 3.86e-04
+2022-06-19 01:43:25,952 INFO [train.py:874] (2/4) Epoch 21, batch 3500, datatang_loss[loss=0.1395, simple_loss=0.2188, pruned_loss=0.0301, over 4929.00 frames.], tot_loss[loss=0.1465, simple_loss=0.2273, pruned_loss=0.03289, over 985978.28 frames.], batch size: 77, aishell_tot_loss[loss=0.1487, simple_loss=0.2345, pruned_loss=0.03144, over 984952.14 frames.], datatang_tot_loss[loss=0.1451, simple_loss=0.2207, pruned_loss=0.03471, over 986500.97 frames.], batch size: 77, lr: 3.86e-04
+2022-06-19 01:43:56,502 INFO [train.py:874] (2/4) Epoch 21, batch 3550, datatang_loss[loss=0.1433, simple_loss=0.2182, pruned_loss=0.03423, over 4951.00 frames.], tot_loss[loss=0.1459, simple_loss=0.2266, pruned_loss=0.03257, over 986016.35 frames.], batch size: 62, aishell_tot_loss[loss=0.1481, simple_loss=0.2339, pruned_loss=0.0312, over 985009.60 frames.], datatang_tot_loss[loss=0.1448, simple_loss=0.2206, pruned_loss=0.03455, over 986545.86 frames.], batch size: 62, lr: 3.86e-04
+2022-06-19 01:44:29,722 INFO [train.py:874] (2/4) Epoch 21, batch 3600, aishell_loss[loss=0.128, simple_loss=0.2009, pruned_loss=0.02749, over 4959.00 frames.], tot_loss[loss=0.1467, simple_loss=0.2271, pruned_loss=0.03314, over 985713.33 frames.], batch size: 25, aishell_tot_loss[loss=0.1484, simple_loss=0.2338, pruned_loss=0.0315, over 984790.79 frames.], datatang_tot_loss[loss=0.1453, simple_loss=0.2211, pruned_loss=0.03475, over 986488.75 frames.], batch size: 25, lr: 3.86e-04
+2022-06-19 01:45:03,501 INFO [train.py:874] (2/4) Epoch 21, batch 3650, datatang_loss[loss=0.1429, simple_loss=0.2046, pruned_loss=0.04058, over 4958.00 frames.], tot_loss[loss=0.1458, simple_loss=0.2262, pruned_loss=0.03272, over 985484.37 frames.], batch size: 37, aishell_tot_loss[loss=0.1484, simple_loss=0.2337, pruned_loss=0.0316, over 984685.39 frames.], datatang_tot_loss[loss=0.1444, simple_loss=0.2204, pruned_loss=0.03418, over 986358.40 frames.], batch size: 37, lr: 3.86e-04
+2022-06-19 01:45:34,393 INFO [train.py:874] (2/4) Epoch 21, batch 3700, datatang_loss[loss=0.1237, simple_loss=0.2093, pruned_loss=0.01903, over 4924.00 frames.], tot_loss[loss=0.1469, simple_loss=0.2276, pruned_loss=0.0331, over 985688.40 frames.], batch size: 71, aishell_tot_loss[loss=0.1489, simple_loss=0.2342, pruned_loss=0.0318, over 984756.06 frames.], datatang_tot_loss[loss=0.145, simple_loss=0.2212, pruned_loss=0.03433, over 986488.87 frames.], batch size: 71, lr: 3.86e-04
+2022-06-19 01:46:07,230 INFO [train.py:874] (2/4) Epoch 21, batch 3750, aishell_loss[loss=0.1331, simple_loss=0.2254, pruned_loss=0.02041, over 4953.00 frames.], tot_loss[loss=0.146, simple_loss=0.227, pruned_loss=0.03246, over 986099.56 frames.], batch size: 45, aishell_tot_loss[loss=0.1487, simple_loss=0.2343, pruned_loss=0.03161, over 985325.71 frames.], datatang_tot_loss[loss=0.1441, simple_loss=0.2205, pruned_loss=0.03384, over 986374.58 frames.], batch size: 45, lr: 3.86e-04
+2022-06-19 01:46:36,114 INFO [train.py:874] (2/4) Epoch 21, batch 3800, aishell_loss[loss=0.165, simple_loss=0.2497, pruned_loss=0.04016, over 4896.00 frames.], tot_loss[loss=0.1467, simple_loss=0.2282, pruned_loss=0.03258, over 985895.35 frames.], batch size: 60, aishell_tot_loss[loss=0.149, simple_loss=0.2347, pruned_loss=0.03163, over 985137.65 frames.], datatang_tot_loss[loss=0.1443, simple_loss=0.2206, pruned_loss=0.03397, over 986453.70 frames.], batch size: 60, lr: 3.86e-04
+2022-06-19 01:47:09,279 INFO [train.py:874] (2/4) Epoch 21, batch 3850, datatang_loss[loss=0.1392, simple_loss=0.22, pruned_loss=0.02927, over 4957.00 frames.], tot_loss[loss=0.1468, simple_loss=0.228, pruned_loss=0.03277, over 985489.30 frames.], batch size: 60, aishell_tot_loss[loss=0.1489, simple_loss=0.2345, pruned_loss=0.03164, over 984922.46 frames.], datatang_tot_loss[loss=0.1446, simple_loss=0.2211, pruned_loss=0.03406, over 986233.26 frames.], batch size: 60, lr: 3.85e-04
+2022-06-19 01:47:38,018 INFO [train.py:874] (2/4) Epoch 21, batch 3900, datatang_loss[loss=0.1309, simple_loss=0.2112, pruned_loss=0.02529, over 4916.00 frames.], tot_loss[loss=0.1471, simple_loss=0.2286, pruned_loss=0.03279, over 985134.56 frames.], batch size: 75, aishell_tot_loss[loss=0.1489, simple_loss=0.2346, pruned_loss=0.03158, over 984750.39 frames.], datatang_tot_loss[loss=0.1448, simple_loss=0.2212, pruned_loss=0.03422, over 986053.05 frames.], batch size: 75, lr: 3.85e-04
+2022-06-19 01:48:11,144 INFO [train.py:874] (2/4) Epoch 21, batch 3950, datatang_loss[loss=0.1536, simple_loss=0.2305, pruned_loss=0.03836, over 4953.00 frames.], tot_loss[loss=0.1472, simple_loss=0.2288, pruned_loss=0.03278, over 984982.90 frames.], batch size: 37, aishell_tot_loss[loss=0.1488, simple_loss=0.2346, pruned_loss=0.03149, over 984741.03 frames.], datatang_tot_loss[loss=0.1451, simple_loss=0.2218, pruned_loss=0.03423, over 985821.40 frames.], batch size: 37, lr: 3.85e-04
+2022-06-19 01:48:40,027 INFO [train.py:874] (2/4) Epoch 21, batch 4000, aishell_loss[loss=0.1507, simple_loss=0.2331, pruned_loss=0.03415, over 4907.00 frames.], tot_loss[loss=0.1466, simple_loss=0.2284, pruned_loss=0.03242, over 984842.80 frames.], batch size: 41, aishell_tot_loss[loss=0.149, simple_loss=0.2348, pruned_loss=0.03164, over 984451.38 frames.], datatang_tot_loss[loss=0.1443, simple_loss=0.2212, pruned_loss=0.0337, over 985914.51 frames.], batch size: 41, lr: 3.85e-04
+2022-06-19 01:48:40,028 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 01:48:57,401 INFO [train.py:914] (2/4) Epoch 21, validation: loss=0.1643, simple_loss=0.2479, pruned_loss=0.04039, over 1622729.00 frames.
+2022-06-19 01:49:25,769 INFO [train.py:874] (2/4) Epoch 21, batch 4050, datatang_loss[loss=0.1278, simple_loss=0.2116, pruned_loss=0.02198, over 4931.00 frames.], tot_loss[loss=0.1469, simple_loss=0.2289, pruned_loss=0.03249, over 984683.76 frames.], batch size: 79, aishell_tot_loss[loss=0.1488, simple_loss=0.2346, pruned_loss=0.03151, over 984308.11 frames.], datatang_tot_loss[loss=0.1448, simple_loss=0.2218, pruned_loss=0.03387, over 985835.05 frames.], batch size: 79, lr: 3.85e-04
+2022-06-19 01:50:39,881 INFO [train.py:874] (2/4) Epoch 22, batch 50, datatang_loss[loss=0.1336, simple_loss=0.2098, pruned_loss=0.02869, over 4963.00 frames.], tot_loss[loss=0.1421, simple_loss=0.2231, pruned_loss=0.03053, over 218206.50 frames.], batch size: 60, aishell_tot_loss[loss=0.1445, simple_loss=0.2305, pruned_loss=0.02922, over 120028.39 frames.], datatang_tot_loss[loss=0.1397, simple_loss=0.2154, pruned_loss=0.03202, over 111801.86 frames.], batch size: 60, lr: 3.76e-04
+2022-06-19 01:51:13,461 INFO [train.py:874] (2/4) Epoch 22, batch 100, datatang_loss[loss=0.146, simple_loss=0.2168, pruned_loss=0.03755, over 4893.00 frames.], tot_loss[loss=0.143, simple_loss=0.2238, pruned_loss=0.0311, over 388325.49 frames.], batch size: 52, aishell_tot_loss[loss=0.1453, simple_loss=0.2304, pruned_loss=0.03006, over 225706.04 frames.], datatang_tot_loss[loss=0.1405, simple_loss=0.2168, pruned_loss=0.03212, over 210940.98 frames.], batch size: 52, lr: 3.76e-04
+2022-06-19 01:51:45,193 INFO [train.py:874] (2/4) Epoch 22, batch 150, aishell_loss[loss=0.1475, simple_loss=0.2313, pruned_loss=0.03185, over 4947.00 frames.], tot_loss[loss=0.1452, simple_loss=0.225, pruned_loss=0.03268, over 520706.83 frames.], batch size: 32, aishell_tot_loss[loss=0.1487, simple_loss=0.2328, pruned_loss=0.03232, over 311956.19 frames.], datatang_tot_loss[loss=0.1411, simple_loss=0.2169, pruned_loss=0.03264, over 305452.84 frames.], batch size: 32, lr: 3.76e-04
+2022-06-19 01:52:17,267 INFO [train.py:874] (2/4) Epoch 22, batch 200, aishell_loss[loss=0.1687, simple_loss=0.247, pruned_loss=0.04518, over 4967.00 frames.], tot_loss[loss=0.1441, simple_loss=0.2245, pruned_loss=0.03184, over 623314.72 frames.], batch size: 44, aishell_tot_loss[loss=0.1473, simple_loss=0.232, pruned_loss=0.03124, over 390833.94 frames.], datatang_tot_loss[loss=0.1407, simple_loss=0.2166, pruned_loss=0.03237, over 385538.86 frames.], batch size: 44, lr: 3.76e-04
+2022-06-19 01:52:50,983 INFO [train.py:874] (2/4) Epoch 22, batch 250, aishell_loss[loss=0.1522, simple_loss=0.2454, pruned_loss=0.02944, over 4963.00 frames.], tot_loss[loss=0.1452, simple_loss=0.2262, pruned_loss=0.03213, over 703802.82 frames.], batch size: 64, aishell_tot_loss[loss=0.1478, simple_loss=0.2329, pruned_loss=0.03134, over 476383.42 frames.], datatang_tot_loss[loss=0.1417, simple_loss=0.2177, pruned_loss=0.03284, over 440275.36 frames.], batch size: 64, lr: 3.76e-04
+2022-06-19 01:53:21,201 INFO [train.py:874] (2/4) Epoch 22, batch 300, datatang_loss[loss=0.1316, simple_loss=0.2, pruned_loss=0.03156, over 4909.00 frames.], tot_loss[loss=0.1453, simple_loss=0.2264, pruned_loss=0.03207, over 765981.95 frames.], batch size: 42, aishell_tot_loss[loss=0.1476, simple_loss=0.233, pruned_loss=0.03108, over 540750.83 frames.], datatang_tot_loss[loss=0.1421, simple_loss=0.218, pruned_loss=0.03308, over 499436.78 frames.], batch size: 42, lr: 3.76e-04
+2022-06-19 01:53:54,110 INFO [train.py:874] (2/4) Epoch 22, batch 350, datatang_loss[loss=0.1355, simple_loss=0.2032, pruned_loss=0.03388, over 4975.00 frames.], tot_loss[loss=0.1457, simple_loss=0.2267, pruned_loss=0.03239, over 814673.44 frames.], batch size: 45, aishell_tot_loss[loss=0.1482, simple_loss=0.2333, pruned_loss=0.03152, over 593163.80 frames.], datatang_tot_loss[loss=0.1424, simple_loss=0.2186, pruned_loss=0.03311, over 556678.08 frames.], batch size: 45, lr: 3.76e-04
+2022-06-19 01:54:25,639 INFO [train.py:874] (2/4) Epoch 22, batch 400, aishell_loss[loss=0.1482, simple_loss=0.2332, pruned_loss=0.03158, over 4955.00 frames.], tot_loss[loss=0.1462, simple_loss=0.2276, pruned_loss=0.03238, over 852742.06 frames.], batch size: 40, aishell_tot_loss[loss=0.1482, simple_loss=0.2337, pruned_loss=0.0313, over 635990.36 frames.], datatang_tot_loss[loss=0.1434, simple_loss=0.22, pruned_loss=0.03337, over 611078.81 frames.], batch size: 40, lr: 3.76e-04
+2022-06-19 01:54:58,625 INFO [train.py:874] (2/4) Epoch 22, batch 450, datatang_loss[loss=0.2377, simple_loss=0.3021, pruned_loss=0.08668, over 4952.00 frames.], tot_loss[loss=0.1469, simple_loss=0.2281, pruned_loss=0.0328, over 882473.32 frames.], batch size: 99, aishell_tot_loss[loss=0.1488, simple_loss=0.2345, pruned_loss=0.03152, over 669357.16 frames.], datatang_tot_loss[loss=0.144, simple_loss=0.2207, pruned_loss=0.03369, over 663646.52 frames.], batch size: 99, lr: 3.76e-04
+2022-06-19 01:55:29,437 INFO [train.py:874] (2/4) Epoch 22, batch 500, datatang_loss[loss=0.1215, simple_loss=0.2058, pruned_loss=0.01864, over 4936.00 frames.], tot_loss[loss=0.1457, simple_loss=0.2271, pruned_loss=0.0321, over 905477.06 frames.], batch size: 79, aishell_tot_loss[loss=0.1494, simple_loss=0.2352, pruned_loss=0.03176, over 706613.69 frames.], datatang_tot_loss[loss=0.1419, simple_loss=0.2188, pruned_loss=0.03254, over 701687.95 frames.], batch size: 79, lr: 3.75e-04
+2022-06-19 01:56:02,059 INFO [train.py:874] (2/4) Epoch 22, batch 550, datatang_loss[loss=0.1537, simple_loss=0.242, pruned_loss=0.03264, over 4930.00 frames.], tot_loss[loss=0.1455, simple_loss=0.2271, pruned_loss=0.03198, over 923298.87 frames.], batch size: 94, aishell_tot_loss[loss=0.1492, simple_loss=0.2352, pruned_loss=0.03159, over 736850.30 frames.], datatang_tot_loss[loss=0.1421, simple_loss=0.2191, pruned_loss=0.03255, over 737818.00 frames.], batch size: 94, lr: 3.75e-04
+2022-06-19 01:56:34,618 INFO [train.py:874] (2/4) Epoch 22, batch 600, aishell_loss[loss=0.1447, simple_loss=0.2309, pruned_loss=0.02924, over 4936.00 frames.], tot_loss[loss=0.1445, simple_loss=0.2259, pruned_loss=0.0316, over 937342.05 frames.], batch size: 32, aishell_tot_loss[loss=0.1486, simple_loss=0.2345, pruned_loss=0.03134, over 764079.30 frames.], datatang_tot_loss[loss=0.1415, simple_loss=0.2184, pruned_loss=0.03227, over 769283.04 frames.], batch size: 32, lr: 3.75e-04
+2022-06-19 01:57:06,339 INFO [train.py:874] (2/4) Epoch 22, batch 650, aishell_loss[loss=0.175, simple_loss=0.2609, pruned_loss=0.04451, over 4975.00 frames.], tot_loss[loss=0.1447, simple_loss=0.226, pruned_loss=0.03165, over 948089.02 frames.], batch size: 61, aishell_tot_loss[loss=0.1486, simple_loss=0.2345, pruned_loss=0.03137, over 788120.48 frames.], datatang_tot_loss[loss=0.1415, simple_loss=0.2186, pruned_loss=0.03223, over 796788.90 frames.], batch size: 61, lr: 3.75e-04
+2022-06-19 01:57:38,735 INFO [train.py:874] (2/4) Epoch 22, batch 700, aishell_loss[loss=0.141, simple_loss=0.241, pruned_loss=0.02045, over 4974.00 frames.], tot_loss[loss=0.1451, simple_loss=0.2267, pruned_loss=0.03177, over 956291.20 frames.], batch size: 44, aishell_tot_loss[loss=0.1487, simple_loss=0.2346, pruned_loss=0.03138, over 811207.61 frames.], datatang_tot_loss[loss=0.1419, simple_loss=0.2192, pruned_loss=0.03232, over 819069.49 frames.], batch size: 44, lr: 3.75e-04
+2022-06-19 01:58:11,423 INFO [train.py:874] (2/4) Epoch 22, batch 750, datatang_loss[loss=0.1814, simple_loss=0.2626, pruned_loss=0.05012, over 4947.00 frames.], tot_loss[loss=0.1461, simple_loss=0.2277, pruned_loss=0.03221, over 963156.56 frames.], batch size: 108, aishell_tot_loss[loss=0.1495, simple_loss=0.2352, pruned_loss=0.03187, over 833444.73 frames.], datatang_tot_loss[loss=0.1421, simple_loss=0.2196, pruned_loss=0.03235, over 837425.21 frames.], batch size: 108, lr: 3.75e-04
+2022-06-19 01:58:43,770 INFO [train.py:874] (2/4) Epoch 22, batch 800, datatang_loss[loss=0.1529, simple_loss=0.225, pruned_loss=0.04036, over 4918.00 frames.], tot_loss[loss=0.1459, simple_loss=0.2274, pruned_loss=0.03222, over 967687.31 frames.], batch size: 64, aishell_tot_loss[loss=0.1493, simple_loss=0.2349, pruned_loss=0.03188, over 849046.26 frames.], datatang_tot_loss[loss=0.1423, simple_loss=0.2199, pruned_loss=0.03237, over 856667.41 frames.], batch size: 64, lr: 3.75e-04
+2022-06-19 01:59:14,945 INFO [train.py:874] (2/4) Epoch 22, batch 850, aishell_loss[loss=0.1327, simple_loss=0.2199, pruned_loss=0.02277, over 4910.00 frames.], tot_loss[loss=0.145, simple_loss=0.2267, pruned_loss=0.03163, over 971506.58 frames.], batch size: 41, aishell_tot_loss[loss=0.1487, simple_loss=0.2344, pruned_loss=0.03153, over 865650.64 frames.], datatang_tot_loss[loss=0.1418, simple_loss=0.2195, pruned_loss=0.03206, over 871213.81 frames.], batch size: 41, lr: 3.75e-04
+2022-06-19 01:59:46,505 INFO [train.py:874] (2/4) Epoch 22, batch 900, aishell_loss[loss=0.1524, simple_loss=0.2418, pruned_loss=0.03146, over 4966.00 frames.], tot_loss[loss=0.1448, simple_loss=0.2262, pruned_loss=0.03171, over 974894.19 frames.], batch size: 51, aishell_tot_loss[loss=0.1483, simple_loss=0.2339, pruned_loss=0.03136, over 878835.97 frames.], datatang_tot_loss[loss=0.142, simple_loss=0.2195, pruned_loss=0.03225, over 885842.77 frames.], batch size: 51, lr: 3.75e-04
+2022-06-19 02:00:23,992 INFO [train.py:874] (2/4) Epoch 22, batch 950, aishell_loss[loss=0.1709, simple_loss=0.2605, pruned_loss=0.04066, over 4862.00 frames.], tot_loss[loss=0.1447, simple_loss=0.2261, pruned_loss=0.03169, over 976596.25 frames.], batch size: 37, aishell_tot_loss[loss=0.1484, simple_loss=0.2339, pruned_loss=0.03142, over 890723.68 frames.], datatang_tot_loss[loss=0.1417, simple_loss=0.2191, pruned_loss=0.03214, over 897591.00 frames.], batch size: 37, lr: 3.74e-04
+2022-06-19 02:00:55,121 INFO [train.py:874] (2/4) Epoch 22, batch 1000, aishell_loss[loss=0.1532, simple_loss=0.2468, pruned_loss=0.02978, over 4982.00 frames.], tot_loss[loss=0.1461, simple_loss=0.2276, pruned_loss=0.03227, over 978867.66 frames.], batch size: 48, aishell_tot_loss[loss=0.1486, simple_loss=0.2344, pruned_loss=0.03145, over 903380.96 frames.], datatang_tot_loss[loss=0.1428, simple_loss=0.2201, pruned_loss=0.03275, over 906830.54 frames.], batch size: 48, lr: 3.74e-04
+2022-06-19 02:00:55,122 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 02:01:12,460 INFO [train.py:914] (2/4) Epoch 22, validation: loss=0.1645, simple_loss=0.2481, pruned_loss=0.04044, over 1622729.00 frames.
+2022-06-19 02:01:46,818 INFO [train.py:874] (2/4) Epoch 22, batch 1050, datatang_loss[loss=0.1273, simple_loss=0.2134, pruned_loss=0.02062, over 4926.00 frames.], tot_loss[loss=0.147, simple_loss=0.2278, pruned_loss=0.03311, over 980763.00 frames.], batch size: 71, aishell_tot_loss[loss=0.1487, simple_loss=0.2344, pruned_loss=0.03152, over 907800.05 frames.], datatang_tot_loss[loss=0.1442, simple_loss=0.2213, pruned_loss=0.03355, over 921224.46 frames.], batch size: 71, lr: 3.74e-04
+2022-06-19 02:02:17,961 INFO [train.py:874] (2/4) Epoch 22, batch 1100, datatang_loss[loss=0.1825, simple_loss=0.2609, pruned_loss=0.05204, over 4917.00 frames.], tot_loss[loss=0.1466, simple_loss=0.2278, pruned_loss=0.0327, over 981869.69 frames.], batch size: 98, aishell_tot_loss[loss=0.1484, simple_loss=0.2342, pruned_loss=0.03125, over 917942.54 frames.], datatang_tot_loss[loss=0.1443, simple_loss=0.2214, pruned_loss=0.03353, over 927976.07 frames.], batch size: 98, lr: 3.74e-04
+2022-06-19 02:02:49,824 INFO [train.py:874] (2/4) Epoch 22, batch 1150, datatang_loss[loss=0.1327, simple_loss=0.2032, pruned_loss=0.03106, over 4976.00 frames.], tot_loss[loss=0.1456, simple_loss=0.2268, pruned_loss=0.03223, over 982447.37 frames.], batch size: 34, aishell_tot_loss[loss=0.1476, simple_loss=0.2335, pruned_loss=0.03085, over 925210.50 frames.], datatang_tot_loss[loss=0.144, simple_loss=0.2211, pruned_loss=0.03346, over 935143.30 frames.], batch size: 34, lr: 3.74e-04
+2022-06-19 02:03:20,660 INFO [train.py:874] (2/4) Epoch 22, batch 1200, aishell_loss[loss=0.1316, simple_loss=0.2263, pruned_loss=0.01847, over 4863.00 frames.], tot_loss[loss=0.1465, simple_loss=0.2278, pruned_loss=0.03256, over 983117.99 frames.], batch size: 37, aishell_tot_loss[loss=0.1482, simple_loss=0.2341, pruned_loss=0.03116, over 933391.10 frames.], datatang_tot_loss[loss=0.1442, simple_loss=0.2212, pruned_loss=0.03358, over 940144.32 frames.], batch size: 37, lr: 3.74e-04
+2022-06-19 02:03:54,478 INFO [train.py:874] (2/4) Epoch 22, batch 1250, datatang_loss[loss=0.1228, simple_loss=0.2, pruned_loss=0.02279, over 4856.00 frames.], tot_loss[loss=0.1474, simple_loss=0.229, pruned_loss=0.03291, over 983367.41 frames.], batch size: 30, aishell_tot_loss[loss=0.1485, simple_loss=0.2345, pruned_loss=0.03126, over 941275.32 frames.], datatang_tot_loss[loss=0.1449, simple_loss=0.2216, pruned_loss=0.03407, over 943646.10 frames.], batch size: 30, lr: 3.74e-04
+2022-06-19 02:04:26,512 INFO [train.py:874] (2/4) Epoch 22, batch 1300, datatang_loss[loss=0.1358, simple_loss=0.207, pruned_loss=0.03233, over 4905.00 frames.], tot_loss[loss=0.147, simple_loss=0.2289, pruned_loss=0.03261, over 983901.45 frames.], batch size: 64, aishell_tot_loss[loss=0.1481, simple_loss=0.2343, pruned_loss=0.03094, over 946913.52 frames.], datatang_tot_loss[loss=0.1451, simple_loss=0.2218, pruned_loss=0.03418, over 948230.16 frames.], batch size: 64, lr: 3.74e-04
+2022-06-19 02:04:58,061 INFO [train.py:874] (2/4) Epoch 22, batch 1350, datatang_loss[loss=0.1214, simple_loss=0.2012, pruned_loss=0.02086, over 4918.00 frames.], tot_loss[loss=0.1476, simple_loss=0.2291, pruned_loss=0.03299, over 984284.74 frames.], batch size: 42, aishell_tot_loss[loss=0.1486, simple_loss=0.2346, pruned_loss=0.03128, over 951687.56 frames.], datatang_tot_loss[loss=0.1452, simple_loss=0.2219, pruned_loss=0.03429, over 952422.49 frames.], batch size: 42, lr: 3.74e-04
+2022-06-19 02:05:30,695 INFO [train.py:874] (2/4) Epoch 22, batch 1400, datatang_loss[loss=0.1468, simple_loss=0.2249, pruned_loss=0.03431, over 4889.00 frames.], tot_loss[loss=0.1479, simple_loss=0.23, pruned_loss=0.03291, over 984597.59 frames.], batch size: 47, aishell_tot_loss[loss=0.1491, simple_loss=0.2354, pruned_loss=0.03136, over 955772.80 frames.], datatang_tot_loss[loss=0.1453, simple_loss=0.2222, pruned_loss=0.0342, over 956272.30 frames.], batch size: 47, lr: 3.73e-04
+2022-06-19 02:06:03,235 INFO [train.py:874] (2/4) Epoch 22, batch 1450, aishell_loss[loss=0.1401, simple_loss=0.2348, pruned_loss=0.02274, over 4890.00 frames.], tot_loss[loss=0.1487, simple_loss=0.231, pruned_loss=0.03319, over 984747.65 frames.], batch size: 42, aishell_tot_loss[loss=0.1493, simple_loss=0.2357, pruned_loss=0.03146, over 959957.85 frames.], datatang_tot_loss[loss=0.146, simple_loss=0.2229, pruned_loss=0.03451, over 958965.52 frames.], batch size: 42, lr: 3.73e-04
+2022-06-19 02:06:34,662 INFO [train.py:874] (2/4) Epoch 22, batch 1500, aishell_loss[loss=0.1617, simple_loss=0.2442, pruned_loss=0.03961, over 4866.00 frames.], tot_loss[loss=0.1481, simple_loss=0.2302, pruned_loss=0.03294, over 985070.47 frames.], batch size: 37, aishell_tot_loss[loss=0.1493, simple_loss=0.2356, pruned_loss=0.03156, over 962939.99 frames.], datatang_tot_loss[loss=0.1455, simple_loss=0.2227, pruned_loss=0.03416, over 962300.02 frames.], batch size: 37, lr: 3.73e-04
+2022-06-19 02:07:06,453 INFO [train.py:874] (2/4) Epoch 22, batch 1550, datatang_loss[loss=0.1342, simple_loss=0.2019, pruned_loss=0.0333, over 4976.00 frames.], tot_loss[loss=0.1488, simple_loss=0.2307, pruned_loss=0.03346, over 985504.68 frames.], batch size: 60, aishell_tot_loss[loss=0.1505, simple_loss=0.2366, pruned_loss=0.03221, over 965775.62 frames.], datatang_tot_loss[loss=0.1453, simple_loss=0.2223, pruned_loss=0.03411, over 965217.52 frames.], batch size: 60, lr: 3.73e-04
+2022-06-19 02:07:39,204 INFO [train.py:874] (2/4) Epoch 22, batch 1600, datatang_loss[loss=0.1573, simple_loss=0.2252, pruned_loss=0.0447, over 4930.00 frames.], tot_loss[loss=0.1482, simple_loss=0.2298, pruned_loss=0.03329, over 984875.43 frames.], batch size: 57, aishell_tot_loss[loss=0.15, simple_loss=0.2359, pruned_loss=0.03206, over 967894.14 frames.], datatang_tot_loss[loss=0.1452, simple_loss=0.2221, pruned_loss=0.03419, over 967129.75 frames.], batch size: 57, lr: 3.73e-04
+2022-06-19 02:08:12,764 INFO [train.py:874] (2/4) Epoch 22, batch 1650, aishell_loss[loss=0.1513, simple_loss=0.2449, pruned_loss=0.02881, over 4950.00 frames.], tot_loss[loss=0.1473, simple_loss=0.2292, pruned_loss=0.03265, over 985355.99 frames.], batch size: 64, aishell_tot_loss[loss=0.1492, simple_loss=0.2354, pruned_loss=0.03151, over 969839.81 frames.], datatang_tot_loss[loss=0.1452, simple_loss=0.2223, pruned_loss=0.03406, over 969781.77 frames.], batch size: 64, lr: 3.73e-04
+2022-06-19 02:08:44,567 INFO [train.py:874] (2/4) Epoch 22, batch 1700, datatang_loss[loss=0.1482, simple_loss=0.2241, pruned_loss=0.0361, over 4926.00 frames.], tot_loss[loss=0.1475, simple_loss=0.2296, pruned_loss=0.03276, over 985613.63 frames.], batch size: 83, aishell_tot_loss[loss=0.1499, simple_loss=0.236, pruned_loss=0.03189, over 972124.73 frames.], datatang_tot_loss[loss=0.1448, simple_loss=0.2219, pruned_loss=0.03382, over 971418.70 frames.], batch size: 83, lr: 3.73e-04
+2022-06-19 02:09:16,515 INFO [train.py:874] (2/4) Epoch 22, batch 1750, datatang_loss[loss=0.1336, simple_loss=0.2014, pruned_loss=0.03286, over 4953.00 frames.], tot_loss[loss=0.1472, simple_loss=0.229, pruned_loss=0.03271, over 985838.60 frames.], batch size: 45, aishell_tot_loss[loss=0.1494, simple_loss=0.2356, pruned_loss=0.03159, over 973774.76 frames.], datatang_tot_loss[loss=0.145, simple_loss=0.2219, pruned_loss=0.03406, over 973269.73 frames.], batch size: 45, lr: 3.73e-04
+2022-06-19 02:09:50,783 INFO [train.py:874] (2/4) Epoch 22, batch 1800, aishell_loss[loss=0.144, simple_loss=0.2385, pruned_loss=0.02478, over 4952.00 frames.], tot_loss[loss=0.1469, simple_loss=0.2288, pruned_loss=0.03248, over 986127.40 frames.], batch size: 40, aishell_tot_loss[loss=0.1498, simple_loss=0.2362, pruned_loss=0.03174, over 974979.27 frames.], datatang_tot_loss[loss=0.1444, simple_loss=0.2215, pruned_loss=0.03359, over 975268.76 frames.], batch size: 40, lr: 3.73e-04
+2022-06-19 02:10:22,907 INFO [train.py:874] (2/4) Epoch 22, batch 1850, datatang_loss[loss=0.1594, simple_loss=0.2367, pruned_loss=0.04106, over 4913.00 frames.], tot_loss[loss=0.147, simple_loss=0.2289, pruned_loss=0.03259, over 985696.92 frames.], batch size: 98, aishell_tot_loss[loss=0.1495, simple_loss=0.2359, pruned_loss=0.0316, over 975939.84 frames.], datatang_tot_loss[loss=0.1447, simple_loss=0.2218, pruned_loss=0.03382, over 976440.56 frames.], batch size: 98, lr: 3.73e-04
+2022-06-19 02:10:56,163 INFO [train.py:874] (2/4) Epoch 22, batch 1900, datatang_loss[loss=0.137, simple_loss=0.2221, pruned_loss=0.02592, over 4952.00 frames.], tot_loss[loss=0.1479, simple_loss=0.2297, pruned_loss=0.03305, over 986051.42 frames.], batch size: 86, aishell_tot_loss[loss=0.1496, simple_loss=0.2359, pruned_loss=0.03164, over 977256.12 frames.], datatang_tot_loss[loss=0.1456, simple_loss=0.2227, pruned_loss=0.03423, over 977729.68 frames.], batch size: 86, lr: 3.72e-04
+2022-06-19 02:11:27,904 INFO [train.py:874] (2/4) Epoch 22, batch 1950, aishell_loss[loss=0.1492, simple_loss=0.2485, pruned_loss=0.02488, over 4962.00 frames.], tot_loss[loss=0.1471, simple_loss=0.2293, pruned_loss=0.03243, over 985784.98 frames.], batch size: 79, aishell_tot_loss[loss=0.1497, simple_loss=0.2363, pruned_loss=0.03157, over 978144.36 frames.], datatang_tot_loss[loss=0.1446, simple_loss=0.2219, pruned_loss=0.03367, over 978572.31 frames.], batch size: 79, lr: 3.72e-04
+2022-06-19 02:12:00,314 INFO [train.py:874] (2/4) Epoch 22, batch 2000, datatang_loss[loss=0.1433, simple_loss=0.2191, pruned_loss=0.0337, over 4920.00 frames.], tot_loss[loss=0.147, simple_loss=0.2289, pruned_loss=0.03252, over 985638.29 frames.], batch size: 73, aishell_tot_loss[loss=0.1494, simple_loss=0.2359, pruned_loss=0.03149, over 979085.73 frames.], datatang_tot_loss[loss=0.1448, simple_loss=0.222, pruned_loss=0.03381, over 979238.25 frames.], batch size: 73, lr: 3.72e-04
+2022-06-19 02:12:00,315 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 02:12:18,451 INFO [train.py:914] (2/4) Epoch 22, validation: loss=0.165, simple_loss=0.2489, pruned_loss=0.04058, over 1622729.00 frames.
+2022-06-19 02:12:50,743 INFO [train.py:874] (2/4) Epoch 22, batch 2050, aishell_loss[loss=0.1512, simple_loss=0.2434, pruned_loss=0.0295, over 4946.00 frames.], tot_loss[loss=0.1467, simple_loss=0.2282, pruned_loss=0.03255, over 985557.21 frames.], batch size: 58, aishell_tot_loss[loss=0.1498, simple_loss=0.236, pruned_loss=0.03175, over 979914.38 frames.], datatang_tot_loss[loss=0.1441, simple_loss=0.2212, pruned_loss=0.03353, over 979840.10 frames.], batch size: 58, lr: 3.72e-04
+2022-06-19 02:13:23,361 INFO [train.py:874] (2/4) Epoch 22, batch 2100, aishell_loss[loss=0.1394, simple_loss=0.2053, pruned_loss=0.03677, over 4957.00 frames.], tot_loss[loss=0.1464, simple_loss=0.2283, pruned_loss=0.03219, over 985514.64 frames.], batch size: 25, aishell_tot_loss[loss=0.1494, simple_loss=0.2358, pruned_loss=0.03153, over 980517.21 frames.], datatang_tot_loss[loss=0.144, simple_loss=0.2213, pruned_loss=0.03333, over 980528.85 frames.], batch size: 25, lr: 3.72e-04
+2022-06-19 02:13:54,813 INFO [train.py:874] (2/4) Epoch 22, batch 2150, datatang_loss[loss=0.1361, simple_loss=0.2013, pruned_loss=0.03542, over 4960.00 frames.], tot_loss[loss=0.147, simple_loss=0.2287, pruned_loss=0.03265, over 986015.71 frames.], batch size: 31, aishell_tot_loss[loss=0.1498, simple_loss=0.2362, pruned_loss=0.03169, over 981263.93 frames.], datatang_tot_loss[loss=0.1443, simple_loss=0.2214, pruned_loss=0.03358, over 981483.49 frames.], batch size: 31, lr: 3.72e-04
+2022-06-19 02:14:25,664 INFO [train.py:874] (2/4) Epoch 22, batch 2200, aishell_loss[loss=0.1137, simple_loss=0.1966, pruned_loss=0.01536, over 4971.00 frames.], tot_loss[loss=0.1468, simple_loss=0.2282, pruned_loss=0.03275, over 985971.85 frames.], batch size: 27, aishell_tot_loss[loss=0.1493, simple_loss=0.2355, pruned_loss=0.0315, over 981856.51 frames.], datatang_tot_loss[loss=0.1446, simple_loss=0.2215, pruned_loss=0.03386, over 981947.14 frames.], batch size: 27, lr: 3.72e-04
+2022-06-19 02:14:59,109 INFO [train.py:874] (2/4) Epoch 22, batch 2250, datatang_loss[loss=0.1352, simple_loss=0.2113, pruned_loss=0.0295, over 4902.00 frames.], tot_loss[loss=0.1466, simple_loss=0.2282, pruned_loss=0.03255, over 985826.54 frames.], batch size: 52, aishell_tot_loss[loss=0.1488, simple_loss=0.2352, pruned_loss=0.03121, over 982113.11 frames.], datatang_tot_loss[loss=0.1448, simple_loss=0.2217, pruned_loss=0.03398, over 982484.17 frames.], batch size: 52, lr: 3.72e-04
+2022-06-19 02:15:30,316 INFO [train.py:874] (2/4) Epoch 22, batch 2300, datatang_loss[loss=0.1309, simple_loss=0.2067, pruned_loss=0.02761, over 4871.00 frames.], tot_loss[loss=0.1459, simple_loss=0.2273, pruned_loss=0.03223, over 985560.73 frames.], batch size: 39, aishell_tot_loss[loss=0.1482, simple_loss=0.2346, pruned_loss=0.03092, over 982564.82 frames.], datatang_tot_loss[loss=0.1446, simple_loss=0.2213, pruned_loss=0.03391, over 982591.89 frames.], batch size: 39, lr: 3.72e-04
+2022-06-19 02:16:02,658 INFO [train.py:874] (2/4) Epoch 22, batch 2350, aishell_loss[loss=0.1544, simple_loss=0.2481, pruned_loss=0.03035, over 4940.00 frames.], tot_loss[loss=0.1458, simple_loss=0.2274, pruned_loss=0.03209, over 985612.34 frames.], batch size: 54, aishell_tot_loss[loss=0.1485, simple_loss=0.2348, pruned_loss=0.03106, over 983036.38 frames.], datatang_tot_loss[loss=0.144, simple_loss=0.2207, pruned_loss=0.03361, over 982866.17 frames.], batch size: 54, lr: 3.72e-04
+2022-06-19 02:16:33,520 INFO [train.py:874] (2/4) Epoch 22, batch 2400, datatang_loss[loss=0.1922, simple_loss=0.2664, pruned_loss=0.05898, over 4962.00 frames.], tot_loss[loss=0.1457, simple_loss=0.2271, pruned_loss=0.03215, over 985707.58 frames.], batch size: 99, aishell_tot_loss[loss=0.1479, simple_loss=0.2342, pruned_loss=0.03082, over 983388.01 frames.], datatang_tot_loss[loss=0.1443, simple_loss=0.2207, pruned_loss=0.0339, over 983241.58 frames.], batch size: 99, lr: 3.71e-04
+2022-06-19 02:17:04,620 INFO [train.py:874] (2/4) Epoch 22, batch 2450, datatang_loss[loss=0.1388, simple_loss=0.214, pruned_loss=0.03179, over 4933.00 frames.], tot_loss[loss=0.146, simple_loss=0.2272, pruned_loss=0.03238, over 985855.43 frames.], batch size: 79, aishell_tot_loss[loss=0.1479, simple_loss=0.2342, pruned_loss=0.03082, over 983621.91 frames.], datatang_tot_loss[loss=0.1445, simple_loss=0.2209, pruned_loss=0.03406, over 983727.32 frames.], batch size: 79, lr: 3.71e-04
+2022-06-19 02:17:37,740 INFO [train.py:874] (2/4) Epoch 22, batch 2500, aishell_loss[loss=0.144, simple_loss=0.2353, pruned_loss=0.02638, over 4860.00 frames.], tot_loss[loss=0.1457, simple_loss=0.227, pruned_loss=0.03221, over 985605.69 frames.], batch size: 38, aishell_tot_loss[loss=0.1472, simple_loss=0.2334, pruned_loss=0.0305, over 983592.19 frames.], datatang_tot_loss[loss=0.145, simple_loss=0.2215, pruned_loss=0.03422, over 984004.36 frames.], batch size: 38, lr: 3.71e-04
+2022-06-19 02:18:09,171 INFO [train.py:874] (2/4) Epoch 22, batch 2550, aishell_loss[loss=0.1376, simple_loss=0.2264, pruned_loss=0.02445, over 4947.00 frames.], tot_loss[loss=0.1452, simple_loss=0.227, pruned_loss=0.03174, over 985731.39 frames.], batch size: 40, aishell_tot_loss[loss=0.1469, simple_loss=0.2333, pruned_loss=0.03027, over 983934.83 frames.], datatang_tot_loss[loss=0.1446, simple_loss=0.2212, pruned_loss=0.03394, over 984219.19 frames.], batch size: 40, lr: 3.71e-04
+2022-06-19 02:18:41,339 INFO [train.py:874] (2/4) Epoch 22, batch 2600, aishell_loss[loss=0.141, simple_loss=0.2271, pruned_loss=0.02749, over 4884.00 frames.], tot_loss[loss=0.145, simple_loss=0.2269, pruned_loss=0.03149, over 985494.47 frames.], batch size: 42, aishell_tot_loss[loss=0.1464, simple_loss=0.233, pruned_loss=0.02995, over 983983.29 frames.], datatang_tot_loss[loss=0.1447, simple_loss=0.2215, pruned_loss=0.03391, over 984308.51 frames.], batch size: 42, lr: 3.71e-04
+2022-06-19 02:19:14,418 INFO [train.py:874] (2/4) Epoch 22, batch 2650, datatang_loss[loss=0.1417, simple_loss=0.2254, pruned_loss=0.02902, over 4933.00 frames.], tot_loss[loss=0.1447, simple_loss=0.2268, pruned_loss=0.03136, over 985437.86 frames.], batch size: 79, aishell_tot_loss[loss=0.1469, simple_loss=0.2332, pruned_loss=0.03027, over 984079.22 frames.], datatang_tot_loss[loss=0.1438, simple_loss=0.2206, pruned_loss=0.03346, over 984481.38 frames.], batch size: 79, lr: 3.71e-04
+2022-06-19 02:19:44,877 INFO [train.py:874] (2/4) Epoch 22, batch 2700, aishell_loss[loss=0.1595, simple_loss=0.2476, pruned_loss=0.03567, over 4951.00 frames.], tot_loss[loss=0.1456, simple_loss=0.2277, pruned_loss=0.0317, over 985457.42 frames.], batch size: 64, aishell_tot_loss[loss=0.1463, simple_loss=0.2327, pruned_loss=0.02998, over 984200.02 frames.], datatang_tot_loss[loss=0.1449, simple_loss=0.2218, pruned_loss=0.03405, over 984659.96 frames.], batch size: 64, lr: 3.71e-04
+2022-06-19 02:20:16,435 INFO [train.py:874] (2/4) Epoch 22, batch 2750, datatang_loss[loss=0.1202, simple_loss=0.2014, pruned_loss=0.01952, over 4936.00 frames.], tot_loss[loss=0.1449, simple_loss=0.2275, pruned_loss=0.03116, over 985237.28 frames.], batch size: 62, aishell_tot_loss[loss=0.1461, simple_loss=0.2326, pruned_loss=0.02979, over 984336.58 frames.], datatang_tot_loss[loss=0.1444, simple_loss=0.2215, pruned_loss=0.03369, over 984543.41 frames.], batch size: 62, lr: 3.71e-04
+2022-06-19 02:20:48,478 INFO [train.py:874] (2/4) Epoch 22, batch 2800, aishell_loss[loss=0.1577, simple_loss=0.2418, pruned_loss=0.03685, over 4974.00 frames.], tot_loss[loss=0.1451, simple_loss=0.2271, pruned_loss=0.03157, over 985237.70 frames.], batch size: 51, aishell_tot_loss[loss=0.146, simple_loss=0.2319, pruned_loss=0.03004, over 984468.81 frames.], datatang_tot_loss[loss=0.1446, simple_loss=0.2216, pruned_loss=0.03381, over 984591.06 frames.], batch size: 51, lr: 3.71e-04
+2022-06-19 02:21:19,579 INFO [train.py:874] (2/4) Epoch 22, batch 2850, aishell_loss[loss=0.1537, simple_loss=0.2456, pruned_loss=0.03085, over 4957.00 frames.], tot_loss[loss=0.1442, simple_loss=0.2266, pruned_loss=0.03091, over 985706.88 frames.], batch size: 61, aishell_tot_loss[loss=0.1461, simple_loss=0.232, pruned_loss=0.0301, over 984827.38 frames.], datatang_tot_loss[loss=0.1435, simple_loss=0.221, pruned_loss=0.03298, over 984898.57 frames.], batch size: 61, lr: 3.70e-04
+2022-06-19 02:21:52,033 INFO [train.py:874] (2/4) Epoch 22, batch 2900, aishell_loss[loss=0.1416, simple_loss=0.226, pruned_loss=0.02864, over 4954.00 frames.], tot_loss[loss=0.1448, simple_loss=0.2273, pruned_loss=0.03117, over 985684.42 frames.], batch size: 31, aishell_tot_loss[loss=0.1466, simple_loss=0.2327, pruned_loss=0.0302, over 984708.46 frames.], datatang_tot_loss[loss=0.1435, simple_loss=0.2211, pruned_loss=0.03295, over 985189.33 frames.], batch size: 31, lr: 3.70e-04
+2022-06-19 02:22:24,845 INFO [train.py:874] (2/4) Epoch 22, batch 2950, aishell_loss[loss=0.1559, simple_loss=0.2438, pruned_loss=0.03404, over 4931.00 frames.], tot_loss[loss=0.145, simple_loss=0.2272, pruned_loss=0.03138, over 985437.40 frames.], batch size: 33, aishell_tot_loss[loss=0.1467, simple_loss=0.2329, pruned_loss=0.03026, over 984603.80 frames.], datatang_tot_loss[loss=0.1435, simple_loss=0.221, pruned_loss=0.03299, over 985197.26 frames.], batch size: 33, lr: 3.70e-04
+2022-06-19 02:22:56,251 INFO [train.py:874] (2/4) Epoch 22, batch 3000, datatang_loss[loss=0.164, simple_loss=0.2359, pruned_loss=0.04604, over 4851.00 frames.], tot_loss[loss=0.1459, simple_loss=0.228, pruned_loss=0.03193, over 984897.23 frames.], batch size: 33, aishell_tot_loss[loss=0.1474, simple_loss=0.2335, pruned_loss=0.03063, over 984165.27 frames.], datatang_tot_loss[loss=0.1438, simple_loss=0.2211, pruned_loss=0.03319, over 985221.27 frames.], batch size: 33, lr: 3.70e-04
+2022-06-19 02:22:56,252 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 02:23:14,258 INFO [train.py:914] (2/4) Epoch 22, validation: loss=0.1638, simple_loss=0.2477, pruned_loss=0.03991, over 1622729.00 frames.
+2022-06-19 02:23:45,734 INFO [train.py:874] (2/4) Epoch 22, batch 3050, datatang_loss[loss=0.1402, simple_loss=0.2257, pruned_loss=0.02737, over 4933.00 frames.], tot_loss[loss=0.1458, simple_loss=0.2278, pruned_loss=0.03191, over 984967.46 frames.], batch size: 94, aishell_tot_loss[loss=0.1474, simple_loss=0.2335, pruned_loss=0.03064, over 984185.53 frames.], datatang_tot_loss[loss=0.1437, simple_loss=0.2211, pruned_loss=0.03314, over 985326.41 frames.], batch size: 94, lr: 3.70e-04
+2022-06-19 02:24:17,447 INFO [train.py:874] (2/4) Epoch 22, batch 3100, aishell_loss[loss=0.1243, simple_loss=0.2163, pruned_loss=0.01618, over 4962.00 frames.], tot_loss[loss=0.1451, simple_loss=0.2271, pruned_loss=0.03161, over 985180.93 frames.], batch size: 27, aishell_tot_loss[loss=0.1473, simple_loss=0.2334, pruned_loss=0.03055, over 984293.61 frames.], datatang_tot_loss[loss=0.1432, simple_loss=0.2207, pruned_loss=0.03288, over 985482.40 frames.], batch size: 27, lr: 3.70e-04
+2022-06-19 02:24:49,584 INFO [train.py:874] (2/4) Epoch 22, batch 3150, aishell_loss[loss=0.1254, simple_loss=0.2085, pruned_loss=0.02115, over 4958.00 frames.], tot_loss[loss=0.1453, simple_loss=0.2274, pruned_loss=0.03154, over 985400.92 frames.], batch size: 30, aishell_tot_loss[loss=0.1473, simple_loss=0.2334, pruned_loss=0.03057, over 984564.59 frames.], datatang_tot_loss[loss=0.1432, simple_loss=0.2207, pruned_loss=0.03285, over 985549.90 frames.], batch size: 30, lr: 3.70e-04
+2022-06-19 02:25:22,930 INFO [train.py:874] (2/4) Epoch 22, batch 3200, datatang_loss[loss=0.1314, simple_loss=0.2122, pruned_loss=0.0253, over 4927.00 frames.], tot_loss[loss=0.1457, simple_loss=0.2276, pruned_loss=0.03192, over 985619.85 frames.], batch size: 79, aishell_tot_loss[loss=0.1477, simple_loss=0.2338, pruned_loss=0.03081, over 984653.19 frames.], datatang_tot_loss[loss=0.1433, simple_loss=0.2208, pruned_loss=0.03294, over 985776.06 frames.], batch size: 79, lr: 3.70e-04
+2022-06-19 02:25:55,796 INFO [train.py:874] (2/4) Epoch 22, batch 3250, datatang_loss[loss=0.1307, simple_loss=0.2171, pruned_loss=0.02215, over 4956.00 frames.], tot_loss[loss=0.1455, simple_loss=0.2272, pruned_loss=0.03194, over 985908.21 frames.], batch size: 86, aishell_tot_loss[loss=0.1479, simple_loss=0.2341, pruned_loss=0.03089, over 984856.54 frames.], datatang_tot_loss[loss=0.143, simple_loss=0.2202, pruned_loss=0.03287, over 985969.75 frames.], batch size: 86, lr: 3.70e-04
+2022-06-19 02:26:28,521 INFO [train.py:874] (2/4) Epoch 22, batch 3300, datatang_loss[loss=0.1615, simple_loss=0.2339, pruned_loss=0.04455, over 4853.00 frames.], tot_loss[loss=0.1465, simple_loss=0.2282, pruned_loss=0.0324, over 985623.03 frames.], batch size: 30, aishell_tot_loss[loss=0.1486, simple_loss=0.2348, pruned_loss=0.03119, over 984904.60 frames.], datatang_tot_loss[loss=0.1433, simple_loss=0.2204, pruned_loss=0.03311, over 985747.16 frames.], batch size: 30, lr: 3.70e-04
+2022-06-19 02:27:02,232 INFO [train.py:874] (2/4) Epoch 22, batch 3350, aishell_loss[loss=0.1488, simple_loss=0.247, pruned_loss=0.02527, over 4956.00 frames.], tot_loss[loss=0.1472, simple_loss=0.2289, pruned_loss=0.03273, over 985745.67 frames.], batch size: 64, aishell_tot_loss[loss=0.1484, simple_loss=0.2346, pruned_loss=0.0311, over 985207.68 frames.], datatang_tot_loss[loss=0.1442, simple_loss=0.2211, pruned_loss=0.03364, over 985663.38 frames.], batch size: 64, lr: 3.69e-04
+2022-06-19 02:27:34,688 INFO [train.py:874] (2/4) Epoch 22, batch 3400, aishell_loss[loss=0.1345, simple_loss=0.2173, pruned_loss=0.02585, over 4873.00 frames.], tot_loss[loss=0.1466, simple_loss=0.2284, pruned_loss=0.03241, over 986012.02 frames.], batch size: 28, aishell_tot_loss[loss=0.1485, simple_loss=0.2349, pruned_loss=0.03108, over 985443.78 frames.], datatang_tot_loss[loss=0.1437, simple_loss=0.2206, pruned_loss=0.0334, over 985773.89 frames.], batch size: 28, lr: 3.69e-04
+2022-06-19 02:28:08,666 INFO [train.py:874] (2/4) Epoch 22, batch 3450, aishell_loss[loss=0.1295, simple_loss=0.2154, pruned_loss=0.02182, over 4824.00 frames.], tot_loss[loss=0.146, simple_loss=0.2275, pruned_loss=0.03223, over 985725.49 frames.], batch size: 29, aishell_tot_loss[loss=0.148, simple_loss=0.2345, pruned_loss=0.03072, over 985060.29 frames.], datatang_tot_loss[loss=0.1439, simple_loss=0.2208, pruned_loss=0.0335, over 985917.47 frames.], batch size: 29, lr: 3.69e-04
+2022-06-19 02:28:40,664 INFO [train.py:874] (2/4) Epoch 22, batch 3500, aishell_loss[loss=0.1566, simple_loss=0.255, pruned_loss=0.02913, over 4962.00 frames.], tot_loss[loss=0.1466, simple_loss=0.2279, pruned_loss=0.03262, over 985883.23 frames.], batch size: 64, aishell_tot_loss[loss=0.1481, simple_loss=0.2349, pruned_loss=0.03062, over 985339.13 frames.], datatang_tot_loss[loss=0.1445, simple_loss=0.2211, pruned_loss=0.03399, over 985851.61 frames.], batch size: 64, lr: 3.69e-04
+2022-06-19 02:29:12,237 INFO [train.py:874] (2/4) Epoch 22, batch 3550, aishell_loss[loss=0.1442, simple_loss=0.2338, pruned_loss=0.02735, over 4911.00 frames.], tot_loss[loss=0.1462, simple_loss=0.2276, pruned_loss=0.03239, over 985846.58 frames.], batch size: 46, aishell_tot_loss[loss=0.148, simple_loss=0.2348, pruned_loss=0.03065, over 985479.86 frames.], datatang_tot_loss[loss=0.1442, simple_loss=0.2207, pruned_loss=0.03382, over 985751.99 frames.], batch size: 46, lr: 3.69e-04
+2022-06-19 02:29:44,436 INFO [train.py:874] (2/4) Epoch 22, batch 3600, datatang_loss[loss=0.1285, simple_loss=0.2009, pruned_loss=0.02801, over 4852.00 frames.], tot_loss[loss=0.1465, simple_loss=0.2274, pruned_loss=0.03284, over 985395.88 frames.], batch size: 25, aishell_tot_loss[loss=0.1476, simple_loss=0.234, pruned_loss=0.0306, over 985426.97 frames.], datatang_tot_loss[loss=0.1449, simple_loss=0.2212, pruned_loss=0.03433, over 985387.65 frames.], batch size: 25, lr: 3.69e-04
+2022-06-19 02:30:17,395 INFO [train.py:874] (2/4) Epoch 22, batch 3650, datatang_loss[loss=0.1558, simple_loss=0.235, pruned_loss=0.03829, over 4953.00 frames.], tot_loss[loss=0.1468, simple_loss=0.2278, pruned_loss=0.03287, over 985553.13 frames.], batch size: 91, aishell_tot_loss[loss=0.1472, simple_loss=0.2336, pruned_loss=0.03043, over 985362.85 frames.], datatang_tot_loss[loss=0.1456, simple_loss=0.2221, pruned_loss=0.0346, over 985621.26 frames.], batch size: 91, lr: 3.69e-04
+2022-06-19 02:30:48,792 INFO [train.py:874] (2/4) Epoch 22, batch 3700, datatang_loss[loss=0.1363, simple_loss=0.2138, pruned_loss=0.02941, over 4954.00 frames.], tot_loss[loss=0.1464, simple_loss=0.2275, pruned_loss=0.03269, over 985696.53 frames.], batch size: 86, aishell_tot_loss[loss=0.147, simple_loss=0.2333, pruned_loss=0.03042, over 985404.10 frames.], datatang_tot_loss[loss=0.1455, simple_loss=0.222, pruned_loss=0.03449, over 985737.33 frames.], batch size: 86, lr: 3.69e-04
+2022-06-19 02:31:19,478 INFO [train.py:874] (2/4) Epoch 22, batch 3750, datatang_loss[loss=0.1242, simple_loss=0.2095, pruned_loss=0.01946, over 4900.00 frames.], tot_loss[loss=0.1466, simple_loss=0.228, pruned_loss=0.03264, over 985464.92 frames.], batch size: 59, aishell_tot_loss[loss=0.1472, simple_loss=0.2335, pruned_loss=0.03047, over 985353.92 frames.], datatang_tot_loss[loss=0.1456, simple_loss=0.222, pruned_loss=0.03456, over 985568.56 frames.], batch size: 59, lr: 3.69e-04
+2022-06-19 02:31:49,498 INFO [train.py:874] (2/4) Epoch 22, batch 3800, aishell_loss[loss=0.1677, simple_loss=0.2405, pruned_loss=0.04746, over 4953.00 frames.], tot_loss[loss=0.1472, simple_loss=0.2284, pruned_loss=0.03301, over 985260.28 frames.], batch size: 31, aishell_tot_loss[loss=0.1473, simple_loss=0.2334, pruned_loss=0.03061, over 985145.75 frames.], datatang_tot_loss[loss=0.1461, simple_loss=0.2226, pruned_loss=0.03485, over 985563.23 frames.], batch size: 31, lr: 3.69e-04
+2022-06-19 02:32:20,319 INFO [train.py:874] (2/4) Epoch 22, batch 3850, datatang_loss[loss=0.1651, simple_loss=0.2441, pruned_loss=0.04302, over 4918.00 frames.], tot_loss[loss=0.1476, simple_loss=0.2289, pruned_loss=0.03318, over 985442.30 frames.], batch size: 83, aishell_tot_loss[loss=0.1473, simple_loss=0.2336, pruned_loss=0.03052, over 985312.62 frames.], datatang_tot_loss[loss=0.1467, simple_loss=0.2232, pruned_loss=0.0351, over 985573.42 frames.], batch size: 83, lr: 3.68e-04
+2022-06-19 02:32:51,368 INFO [train.py:874] (2/4) Epoch 22, batch 3900, aishell_loss[loss=0.1457, simple_loss=0.2313, pruned_loss=0.03007, over 4911.00 frames.], tot_loss[loss=0.147, simple_loss=0.2285, pruned_loss=0.03276, over 985622.87 frames.], batch size: 52, aishell_tot_loss[loss=0.1475, simple_loss=0.234, pruned_loss=0.0305, over 985419.09 frames.], datatang_tot_loss[loss=0.146, simple_loss=0.2229, pruned_loss=0.03461, over 985635.49 frames.], batch size: 52, lr: 3.68e-04
+2022-06-19 02:33:21,709 INFO [train.py:874] (2/4) Epoch 22, batch 3950, aishell_loss[loss=0.1396, simple_loss=0.2249, pruned_loss=0.02717, over 4856.00 frames.], tot_loss[loss=0.1467, simple_loss=0.2283, pruned_loss=0.03256, over 985987.46 frames.], batch size: 37, aishell_tot_loss[loss=0.1476, simple_loss=0.2339, pruned_loss=0.03069, over 985525.65 frames.], datatang_tot_loss[loss=0.1456, simple_loss=0.2225, pruned_loss=0.03436, over 985951.38 frames.], batch size: 37, lr: 3.68e-04
+2022-06-19 02:33:51,300 INFO [train.py:874] (2/4) Epoch 22, batch 4000, aishell_loss[loss=0.1524, simple_loss=0.2338, pruned_loss=0.03553, over 4900.00 frames.], tot_loss[loss=0.1458, simple_loss=0.2274, pruned_loss=0.03214, over 985608.56 frames.], batch size: 34, aishell_tot_loss[loss=0.1477, simple_loss=0.2338, pruned_loss=0.03084, over 985396.02 frames.], datatang_tot_loss[loss=0.1446, simple_loss=0.2217, pruned_loss=0.03375, over 985736.88 frames.], batch size: 34, lr: 3.68e-04
+2022-06-19 02:33:51,301 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 02:34:08,825 INFO [train.py:914] (2/4) Epoch 22, validation: loss=0.165, simple_loss=0.2485, pruned_loss=0.04074, over 1622729.00 frames.
+2022-06-19 02:34:38,020 INFO [train.py:874] (2/4) Epoch 22, batch 4050, aishell_loss[loss=0.1419, simple_loss=0.2345, pruned_loss=0.02463, over 4920.00 frames.], tot_loss[loss=0.1463, simple_loss=0.2282, pruned_loss=0.03225, over 985481.75 frames.], batch size: 41, aishell_tot_loss[loss=0.1483, simple_loss=0.2343, pruned_loss=0.03119, over 985083.34 frames.], datatang_tot_loss[loss=0.1444, simple_loss=0.2218, pruned_loss=0.03352, over 985938.66 frames.], batch size: 41, lr: 3.68e-04
+2022-06-19 02:36:04,872 INFO [train.py:874] (2/4) Epoch 23, batch 50, datatang_loss[loss=0.1251, simple_loss=0.1945, pruned_loss=0.02782, over 4978.00 frames.], tot_loss[loss=0.1442, simple_loss=0.2236, pruned_loss=0.03241, over 218476.59 frames.], batch size: 37, aishell_tot_loss[loss=0.1487, simple_loss=0.2345, pruned_loss=0.03146, over 107022.93 frames.], datatang_tot_loss[loss=0.1406, simple_loss=0.2145, pruned_loss=0.03332, over 125017.73 frames.], batch size: 37, lr: 3.60e-04
+2022-06-19 02:36:35,937 INFO [train.py:874] (2/4) Epoch 23, batch 100, datatang_loss[loss=0.1456, simple_loss=0.2322, pruned_loss=0.02951, over 4957.00 frames.], tot_loss[loss=0.1421, simple_loss=0.223, pruned_loss=0.03065, over 388546.67 frames.], batch size: 99, aishell_tot_loss[loss=0.1487, simple_loss=0.2347, pruned_loss=0.03135, over 210382.55 frames.], datatang_tot_loss[loss=0.1363, simple_loss=0.2121, pruned_loss=0.03025, over 226511.98 frames.], batch size: 99, lr: 3.60e-04
+2022-06-19 02:37:08,832 INFO [train.py:874] (2/4) Epoch 23, batch 150, aishell_loss[loss=0.1514, simple_loss=0.2448, pruned_loss=0.029, over 4910.00 frames.], tot_loss[loss=0.1422, simple_loss=0.223, pruned_loss=0.03074, over 521119.19 frames.], batch size: 77, aishell_tot_loss[loss=0.148, simple_loss=0.2338, pruned_loss=0.03111, over 280800.77 frames.], datatang_tot_loss[loss=0.1379, simple_loss=0.2145, pruned_loss=0.03062, over 335984.39 frames.], batch size: 77, lr: 3.60e-04
+2022-06-19 02:37:40,901 INFO [train.py:874] (2/4) Epoch 23, batch 200, aishell_loss[loss=0.117, simple_loss=0.1877, pruned_loss=0.02313, over 4955.00 frames.], tot_loss[loss=0.1428, simple_loss=0.2246, pruned_loss=0.03052, over 624520.50 frames.], batch size: 25, aishell_tot_loss[loss=0.1471, simple_loss=0.2331, pruned_loss=0.03052, over 388833.77 frames.], datatang_tot_loss[loss=0.1383, simple_loss=0.2153, pruned_loss=0.03069, over 388902.69 frames.], batch size: 25, lr: 3.60e-04
+2022-06-19 02:38:12,796 INFO [train.py:874] (2/4) Epoch 23, batch 250, aishell_loss[loss=0.1167, simple_loss=0.1949, pruned_loss=0.01921, over 4947.00 frames.], tot_loss[loss=0.1425, simple_loss=0.2247, pruned_loss=0.03017, over 704664.02 frames.], batch size: 25, aishell_tot_loss[loss=0.1473, simple_loss=0.234, pruned_loss=0.03031, over 459039.76 frames.], datatang_tot_loss[loss=0.1377, simple_loss=0.2148, pruned_loss=0.03028, over 459359.12 frames.], batch size: 25, lr: 3.60e-04
+2022-06-19 02:38:44,865 INFO [train.py:874] (2/4) Epoch 23, batch 300, datatang_loss[loss=0.1519, simple_loss=0.2219, pruned_loss=0.04095, over 4953.00 frames.], tot_loss[loss=0.1439, simple_loss=0.2253, pruned_loss=0.03126, over 766868.88 frames.], batch size: 55, aishell_tot_loss[loss=0.1475, simple_loss=0.234, pruned_loss=0.03053, over 518427.68 frames.], datatang_tot_loss[loss=0.1396, simple_loss=0.216, pruned_loss=0.03167, over 523880.35 frames.], batch size: 55, lr: 3.60e-04
+2022-06-19 02:39:17,363 INFO [train.py:874] (2/4) Epoch 23, batch 350, aishell_loss[loss=0.1252, simple_loss=0.2233, pruned_loss=0.01355, over 4900.00 frames.], tot_loss[loss=0.1435, simple_loss=0.225, pruned_loss=0.031, over 815007.23 frames.], batch size: 60, aishell_tot_loss[loss=0.1466, simple_loss=0.2333, pruned_loss=0.02997, over 577418.47 frames.], datatang_tot_loss[loss=0.14, simple_loss=0.2162, pruned_loss=0.03193, over 573931.74 frames.], batch size: 60, lr: 3.59e-04
+2022-06-19 02:39:48,705 INFO [train.py:874] (2/4) Epoch 23, batch 400, datatang_loss[loss=0.1515, simple_loss=0.2267, pruned_loss=0.03814, over 4954.00 frames.], tot_loss[loss=0.1426, simple_loss=0.2243, pruned_loss=0.03042, over 852841.22 frames.], batch size: 86, aishell_tot_loss[loss=0.1461, simple_loss=0.233, pruned_loss=0.02955, over 618265.34 frames.], datatang_tot_loss[loss=0.1396, simple_loss=0.2162, pruned_loss=0.03147, over 629589.58 frames.], batch size: 86, lr: 3.59e-04
+2022-06-19 02:40:20,256 INFO [train.py:874] (2/4) Epoch 23, batch 450, datatang_loss[loss=0.1415, simple_loss=0.2065, pruned_loss=0.0383, over 4963.00 frames.], tot_loss[loss=0.1436, simple_loss=0.2255, pruned_loss=0.03086, over 882106.76 frames.], batch size: 37, aishell_tot_loss[loss=0.1467, simple_loss=0.2338, pruned_loss=0.02986, over 669511.20 frames.], datatang_tot_loss[loss=0.14, simple_loss=0.2163, pruned_loss=0.03181, over 663387.01 frames.], batch size: 37, lr: 3.59e-04
+2022-06-19 02:40:53,009 INFO [train.py:874] (2/4) Epoch 23, batch 500, datatang_loss[loss=0.1288, simple_loss=0.1999, pruned_loss=0.02879, over 4971.00 frames.], tot_loss[loss=0.1442, simple_loss=0.2261, pruned_loss=0.03112, over 904960.00 frames.], batch size: 40, aishell_tot_loss[loss=0.1469, simple_loss=0.2337, pruned_loss=0.03009, over 710697.00 frames.], datatang_tot_loss[loss=0.1405, simple_loss=0.2171, pruned_loss=0.03198, over 697134.56 frames.], batch size: 40, lr: 3.59e-04
+2022-06-19 02:41:25,231 INFO [train.py:874] (2/4) Epoch 23, batch 550, datatang_loss[loss=0.1641, simple_loss=0.2185, pruned_loss=0.05491, over 4940.00 frames.], tot_loss[loss=0.1456, simple_loss=0.2276, pruned_loss=0.03183, over 922775.10 frames.], batch size: 50, aishell_tot_loss[loss=0.1474, simple_loss=0.2344, pruned_loss=0.03027, over 744261.84 frames.], datatang_tot_loss[loss=0.1421, simple_loss=0.2187, pruned_loss=0.0328, over 729785.36 frames.], batch size: 50, lr: 3.59e-04
+2022-06-19 02:41:56,286 INFO [train.py:874] (2/4) Epoch 23, batch 600, aishell_loss[loss=0.129, simple_loss=0.2029, pruned_loss=0.02758, over 4947.00 frames.], tot_loss[loss=0.1461, simple_loss=0.228, pruned_loss=0.03213, over 936535.51 frames.], batch size: 31, aishell_tot_loss[loss=0.1478, simple_loss=0.2346, pruned_loss=0.03052, over 772496.51 frames.], datatang_tot_loss[loss=0.1426, simple_loss=0.2192, pruned_loss=0.03303, over 759929.30 frames.], batch size: 31, lr: 3.59e-04
+2022-06-19 02:42:28,257 INFO [train.py:874] (2/4) Epoch 23, batch 650, datatang_loss[loss=0.1445, simple_loss=0.2203, pruned_loss=0.03435, over 4851.00 frames.], tot_loss[loss=0.1456, simple_loss=0.2273, pruned_loss=0.03194, over 947281.95 frames.], batch size: 30, aishell_tot_loss[loss=0.1472, simple_loss=0.2339, pruned_loss=0.03028, over 799312.27 frames.], datatang_tot_loss[loss=0.1428, simple_loss=0.2192, pruned_loss=0.03313, over 784537.74 frames.], batch size: 30, lr: 3.59e-04
+2022-06-19 02:42:59,739 INFO [train.py:874] (2/4) Epoch 23, batch 700, aishell_loss[loss=0.1395, simple_loss=0.2311, pruned_loss=0.02397, over 4962.00 frames.], tot_loss[loss=0.1461, simple_loss=0.2276, pruned_loss=0.03232, over 956104.49 frames.], batch size: 61, aishell_tot_loss[loss=0.1476, simple_loss=0.2341, pruned_loss=0.03058, over 820559.33 frames.], datatang_tot_loss[loss=0.1433, simple_loss=0.2197, pruned_loss=0.03339, over 809315.35 frames.], batch size: 61, lr: 3.59e-04
+2022-06-19 02:43:32,099 INFO [train.py:874] (2/4) Epoch 23, batch 750, datatang_loss[loss=0.1838, simple_loss=0.2442, pruned_loss=0.06173, over 4938.00 frames.], tot_loss[loss=0.1458, simple_loss=0.227, pruned_loss=0.03231, over 962674.59 frames.], batch size: 69, aishell_tot_loss[loss=0.1473, simple_loss=0.2335, pruned_loss=0.03055, over 838411.72 frames.], datatang_tot_loss[loss=0.1434, simple_loss=0.2199, pruned_loss=0.03347, over 831780.62 frames.], batch size: 69, lr: 3.59e-04
+2022-06-19 02:44:03,149 INFO [train.py:874] (2/4) Epoch 23, batch 800, aishell_loss[loss=0.1575, simple_loss=0.2512, pruned_loss=0.03191, over 4911.00 frames.], tot_loss[loss=0.1446, simple_loss=0.2257, pruned_loss=0.03173, over 967373.97 frames.], batch size: 78, aishell_tot_loss[loss=0.1464, simple_loss=0.2325, pruned_loss=0.03013, over 855474.91 frames.], datatang_tot_loss[loss=0.143, simple_loss=0.2194, pruned_loss=0.03328, over 849771.56 frames.], batch size: 78, lr: 3.59e-04
+2022-06-19 02:44:41,818 INFO [train.py:874] (2/4) Epoch 23, batch 850, aishell_loss[loss=0.1419, simple_loss=0.2413, pruned_loss=0.02127, over 4942.00 frames.], tot_loss[loss=0.1451, simple_loss=0.2265, pruned_loss=0.03178, over 971626.95 frames.], batch size: 64, aishell_tot_loss[loss=0.1468, simple_loss=0.233, pruned_loss=0.03029, over 873888.64 frames.], datatang_tot_loss[loss=0.143, simple_loss=0.2195, pruned_loss=0.03327, over 862704.12 frames.], batch size: 64, lr: 3.58e-04
+2022-06-19 02:45:13,060 INFO [train.py:874] (2/4) Epoch 23, batch 900, datatang_loss[loss=0.1434, simple_loss=0.2204, pruned_loss=0.03322, over 4913.00 frames.], tot_loss[loss=0.1451, simple_loss=0.2266, pruned_loss=0.03181, over 974605.20 frames.], batch size: 75, aishell_tot_loss[loss=0.1469, simple_loss=0.233, pruned_loss=0.03038, over 886859.94 frames.], datatang_tot_loss[loss=0.143, simple_loss=0.2196, pruned_loss=0.03322, over 877242.62 frames.], batch size: 75, lr: 3.58e-04
+2022-06-19 02:45:44,642 INFO [train.py:874] (2/4) Epoch 23, batch 950, datatang_loss[loss=0.1511, simple_loss=0.2205, pruned_loss=0.04088, over 4961.00 frames.], tot_loss[loss=0.1449, simple_loss=0.2267, pruned_loss=0.0316, over 977136.66 frames.], batch size: 86, aishell_tot_loss[loss=0.1466, simple_loss=0.2329, pruned_loss=0.03014, over 900646.64 frames.], datatang_tot_loss[loss=0.1431, simple_loss=0.2196, pruned_loss=0.03327, over 887690.90 frames.], batch size: 86, lr: 3.58e-04
+2022-06-19 02:46:17,102 INFO [train.py:874] (2/4) Epoch 23, batch 1000, datatang_loss[loss=0.1276, simple_loss=0.2051, pruned_loss=0.02509, over 4918.00 frames.], tot_loss[loss=0.1448, simple_loss=0.227, pruned_loss=0.03136, over 978850.46 frames.], batch size: 64, aishell_tot_loss[loss=0.1467, simple_loss=0.2332, pruned_loss=0.03012, over 911237.71 frames.], datatang_tot_loss[loss=0.1428, simple_loss=0.2196, pruned_loss=0.03301, over 898340.34 frames.], batch size: 64, lr: 3.58e-04
+2022-06-19 02:46:17,102 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 02:46:34,398 INFO [train.py:914] (2/4) Epoch 23, validation: loss=0.165, simple_loss=0.2485, pruned_loss=0.04077, over 1622729.00 frames.
+2022-06-19 02:47:05,337 INFO [train.py:874] (2/4) Epoch 23, batch 1050, datatang_loss[loss=0.1648, simple_loss=0.2398, pruned_loss=0.04495, over 4961.00 frames.], tot_loss[loss=0.1446, simple_loss=0.2266, pruned_loss=0.03131, over 980434.37 frames.], batch size: 91, aishell_tot_loss[loss=0.1458, simple_loss=0.2324, pruned_loss=0.02958, over 918928.70 frames.], datatang_tot_loss[loss=0.1436, simple_loss=0.2204, pruned_loss=0.0334, over 909935.34 frames.], batch size: 91, lr: 3.58e-04
+2022-06-19 02:47:36,040 INFO [train.py:874] (2/4) Epoch 23, batch 1100, datatang_loss[loss=0.126, simple_loss=0.2138, pruned_loss=0.01905, over 4948.00 frames.], tot_loss[loss=0.1444, simple_loss=0.226, pruned_loss=0.03137, over 981575.63 frames.], batch size: 62, aishell_tot_loss[loss=0.1462, simple_loss=0.2328, pruned_loss=0.02982, over 925756.71 frames.], datatang_tot_loss[loss=0.1429, simple_loss=0.2195, pruned_loss=0.03313, over 919973.06 frames.], batch size: 62, lr: 3.58e-04
+2022-06-19 02:48:07,576 INFO [train.py:874] (2/4) Epoch 23, batch 1150, aishell_loss[loss=0.1363, simple_loss=0.2285, pruned_loss=0.02207, over 4948.00 frames.], tot_loss[loss=0.1449, simple_loss=0.2264, pruned_loss=0.03166, over 982582.66 frames.], batch size: 54, aishell_tot_loss[loss=0.1464, simple_loss=0.2327, pruned_loss=0.03008, over 933025.24 frames.], datatang_tot_loss[loss=0.1432, simple_loss=0.22, pruned_loss=0.03319, over 927568.69 frames.], batch size: 54, lr: 3.58e-04
+2022-06-19 02:48:40,912 INFO [train.py:874] (2/4) Epoch 23, batch 1200, datatang_loss[loss=0.1226, simple_loss=0.2023, pruned_loss=0.02151, over 4919.00 frames.], tot_loss[loss=0.1448, simple_loss=0.2264, pruned_loss=0.03161, over 983379.84 frames.], batch size: 57, aishell_tot_loss[loss=0.146, simple_loss=0.2324, pruned_loss=0.02979, over 938799.81 frames.], datatang_tot_loss[loss=0.1436, simple_loss=0.2204, pruned_loss=0.03339, over 935004.71 frames.], batch size: 57, lr: 3.58e-04
+2022-06-19 02:49:12,894 INFO [train.py:874] (2/4) Epoch 23, batch 1250, datatang_loss[loss=0.1349, simple_loss=0.212, pruned_loss=0.02894, over 4934.00 frames.], tot_loss[loss=0.1445, simple_loss=0.2264, pruned_loss=0.03135, over 983539.70 frames.], batch size: 79, aishell_tot_loss[loss=0.1457, simple_loss=0.2323, pruned_loss=0.02959, over 944226.77 frames.], datatang_tot_loss[loss=0.1435, simple_loss=0.2204, pruned_loss=0.03335, over 940663.88 frames.], batch size: 79, lr: 3.58e-04
+2022-06-19 02:49:43,400 INFO [train.py:874] (2/4) Epoch 23, batch 1300, aishell_loss[loss=0.1694, simple_loss=0.2595, pruned_loss=0.03964, over 4952.00 frames.], tot_loss[loss=0.1454, simple_loss=0.2273, pruned_loss=0.03171, over 983657.06 frames.], batch size: 68, aishell_tot_loss[loss=0.1458, simple_loss=0.2326, pruned_loss=0.02949, over 949828.26 frames.], datatang_tot_loss[loss=0.1443, simple_loss=0.2208, pruned_loss=0.03389, over 944766.00 frames.], batch size: 68, lr: 3.58e-04
+2022-06-19 02:50:15,877 INFO [train.py:874] (2/4) Epoch 23, batch 1350, aishell_loss[loss=0.1555, simple_loss=0.2398, pruned_loss=0.03564, over 4921.00 frames.], tot_loss[loss=0.145, simple_loss=0.2266, pruned_loss=0.03172, over 983966.85 frames.], batch size: 33, aishell_tot_loss[loss=0.1462, simple_loss=0.2329, pruned_loss=0.02973, over 953035.25 frames.], datatang_tot_loss[loss=0.1436, simple_loss=0.2201, pruned_loss=0.03357, over 950549.36 frames.], batch size: 33, lr: 3.57e-04
+2022-06-19 02:50:47,390 INFO [train.py:874] (2/4) Epoch 23, batch 1400, datatang_loss[loss=0.1839, simple_loss=0.256, pruned_loss=0.05585, over 4910.00 frames.], tot_loss[loss=0.1454, simple_loss=0.2272, pruned_loss=0.03181, over 984209.36 frames.], batch size: 108, aishell_tot_loss[loss=0.1465, simple_loss=0.2333, pruned_loss=0.02981, over 956765.53 frames.], datatang_tot_loss[loss=0.1438, simple_loss=0.2203, pruned_loss=0.0336, over 954629.07 frames.], batch size: 108, lr: 3.57e-04
+2022-06-19 02:51:19,371 INFO [train.py:874] (2/4) Epoch 23, batch 1450, aishell_loss[loss=0.1434, simple_loss=0.2225, pruned_loss=0.03221, over 4988.00 frames.], tot_loss[loss=0.1452, simple_loss=0.2266, pruned_loss=0.03184, over 984537.12 frames.], batch size: 30, aishell_tot_loss[loss=0.1462, simple_loss=0.2329, pruned_loss=0.02974, over 960335.49 frames.], datatang_tot_loss[loss=0.1439, simple_loss=0.2203, pruned_loss=0.03371, over 958128.07 frames.], batch size: 30, lr: 3.57e-04
+2022-06-19 02:51:53,281 INFO [train.py:874] (2/4) Epoch 23, batch 1500, aishell_loss[loss=0.1432, simple_loss=0.2241, pruned_loss=0.03116, over 4888.00 frames.], tot_loss[loss=0.1444, simple_loss=0.2257, pruned_loss=0.03156, over 984883.55 frames.], batch size: 42, aishell_tot_loss[loss=0.1459, simple_loss=0.2326, pruned_loss=0.02961, over 963374.30 frames.], datatang_tot_loss[loss=0.1434, simple_loss=0.2196, pruned_loss=0.03356, over 961425.85 frames.], batch size: 42, lr: 3.57e-04
+2022-06-19 02:52:24,980 INFO [train.py:874] (2/4) Epoch 23, batch 1550, aishell_loss[loss=0.1528, simple_loss=0.2364, pruned_loss=0.03461, over 4933.00 frames.], tot_loss[loss=0.1454, simple_loss=0.2269, pruned_loss=0.0319, over 984914.25 frames.], batch size: 58, aishell_tot_loss[loss=0.1468, simple_loss=0.2335, pruned_loss=0.03007, over 966626.55 frames.], datatang_tot_loss[loss=0.1433, simple_loss=0.2195, pruned_loss=0.03358, over 963412.81 frames.], batch size: 58, lr: 3.57e-04
+2022-06-19 02:52:56,564 INFO [train.py:874] (2/4) Epoch 23, batch 1600, datatang_loss[loss=0.1206, simple_loss=0.2066, pruned_loss=0.01728, over 4923.00 frames.], tot_loss[loss=0.1457, simple_loss=0.2273, pruned_loss=0.03204, over 984922.76 frames.], batch size: 75, aishell_tot_loss[loss=0.1471, simple_loss=0.2337, pruned_loss=0.03026, over 968067.72 frames.], datatang_tot_loss[loss=0.1435, simple_loss=0.22, pruned_loss=0.03347, over 966698.49 frames.], batch size: 75, lr: 3.57e-04
+2022-06-19 02:53:27,751 INFO [train.py:874] (2/4) Epoch 23, batch 1650, datatang_loss[loss=0.1599, simple_loss=0.2361, pruned_loss=0.04188, over 4973.00 frames.], tot_loss[loss=0.1464, simple_loss=0.2282, pruned_loss=0.03235, over 985181.72 frames.], batch size: 45, aishell_tot_loss[loss=0.1476, simple_loss=0.2342, pruned_loss=0.03051, over 970266.25 frames.], datatang_tot_loss[loss=0.1438, simple_loss=0.2204, pruned_loss=0.03363, over 968883.61 frames.], batch size: 45, lr: 3.57e-04
+2022-06-19 02:54:00,679 INFO [train.py:874] (2/4) Epoch 23, batch 1700, aishell_loss[loss=0.1463, simple_loss=0.2314, pruned_loss=0.03062, over 4971.00 frames.], tot_loss[loss=0.1463, simple_loss=0.2285, pruned_loss=0.03205, over 985197.86 frames.], batch size: 38, aishell_tot_loss[loss=0.1477, simple_loss=0.2345, pruned_loss=0.03045, over 972208.89 frames.], datatang_tot_loss[loss=0.1436, simple_loss=0.2203, pruned_loss=0.03346, over 970604.87 frames.], batch size: 38, lr: 3.57e-04
+2022-06-19 02:54:33,023 INFO [train.py:874] (2/4) Epoch 23, batch 1750, aishell_loss[loss=0.1244, simple_loss=0.1994, pruned_loss=0.02474, over 4833.00 frames.], tot_loss[loss=0.145, simple_loss=0.2273, pruned_loss=0.03138, over 984930.36 frames.], batch size: 24, aishell_tot_loss[loss=0.147, simple_loss=0.234, pruned_loss=0.03, over 973337.59 frames.], datatang_tot_loss[loss=0.1432, simple_loss=0.2199, pruned_loss=0.03321, over 972443.37 frames.], batch size: 24, lr: 3.57e-04
+2022-06-19 02:55:05,061 INFO [train.py:874] (2/4) Epoch 23, batch 1800, aishell_loss[loss=0.1624, simple_loss=0.2524, pruned_loss=0.03623, over 4858.00 frames.], tot_loss[loss=0.1454, simple_loss=0.2277, pruned_loss=0.03155, over 984944.60 frames.], batch size: 36, aishell_tot_loss[loss=0.1472, simple_loss=0.2341, pruned_loss=0.03013, over 974277.81 frames.], datatang_tot_loss[loss=0.1434, simple_loss=0.2204, pruned_loss=0.03319, over 974340.63 frames.], batch size: 36, lr: 3.57e-04
+2022-06-19 02:55:38,717 INFO [train.py:874] (2/4) Epoch 23, batch 1850, datatang_loss[loss=0.132, simple_loss=0.2088, pruned_loss=0.02757, over 4976.00 frames.], tot_loss[loss=0.1456, simple_loss=0.2278, pruned_loss=0.03164, over 985111.54 frames.], batch size: 45, aishell_tot_loss[loss=0.1474, simple_loss=0.2344, pruned_loss=0.03021, over 975420.47 frames.], datatang_tot_loss[loss=0.1434, simple_loss=0.2204, pruned_loss=0.03319, over 975867.34 frames.], batch size: 45, lr: 3.57e-04
+2022-06-19 02:56:09,932 INFO [train.py:874] (2/4) Epoch 23, batch 1900, aishell_loss[loss=0.1196, simple_loss=0.2053, pruned_loss=0.01692, over 4961.00 frames.], tot_loss[loss=0.1458, simple_loss=0.2281, pruned_loss=0.03176, over 985250.77 frames.], batch size: 27, aishell_tot_loss[loss=0.1477, simple_loss=0.2348, pruned_loss=0.03027, over 976731.25 frames.], datatang_tot_loss[loss=0.1434, simple_loss=0.2203, pruned_loss=0.03324, over 976959.59 frames.], batch size: 27, lr: 3.56e-04
+2022-06-19 02:56:41,439 INFO [train.py:874] (2/4) Epoch 23, batch 1950, datatang_loss[loss=0.1473, simple_loss=0.2113, pruned_loss=0.04171, over 4927.00 frames.], tot_loss[loss=0.1459, simple_loss=0.2274, pruned_loss=0.03218, over 985286.87 frames.], batch size: 57, aishell_tot_loss[loss=0.1476, simple_loss=0.2343, pruned_loss=0.03045, over 977481.53 frames.], datatang_tot_loss[loss=0.1437, simple_loss=0.2205, pruned_loss=0.03343, over 978225.29 frames.], batch size: 57, lr: 3.56e-04
+2022-06-19 02:57:14,387 INFO [train.py:874] (2/4) Epoch 23, batch 2000, datatang_loss[loss=0.1373, simple_loss=0.2152, pruned_loss=0.02972, over 4930.00 frames.], tot_loss[loss=0.146, simple_loss=0.2276, pruned_loss=0.03219, over 985553.14 frames.], batch size: 69, aishell_tot_loss[loss=0.1474, simple_loss=0.2341, pruned_loss=0.03033, over 978438.44 frames.], datatang_tot_loss[loss=0.1443, simple_loss=0.2214, pruned_loss=0.03355, over 979293.07 frames.], batch size: 69, lr: 3.56e-04
+2022-06-19 02:57:14,388 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 02:57:32,299 INFO [train.py:914] (2/4) Epoch 23, validation: loss=0.1648, simple_loss=0.2485, pruned_loss=0.04055, over 1622729.00 frames.
+2022-06-19 02:58:03,738 INFO [train.py:874] (2/4) Epoch 23, batch 2050, datatang_loss[loss=0.1332, simple_loss=0.2164, pruned_loss=0.02497, over 4945.00 frames.], tot_loss[loss=0.1454, simple_loss=0.2272, pruned_loss=0.03178, over 985497.78 frames.], batch size: 69, aishell_tot_loss[loss=0.1477, simple_loss=0.2344, pruned_loss=0.0305, over 979094.06 frames.], datatang_tot_loss[loss=0.1433, simple_loss=0.2206, pruned_loss=0.03303, over 980170.63 frames.], batch size: 69, lr: 3.56e-04
+2022-06-19 02:58:35,257 INFO [train.py:874] (2/4) Epoch 23, batch 2100, aishell_loss[loss=0.1554, simple_loss=0.2401, pruned_loss=0.03537, over 4883.00 frames.], tot_loss[loss=0.145, simple_loss=0.2265, pruned_loss=0.03178, over 985393.77 frames.], batch size: 42, aishell_tot_loss[loss=0.1473, simple_loss=0.2338, pruned_loss=0.03041, over 979656.60 frames.], datatang_tot_loss[loss=0.1433, simple_loss=0.2205, pruned_loss=0.03311, over 980854.04 frames.], batch size: 42, lr: 3.56e-04
+2022-06-19 02:59:07,976 INFO [train.py:874] (2/4) Epoch 23, batch 2150, datatang_loss[loss=0.1636, simple_loss=0.2359, pruned_loss=0.04571, over 4978.00 frames.], tot_loss[loss=0.145, simple_loss=0.2265, pruned_loss=0.0318, over 985641.17 frames.], batch size: 45, aishell_tot_loss[loss=0.1472, simple_loss=0.2337, pruned_loss=0.03039, over 980182.95 frames.], datatang_tot_loss[loss=0.1434, simple_loss=0.2207, pruned_loss=0.03306, over 981764.56 frames.], batch size: 45, lr: 3.56e-04
+2022-06-19 02:59:40,141 INFO [train.py:874] (2/4) Epoch 23, batch 2200, datatang_loss[loss=0.1531, simple_loss=0.2157, pruned_loss=0.04529, over 4943.00 frames.], tot_loss[loss=0.1455, simple_loss=0.2269, pruned_loss=0.03203, over 985873.45 frames.], batch size: 26, aishell_tot_loss[loss=0.1475, simple_loss=0.2338, pruned_loss=0.03057, over 980862.96 frames.], datatang_tot_loss[loss=0.1436, simple_loss=0.2208, pruned_loss=0.03316, over 982432.69 frames.], batch size: 26, lr: 3.56e-04
+2022-06-19 03:00:11,257 INFO [train.py:874] (2/4) Epoch 23, batch 2250, datatang_loss[loss=0.1465, simple_loss=0.2267, pruned_loss=0.03314, over 4904.00 frames.], tot_loss[loss=0.1451, simple_loss=0.2264, pruned_loss=0.03188, over 986080.36 frames.], batch size: 52, aishell_tot_loss[loss=0.1476, simple_loss=0.2338, pruned_loss=0.03071, over 981464.21 frames.], datatang_tot_loss[loss=0.1431, simple_loss=0.2205, pruned_loss=0.03287, over 983016.01 frames.], batch size: 52, lr: 3.56e-04
+2022-06-19 03:00:43,962 INFO [train.py:874] (2/4) Epoch 23, batch 2300, aishell_loss[loss=0.1289, simple_loss=0.2183, pruned_loss=0.01975, over 4873.00 frames.], tot_loss[loss=0.1452, simple_loss=0.2266, pruned_loss=0.03194, over 985666.14 frames.], batch size: 28, aishell_tot_loss[loss=0.1475, simple_loss=0.2337, pruned_loss=0.03065, over 981575.39 frames.], datatang_tot_loss[loss=0.1434, simple_loss=0.2208, pruned_loss=0.033, over 983373.82 frames.], batch size: 28, lr: 3.56e-04
+2022-06-19 03:01:16,190 INFO [train.py:874] (2/4) Epoch 23, batch 2350, aishell_loss[loss=0.1389, simple_loss=0.2168, pruned_loss=0.03057, over 4982.00 frames.], tot_loss[loss=0.1444, simple_loss=0.2257, pruned_loss=0.03161, over 985686.17 frames.], batch size: 27, aishell_tot_loss[loss=0.1473, simple_loss=0.2335, pruned_loss=0.0306, over 982021.46 frames.], datatang_tot_loss[loss=0.1427, simple_loss=0.22, pruned_loss=0.0327, over 983660.61 frames.], batch size: 27, lr: 3.56e-04
+2022-06-19 03:01:47,602 INFO [train.py:874] (2/4) Epoch 23, batch 2400, aishell_loss[loss=0.13, simple_loss=0.2168, pruned_loss=0.02159, over 4892.00 frames.], tot_loss[loss=0.144, simple_loss=0.2253, pruned_loss=0.03134, over 985680.95 frames.], batch size: 28, aishell_tot_loss[loss=0.1477, simple_loss=0.2339, pruned_loss=0.03073, over 982389.38 frames.], datatang_tot_loss[loss=0.1418, simple_loss=0.2192, pruned_loss=0.03223, over 983924.41 frames.], batch size: 28, lr: 3.55e-04
+2022-06-19 03:02:18,516 INFO [train.py:874] (2/4) Epoch 23, batch 2450, aishell_loss[loss=0.1377, simple_loss=0.2301, pruned_loss=0.02259, over 4941.00 frames.], tot_loss[loss=0.1443, simple_loss=0.2261, pruned_loss=0.03129, over 986147.10 frames.], batch size: 54, aishell_tot_loss[loss=0.1476, simple_loss=0.234, pruned_loss=0.0306, over 983089.96 frames.], datatang_tot_loss[loss=0.142, simple_loss=0.2193, pruned_loss=0.03231, over 984359.09 frames.], batch size: 54, lr: 3.55e-04
+2022-06-19 03:02:51,512 INFO [train.py:874] (2/4) Epoch 23, batch 2500, datatang_loss[loss=0.215, simple_loss=0.2773, pruned_loss=0.07637, over 4944.00 frames.], tot_loss[loss=0.1446, simple_loss=0.2263, pruned_loss=0.0314, over 986085.51 frames.], batch size: 109, aishell_tot_loss[loss=0.1476, simple_loss=0.2342, pruned_loss=0.03044, over 983332.87 frames.], datatang_tot_loss[loss=0.1422, simple_loss=0.2194, pruned_loss=0.03249, over 984590.28 frames.], batch size: 109, lr: 3.55e-04
+2022-06-19 03:03:23,712 INFO [train.py:874] (2/4) Epoch 23, batch 2550, datatang_loss[loss=0.1578, simple_loss=0.236, pruned_loss=0.03977, over 4962.00 frames.], tot_loss[loss=0.1448, simple_loss=0.2269, pruned_loss=0.03135, over 986365.40 frames.], batch size: 55, aishell_tot_loss[loss=0.1476, simple_loss=0.2347, pruned_loss=0.03025, over 983737.22 frames.], datatang_tot_loss[loss=0.1425, simple_loss=0.2198, pruned_loss=0.03259, over 984970.31 frames.], batch size: 55, lr: 3.55e-04
+2022-06-19 03:03:55,788 INFO [train.py:874] (2/4) Epoch 23, batch 2600, aishell_loss[loss=0.1615, simple_loss=0.252, pruned_loss=0.03553, over 4875.00 frames.], tot_loss[loss=0.1453, simple_loss=0.2272, pruned_loss=0.03167, over 986118.86 frames.], batch size: 42, aishell_tot_loss[loss=0.1472, simple_loss=0.2344, pruned_loss=0.03005, over 983977.33 frames.], datatang_tot_loss[loss=0.1433, simple_loss=0.2204, pruned_loss=0.03306, over 984933.19 frames.], batch size: 42, lr: 3.55e-04
+2022-06-19 03:04:28,495 INFO [train.py:874] (2/4) Epoch 23, batch 2650, datatang_loss[loss=0.1358, simple_loss=0.2139, pruned_loss=0.02882, over 4924.00 frames.], tot_loss[loss=0.146, simple_loss=0.2278, pruned_loss=0.03207, over 986207.82 frames.], batch size: 81, aishell_tot_loss[loss=0.1481, simple_loss=0.2349, pruned_loss=0.03066, over 984286.80 frames.], datatang_tot_loss[loss=0.1431, simple_loss=0.2203, pruned_loss=0.03296, over 985113.82 frames.], batch size: 81, lr: 3.55e-04
+2022-06-19 03:05:00,011 INFO [train.py:874] (2/4) Epoch 23, batch 2700, aishell_loss[loss=0.1584, simple_loss=0.2455, pruned_loss=0.0356, over 4930.00 frames.], tot_loss[loss=0.1466, simple_loss=0.2281, pruned_loss=0.03253, over 986048.13 frames.], batch size: 58, aishell_tot_loss[loss=0.1477, simple_loss=0.2345, pruned_loss=0.03052, over 984206.73 frames.], datatang_tot_loss[loss=0.1442, simple_loss=0.2211, pruned_loss=0.03365, over 985388.05 frames.], batch size: 58, lr: 3.55e-04
+2022-06-19 03:05:31,311 INFO [train.py:874] (2/4) Epoch 23, batch 2750, aishell_loss[loss=0.1157, simple_loss=0.1983, pruned_loss=0.01654, over 4944.00 frames.], tot_loss[loss=0.1457, simple_loss=0.2273, pruned_loss=0.03201, over 986138.15 frames.], batch size: 25, aishell_tot_loss[loss=0.1469, simple_loss=0.2337, pruned_loss=0.0301, over 984397.82 frames.], datatang_tot_loss[loss=0.1442, simple_loss=0.2211, pruned_loss=0.03364, over 985601.03 frames.], batch size: 25, lr: 3.55e-04
+2022-06-19 03:06:03,718 INFO [train.py:874] (2/4) Epoch 23, batch 2800, aishell_loss[loss=0.1281, simple_loss=0.2215, pruned_loss=0.01738, over 4940.00 frames.], tot_loss[loss=0.1459, simple_loss=0.2273, pruned_loss=0.03223, over 986058.54 frames.], batch size: 45, aishell_tot_loss[loss=0.1472, simple_loss=0.2338, pruned_loss=0.0303, over 984346.77 frames.], datatang_tot_loss[loss=0.1443, simple_loss=0.2212, pruned_loss=0.03368, over 985822.34 frames.], batch size: 45, lr: 3.55e-04
+2022-06-19 03:06:35,782 INFO [train.py:874] (2/4) Epoch 23, batch 2850, aishell_loss[loss=0.1205, simple_loss=0.2036, pruned_loss=0.01868, over 4976.00 frames.], tot_loss[loss=0.1458, simple_loss=0.2273, pruned_loss=0.03214, over 985921.84 frames.], batch size: 27, aishell_tot_loss[loss=0.1471, simple_loss=0.2339, pruned_loss=0.03015, over 984340.79 frames.], datatang_tot_loss[loss=0.1443, simple_loss=0.2209, pruned_loss=0.03381, over 985919.13 frames.], batch size: 27, lr: 3.55e-04
+2022-06-19 03:07:06,069 INFO [train.py:874] (2/4) Epoch 23, batch 2900, aishell_loss[loss=0.1381, simple_loss=0.2279, pruned_loss=0.02411, over 4936.00 frames.], tot_loss[loss=0.146, simple_loss=0.2279, pruned_loss=0.03209, over 985769.43 frames.], batch size: 49, aishell_tot_loss[loss=0.1476, simple_loss=0.2346, pruned_loss=0.03031, over 984444.24 frames.], datatang_tot_loss[loss=0.1441, simple_loss=0.221, pruned_loss=0.03361, over 985835.34 frames.], batch size: 49, lr: 3.55e-04
+2022-06-19 03:07:38,864 INFO [train.py:874] (2/4) Epoch 23, batch 2950, datatang_loss[loss=0.1468, simple_loss=0.2321, pruned_loss=0.03074, over 4914.00 frames.], tot_loss[loss=0.1461, simple_loss=0.2278, pruned_loss=0.03216, over 985821.67 frames.], batch size: 42, aishell_tot_loss[loss=0.1475, simple_loss=0.2342, pruned_loss=0.03038, over 984842.60 frames.], datatang_tot_loss[loss=0.1443, simple_loss=0.2212, pruned_loss=0.0337, over 985644.12 frames.], batch size: 42, lr: 3.54e-04
+2022-06-19 03:08:11,512 INFO [train.py:874] (2/4) Epoch 23, batch 3000, aishell_loss[loss=0.1382, simple_loss=0.2181, pruned_loss=0.02918, over 4873.00 frames.], tot_loss[loss=0.1462, simple_loss=0.2281, pruned_loss=0.03216, over 985836.33 frames.], batch size: 35, aishell_tot_loss[loss=0.1475, simple_loss=0.2343, pruned_loss=0.03035, over 984782.87 frames.], datatang_tot_loss[loss=0.1444, simple_loss=0.2212, pruned_loss=0.03379, over 985861.41 frames.], batch size: 35, lr: 3.54e-04
+2022-06-19 03:08:11,513 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 03:08:29,176 INFO [train.py:914] (2/4) Epoch 23, validation: loss=0.1651, simple_loss=0.2488, pruned_loss=0.04074, over 1622729.00 frames.
+2022-06-19 03:09:00,948 INFO [train.py:874] (2/4) Epoch 23, batch 3050, datatang_loss[loss=0.2131, simple_loss=0.2777, pruned_loss=0.07419, over 4950.00 frames.], tot_loss[loss=0.1455, simple_loss=0.2276, pruned_loss=0.03171, over 985757.51 frames.], batch size: 108, aishell_tot_loss[loss=0.1469, simple_loss=0.2338, pruned_loss=0.03004, over 984904.16 frames.], datatang_tot_loss[loss=0.1442, simple_loss=0.221, pruned_loss=0.03374, over 985812.55 frames.], batch size: 108, lr: 3.54e-04
+2022-06-19 03:09:32,348 INFO [train.py:874] (2/4) Epoch 23, batch 3100, datatang_loss[loss=0.123, simple_loss=0.2067, pruned_loss=0.01969, over 4950.00 frames.], tot_loss[loss=0.145, simple_loss=0.2271, pruned_loss=0.03139, over 986096.75 frames.], batch size: 55, aishell_tot_loss[loss=0.1469, simple_loss=0.2336, pruned_loss=0.0301, over 985474.96 frames.], datatang_tot_loss[loss=0.1436, simple_loss=0.2204, pruned_loss=0.03337, over 985708.37 frames.], batch size: 55, lr: 3.54e-04
+2022-06-19 03:10:04,560 INFO [train.py:874] (2/4) Epoch 23, batch 3150, datatang_loss[loss=0.1195, simple_loss=0.1978, pruned_loss=0.02061, over 4939.00 frames.], tot_loss[loss=0.1451, simple_loss=0.227, pruned_loss=0.03158, over 986039.75 frames.], batch size: 79, aishell_tot_loss[loss=0.1473, simple_loss=0.2339, pruned_loss=0.03031, over 985587.29 frames.], datatang_tot_loss[loss=0.1433, simple_loss=0.2201, pruned_loss=0.03324, over 985643.28 frames.], batch size: 79, lr: 3.54e-04
+2022-06-19 03:10:38,020 INFO [train.py:874] (2/4) Epoch 23, batch 3200, datatang_loss[loss=0.1462, simple_loss=0.2189, pruned_loss=0.03675, over 4928.00 frames.], tot_loss[loss=0.1455, simple_loss=0.2276, pruned_loss=0.03175, over 986170.86 frames.], batch size: 62, aishell_tot_loss[loss=0.1475, simple_loss=0.2343, pruned_loss=0.03036, over 985684.91 frames.], datatang_tot_loss[loss=0.1435, simple_loss=0.2203, pruned_loss=0.03335, over 985778.98 frames.], batch size: 62, lr: 3.54e-04
+2022-06-19 03:11:09,275 INFO [train.py:874] (2/4) Epoch 23, batch 3250, datatang_loss[loss=0.1265, simple_loss=0.1982, pruned_loss=0.0274, over 4955.00 frames.], tot_loss[loss=0.1445, simple_loss=0.2268, pruned_loss=0.03112, over 986391.77 frames.], batch size: 60, aishell_tot_loss[loss=0.1473, simple_loss=0.2343, pruned_loss=0.03019, over 986133.91 frames.], datatang_tot_loss[loss=0.1426, simple_loss=0.2195, pruned_loss=0.03288, over 985663.87 frames.], batch size: 60, lr: 3.54e-04
+2022-06-19 03:11:41,513 INFO [train.py:874] (2/4) Epoch 23, batch 3300, aishell_loss[loss=0.1418, simple_loss=0.225, pruned_loss=0.02931, over 4907.00 frames.], tot_loss[loss=0.1447, simple_loss=0.2271, pruned_loss=0.03112, over 986331.62 frames.], batch size: 34, aishell_tot_loss[loss=0.1477, simple_loss=0.2346, pruned_loss=0.03041, over 986062.87 frames.], datatang_tot_loss[loss=0.1423, simple_loss=0.2193, pruned_loss=0.0326, over 985776.51 frames.], batch size: 34, lr: 3.54e-04
+2022-06-19 03:12:14,928 INFO [train.py:874] (2/4) Epoch 23, batch 3350, aishell_loss[loss=0.1451, simple_loss=0.2286, pruned_loss=0.03074, over 4962.00 frames.], tot_loss[loss=0.1444, simple_loss=0.2269, pruned_loss=0.03101, over 986470.10 frames.], batch size: 61, aishell_tot_loss[loss=0.1477, simple_loss=0.2346, pruned_loss=0.03041, over 986236.96 frames.], datatang_tot_loss[loss=0.142, simple_loss=0.2191, pruned_loss=0.03241, over 985855.13 frames.], batch size: 61, lr: 3.54e-04
+2022-06-19 03:12:46,724 INFO [train.py:874] (2/4) Epoch 23, batch 3400, datatang_loss[loss=0.1325, simple_loss=0.2131, pruned_loss=0.02593, over 4942.00 frames.], tot_loss[loss=0.1429, simple_loss=0.2256, pruned_loss=0.03014, over 985969.32 frames.], batch size: 69, aishell_tot_loss[loss=0.1469, simple_loss=0.234, pruned_loss=0.02987, over 985914.11 frames.], datatang_tot_loss[loss=0.1412, simple_loss=0.2186, pruned_loss=0.03189, over 985725.61 frames.], batch size: 69, lr: 3.54e-04
+2022-06-19 03:13:18,806 INFO [train.py:874] (2/4) Epoch 23, batch 3450, aishell_loss[loss=0.157, simple_loss=0.2368, pruned_loss=0.03866, over 4919.00 frames.], tot_loss[loss=0.1432, simple_loss=0.2258, pruned_loss=0.03036, over 986102.44 frames.], batch size: 33, aishell_tot_loss[loss=0.1469, simple_loss=0.2341, pruned_loss=0.02987, over 985833.11 frames.], datatang_tot_loss[loss=0.1412, simple_loss=0.2186, pruned_loss=0.03192, over 985990.61 frames.], batch size: 33, lr: 3.54e-04
+2022-06-19 03:13:51,956 INFO [train.py:874] (2/4) Epoch 23, batch 3500, datatang_loss[loss=0.1349, simple_loss=0.2181, pruned_loss=0.02584, over 4928.00 frames.], tot_loss[loss=0.144, simple_loss=0.2262, pruned_loss=0.03095, over 985709.99 frames.], batch size: 79, aishell_tot_loss[loss=0.148, simple_loss=0.235, pruned_loss=0.03043, over 985720.65 frames.], datatang_tot_loss[loss=0.141, simple_loss=0.2184, pruned_loss=0.03183, over 985731.04 frames.], batch size: 79, lr: 3.53e-04
+2022-06-19 03:14:23,655 INFO [train.py:874] (2/4) Epoch 23, batch 3550, datatang_loss[loss=0.184, simple_loss=0.2586, pruned_loss=0.05465, over 4947.00 frames.], tot_loss[loss=0.1449, simple_loss=0.2269, pruned_loss=0.03142, over 985757.05 frames.], batch size: 45, aishell_tot_loss[loss=0.1481, simple_loss=0.2352, pruned_loss=0.03053, over 985677.79 frames.], datatang_tot_loss[loss=0.1418, simple_loss=0.2192, pruned_loss=0.03216, over 985818.73 frames.], batch size: 45, lr: 3.53e-04
+2022-06-19 03:14:55,474 INFO [train.py:874] (2/4) Epoch 23, batch 3600, aishell_loss[loss=0.1795, simple_loss=0.2679, pruned_loss=0.04559, over 4852.00 frames.], tot_loss[loss=0.1449, simple_loss=0.2269, pruned_loss=0.03145, over 985538.37 frames.], batch size: 37, aishell_tot_loss[loss=0.1486, simple_loss=0.2355, pruned_loss=0.03078, over 985411.67 frames.], datatang_tot_loss[loss=0.1415, simple_loss=0.2191, pruned_loss=0.03191, over 985852.83 frames.], batch size: 37, lr: 3.53e-04
+2022-06-19 03:15:29,862 INFO [train.py:874] (2/4) Epoch 23, batch 3650, aishell_loss[loss=0.1302, simple_loss=0.2227, pruned_loss=0.01889, over 4971.00 frames.], tot_loss[loss=0.145, simple_loss=0.2267, pruned_loss=0.03161, over 985387.47 frames.], batch size: 30, aishell_tot_loss[loss=0.1488, simple_loss=0.2356, pruned_loss=0.03097, over 985121.69 frames.], datatang_tot_loss[loss=0.1414, simple_loss=0.2189, pruned_loss=0.0319, over 985959.07 frames.], batch size: 30, lr: 3.53e-04
+2022-06-19 03:16:01,414 INFO [train.py:874] (2/4) Epoch 23, batch 3700, datatang_loss[loss=0.1203, simple_loss=0.2001, pruned_loss=0.02029, over 4959.00 frames.], tot_loss[loss=0.1439, simple_loss=0.2254, pruned_loss=0.03118, over 985380.69 frames.], batch size: 86, aishell_tot_loss[loss=0.1484, simple_loss=0.2353, pruned_loss=0.0308, over 984901.37 frames.], datatang_tot_loss[loss=0.1408, simple_loss=0.2183, pruned_loss=0.03163, over 986095.81 frames.], batch size: 86, lr: 3.53e-04
+2022-06-19 03:16:34,256 INFO [train.py:874] (2/4) Epoch 23, batch 3750, datatang_loss[loss=0.128, simple_loss=0.2173, pruned_loss=0.01938, over 4853.00 frames.], tot_loss[loss=0.1439, simple_loss=0.2253, pruned_loss=0.03121, over 985050.77 frames.], batch size: 36, aishell_tot_loss[loss=0.1483, simple_loss=0.2349, pruned_loss=0.03088, over 984683.82 frames.], datatang_tot_loss[loss=0.1407, simple_loss=0.2182, pruned_loss=0.03157, over 985934.98 frames.], batch size: 36, lr: 3.53e-04
+2022-06-19 03:17:06,327 INFO [train.py:874] (2/4) Epoch 23, batch 3800, datatang_loss[loss=0.1358, simple_loss=0.2115, pruned_loss=0.03003, over 4924.00 frames.], tot_loss[loss=0.1431, simple_loss=0.2245, pruned_loss=0.03078, over 984741.37 frames.], batch size: 71, aishell_tot_loss[loss=0.1481, simple_loss=0.2346, pruned_loss=0.03081, over 984574.74 frames.], datatang_tot_loss[loss=0.1398, simple_loss=0.2172, pruned_loss=0.03119, over 985656.30 frames.], batch size: 71, lr: 3.53e-04
+2022-06-19 03:17:37,346 INFO [train.py:874] (2/4) Epoch 23, batch 3850, aishell_loss[loss=0.1339, simple_loss=0.229, pruned_loss=0.01944, over 4950.00 frames.], tot_loss[loss=0.1438, simple_loss=0.226, pruned_loss=0.03086, over 985067.65 frames.], batch size: 56, aishell_tot_loss[loss=0.1482, simple_loss=0.235, pruned_loss=0.03071, over 984860.16 frames.], datatang_tot_loss[loss=0.1401, simple_loss=0.2175, pruned_loss=0.03132, over 985658.64 frames.], batch size: 56, lr: 3.53e-04
+2022-06-19 03:18:07,536 INFO [train.py:874] (2/4) Epoch 23, batch 3900, aishell_loss[loss=0.1417, simple_loss=0.2298, pruned_loss=0.02681, over 4966.00 frames.], tot_loss[loss=0.1441, simple_loss=0.2262, pruned_loss=0.03097, over 985256.39 frames.], batch size: 51, aishell_tot_loss[loss=0.148, simple_loss=0.2347, pruned_loss=0.03064, over 984776.62 frames.], datatang_tot_loss[loss=0.1405, simple_loss=0.218, pruned_loss=0.03147, over 985870.47 frames.], batch size: 51, lr: 3.53e-04
+2022-06-19 03:18:38,102 INFO [train.py:874] (2/4) Epoch 23, batch 3950, datatang_loss[loss=0.153, simple_loss=0.2252, pruned_loss=0.04042, over 4960.00 frames.], tot_loss[loss=0.1442, simple_loss=0.2265, pruned_loss=0.03094, over 985349.08 frames.], batch size: 86, aishell_tot_loss[loss=0.1477, simple_loss=0.2343, pruned_loss=0.03054, over 984701.59 frames.], datatang_tot_loss[loss=0.1407, simple_loss=0.2183, pruned_loss=0.03153, over 986057.64 frames.], batch size: 86, lr: 3.53e-04
+2022-06-19 03:19:10,313 INFO [train.py:874] (2/4) Epoch 23, batch 4000, aishell_loss[loss=0.1562, simple_loss=0.2398, pruned_loss=0.03637, over 4961.00 frames.], tot_loss[loss=0.1448, simple_loss=0.2269, pruned_loss=0.03135, over 984876.69 frames.], batch size: 31, aishell_tot_loss[loss=0.1478, simple_loss=0.2342, pruned_loss=0.03067, over 984215.77 frames.], datatang_tot_loss[loss=0.1412, simple_loss=0.2188, pruned_loss=0.03181, over 986044.59 frames.], batch size: 31, lr: 3.52e-04
+2022-06-19 03:19:10,314 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 03:19:27,333 INFO [train.py:914] (2/4) Epoch 23, validation: loss=0.166, simple_loss=0.2494, pruned_loss=0.04127, over 1622729.00 frames.
+2022-06-19 03:19:59,953 INFO [train.py:874] (2/4) Epoch 23, batch 4050, datatang_loss[loss=0.1683, simple_loss=0.2438, pruned_loss=0.04634, over 4945.00 frames.], tot_loss[loss=0.1444, simple_loss=0.2263, pruned_loss=0.03123, over 985065.79 frames.], batch size: 91, aishell_tot_loss[loss=0.1476, simple_loss=0.234, pruned_loss=0.03062, over 984354.94 frames.], datatang_tot_loss[loss=0.1412, simple_loss=0.2191, pruned_loss=0.0317, over 985968.54 frames.], batch size: 91, lr: 3.52e-04
+2022-06-19 03:20:30,739 INFO [train.py:874] (2/4) Epoch 23, batch 4100, datatang_loss[loss=0.1473, simple_loss=0.2283, pruned_loss=0.03311, over 4955.00 frames.], tot_loss[loss=0.1454, simple_loss=0.2269, pruned_loss=0.03197, over 985085.71 frames.], batch size: 91, aishell_tot_loss[loss=0.1479, simple_loss=0.2344, pruned_loss=0.03071, over 984360.44 frames.], datatang_tot_loss[loss=0.1421, simple_loss=0.2194, pruned_loss=0.03238, over 985939.31 frames.], batch size: 91, lr: 3.52e-04
+2022-06-19 03:21:01,185 INFO [train.py:874] (2/4) Epoch 23, batch 4150, datatang_loss[loss=0.1501, simple_loss=0.2182, pruned_loss=0.04105, over 4969.00 frames.], tot_loss[loss=0.1457, simple_loss=0.2272, pruned_loss=0.03207, over 984876.73 frames.], batch size: 45, aishell_tot_loss[loss=0.148, simple_loss=0.2345, pruned_loss=0.03076, over 984104.36 frames.], datatang_tot_loss[loss=0.1423, simple_loss=0.2195, pruned_loss=0.03256, over 985994.19 frames.], batch size: 45, lr: 3.52e-04
+2022-06-19 03:22:22,965 INFO [train.py:874] (2/4) Epoch 24, batch 50, aishell_loss[loss=0.1152, simple_loss=0.2019, pruned_loss=0.0143, over 4976.00 frames.], tot_loss[loss=0.1396, simple_loss=0.2206, pruned_loss=0.02935, over 218325.92 frames.], batch size: 30, aishell_tot_loss[loss=0.1458, simple_loss=0.2315, pruned_loss=0.03009, over 124517.96 frames.], datatang_tot_loss[loss=0.1324, simple_loss=0.2078, pruned_loss=0.02852, over 107385.41 frames.], batch size: 30, lr: 3.45e-04
+2022-06-19 03:22:55,070 INFO [train.py:874] (2/4) Epoch 24, batch 100, datatang_loss[loss=0.1283, simple_loss=0.2116, pruned_loss=0.02252, over 4969.00 frames.], tot_loss[loss=0.1413, simple_loss=0.2235, pruned_loss=0.02951, over 388510.99 frames.], batch size: 60, aishell_tot_loss[loss=0.1472, simple_loss=0.2338, pruned_loss=0.03032, over 237399.33 frames.], datatang_tot_loss[loss=0.1335, simple_loss=0.2103, pruned_loss=0.02837, over 199039.70 frames.], batch size: 60, lr: 3.45e-04
+2022-06-19 03:23:26,188 INFO [train.py:874] (2/4) Epoch 24, batch 150, aishell_loss[loss=0.1185, simple_loss=0.1955, pruned_loss=0.02075, over 4792.00 frames.], tot_loss[loss=0.1412, simple_loss=0.2236, pruned_loss=0.02941, over 520638.97 frames.], batch size: 24, aishell_tot_loss[loss=0.1462, simple_loss=0.2329, pruned_loss=0.02973, over 335091.52 frames.], datatang_tot_loss[loss=0.1348, simple_loss=0.2117, pruned_loss=0.02896, over 281218.93 frames.], batch size: 24, lr: 3.45e-04
+2022-06-19 03:23:59,934 INFO [train.py:874] (2/4) Epoch 24, batch 200, datatang_loss[loss=0.1146, simple_loss=0.1899, pruned_loss=0.01968, over 4891.00 frames.], tot_loss[loss=0.1415, simple_loss=0.2234, pruned_loss=0.02976, over 623595.09 frames.], batch size: 52, aishell_tot_loss[loss=0.1472, simple_loss=0.234, pruned_loss=0.03014, over 408670.00 frames.], datatang_tot_loss[loss=0.1349, simple_loss=0.2114, pruned_loss=0.02919, over 367310.62 frames.], batch size: 52, lr: 3.45e-04
+2022-06-19 03:24:28,464 INFO [train.py:874] (2/4) Epoch 24, batch 250, aishell_loss[loss=0.1397, simple_loss=0.2264, pruned_loss=0.0265, over 4929.00 frames.], tot_loss[loss=0.1421, simple_loss=0.2238, pruned_loss=0.03022, over 703433.83 frames.], batch size: 49, aishell_tot_loss[loss=0.1472, simple_loss=0.2332, pruned_loss=0.03058, over 483855.45 frames.], datatang_tot_loss[loss=0.1358, simple_loss=0.2126, pruned_loss=0.02953, over 431806.10 frames.], batch size: 49, lr: 3.44e-04
+2022-06-19 03:25:02,110 INFO [train.py:874] (2/4) Epoch 24, batch 300, datatang_loss[loss=0.1425, simple_loss=0.2298, pruned_loss=0.02759, over 4955.00 frames.], tot_loss[loss=0.1426, simple_loss=0.224, pruned_loss=0.0306, over 765825.48 frames.], batch size: 91, aishell_tot_loss[loss=0.1476, simple_loss=0.2337, pruned_loss=0.03075, over 526689.92 frames.], datatang_tot_loss[loss=0.1372, simple_loss=0.2141, pruned_loss=0.03011, over 514161.05 frames.], batch size: 91, lr: 3.44e-04
+2022-06-19 03:25:34,574 INFO [train.py:874] (2/4) Epoch 24, batch 350, aishell_loss[loss=0.1668, simple_loss=0.259, pruned_loss=0.03734, over 4949.00 frames.], tot_loss[loss=0.142, simple_loss=0.2236, pruned_loss=0.03026, over 814076.53 frames.], batch size: 56, aishell_tot_loss[loss=0.1467, simple_loss=0.2328, pruned_loss=0.03034, over 582274.22 frames.], datatang_tot_loss[loss=0.1372, simple_loss=0.2142, pruned_loss=0.0301, over 567600.90 frames.], batch size: 56, lr: 3.44e-04
+2022-06-19 03:26:05,101 INFO [train.py:874] (2/4) Epoch 24, batch 400, datatang_loss[loss=0.1432, simple_loss=0.2135, pruned_loss=0.03643, over 4913.00 frames.], tot_loss[loss=0.1418, simple_loss=0.2233, pruned_loss=0.03014, over 852008.85 frames.], batch size: 75, aishell_tot_loss[loss=0.1469, simple_loss=0.233, pruned_loss=0.03041, over 622245.90 frames.], datatang_tot_loss[loss=0.137, simple_loss=0.2142, pruned_loss=0.02988, over 624362.03 frames.], batch size: 75, lr: 3.44e-04
+2022-06-19 03:26:39,127 INFO [train.py:874] (2/4) Epoch 24, batch 450, datatang_loss[loss=0.1216, simple_loss=0.1853, pruned_loss=0.02898, over 4937.00 frames.], tot_loss[loss=0.1418, simple_loss=0.2228, pruned_loss=0.03045, over 881631.17 frames.], batch size: 34, aishell_tot_loss[loss=0.146, simple_loss=0.2316, pruned_loss=0.03015, over 653504.34 frames.], datatang_tot_loss[loss=0.1384, simple_loss=0.2156, pruned_loss=0.03055, over 677996.67 frames.], batch size: 34, lr: 3.44e-04
+2022-06-19 03:27:12,087 INFO [train.py:874] (2/4) Epoch 24, batch 500, aishell_loss[loss=0.1587, simple_loss=0.2497, pruned_loss=0.03383, over 4911.00 frames.], tot_loss[loss=0.1422, simple_loss=0.2238, pruned_loss=0.0303, over 904965.87 frames.], batch size: 78, aishell_tot_loss[loss=0.1459, simple_loss=0.2319, pruned_loss=0.02995, over 698529.09 frames.], datatang_tot_loss[loss=0.1386, simple_loss=0.216, pruned_loss=0.03061, over 708940.09 frames.], batch size: 78, lr: 3.44e-04
+2022-06-19 03:27:40,939 INFO [train.py:874] (2/4) Epoch 24, batch 550, datatang_loss[loss=0.1525, simple_loss=0.2287, pruned_loss=0.0381, over 4964.00 frames.], tot_loss[loss=0.1431, simple_loss=0.2246, pruned_loss=0.03079, over 923130.77 frames.], batch size: 55, aishell_tot_loss[loss=0.1462, simple_loss=0.2323, pruned_loss=0.03007, over 733627.86 frames.], datatang_tot_loss[loss=0.1394, simple_loss=0.2165, pruned_loss=0.03112, over 740593.76 frames.], batch size: 55, lr: 3.44e-04
+2022-06-19 03:28:15,527 INFO [train.py:874] (2/4) Epoch 24, batch 600, datatang_loss[loss=0.1215, simple_loss=0.2062, pruned_loss=0.01839, over 4904.00 frames.], tot_loss[loss=0.1425, simple_loss=0.224, pruned_loss=0.03055, over 936669.18 frames.], batch size: 64, aishell_tot_loss[loss=0.1451, simple_loss=0.2311, pruned_loss=0.02956, over 762957.02 frames.], datatang_tot_loss[loss=0.1398, simple_loss=0.2169, pruned_loss=0.03136, over 769446.07 frames.], batch size: 64, lr: 3.44e-04
+2022-06-19 03:28:47,967 INFO [train.py:874] (2/4) Epoch 24, batch 650, aishell_loss[loss=0.1351, simple_loss=0.2189, pruned_loss=0.02561, over 4979.00 frames.], tot_loss[loss=0.1434, simple_loss=0.2253, pruned_loss=0.03072, over 947695.15 frames.], batch size: 39, aishell_tot_loss[loss=0.1458, simple_loss=0.232, pruned_loss=0.02984, over 795957.75 frames.], datatang_tot_loss[loss=0.14, simple_loss=0.2172, pruned_loss=0.03139, over 788286.12 frames.], batch size: 39, lr: 3.44e-04
+2022-06-19 03:29:24,091 INFO [train.py:874] (2/4) Epoch 24, batch 700, aishell_loss[loss=0.144, simple_loss=0.2322, pruned_loss=0.0279, over 4959.00 frames.], tot_loss[loss=0.1431, simple_loss=0.225, pruned_loss=0.03063, over 955995.47 frames.], batch size: 56, aishell_tot_loss[loss=0.1452, simple_loss=0.2313, pruned_loss=0.02956, over 818787.63 frames.], datatang_tot_loss[loss=0.1403, simple_loss=0.2175, pruned_loss=0.03158, over 810887.61 frames.], batch size: 56, lr: 3.44e-04
+2022-06-19 03:29:56,689 INFO [train.py:874] (2/4) Epoch 24, batch 750, aishell_loss[loss=0.1451, simple_loss=0.2417, pruned_loss=0.02423, over 4932.00 frames.], tot_loss[loss=0.1435, simple_loss=0.2248, pruned_loss=0.03104, over 962852.09 frames.], batch size: 68, aishell_tot_loss[loss=0.1452, simple_loss=0.2313, pruned_loss=0.02953, over 834709.69 frames.], datatang_tot_loss[loss=0.141, simple_loss=0.2179, pruned_loss=0.03207, over 835581.56 frames.], batch size: 68, lr: 3.44e-04
+2022-06-19 03:30:29,821 INFO [train.py:874] (2/4) Epoch 24, batch 800, datatang_loss[loss=0.1369, simple_loss=0.2155, pruned_loss=0.02911, over 4929.00 frames.], tot_loss[loss=0.1446, simple_loss=0.2265, pruned_loss=0.0313, over 968014.86 frames.], batch size: 62, aishell_tot_loss[loss=0.146, simple_loss=0.2322, pruned_loss=0.0299, over 855822.31 frames.], datatang_tot_loss[loss=0.1415, simple_loss=0.2187, pruned_loss=0.03215, over 849972.76 frames.], batch size: 62, lr: 3.44e-04
+2022-06-19 03:30:59,441 INFO [train.py:874] (2/4) Epoch 24, batch 850, aishell_loss[loss=0.1722, simple_loss=0.2536, pruned_loss=0.04543, over 4922.00 frames.], tot_loss[loss=0.1454, simple_loss=0.2271, pruned_loss=0.03189, over 971874.10 frames.], batch size: 46, aishell_tot_loss[loss=0.1464, simple_loss=0.2322, pruned_loss=0.03029, over 870393.36 frames.], datatang_tot_loss[loss=0.1424, simple_loss=0.2197, pruned_loss=0.03251, over 866649.25 frames.], batch size: 46, lr: 3.43e-04
+2022-06-19 03:31:31,307 INFO [train.py:874] (2/4) Epoch 24, batch 900, aishell_loss[loss=0.1219, simple_loss=0.1924, pruned_loss=0.02573, over 4957.00 frames.], tot_loss[loss=0.1444, simple_loss=0.2261, pruned_loss=0.0313, over 974502.70 frames.], batch size: 25, aishell_tot_loss[loss=0.1465, simple_loss=0.2325, pruned_loss=0.03024, over 884789.38 frames.], datatang_tot_loss[loss=0.1412, simple_loss=0.2184, pruned_loss=0.032, over 879341.18 frames.], batch size: 25, lr: 3.43e-04
+2022-06-19 03:31:59,208 INFO [train.py:874] (2/4) Epoch 24, batch 950, aishell_loss[loss=0.1466, simple_loss=0.2336, pruned_loss=0.02975, over 4921.00 frames.], tot_loss[loss=0.1448, simple_loss=0.227, pruned_loss=0.03133, over 976672.66 frames.], batch size: 33, aishell_tot_loss[loss=0.147, simple_loss=0.2335, pruned_loss=0.03026, over 896937.44 frames.], datatang_tot_loss[loss=0.1413, simple_loss=0.2185, pruned_loss=0.03206, over 891265.19 frames.], batch size: 33, lr: 3.43e-04
+2022-06-19 03:32:28,606 INFO [train.py:874] (2/4) Epoch 24, batch 1000, datatang_loss[loss=0.121, simple_loss=0.1982, pruned_loss=0.02189, over 4894.00 frames.], tot_loss[loss=0.145, simple_loss=0.2272, pruned_loss=0.03138, over 978662.56 frames.], batch size: 52, aishell_tot_loss[loss=0.147, simple_loss=0.2336, pruned_loss=0.03021, over 907449.41 frames.], datatang_tot_loss[loss=0.1417, simple_loss=0.2189, pruned_loss=0.03222, over 902314.88 frames.], batch size: 52, lr: 3.43e-04
+2022-06-19 03:32:28,606 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 03:32:45,104 INFO [train.py:914] (2/4) Epoch 24, validation: loss=0.1641, simple_loss=0.2485, pruned_loss=0.03987, over 1622729.00 frames.
+2022-06-19 03:33:11,281 INFO [train.py:874] (2/4) Epoch 24, batch 1050, aishell_loss[loss=0.1479, simple_loss=0.2396, pruned_loss=0.0281, over 4976.00 frames.], tot_loss[loss=0.1447, simple_loss=0.227, pruned_loss=0.03117, over 980869.33 frames.], batch size: 39, aishell_tot_loss[loss=0.1466, simple_loss=0.2332, pruned_loss=0.02999, over 918860.60 frames.], datatang_tot_loss[loss=0.1418, simple_loss=0.219, pruned_loss=0.0323, over 910470.41 frames.], batch size: 39, lr: 3.43e-04
+2022-06-19 03:33:42,075 INFO [train.py:874] (2/4) Epoch 24, batch 1100, aishell_loss[loss=0.1758, simple_loss=0.263, pruned_loss=0.04434, over 4916.00 frames.], tot_loss[loss=0.1449, simple_loss=0.2274, pruned_loss=0.03114, over 981894.07 frames.], batch size: 78, aishell_tot_loss[loss=0.147, simple_loss=0.2339, pruned_loss=0.03005, over 926957.48 frames.], datatang_tot_loss[loss=0.1417, simple_loss=0.219, pruned_loss=0.03221, over 919052.94 frames.], batch size: 78, lr: 3.43e-04
+2022-06-19 03:34:11,131 INFO [train.py:874] (2/4) Epoch 24, batch 1150, aishell_loss[loss=0.1612, simple_loss=0.2442, pruned_loss=0.03908, over 4882.00 frames.], tot_loss[loss=0.1455, simple_loss=0.2282, pruned_loss=0.03143, over 982860.91 frames.], batch size: 47, aishell_tot_loss[loss=0.1475, simple_loss=0.2344, pruned_loss=0.0303, over 934832.93 frames.], datatang_tot_loss[loss=0.142, simple_loss=0.2193, pruned_loss=0.03234, over 925923.87 frames.], batch size: 47, lr: 3.43e-04
+2022-06-19 03:34:38,365 INFO [train.py:874] (2/4) Epoch 24, batch 1200, datatang_loss[loss=0.1437, simple_loss=0.2246, pruned_loss=0.03139, over 4919.00 frames.], tot_loss[loss=0.1445, simple_loss=0.2268, pruned_loss=0.03109, over 983831.57 frames.], batch size: 81, aishell_tot_loss[loss=0.1468, simple_loss=0.2335, pruned_loss=0.03003, over 940166.73 frames.], datatang_tot_loss[loss=0.1418, simple_loss=0.2192, pruned_loss=0.03223, over 934129.75 frames.], batch size: 81, lr: 3.43e-04
+2022-06-19 03:35:08,060 INFO [train.py:874] (2/4) Epoch 24, batch 1250, datatang_loss[loss=0.1566, simple_loss=0.2129, pruned_loss=0.05013, over 4971.00 frames.], tot_loss[loss=0.1445, simple_loss=0.2264, pruned_loss=0.03127, over 984433.62 frames.], batch size: 26, aishell_tot_loss[loss=0.1475, simple_loss=0.2341, pruned_loss=0.03041, over 944262.47 frames.], datatang_tot_loss[loss=0.1414, simple_loss=0.2187, pruned_loss=0.03198, over 941775.40 frames.], batch size: 26, lr: 3.43e-04
+2022-06-19 03:35:37,361 INFO [train.py:874] (2/4) Epoch 24, batch 1300, datatang_loss[loss=0.1512, simple_loss=0.2286, pruned_loss=0.03685, over 4930.00 frames.], tot_loss[loss=0.1444, simple_loss=0.2262, pruned_loss=0.03129, over 984503.04 frames.], batch size: 62, aishell_tot_loss[loss=0.1468, simple_loss=0.2334, pruned_loss=0.03013, over 948723.92 frames.], datatang_tot_loss[loss=0.142, simple_loss=0.2193, pruned_loss=0.03229, over 947153.01 frames.], batch size: 62, lr: 3.43e-04
+2022-06-19 03:36:04,181 INFO [train.py:874] (2/4) Epoch 24, batch 1350, aishell_loss[loss=0.1728, simple_loss=0.2584, pruned_loss=0.04362, over 4909.00 frames.], tot_loss[loss=0.1446, simple_loss=0.2271, pruned_loss=0.03103, over 985034.41 frames.], batch size: 78, aishell_tot_loss[loss=0.1467, simple_loss=0.2338, pruned_loss=0.0298, over 953743.74 frames.], datatang_tot_loss[loss=0.1422, simple_loss=0.2196, pruned_loss=0.0324, over 951284.54 frames.], batch size: 78, lr: 3.43e-04
+2022-06-19 03:36:34,124 INFO [train.py:874] (2/4) Epoch 24, batch 1400, aishell_loss[loss=0.1692, simple_loss=0.2599, pruned_loss=0.03923, over 4923.00 frames.], tot_loss[loss=0.145, simple_loss=0.2275, pruned_loss=0.03128, over 984992.27 frames.], batch size: 33, aishell_tot_loss[loss=0.1473, simple_loss=0.2345, pruned_loss=0.03004, over 957077.04 frames.], datatang_tot_loss[loss=0.1421, simple_loss=0.2195, pruned_loss=0.03237, over 955580.05 frames.], batch size: 33, lr: 3.42e-04
+2022-06-19 03:37:01,618 INFO [train.py:874] (2/4) Epoch 24, batch 1450, datatang_loss[loss=0.1507, simple_loss=0.2356, pruned_loss=0.03285, over 4957.00 frames.], tot_loss[loss=0.1454, simple_loss=0.2279, pruned_loss=0.03146, over 984861.91 frames.], batch size: 99, aishell_tot_loss[loss=0.1475, simple_loss=0.2347, pruned_loss=0.03017, over 960368.19 frames.], datatang_tot_loss[loss=0.1423, simple_loss=0.2198, pruned_loss=0.03245, over 958899.31 frames.], batch size: 99, lr: 3.42e-04
+2022-06-19 03:37:30,958 INFO [train.py:874] (2/4) Epoch 24, batch 1500, aishell_loss[loss=0.1507, simple_loss=0.2371, pruned_loss=0.03216, over 4914.00 frames.], tot_loss[loss=0.1451, simple_loss=0.2273, pruned_loss=0.03148, over 984498.00 frames.], batch size: 41, aishell_tot_loss[loss=0.1477, simple_loss=0.2348, pruned_loss=0.03033, over 962744.06 frames.], datatang_tot_loss[loss=0.1421, simple_loss=0.2195, pruned_loss=0.03233, over 962090.25 frames.], batch size: 41, lr: 3.42e-04
+2022-06-19 03:38:00,676 INFO [train.py:874] (2/4) Epoch 24, batch 1550, datatang_loss[loss=0.136, simple_loss=0.2092, pruned_loss=0.03147, over 4910.00 frames.], tot_loss[loss=0.1449, simple_loss=0.2268, pruned_loss=0.03151, over 984731.38 frames.], batch size: 47, aishell_tot_loss[loss=0.1479, simple_loss=0.2348, pruned_loss=0.03049, over 965280.49 frames.], datatang_tot_loss[loss=0.1417, simple_loss=0.219, pruned_loss=0.0322, over 965004.67 frames.], batch size: 47, lr: 3.42e-04
+2022-06-19 03:38:29,166 INFO [train.py:874] (2/4) Epoch 24, batch 1600, datatang_loss[loss=0.136, simple_loss=0.2114, pruned_loss=0.03031, over 4945.00 frames.], tot_loss[loss=0.1451, simple_loss=0.2272, pruned_loss=0.03149, over 985120.34 frames.], batch size: 69, aishell_tot_loss[loss=0.1475, simple_loss=0.2346, pruned_loss=0.03019, over 967700.54 frames.], datatang_tot_loss[loss=0.1424, simple_loss=0.2197, pruned_loss=0.03252, over 967610.77 frames.], batch size: 69, lr: 3.42e-04
+2022-06-19 03:38:56,950 INFO [train.py:874] (2/4) Epoch 24, batch 1650, aishell_loss[loss=0.1613, simple_loss=0.2455, pruned_loss=0.03855, over 4857.00 frames.], tot_loss[loss=0.1454, simple_loss=0.2275, pruned_loss=0.03168, over 984805.56 frames.], batch size: 37, aishell_tot_loss[loss=0.1478, simple_loss=0.2349, pruned_loss=0.03036, over 969580.44 frames.], datatang_tot_loss[loss=0.1424, simple_loss=0.2197, pruned_loss=0.03259, over 969512.52 frames.], batch size: 37, lr: 3.42e-04
+2022-06-19 03:39:27,572 INFO [train.py:874] (2/4) Epoch 24, batch 1700, datatang_loss[loss=0.1288, simple_loss=0.2126, pruned_loss=0.02245, over 4922.00 frames.], tot_loss[loss=0.1453, simple_loss=0.227, pruned_loss=0.03179, over 985219.68 frames.], batch size: 77, aishell_tot_loss[loss=0.1479, simple_loss=0.2348, pruned_loss=0.0305, over 970814.44 frames.], datatang_tot_loss[loss=0.1425, simple_loss=0.2198, pruned_loss=0.03257, over 972283.26 frames.], batch size: 77, lr: 3.42e-04
+2022-06-19 03:39:57,074 INFO [train.py:874] (2/4) Epoch 24, batch 1750, aishell_loss[loss=0.1576, simple_loss=0.2536, pruned_loss=0.03081, over 4970.00 frames.], tot_loss[loss=0.1449, simple_loss=0.2263, pruned_loss=0.03171, over 985462.91 frames.], batch size: 69, aishell_tot_loss[loss=0.1479, simple_loss=0.2346, pruned_loss=0.03056, over 972083.68 frames.], datatang_tot_loss[loss=0.1423, simple_loss=0.2197, pruned_loss=0.03242, over 974426.82 frames.], batch size: 69, lr: 3.42e-04
+2022-06-19 03:40:24,390 INFO [train.py:874] (2/4) Epoch 24, batch 1800, datatang_loss[loss=0.1179, simple_loss=0.1936, pruned_loss=0.0211, over 4920.00 frames.], tot_loss[loss=0.1449, simple_loss=0.2264, pruned_loss=0.03176, over 985513.58 frames.], batch size: 75, aishell_tot_loss[loss=0.148, simple_loss=0.2347, pruned_loss=0.03059, over 973268.34 frames.], datatang_tot_loss[loss=0.1424, simple_loss=0.2198, pruned_loss=0.03247, over 976142.13 frames.], batch size: 75, lr: 3.42e-04
+2022-06-19 03:40:54,927 INFO [train.py:874] (2/4) Epoch 24, batch 1850, datatang_loss[loss=0.1238, simple_loss=0.2052, pruned_loss=0.02117, over 4932.00 frames.], tot_loss[loss=0.1448, simple_loss=0.2261, pruned_loss=0.03168, over 984838.13 frames.], batch size: 73, aishell_tot_loss[loss=0.1481, simple_loss=0.2349, pruned_loss=0.0307, over 974286.37 frames.], datatang_tot_loss[loss=0.142, simple_loss=0.2193, pruned_loss=0.03233, over 976968.19 frames.], batch size: 73, lr: 3.42e-04
+2022-06-19 03:41:25,559 INFO [train.py:874] (2/4) Epoch 24, batch 1900, datatang_loss[loss=0.1412, simple_loss=0.218, pruned_loss=0.03221, over 4906.00 frames.], tot_loss[loss=0.1454, simple_loss=0.2273, pruned_loss=0.03171, over 985074.42 frames.], batch size: 47, aishell_tot_loss[loss=0.1485, simple_loss=0.2356, pruned_loss=0.03072, over 975656.72 frames.], datatang_tot_loss[loss=0.1421, simple_loss=0.2194, pruned_loss=0.0324, over 978039.85 frames.], batch size: 47, lr: 3.42e-04
+2022-06-19 03:41:52,661 INFO [train.py:874] (2/4) Epoch 24, batch 1950, datatang_loss[loss=0.1511, simple_loss=0.223, pruned_loss=0.03966, over 4932.00 frames.], tot_loss[loss=0.1454, simple_loss=0.2272, pruned_loss=0.03181, over 985311.53 frames.], batch size: 57, aishell_tot_loss[loss=0.1487, simple_loss=0.2359, pruned_loss=0.03077, over 976502.20 frames.], datatang_tot_loss[loss=0.1422, simple_loss=0.2194, pruned_loss=0.03244, over 979289.08 frames.], batch size: 57, lr: 3.41e-04
+2022-06-19 03:42:22,512 INFO [train.py:874] (2/4) Epoch 24, batch 2000, datatang_loss[loss=0.1454, simple_loss=0.2272, pruned_loss=0.03178, over 4926.00 frames.], tot_loss[loss=0.1456, simple_loss=0.2273, pruned_loss=0.03188, over 985323.75 frames.], batch size: 57, aishell_tot_loss[loss=0.1487, simple_loss=0.236, pruned_loss=0.03072, over 977226.18 frames.], datatang_tot_loss[loss=0.1424, simple_loss=0.2197, pruned_loss=0.03259, over 980263.12 frames.], batch size: 57, lr: 3.41e-04
+2022-06-19 03:42:22,513 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 03:42:39,280 INFO [train.py:914] (2/4) Epoch 24, validation: loss=0.1651, simple_loss=0.2488, pruned_loss=0.04067, over 1622729.00 frames.
+2022-06-19 03:43:08,908 INFO [train.py:874] (2/4) Epoch 24, batch 2050, aishell_loss[loss=0.1745, simple_loss=0.2523, pruned_loss=0.04829, over 4982.00 frames.], tot_loss[loss=0.1462, simple_loss=0.2278, pruned_loss=0.03234, over 985405.54 frames.], batch size: 39, aishell_tot_loss[loss=0.1494, simple_loss=0.2362, pruned_loss=0.03134, over 978221.65 frames.], datatang_tot_loss[loss=0.1424, simple_loss=0.2197, pruned_loss=0.03251, over 980925.39 frames.], batch size: 39, lr: 3.41e-04
+2022-06-19 03:43:38,477 INFO [train.py:874] (2/4) Epoch 24, batch 2100, datatang_loss[loss=0.1385, simple_loss=0.2188, pruned_loss=0.02907, over 4883.00 frames.], tot_loss[loss=0.146, simple_loss=0.2278, pruned_loss=0.03213, over 985435.29 frames.], batch size: 47, aishell_tot_loss[loss=0.1494, simple_loss=0.2362, pruned_loss=0.03128, over 979179.97 frames.], datatang_tot_loss[loss=0.1424, simple_loss=0.2199, pruned_loss=0.03244, over 981349.35 frames.], batch size: 47, lr: 3.41e-04
+2022-06-19 03:44:05,029 INFO [train.py:874] (2/4) Epoch 24, batch 2150, aishell_loss[loss=0.1446, simple_loss=0.227, pruned_loss=0.03111, over 4939.00 frames.], tot_loss[loss=0.1457, simple_loss=0.2275, pruned_loss=0.0319, over 985747.18 frames.], batch size: 31, aishell_tot_loss[loss=0.149, simple_loss=0.2358, pruned_loss=0.03111, over 980091.37 frames.], datatang_tot_loss[loss=0.1424, simple_loss=0.2201, pruned_loss=0.03241, over 981974.81 frames.], batch size: 31, lr: 3.41e-04
+2022-06-19 03:44:35,445 INFO [train.py:874] (2/4) Epoch 24, batch 2200, aishell_loss[loss=0.1613, simple_loss=0.2528, pruned_loss=0.03486, over 4954.00 frames.], tot_loss[loss=0.1448, simple_loss=0.2269, pruned_loss=0.03132, over 985722.51 frames.], batch size: 56, aishell_tot_loss[loss=0.1483, simple_loss=0.2353, pruned_loss=0.03069, over 980822.87 frames.], datatang_tot_loss[loss=0.142, simple_loss=0.2195, pruned_loss=0.03226, over 982350.89 frames.], batch size: 56, lr: 3.41e-04
+2022-06-19 03:45:05,761 INFO [train.py:874] (2/4) Epoch 24, batch 2250, datatang_loss[loss=0.1384, simple_loss=0.2181, pruned_loss=0.02937, over 4926.00 frames.], tot_loss[loss=0.145, simple_loss=0.227, pruned_loss=0.03147, over 985681.21 frames.], batch size: 71, aishell_tot_loss[loss=0.1484, simple_loss=0.2354, pruned_loss=0.03065, over 981370.24 frames.], datatang_tot_loss[loss=0.1422, simple_loss=0.2196, pruned_loss=0.03238, over 982712.26 frames.], batch size: 71, lr: 3.41e-04
+2022-06-19 03:45:33,372 INFO [train.py:874] (2/4) Epoch 24, batch 2300, aishell_loss[loss=0.1453, simple_loss=0.2383, pruned_loss=0.02615, over 4951.00 frames.], tot_loss[loss=0.1448, simple_loss=0.2267, pruned_loss=0.03147, over 985399.96 frames.], batch size: 45, aishell_tot_loss[loss=0.1483, simple_loss=0.2354, pruned_loss=0.03065, over 981526.89 frames.], datatang_tot_loss[loss=0.1421, simple_loss=0.2195, pruned_loss=0.03233, over 983087.80 frames.], batch size: 45, lr: 3.41e-04
+2022-06-19 03:46:03,314 INFO [train.py:874] (2/4) Epoch 24, batch 2350, datatang_loss[loss=0.136, simple_loss=0.2139, pruned_loss=0.02905, over 4954.00 frames.], tot_loss[loss=0.145, simple_loss=0.2269, pruned_loss=0.0315, over 985831.62 frames.], batch size: 45, aishell_tot_loss[loss=0.1483, simple_loss=0.2354, pruned_loss=0.0306, over 982257.06 frames.], datatang_tot_loss[loss=0.1422, simple_loss=0.2196, pruned_loss=0.0324, over 983546.11 frames.], batch size: 45, lr: 3.41e-04
+2022-06-19 03:46:33,437 INFO [train.py:874] (2/4) Epoch 24, batch 2400, datatang_loss[loss=0.1371, simple_loss=0.2167, pruned_loss=0.02868, over 4953.00 frames.], tot_loss[loss=0.144, simple_loss=0.2261, pruned_loss=0.0309, over 985479.62 frames.], batch size: 69, aishell_tot_loss[loss=0.1481, simple_loss=0.2352, pruned_loss=0.03047, over 982563.09 frames.], datatang_tot_loss[loss=0.1414, simple_loss=0.219, pruned_loss=0.03189, over 983541.62 frames.], batch size: 69, lr: 3.41e-04
+2022-06-19 03:46:59,464 INFO [train.py:874] (2/4) Epoch 24, batch 2450, datatang_loss[loss=0.1414, simple_loss=0.2269, pruned_loss=0.02793, over 4946.00 frames.], tot_loss[loss=0.1435, simple_loss=0.2259, pruned_loss=0.03055, over 985396.77 frames.], batch size: 88, aishell_tot_loss[loss=0.1478, simple_loss=0.2349, pruned_loss=0.03029, over 982769.87 frames.], datatang_tot_loss[loss=0.141, simple_loss=0.2188, pruned_loss=0.03164, over 983826.94 frames.], batch size: 88, lr: 3.41e-04
+2022-06-19 03:47:29,041 INFO [train.py:874] (2/4) Epoch 24, batch 2500, aishell_loss[loss=0.1554, simple_loss=0.2395, pruned_loss=0.03567, over 4901.00 frames.], tot_loss[loss=0.143, simple_loss=0.2252, pruned_loss=0.03046, over 985408.56 frames.], batch size: 34, aishell_tot_loss[loss=0.1473, simple_loss=0.2343, pruned_loss=0.03016, over 982968.87 frames.], datatang_tot_loss[loss=0.1409, simple_loss=0.2186, pruned_loss=0.03156, over 984121.29 frames.], batch size: 34, lr: 3.41e-04
+2022-06-19 03:47:57,736 INFO [train.py:874] (2/4) Epoch 24, batch 2550, aishell_loss[loss=0.1571, simple_loss=0.2438, pruned_loss=0.03521, over 4938.00 frames.], tot_loss[loss=0.1431, simple_loss=0.2257, pruned_loss=0.03028, over 985265.27 frames.], batch size: 58, aishell_tot_loss[loss=0.1471, simple_loss=0.2342, pruned_loss=0.02999, over 983129.52 frames.], datatang_tot_loss[loss=0.1407, simple_loss=0.2184, pruned_loss=0.03149, over 984296.38 frames.], batch size: 58, lr: 3.40e-04
+2022-06-19 03:48:24,737 INFO [train.py:874] (2/4) Epoch 24, batch 2600, aishell_loss[loss=0.1281, simple_loss=0.2178, pruned_loss=0.01915, over 4972.00 frames.], tot_loss[loss=0.1432, simple_loss=0.2258, pruned_loss=0.03028, over 985470.44 frames.], batch size: 44, aishell_tot_loss[loss=0.1465, simple_loss=0.2336, pruned_loss=0.02974, over 983463.55 frames.], datatang_tot_loss[loss=0.1411, simple_loss=0.2187, pruned_loss=0.03168, over 984551.35 frames.], batch size: 44, lr: 3.40e-04
+2022-06-19 03:48:54,375 INFO [train.py:874] (2/4) Epoch 24, batch 2650, aishell_loss[loss=0.1355, simple_loss=0.2249, pruned_loss=0.02306, over 4885.00 frames.], tot_loss[loss=0.1428, simple_loss=0.2252, pruned_loss=0.03016, over 985263.35 frames.], batch size: 34, aishell_tot_loss[loss=0.1463, simple_loss=0.2333, pruned_loss=0.0297, over 983339.71 frames.], datatang_tot_loss[loss=0.1407, simple_loss=0.2184, pruned_loss=0.03148, over 984779.44 frames.], batch size: 34, lr: 3.40e-04
+2022-06-19 03:49:23,859 INFO [train.py:874] (2/4) Epoch 24, batch 2700, datatang_loss[loss=0.1452, simple_loss=0.2192, pruned_loss=0.03561, over 4905.00 frames.], tot_loss[loss=0.1428, simple_loss=0.225, pruned_loss=0.03028, over 985542.83 frames.], batch size: 77, aishell_tot_loss[loss=0.1461, simple_loss=0.233, pruned_loss=0.02958, over 983873.90 frames.], datatang_tot_loss[loss=0.1408, simple_loss=0.2185, pruned_loss=0.03157, over 984806.48 frames.], batch size: 77, lr: 3.40e-04
+2022-06-19 03:49:50,611 INFO [train.py:874] (2/4) Epoch 24, batch 2750, aishell_loss[loss=0.1442, simple_loss=0.2192, pruned_loss=0.03462, over 4924.00 frames.], tot_loss[loss=0.1449, simple_loss=0.2272, pruned_loss=0.03127, over 985824.41 frames.], batch size: 33, aishell_tot_loss[loss=0.1476, simple_loss=0.2344, pruned_loss=0.03037, over 984192.46 frames.], datatang_tot_loss[loss=0.1412, simple_loss=0.2188, pruned_loss=0.03182, over 985094.91 frames.], batch size: 33, lr: 3.40e-04
+2022-06-19 03:50:20,308 INFO [train.py:874] (2/4) Epoch 24, batch 2800, aishell_loss[loss=0.1368, simple_loss=0.2312, pruned_loss=0.02115, over 4883.00 frames.], tot_loss[loss=0.1442, simple_loss=0.2267, pruned_loss=0.0309, over 985747.91 frames.], batch size: 42, aishell_tot_loss[loss=0.1475, simple_loss=0.2344, pruned_loss=0.03026, over 984553.96 frames.], datatang_tot_loss[loss=0.1408, simple_loss=0.2185, pruned_loss=0.03157, over 984939.24 frames.], batch size: 42, lr: 3.40e-04
+2022-06-19 03:50:47,473 INFO [train.py:874] (2/4) Epoch 24, batch 2850, aishell_loss[loss=0.1684, simple_loss=0.2596, pruned_loss=0.03864, over 4898.00 frames.], tot_loss[loss=0.1439, simple_loss=0.2257, pruned_loss=0.03102, over 985944.43 frames.], batch size: 34, aishell_tot_loss[loss=0.1468, simple_loss=0.2337, pruned_loss=0.02998, over 984689.21 frames.], datatang_tot_loss[loss=0.1412, simple_loss=0.2185, pruned_loss=0.03192, over 985231.55 frames.], batch size: 34, lr: 3.40e-04
+2022-06-19 03:51:16,575 INFO [train.py:874] (2/4) Epoch 24, batch 2900, aishell_loss[loss=0.1479, simple_loss=0.2385, pruned_loss=0.02866, over 4891.00 frames.], tot_loss[loss=0.1434, simple_loss=0.2252, pruned_loss=0.03077, over 985840.01 frames.], batch size: 47, aishell_tot_loss[loss=0.1462, simple_loss=0.233, pruned_loss=0.02968, over 984644.76 frames.], datatang_tot_loss[loss=0.1412, simple_loss=0.2183, pruned_loss=0.032, over 985423.08 frames.], batch size: 47, lr: 3.40e-04
+2022-06-19 03:51:46,002 INFO [train.py:874] (2/4) Epoch 24, batch 2950, datatang_loss[loss=0.1535, simple_loss=0.2299, pruned_loss=0.03854, over 4919.00 frames.], tot_loss[loss=0.1431, simple_loss=0.2249, pruned_loss=0.03069, over 985913.81 frames.], batch size: 75, aishell_tot_loss[loss=0.1462, simple_loss=0.233, pruned_loss=0.02972, over 984865.22 frames.], datatang_tot_loss[loss=0.1407, simple_loss=0.2178, pruned_loss=0.03184, over 985476.28 frames.], batch size: 75, lr: 3.40e-04
+2022-06-19 03:52:13,951 INFO [train.py:874] (2/4) Epoch 24, batch 3000, aishell_loss[loss=0.1359, simple_loss=0.2336, pruned_loss=0.01906, over 4865.00 frames.], tot_loss[loss=0.1436, simple_loss=0.2256, pruned_loss=0.03079, over 986269.85 frames.], batch size: 37, aishell_tot_loss[loss=0.1462, simple_loss=0.2332, pruned_loss=0.02957, over 985147.96 frames.], datatang_tot_loss[loss=0.1412, simple_loss=0.2183, pruned_loss=0.03206, over 985756.48 frames.], batch size: 37, lr: 3.40e-04
+2022-06-19 03:52:13,952 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 03:52:29,516 INFO [train.py:914] (2/4) Epoch 24, validation: loss=0.1647, simple_loss=0.2484, pruned_loss=0.04048, over 1622729.00 frames.
+2022-06-19 03:52:58,013 INFO [train.py:874] (2/4) Epoch 24, batch 3050, aishell_loss[loss=0.1388, simple_loss=0.2326, pruned_loss=0.02253, over 4941.00 frames.], tot_loss[loss=0.1428, simple_loss=0.2248, pruned_loss=0.03045, over 986200.20 frames.], batch size: 64, aishell_tot_loss[loss=0.1458, simple_loss=0.2327, pruned_loss=0.02942, over 985199.36 frames.], datatang_tot_loss[loss=0.1408, simple_loss=0.2178, pruned_loss=0.03187, over 985833.44 frames.], batch size: 64, lr: 3.40e-04
+2022-06-19 03:53:24,993 INFO [train.py:874] (2/4) Epoch 24, batch 3100, aishell_loss[loss=0.1341, simple_loss=0.2185, pruned_loss=0.02489, over 4877.00 frames.], tot_loss[loss=0.1427, simple_loss=0.2244, pruned_loss=0.03048, over 985962.78 frames.], batch size: 28, aishell_tot_loss[loss=0.1458, simple_loss=0.2326, pruned_loss=0.02946, over 985218.25 frames.], datatang_tot_loss[loss=0.1406, simple_loss=0.2176, pruned_loss=0.03182, over 985751.09 frames.], batch size: 28, lr: 3.39e-04
+2022-06-19 03:53:54,429 INFO [train.py:874] (2/4) Epoch 24, batch 3150, aishell_loss[loss=0.1756, simple_loss=0.2496, pruned_loss=0.05079, over 4952.00 frames.], tot_loss[loss=0.1431, simple_loss=0.2251, pruned_loss=0.03058, over 986090.86 frames.], batch size: 31, aishell_tot_loss[loss=0.1457, simple_loss=0.2328, pruned_loss=0.02937, over 985361.86 frames.], datatang_tot_loss[loss=0.1408, simple_loss=0.2177, pruned_loss=0.03201, over 985868.73 frames.], batch size: 31, lr: 3.39e-04
+2022-06-19 03:54:21,501 INFO [train.py:874] (2/4) Epoch 24, batch 3200, datatang_loss[loss=0.1508, simple_loss=0.2337, pruned_loss=0.03397, over 4938.00 frames.], tot_loss[loss=0.1432, simple_loss=0.2255, pruned_loss=0.03045, over 986035.96 frames.], batch size: 88, aishell_tot_loss[loss=0.1455, simple_loss=0.2325, pruned_loss=0.02924, over 985443.56 frames.], datatang_tot_loss[loss=0.141, simple_loss=0.218, pruned_loss=0.032, over 985860.61 frames.], batch size: 88, lr: 3.39e-04
+2022-06-19 03:54:50,560 INFO [train.py:874] (2/4) Epoch 24, batch 3250, datatang_loss[loss=0.1209, simple_loss=0.2005, pruned_loss=0.02063, over 4917.00 frames.], tot_loss[loss=0.1432, simple_loss=0.2254, pruned_loss=0.03045, over 985972.45 frames.], batch size: 75, aishell_tot_loss[loss=0.1456, simple_loss=0.2326, pruned_loss=0.02933, over 985345.12 frames.], datatang_tot_loss[loss=0.1408, simple_loss=0.2178, pruned_loss=0.03189, over 985992.07 frames.], batch size: 75, lr: 3.39e-04
+2022-06-19 03:55:21,753 INFO [train.py:874] (2/4) Epoch 24, batch 3300, aishell_loss[loss=0.1713, simple_loss=0.2531, pruned_loss=0.04478, over 4854.00 frames.], tot_loss[loss=0.1432, simple_loss=0.2258, pruned_loss=0.03033, over 986072.23 frames.], batch size: 36, aishell_tot_loss[loss=0.1457, simple_loss=0.2329, pruned_loss=0.02921, over 985592.89 frames.], datatang_tot_loss[loss=0.1408, simple_loss=0.218, pruned_loss=0.03182, over 985920.72 frames.], batch size: 36, lr: 3.39e-04
+2022-06-19 03:55:47,661 INFO [train.py:874] (2/4) Epoch 24, batch 3350, aishell_loss[loss=0.1531, simple_loss=0.2342, pruned_loss=0.03597, over 4904.00 frames.], tot_loss[loss=0.1443, simple_loss=0.227, pruned_loss=0.03082, over 986389.11 frames.], batch size: 34, aishell_tot_loss[loss=0.1464, simple_loss=0.2336, pruned_loss=0.02958, over 985858.50 frames.], datatang_tot_loss[loss=0.1411, simple_loss=0.2182, pruned_loss=0.03198, over 986066.90 frames.], batch size: 34, lr: 3.39e-04
+2022-06-19 03:56:18,372 INFO [train.py:874] (2/4) Epoch 24, batch 3400, aishell_loss[loss=0.1348, simple_loss=0.2259, pruned_loss=0.02178, over 4912.00 frames.], tot_loss[loss=0.1442, simple_loss=0.2267, pruned_loss=0.03083, over 986600.51 frames.], batch size: 52, aishell_tot_loss[loss=0.1467, simple_loss=0.2339, pruned_loss=0.02974, over 985998.98 frames.], datatang_tot_loss[loss=0.1407, simple_loss=0.2177, pruned_loss=0.03184, over 986261.17 frames.], batch size: 52, lr: 3.39e-04
+2022-06-19 03:56:47,581 INFO [train.py:874] (2/4) Epoch 24, batch 3450, aishell_loss[loss=0.1517, simple_loss=0.242, pruned_loss=0.03071, over 4952.00 frames.], tot_loss[loss=0.1439, simple_loss=0.2258, pruned_loss=0.03095, over 986476.67 frames.], batch size: 64, aishell_tot_loss[loss=0.1466, simple_loss=0.2336, pruned_loss=0.02975, over 985948.64 frames.], datatang_tot_loss[loss=0.1407, simple_loss=0.2176, pruned_loss=0.03192, over 986301.26 frames.], batch size: 64, lr: 3.39e-04
+2022-06-19 03:57:14,728 INFO [train.py:874] (2/4) Epoch 24, batch 3500, aishell_loss[loss=0.1765, simple_loss=0.2517, pruned_loss=0.05064, over 4913.00 frames.], tot_loss[loss=0.1442, simple_loss=0.2256, pruned_loss=0.03142, over 986116.00 frames.], batch size: 33, aishell_tot_loss[loss=0.1472, simple_loss=0.234, pruned_loss=0.03016, over 985941.49 frames.], datatang_tot_loss[loss=0.1406, simple_loss=0.2173, pruned_loss=0.032, over 986020.68 frames.], batch size: 33, lr: 3.39e-04
+2022-06-19 03:57:44,566 INFO [train.py:874] (2/4) Epoch 24, batch 3550, datatang_loss[loss=0.1776, simple_loss=0.2526, pruned_loss=0.05134, over 4927.00 frames.], tot_loss[loss=0.1445, simple_loss=0.2263, pruned_loss=0.03142, over 985916.77 frames.], batch size: 108, aishell_tot_loss[loss=0.1475, simple_loss=0.2346, pruned_loss=0.03023, over 985880.97 frames.], datatang_tot_loss[loss=0.1409, simple_loss=0.218, pruned_loss=0.03193, over 985893.25 frames.], batch size: 108, lr: 3.39e-04
+2022-06-19 03:58:14,376 INFO [train.py:874] (2/4) Epoch 24, batch 3600, datatang_loss[loss=0.1844, simple_loss=0.2429, pruned_loss=0.0629, over 4920.00 frames.], tot_loss[loss=0.1451, simple_loss=0.2265, pruned_loss=0.0319, over 985827.00 frames.], batch size: 81, aishell_tot_loss[loss=0.1476, simple_loss=0.2344, pruned_loss=0.03037, over 985869.29 frames.], datatang_tot_loss[loss=0.1416, simple_loss=0.2185, pruned_loss=0.03235, over 985812.18 frames.], batch size: 81, lr: 3.39e-04
+2022-06-19 03:58:41,412 INFO [train.py:874] (2/4) Epoch 24, batch 3650, aishell_loss[loss=0.1604, simple_loss=0.2529, pruned_loss=0.03396, over 4933.00 frames.], tot_loss[loss=0.1444, simple_loss=0.2261, pruned_loss=0.03133, over 985566.37 frames.], batch size: 49, aishell_tot_loss[loss=0.1473, simple_loss=0.2342, pruned_loss=0.03021, over 985564.43 frames.], datatang_tot_loss[loss=0.1411, simple_loss=0.2181, pruned_loss=0.03204, over 985826.28 frames.], batch size: 49, lr: 3.39e-04
+2022-06-19 03:59:11,567 INFO [train.py:874] (2/4) Epoch 24, batch 3700, aishell_loss[loss=0.1474, simple_loss=0.234, pruned_loss=0.03041, over 4936.00 frames.], tot_loss[loss=0.1441, simple_loss=0.2257, pruned_loss=0.03123, over 985673.13 frames.], batch size: 49, aishell_tot_loss[loss=0.1473, simple_loss=0.2343, pruned_loss=0.0302, over 985526.90 frames.], datatang_tot_loss[loss=0.1409, simple_loss=0.2179, pruned_loss=0.03198, over 985928.24 frames.], batch size: 49, lr: 3.38e-04
+2022-06-19 03:59:40,796 INFO [train.py:874] (2/4) Epoch 24, batch 3750, aishell_loss[loss=0.154, simple_loss=0.2468, pruned_loss=0.0306, over 4914.00 frames.], tot_loss[loss=0.1437, simple_loss=0.2256, pruned_loss=0.0309, over 985858.41 frames.], batch size: 68, aishell_tot_loss[loss=0.1475, simple_loss=0.2345, pruned_loss=0.0302, over 985531.68 frames.], datatang_tot_loss[loss=0.1404, simple_loss=0.2174, pruned_loss=0.03167, over 986132.88 frames.], batch size: 68, lr: 3.38e-04
+2022-06-19 04:00:08,468 INFO [train.py:874] (2/4) Epoch 24, batch 3800, datatang_loss[loss=0.1387, simple_loss=0.2167, pruned_loss=0.03037, over 4919.00 frames.], tot_loss[loss=0.1431, simple_loss=0.225, pruned_loss=0.03057, over 985802.03 frames.], batch size: 64, aishell_tot_loss[loss=0.1471, simple_loss=0.2343, pruned_loss=0.03, over 985522.38 frames.], datatang_tot_loss[loss=0.1401, simple_loss=0.2171, pruned_loss=0.0315, over 986080.50 frames.], batch size: 64, lr: 3.38e-04
+2022-06-19 04:00:37,995 INFO [train.py:874] (2/4) Epoch 24, batch 3850, datatang_loss[loss=0.1584, simple_loss=0.2356, pruned_loss=0.04054, over 4863.00 frames.], tot_loss[loss=0.1441, simple_loss=0.2261, pruned_loss=0.03104, over 985776.26 frames.], batch size: 39, aishell_tot_loss[loss=0.1475, simple_loss=0.2347, pruned_loss=0.03012, over 985458.00 frames.], datatang_tot_loss[loss=0.1405, simple_loss=0.2173, pruned_loss=0.03187, over 986115.40 frames.], batch size: 39, lr: 3.38e-04
+2022-06-19 04:01:05,126 INFO [train.py:874] (2/4) Epoch 24, batch 3900, aishell_loss[loss=0.158, simple_loss=0.2395, pruned_loss=0.03828, over 4864.00 frames.], tot_loss[loss=0.1437, simple_loss=0.2258, pruned_loss=0.03078, over 985907.15 frames.], batch size: 37, aishell_tot_loss[loss=0.148, simple_loss=0.2353, pruned_loss=0.03034, over 985519.26 frames.], datatang_tot_loss[loss=0.1397, simple_loss=0.2167, pruned_loss=0.03138, over 986189.00 frames.], batch size: 37, lr: 3.38e-04
+2022-06-19 04:01:33,261 INFO [train.py:874] (2/4) Epoch 24, batch 3950, aishell_loss[loss=0.1582, simple_loss=0.249, pruned_loss=0.03372, over 4865.00 frames.], tot_loss[loss=0.1428, simple_loss=0.2251, pruned_loss=0.03027, over 985511.00 frames.], batch size: 37, aishell_tot_loss[loss=0.148, simple_loss=0.2354, pruned_loss=0.03032, over 985278.15 frames.], datatang_tot_loss[loss=0.1388, simple_loss=0.216, pruned_loss=0.03082, over 986006.79 frames.], batch size: 37, lr: 3.38e-04
+2022-06-19 04:01:59,973 INFO [train.py:874] (2/4) Epoch 24, batch 4000, aishell_loss[loss=0.1295, simple_loss=0.2005, pruned_loss=0.02924, over 4820.00 frames.], tot_loss[loss=0.1434, simple_loss=0.2255, pruned_loss=0.0306, over 985670.18 frames.], batch size: 24, aishell_tot_loss[loss=0.1476, simple_loss=0.2346, pruned_loss=0.03023, over 985383.90 frames.], datatang_tot_loss[loss=0.1396, simple_loss=0.2167, pruned_loss=0.03119, over 986066.27 frames.], batch size: 24, lr: 3.38e-04
+2022-06-19 04:01:59,974 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 04:02:15,553 INFO [train.py:914] (2/4) Epoch 24, validation: loss=0.1649, simple_loss=0.2488, pruned_loss=0.04056, over 1622729.00 frames.
+2022-06-19 04:02:42,070 INFO [train.py:874] (2/4) Epoch 24, batch 4050, aishell_loss[loss=0.1479, simple_loss=0.2344, pruned_loss=0.03076, over 4960.00 frames.], tot_loss[loss=0.1434, simple_loss=0.2252, pruned_loss=0.03079, over 985906.46 frames.], batch size: 40, aishell_tot_loss[loss=0.1476, simple_loss=0.2345, pruned_loss=0.03031, over 985446.02 frames.], datatang_tot_loss[loss=0.1395, simple_loss=0.2163, pruned_loss=0.0313, over 986241.85 frames.], batch size: 40, lr: 3.38e-04
+2022-06-19 04:03:10,320 INFO [train.py:874] (2/4) Epoch 24, batch 4100, datatang_loss[loss=0.1225, simple_loss=0.2059, pruned_loss=0.01959, over 4934.00 frames.], tot_loss[loss=0.1433, simple_loss=0.2254, pruned_loss=0.03062, over 985978.82 frames.], batch size: 79, aishell_tot_loss[loss=0.1477, simple_loss=0.2347, pruned_loss=0.0304, over 985571.59 frames.], datatang_tot_loss[loss=0.1392, simple_loss=0.2163, pruned_loss=0.03101, over 986211.25 frames.], batch size: 79, lr: 3.38e-04
+2022-06-19 04:03:36,843 INFO [train.py:874] (2/4) Epoch 24, batch 4150, aishell_loss[loss=0.1565, simple_loss=0.246, pruned_loss=0.03351, over 4934.00 frames.], tot_loss[loss=0.143, simple_loss=0.2251, pruned_loss=0.03047, over 985661.10 frames.], batch size: 49, aishell_tot_loss[loss=0.1476, simple_loss=0.2345, pruned_loss=0.03041, over 985331.76 frames.], datatang_tot_loss[loss=0.1389, simple_loss=0.2163, pruned_loss=0.0308, over 986123.28 frames.], batch size: 49, lr: 3.38e-04
+2022-06-19 04:04:55,975 INFO [train.py:874] (2/4) Epoch 25, batch 50, datatang_loss[loss=0.1104, simple_loss=0.1822, pruned_loss=0.01933, over 4897.00 frames.], tot_loss[loss=0.1386, simple_loss=0.222, pruned_loss=0.02758, over 218114.72 frames.], batch size: 42, aishell_tot_loss[loss=0.1443, simple_loss=0.2319, pruned_loss=0.02837, over 141550.12 frames.], datatang_tot_loss[loss=0.1296, simple_loss=0.2064, pruned_loss=0.0264, over 89415.08 frames.], batch size: 42, lr: 3.31e-04
+2022-06-19 04:05:25,165 INFO [train.py:874] (2/4) Epoch 25, batch 100, datatang_loss[loss=0.1622, simple_loss=0.2396, pruned_loss=0.04246, over 4951.00 frames.], tot_loss[loss=0.1392, simple_loss=0.2214, pruned_loss=0.02855, over 388081.36 frames.], batch size: 99, aishell_tot_loss[loss=0.1454, simple_loss=0.2326, pruned_loss=0.02909, over 233204.92 frames.], datatang_tot_loss[loss=0.1321, simple_loss=0.2085, pruned_loss=0.02782, over 202916.88 frames.], batch size: 99, lr: 3.31e-04
+2022-06-19 04:05:54,549 INFO [train.py:874] (2/4) Epoch 25, batch 150, datatang_loss[loss=0.1252, simple_loss=0.2029, pruned_loss=0.02371, over 4919.00 frames.], tot_loss[loss=0.1382, simple_loss=0.2194, pruned_loss=0.02853, over 520357.29 frames.], batch size: 83, aishell_tot_loss[loss=0.146, simple_loss=0.2322, pruned_loss=0.02994, over 304728.11 frames.], datatang_tot_loss[loss=0.131, simple_loss=0.2078, pruned_loss=0.02709, over 312241.09 frames.], batch size: 83, lr: 3.31e-04
+2022-06-19 04:06:21,629 INFO [train.py:874] (2/4) Epoch 25, batch 200, aishell_loss[loss=0.1338, simple_loss=0.2191, pruned_loss=0.02427, over 4975.00 frames.], tot_loss[loss=0.1381, simple_loss=0.2198, pruned_loss=0.02822, over 623112.09 frames.], batch size: 51, aishell_tot_loss[loss=0.145, simple_loss=0.2317, pruned_loss=0.0292, over 393673.43 frames.], datatang_tot_loss[loss=0.1311, simple_loss=0.2078, pruned_loss=0.02723, over 382348.19 frames.], batch size: 51, lr: 3.31e-04
+2022-06-19 04:06:50,343 INFO [train.py:874] (2/4) Epoch 25, batch 250, datatang_loss[loss=0.1441, simple_loss=0.225, pruned_loss=0.03164, over 4931.00 frames.], tot_loss[loss=0.1386, simple_loss=0.2205, pruned_loss=0.02841, over 703556.22 frames.], batch size: 94, aishell_tot_loss[loss=0.1455, simple_loss=0.2322, pruned_loss=0.02938, over 457961.95 frames.], datatang_tot_loss[loss=0.1318, simple_loss=0.2089, pruned_loss=0.02739, over 458951.99 frames.], batch size: 94, lr: 3.31e-04
+2022-06-19 04:07:19,530 INFO [train.py:874] (2/4) Epoch 25, batch 300, datatang_loss[loss=0.1598, simple_loss=0.2244, pruned_loss=0.0476, over 4965.00 frames.], tot_loss[loss=0.1409, simple_loss=0.2225, pruned_loss=0.02965, over 766270.23 frames.], batch size: 45, aishell_tot_loss[loss=0.147, simple_loss=0.2335, pruned_loss=0.03025, over 520184.78 frames.], datatang_tot_loss[loss=0.1338, simple_loss=0.2107, pruned_loss=0.02843, over 521085.59 frames.], batch size: 45, lr: 3.30e-04
+2022-06-19 04:07:46,662 INFO [train.py:874] (2/4) Epoch 25, batch 350, datatang_loss[loss=0.1638, simple_loss=0.232, pruned_loss=0.04776, over 4889.00 frames.], tot_loss[loss=0.1414, simple_loss=0.2231, pruned_loss=0.02985, over 814509.75 frames.], batch size: 47, aishell_tot_loss[loss=0.1464, simple_loss=0.2328, pruned_loss=0.03005, over 582660.94 frames.], datatang_tot_loss[loss=0.1349, simple_loss=0.2117, pruned_loss=0.02901, over 567581.97 frames.], batch size: 47, lr: 3.30e-04
+2022-06-19 04:08:15,442 INFO [train.py:874] (2/4) Epoch 25, batch 400, datatang_loss[loss=0.128, simple_loss=0.2074, pruned_loss=0.0243, over 4927.00 frames.], tot_loss[loss=0.141, simple_loss=0.2232, pruned_loss=0.02945, over 852745.53 frames.], batch size: 71, aishell_tot_loss[loss=0.1465, simple_loss=0.233, pruned_loss=0.02998, over 624903.83 frames.], datatang_tot_loss[loss=0.1349, simple_loss=0.2124, pruned_loss=0.02866, over 622465.74 frames.], batch size: 71, lr: 3.30e-04
+2022-06-19 04:08:43,830 INFO [train.py:874] (2/4) Epoch 25, batch 450, datatang_loss[loss=0.1341, simple_loss=0.2126, pruned_loss=0.02776, over 4916.00 frames.], tot_loss[loss=0.141, simple_loss=0.2229, pruned_loss=0.02953, over 881992.37 frames.], batch size: 57, aishell_tot_loss[loss=0.1462, simple_loss=0.2327, pruned_loss=0.02989, over 657541.59 frames.], datatang_tot_loss[loss=0.1356, simple_loss=0.2132, pruned_loss=0.02896, over 674644.92 frames.], batch size: 57, lr: 3.30e-04
+2022-06-19 04:09:15,236 INFO [train.py:874] (2/4) Epoch 25, batch 500, aishell_loss[loss=0.1541, simple_loss=0.2322, pruned_loss=0.038, over 4929.00 frames.], tot_loss[loss=0.1418, simple_loss=0.2239, pruned_loss=0.02982, over 905173.97 frames.], batch size: 33, aishell_tot_loss[loss=0.1463, simple_loss=0.2331, pruned_loss=0.02978, over 697658.14 frames.], datatang_tot_loss[loss=0.1365, simple_loss=0.2141, pruned_loss=0.02951, over 710074.84 frames.], batch size: 33, lr: 3.30e-04
+2022-06-19 04:09:44,018 INFO [train.py:874] (2/4) Epoch 25, batch 550, aishell_loss[loss=0.1961, simple_loss=0.296, pruned_loss=0.0481, over 4944.00 frames.], tot_loss[loss=0.1424, simple_loss=0.2244, pruned_loss=0.03023, over 923056.17 frames.], batch size: 32, aishell_tot_loss[loss=0.1464, simple_loss=0.2331, pruned_loss=0.02981, over 732640.59 frames.], datatang_tot_loss[loss=0.1374, simple_loss=0.2147, pruned_loss=0.03009, over 741569.55 frames.], batch size: 32, lr: 3.30e-04
+2022-06-19 04:10:12,972 INFO [train.py:874] (2/4) Epoch 25, batch 600, aishell_loss[loss=0.1522, simple_loss=0.2398, pruned_loss=0.03229, over 4937.00 frames.], tot_loss[loss=0.1417, simple_loss=0.2238, pruned_loss=0.02984, over 936816.11 frames.], batch size: 58, aishell_tot_loss[loss=0.1457, simple_loss=0.2324, pruned_loss=0.02947, over 766772.71 frames.], datatang_tot_loss[loss=0.1372, simple_loss=0.2144, pruned_loss=0.03003, over 765938.23 frames.], batch size: 58, lr: 3.30e-04
+2022-06-19 04:10:41,776 INFO [train.py:874] (2/4) Epoch 25, batch 650, aishell_loss[loss=0.1377, simple_loss=0.2304, pruned_loss=0.02247, over 4974.00 frames.], tot_loss[loss=0.142, simple_loss=0.2245, pruned_loss=0.02973, over 947971.50 frames.], batch size: 51, aishell_tot_loss[loss=0.1454, simple_loss=0.2322, pruned_loss=0.02928, over 799305.37 frames.], datatang_tot_loss[loss=0.1375, simple_loss=0.2149, pruned_loss=0.03011, over 785148.01 frames.], batch size: 51, lr: 3.30e-04
+2022-06-19 04:11:11,129 INFO [train.py:874] (2/4) Epoch 25, batch 700, datatang_loss[loss=0.1389, simple_loss=0.2287, pruned_loss=0.02456, over 4952.00 frames.], tot_loss[loss=0.1418, simple_loss=0.2247, pruned_loss=0.0295, over 955960.45 frames.], batch size: 86, aishell_tot_loss[loss=0.1452, simple_loss=0.2322, pruned_loss=0.02908, over 821639.85 frames.], datatang_tot_loss[loss=0.1376, simple_loss=0.2152, pruned_loss=0.03003, over 807929.70 frames.], batch size: 86, lr: 3.30e-04
+2022-06-19 04:11:37,448 INFO [train.py:874] (2/4) Epoch 25, batch 750, datatang_loss[loss=0.1601, simple_loss=0.2369, pruned_loss=0.04161, over 4955.00 frames.], tot_loss[loss=0.1416, simple_loss=0.2244, pruned_loss=0.02943, over 962603.85 frames.], batch size: 91, aishell_tot_loss[loss=0.1448, simple_loss=0.2318, pruned_loss=0.02888, over 842240.50 frames.], datatang_tot_loss[loss=0.1378, simple_loss=0.2153, pruned_loss=0.03011, over 827499.86 frames.], batch size: 91, lr: 3.30e-04
+2022-06-19 04:12:06,290 INFO [train.py:874] (2/4) Epoch 25, batch 800, datatang_loss[loss=0.1381, simple_loss=0.2183, pruned_loss=0.02899, over 4895.00 frames.], tot_loss[loss=0.1411, simple_loss=0.2235, pruned_loss=0.02934, over 968018.06 frames.], batch size: 52, aishell_tot_loss[loss=0.1449, simple_loss=0.2319, pruned_loss=0.02901, over 857436.07 frames.], datatang_tot_loss[loss=0.1372, simple_loss=0.2148, pruned_loss=0.02984, over 848311.91 frames.], batch size: 52, lr: 3.30e-04
+2022-06-19 04:12:35,214 INFO [train.py:874] (2/4) Epoch 25, batch 850, datatang_loss[loss=0.1392, simple_loss=0.2165, pruned_loss=0.03091, over 4976.00 frames.], tot_loss[loss=0.1417, simple_loss=0.2242, pruned_loss=0.02959, over 971702.18 frames.], batch size: 31, aishell_tot_loss[loss=0.1449, simple_loss=0.232, pruned_loss=0.02893, over 871303.75 frames.], datatang_tot_loss[loss=0.1381, simple_loss=0.2157, pruned_loss=0.0302, over 865545.39 frames.], batch size: 31, lr: 3.30e-04
+2022-06-19 04:13:03,057 INFO [train.py:874] (2/4) Epoch 25, batch 900, datatang_loss[loss=0.1306, simple_loss=0.2209, pruned_loss=0.02016, over 4923.00 frames.], tot_loss[loss=0.1415, simple_loss=0.2237, pruned_loss=0.02961, over 974529.82 frames.], batch size: 57, aishell_tot_loss[loss=0.1448, simple_loss=0.2315, pruned_loss=0.02902, over 883161.86 frames.], datatang_tot_loss[loss=0.1382, simple_loss=0.216, pruned_loss=0.03016, over 881058.56 frames.], batch size: 57, lr: 3.29e-04
+2022-06-19 04:13:30,842 INFO [train.py:874] (2/4) Epoch 25, batch 950, aishell_loss[loss=0.143, simple_loss=0.2333, pruned_loss=0.0264, over 4949.00 frames.], tot_loss[loss=0.1425, simple_loss=0.2248, pruned_loss=0.03007, over 976911.07 frames.], batch size: 64, aishell_tot_loss[loss=0.1452, simple_loss=0.232, pruned_loss=0.02921, over 894649.49 frames.], datatang_tot_loss[loss=0.1389, simple_loss=0.2169, pruned_loss=0.0305, over 893875.01 frames.], batch size: 64, lr: 3.29e-04
+2022-06-19 04:14:00,698 INFO [train.py:874] (2/4) Epoch 25, batch 1000, datatang_loss[loss=0.1236, simple_loss=0.2064, pruned_loss=0.02043, over 4931.00 frames.], tot_loss[loss=0.1422, simple_loss=0.2249, pruned_loss=0.02979, over 978829.77 frames.], batch size: 71, aishell_tot_loss[loss=0.1451, simple_loss=0.2317, pruned_loss=0.02918, over 907892.50 frames.], datatang_tot_loss[loss=0.1387, simple_loss=0.2168, pruned_loss=0.03031, over 902041.54 frames.], batch size: 71, lr: 3.29e-04
+2022-06-19 04:14:00,698 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 04:14:17,421 INFO [train.py:914] (2/4) Epoch 25, validation: loss=0.1644, simple_loss=0.248, pruned_loss=0.0404, over 1622729.00 frames.
+2022-06-19 04:14:45,770 INFO [train.py:874] (2/4) Epoch 25, batch 1050, datatang_loss[loss=0.1282, simple_loss=0.212, pruned_loss=0.02225, over 4950.00 frames.], tot_loss[loss=0.1424, simple_loss=0.2246, pruned_loss=0.03015, over 980479.55 frames.], batch size: 69, aishell_tot_loss[loss=0.1448, simple_loss=0.231, pruned_loss=0.02935, over 917858.65 frames.], datatang_tot_loss[loss=0.1391, simple_loss=0.2172, pruned_loss=0.03056, over 911119.39 frames.], batch size: 69, lr: 3.29e-04
+2022-06-19 04:15:14,183 INFO [train.py:874] (2/4) Epoch 25, batch 1100, aishell_loss[loss=0.1448, simple_loss=0.2312, pruned_loss=0.02921, over 4895.00 frames.], tot_loss[loss=0.1429, simple_loss=0.2253, pruned_loss=0.03024, over 981411.49 frames.], batch size: 50, aishell_tot_loss[loss=0.1447, simple_loss=0.2311, pruned_loss=0.02912, over 927640.51 frames.], datatang_tot_loss[loss=0.1397, simple_loss=0.2175, pruned_loss=0.03098, over 917590.21 frames.], batch size: 50, lr: 3.29e-04
+2022-06-19 04:15:41,601 INFO [train.py:874] (2/4) Epoch 25, batch 1150, aishell_loss[loss=0.134, simple_loss=0.2282, pruned_loss=0.01996, over 4951.00 frames.], tot_loss[loss=0.1429, simple_loss=0.2254, pruned_loss=0.03019, over 982547.41 frames.], batch size: 54, aishell_tot_loss[loss=0.145, simple_loss=0.2317, pruned_loss=0.02918, over 934799.83 frames.], datatang_tot_loss[loss=0.1395, simple_loss=0.2171, pruned_loss=0.03093, over 925452.12 frames.], batch size: 54, lr: 3.29e-04
+2022-06-19 04:16:10,677 INFO [train.py:874] (2/4) Epoch 25, batch 1200, aishell_loss[loss=0.1568, simple_loss=0.237, pruned_loss=0.03826, over 4921.00 frames.], tot_loss[loss=0.1438, simple_loss=0.2263, pruned_loss=0.03062, over 983181.65 frames.], batch size: 41, aishell_tot_loss[loss=0.1457, simple_loss=0.2323, pruned_loss=0.02955, over 941348.96 frames.], datatang_tot_loss[loss=0.1399, simple_loss=0.2176, pruned_loss=0.0311, over 931783.99 frames.], batch size: 41, lr: 3.29e-04
+2022-06-19 04:16:37,442 INFO [train.py:874] (2/4) Epoch 25, batch 1250, datatang_loss[loss=0.1357, simple_loss=0.2072, pruned_loss=0.03212, over 4945.00 frames.], tot_loss[loss=0.1442, simple_loss=0.2265, pruned_loss=0.03097, over 983466.12 frames.], batch size: 55, aishell_tot_loss[loss=0.1459, simple_loss=0.2324, pruned_loss=0.02974, over 946665.38 frames.], datatang_tot_loss[loss=0.1404, simple_loss=0.2179, pruned_loss=0.0314, over 937664.31 frames.], batch size: 55, lr: 3.29e-04
+2022-06-19 04:17:06,238 INFO [train.py:874] (2/4) Epoch 25, batch 1300, aishell_loss[loss=0.1515, simple_loss=0.2385, pruned_loss=0.03223, over 4879.00 frames.], tot_loss[loss=0.1453, simple_loss=0.2274, pruned_loss=0.03163, over 983681.27 frames.], batch size: 42, aishell_tot_loss[loss=0.1466, simple_loss=0.233, pruned_loss=0.03011, over 950475.49 frames.], datatang_tot_loss[loss=0.1412, simple_loss=0.2188, pruned_loss=0.03179, over 943914.46 frames.], batch size: 42, lr: 3.29e-04
+2022-06-19 04:17:37,327 INFO [train.py:874] (2/4) Epoch 25, batch 1350, datatang_loss[loss=0.139, simple_loss=0.2117, pruned_loss=0.03315, over 4992.00 frames.], tot_loss[loss=0.145, simple_loss=0.2268, pruned_loss=0.03161, over 983945.95 frames.], batch size: 24, aishell_tot_loss[loss=0.1463, simple_loss=0.2327, pruned_loss=0.02997, over 953943.94 frames.], datatang_tot_loss[loss=0.1416, simple_loss=0.219, pruned_loss=0.03206, over 949372.01 frames.], batch size: 24, lr: 3.29e-04
+2022-06-19 04:18:04,165 INFO [train.py:874] (2/4) Epoch 25, batch 1400, datatang_loss[loss=0.1497, simple_loss=0.2332, pruned_loss=0.03312, over 4859.00 frames.], tot_loss[loss=0.1435, simple_loss=0.2254, pruned_loss=0.03084, over 984374.72 frames.], batch size: 30, aishell_tot_loss[loss=0.1461, simple_loss=0.2325, pruned_loss=0.02981, over 957459.75 frames.], datatang_tot_loss[loss=0.1405, simple_loss=0.218, pruned_loss=0.0315, over 953960.41 frames.], batch size: 30, lr: 3.29e-04
+2022-06-19 04:18:32,335 INFO [train.py:874] (2/4) Epoch 25, batch 1450, datatang_loss[loss=0.1494, simple_loss=0.2386, pruned_loss=0.03012, over 4926.00 frames.], tot_loss[loss=0.1429, simple_loss=0.2247, pruned_loss=0.03051, over 984845.98 frames.], batch size: 94, aishell_tot_loss[loss=0.1455, simple_loss=0.2318, pruned_loss=0.02957, over 960398.89 frames.], datatang_tot_loss[loss=0.1404, simple_loss=0.2181, pruned_loss=0.03141, over 958289.91 frames.], batch size: 94, lr: 3.29e-04
+2022-06-19 04:19:02,036 INFO [train.py:874] (2/4) Epoch 25, batch 1500, aishell_loss[loss=0.1608, simple_loss=0.2492, pruned_loss=0.03614, over 4933.00 frames.], tot_loss[loss=0.1443, simple_loss=0.2258, pruned_loss=0.03136, over 984590.06 frames.], batch size: 49, aishell_tot_loss[loss=0.1457, simple_loss=0.2322, pruned_loss=0.02962, over 962707.61 frames.], datatang_tot_loss[loss=0.1418, simple_loss=0.219, pruned_loss=0.03225, over 961732.41 frames.], batch size: 49, lr: 3.28e-04
+2022-06-19 04:19:29,738 INFO [train.py:874] (2/4) Epoch 25, batch 1550, datatang_loss[loss=0.1384, simple_loss=0.215, pruned_loss=0.03087, over 4952.00 frames.], tot_loss[loss=0.1441, simple_loss=0.2252, pruned_loss=0.03149, over 985075.69 frames.], batch size: 86, aishell_tot_loss[loss=0.1458, simple_loss=0.2319, pruned_loss=0.02983, over 965273.66 frames.], datatang_tot_loss[loss=0.1417, simple_loss=0.2189, pruned_loss=0.03224, over 964959.65 frames.], batch size: 86, lr: 3.28e-04
+2022-06-19 04:19:59,185 INFO [train.py:874] (2/4) Epoch 25, batch 1600, aishell_loss[loss=0.1618, simple_loss=0.2471, pruned_loss=0.03823, over 4940.00 frames.], tot_loss[loss=0.1433, simple_loss=0.2245, pruned_loss=0.03109, over 985221.17 frames.], batch size: 45, aishell_tot_loss[loss=0.1451, simple_loss=0.231, pruned_loss=0.0296, over 967952.86 frames.], datatang_tot_loss[loss=0.1416, simple_loss=0.2188, pruned_loss=0.03218, over 967122.16 frames.], batch size: 45, lr: 3.28e-04
+2022-06-19 04:20:27,787 INFO [train.py:874] (2/4) Epoch 25, batch 1650, aishell_loss[loss=0.152, simple_loss=0.2395, pruned_loss=0.03223, over 4860.00 frames.], tot_loss[loss=0.143, simple_loss=0.2246, pruned_loss=0.03068, over 985460.10 frames.], batch size: 36, aishell_tot_loss[loss=0.1449, simple_loss=0.2311, pruned_loss=0.02931, over 970042.39 frames.], datatang_tot_loss[loss=0.1415, simple_loss=0.2189, pruned_loss=0.03206, over 969455.98 frames.], batch size: 36, lr: 3.28e-04
+2022-06-19 04:20:55,474 INFO [train.py:874] (2/4) Epoch 25, batch 1700, datatang_loss[loss=0.1213, simple_loss=0.2016, pruned_loss=0.02054, over 4942.00 frames.], tot_loss[loss=0.1431, simple_loss=0.2247, pruned_loss=0.03078, over 985256.15 frames.], batch size: 69, aishell_tot_loss[loss=0.1451, simple_loss=0.2313, pruned_loss=0.02943, over 971657.38 frames.], datatang_tot_loss[loss=0.1414, simple_loss=0.2187, pruned_loss=0.03204, over 971332.77 frames.], batch size: 69, lr: 3.28e-04
+2022-06-19 04:21:24,499 INFO [train.py:874] (2/4) Epoch 25, batch 1750, aishell_loss[loss=0.1606, simple_loss=0.2504, pruned_loss=0.03539, over 4883.00 frames.], tot_loss[loss=0.1432, simple_loss=0.2249, pruned_loss=0.03073, over 985008.49 frames.], batch size: 47, aishell_tot_loss[loss=0.1446, simple_loss=0.2308, pruned_loss=0.02917, over 972807.84 frames.], datatang_tot_loss[loss=0.142, simple_loss=0.2194, pruned_loss=0.03224, over 973166.59 frames.], batch size: 47, lr: 3.28e-04
+2022-06-19 04:21:53,973 INFO [train.py:874] (2/4) Epoch 25, batch 1800, datatang_loss[loss=0.1457, simple_loss=0.2201, pruned_loss=0.03568, over 4949.00 frames.], tot_loss[loss=0.1432, simple_loss=0.2244, pruned_loss=0.03095, over 985522.03 frames.], batch size: 67, aishell_tot_loss[loss=0.1447, simple_loss=0.2309, pruned_loss=0.02921, over 974094.79 frames.], datatang_tot_loss[loss=0.1419, simple_loss=0.2191, pruned_loss=0.03232, over 975234.00 frames.], batch size: 67, lr: 3.28e-04
+2022-06-19 04:22:21,497 INFO [train.py:874] (2/4) Epoch 25, batch 1850, datatang_loss[loss=0.1475, simple_loss=0.2273, pruned_loss=0.03389, over 4937.00 frames.], tot_loss[loss=0.1436, simple_loss=0.2253, pruned_loss=0.03094, over 985281.55 frames.], batch size: 79, aishell_tot_loss[loss=0.1452, simple_loss=0.2314, pruned_loss=0.02951, over 975628.72 frames.], datatang_tot_loss[loss=0.1417, simple_loss=0.2192, pruned_loss=0.0321, over 976030.20 frames.], batch size: 79, lr: 3.28e-04
+2022-06-19 04:22:52,302 INFO [train.py:874] (2/4) Epoch 25, batch 1900, datatang_loss[loss=0.1378, simple_loss=0.2158, pruned_loss=0.02992, over 4964.00 frames.], tot_loss[loss=0.1436, simple_loss=0.2255, pruned_loss=0.03082, over 985293.33 frames.], batch size: 60, aishell_tot_loss[loss=0.1459, simple_loss=0.2321, pruned_loss=0.02989, over 976948.42 frames.], datatang_tot_loss[loss=0.141, simple_loss=0.2186, pruned_loss=0.03167, over 976947.20 frames.], batch size: 60, lr: 3.28e-04
+2022-06-19 04:23:20,040 INFO [train.py:874] (2/4) Epoch 25, batch 1950, aishell_loss[loss=0.147, simple_loss=0.2402, pruned_loss=0.02692, over 4945.00 frames.], tot_loss[loss=0.1438, simple_loss=0.2261, pruned_loss=0.03077, over 985051.45 frames.], batch size: 49, aishell_tot_loss[loss=0.1464, simple_loss=0.2328, pruned_loss=0.02998, over 977805.65 frames.], datatang_tot_loss[loss=0.1408, simple_loss=0.2184, pruned_loss=0.03154, over 977792.96 frames.], batch size: 49, lr: 3.28e-04
+2022-06-19 04:23:47,731 INFO [train.py:874] (2/4) Epoch 25, batch 2000, datatang_loss[loss=0.1452, simple_loss=0.2326, pruned_loss=0.02891, over 4935.00 frames.], tot_loss[loss=0.1436, simple_loss=0.2262, pruned_loss=0.03052, over 985107.32 frames.], batch size: 62, aishell_tot_loss[loss=0.1466, simple_loss=0.2332, pruned_loss=0.03005, over 978728.09 frames.], datatang_tot_loss[loss=0.1403, simple_loss=0.2181, pruned_loss=0.03122, over 978636.32 frames.], batch size: 62, lr: 3.28e-04
+2022-06-19 04:23:47,732 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 04:24:04,380 INFO [train.py:914] (2/4) Epoch 25, validation: loss=0.1644, simple_loss=0.2485, pruned_loss=0.04013, over 1622729.00 frames.
+2022-06-19 04:24:31,264 INFO [train.py:874] (2/4) Epoch 25, batch 2050, aishell_loss[loss=0.1451, simple_loss=0.2243, pruned_loss=0.03294, over 4967.00 frames.], tot_loss[loss=0.1428, simple_loss=0.2256, pruned_loss=0.03004, over 985197.57 frames.], batch size: 44, aishell_tot_loss[loss=0.1458, simple_loss=0.2324, pruned_loss=0.02961, over 979722.83 frames.], datatang_tot_loss[loss=0.1401, simple_loss=0.2178, pruned_loss=0.03118, over 979241.81 frames.], batch size: 44, lr: 3.28e-04
+2022-06-19 04:24:59,612 INFO [train.py:874] (2/4) Epoch 25, batch 2100, aishell_loss[loss=0.1436, simple_loss=0.2356, pruned_loss=0.02579, over 4945.00 frames.], tot_loss[loss=0.1428, simple_loss=0.2258, pruned_loss=0.02991, over 985537.80 frames.], batch size: 49, aishell_tot_loss[loss=0.1455, simple_loss=0.2323, pruned_loss=0.02931, over 980656.14 frames.], datatang_tot_loss[loss=0.1403, simple_loss=0.218, pruned_loss=0.03129, over 980009.39 frames.], batch size: 49, lr: 3.28e-04
+2022-06-19 04:25:26,934 INFO [train.py:874] (2/4) Epoch 25, batch 2150, datatang_loss[loss=0.1158, simple_loss=0.1898, pruned_loss=0.02088, over 4921.00 frames.], tot_loss[loss=0.1432, simple_loss=0.226, pruned_loss=0.0302, over 985722.50 frames.], batch size: 31, aishell_tot_loss[loss=0.1455, simple_loss=0.2324, pruned_loss=0.02932, over 981224.04 frames.], datatang_tot_loss[loss=0.1407, simple_loss=0.2185, pruned_loss=0.03146, over 980876.50 frames.], batch size: 31, lr: 3.27e-04
+2022-06-19 04:25:55,459 INFO [train.py:874] (2/4) Epoch 25, batch 2200, datatang_loss[loss=0.143, simple_loss=0.2238, pruned_loss=0.03105, over 4919.00 frames.], tot_loss[loss=0.144, simple_loss=0.2266, pruned_loss=0.03064, over 985516.78 frames.], batch size: 57, aishell_tot_loss[loss=0.1459, simple_loss=0.2328, pruned_loss=0.02952, over 981577.17 frames.], datatang_tot_loss[loss=0.1411, simple_loss=0.2189, pruned_loss=0.03167, over 981414.29 frames.], batch size: 57, lr: 3.27e-04
+2022-06-19 04:26:24,822 INFO [train.py:874] (2/4) Epoch 25, batch 2250, aishell_loss[loss=0.133, simple_loss=0.2235, pruned_loss=0.02126, over 4932.00 frames.], tot_loss[loss=0.144, simple_loss=0.2265, pruned_loss=0.03078, over 985424.00 frames.], batch size: 58, aishell_tot_loss[loss=0.1456, simple_loss=0.2324, pruned_loss=0.02938, over 981858.21 frames.], datatang_tot_loss[loss=0.1416, simple_loss=0.2194, pruned_loss=0.03194, over 981981.59 frames.], batch size: 58, lr: 3.27e-04
+2022-06-19 04:26:54,169 INFO [train.py:874] (2/4) Epoch 25, batch 2300, datatang_loss[loss=0.1578, simple_loss=0.2421, pruned_loss=0.03677, over 4917.00 frames.], tot_loss[loss=0.1447, simple_loss=0.227, pruned_loss=0.03124, over 985437.71 frames.], batch size: 98, aishell_tot_loss[loss=0.1462, simple_loss=0.2329, pruned_loss=0.02971, over 982228.03 frames.], datatang_tot_loss[loss=0.1419, simple_loss=0.2196, pruned_loss=0.03211, over 982457.64 frames.], batch size: 98, lr: 3.27e-04
+2022-06-19 04:27:21,751 INFO [train.py:874] (2/4) Epoch 25, batch 2350, aishell_loss[loss=0.1528, simple_loss=0.2434, pruned_loss=0.0311, over 4900.00 frames.], tot_loss[loss=0.1439, simple_loss=0.2265, pruned_loss=0.03063, over 985457.14 frames.], batch size: 60, aishell_tot_loss[loss=0.1463, simple_loss=0.2333, pruned_loss=0.02968, over 982526.47 frames.], datatang_tot_loss[loss=0.141, simple_loss=0.2189, pruned_loss=0.03153, over 982894.35 frames.], batch size: 60, lr: 3.27e-04
+2022-06-19 04:27:52,155 INFO [train.py:874] (2/4) Epoch 25, batch 2400, aishell_loss[loss=0.1544, simple_loss=0.2453, pruned_loss=0.03171, over 4959.00 frames.], tot_loss[loss=0.1437, simple_loss=0.2262, pruned_loss=0.03059, over 985577.14 frames.], batch size: 64, aishell_tot_loss[loss=0.1461, simple_loss=0.2331, pruned_loss=0.0295, over 983020.33 frames.], datatang_tot_loss[loss=0.1413, simple_loss=0.2193, pruned_loss=0.03162, over 983147.45 frames.], batch size: 64, lr: 3.27e-04
+2022-06-19 04:28:20,096 INFO [train.py:874] (2/4) Epoch 25, batch 2450, aishell_loss[loss=0.1093, simple_loss=0.1972, pruned_loss=0.01072, over 4968.00 frames.], tot_loss[loss=0.1439, simple_loss=0.2263, pruned_loss=0.03073, over 985409.63 frames.], batch size: 30, aishell_tot_loss[loss=0.1457, simple_loss=0.2327, pruned_loss=0.02939, over 982893.31 frames.], datatang_tot_loss[loss=0.1418, simple_loss=0.2198, pruned_loss=0.0319, over 983678.99 frames.], batch size: 30, lr: 3.27e-04
+2022-06-19 04:28:48,388 INFO [train.py:874] (2/4) Epoch 25, batch 2500, aishell_loss[loss=0.1577, simple_loss=0.2439, pruned_loss=0.03581, over 4929.00 frames.], tot_loss[loss=0.1442, simple_loss=0.2262, pruned_loss=0.03111, over 984991.20 frames.], batch size: 33, aishell_tot_loss[loss=0.1457, simple_loss=0.2321, pruned_loss=0.0296, over 982727.23 frames.], datatang_tot_loss[loss=0.1423, simple_loss=0.2202, pruned_loss=0.03213, over 983914.09 frames.], batch size: 33, lr: 3.27e-04
+2022-06-19 04:29:17,543 INFO [train.py:874] (2/4) Epoch 25, batch 2550, aishell_loss[loss=0.154, simple_loss=0.2426, pruned_loss=0.03269, over 4929.00 frames.], tot_loss[loss=0.1442, simple_loss=0.2263, pruned_loss=0.03105, over 984945.46 frames.], batch size: 49, aishell_tot_loss[loss=0.1463, simple_loss=0.2328, pruned_loss=0.02987, over 982779.71 frames.], datatang_tot_loss[loss=0.1415, simple_loss=0.2193, pruned_loss=0.0319, over 984247.53 frames.], batch size: 49, lr: 3.27e-04
+2022-06-19 04:29:45,264 INFO [train.py:874] (2/4) Epoch 25, batch 2600, aishell_loss[loss=0.1287, simple_loss=0.2142, pruned_loss=0.02161, over 4863.00 frames.], tot_loss[loss=0.1437, simple_loss=0.2261, pruned_loss=0.03062, over 985143.75 frames.], batch size: 28, aishell_tot_loss[loss=0.146, simple_loss=0.2328, pruned_loss=0.02959, over 983137.34 frames.], datatang_tot_loss[loss=0.1413, simple_loss=0.2191, pruned_loss=0.03177, over 984447.23 frames.], batch size: 28, lr: 3.27e-04
+2022-06-19 04:30:13,908 INFO [train.py:874] (2/4) Epoch 25, batch 2650, aishell_loss[loss=0.15, simple_loss=0.2316, pruned_loss=0.03426, over 4957.00 frames.], tot_loss[loss=0.1433, simple_loss=0.2258, pruned_loss=0.03038, over 985250.91 frames.], batch size: 40, aishell_tot_loss[loss=0.146, simple_loss=0.2327, pruned_loss=0.02967, over 983293.18 frames.], datatang_tot_loss[loss=0.141, simple_loss=0.2191, pruned_loss=0.0314, over 984714.17 frames.], batch size: 40, lr: 3.27e-04
+2022-06-19 04:30:43,849 INFO [train.py:874] (2/4) Epoch 25, batch 2700, aishell_loss[loss=0.1599, simple_loss=0.251, pruned_loss=0.03441, over 4915.00 frames.], tot_loss[loss=0.1435, simple_loss=0.226, pruned_loss=0.03053, over 985109.26 frames.], batch size: 46, aishell_tot_loss[loss=0.146, simple_loss=0.2327, pruned_loss=0.02959, over 983267.10 frames.], datatang_tot_loss[loss=0.1412, simple_loss=0.2191, pruned_loss=0.03161, over 984903.34 frames.], batch size: 46, lr: 3.27e-04
+2022-06-19 04:31:12,109 INFO [train.py:874] (2/4) Epoch 25, batch 2750, datatang_loss[loss=0.1261, simple_loss=0.2029, pruned_loss=0.0246, over 4922.00 frames.], tot_loss[loss=0.1432, simple_loss=0.2254, pruned_loss=0.03048, over 985239.00 frames.], batch size: 64, aishell_tot_loss[loss=0.146, simple_loss=0.2326, pruned_loss=0.02969, over 983604.39 frames.], datatang_tot_loss[loss=0.1408, simple_loss=0.2187, pruned_loss=0.03146, over 984931.10 frames.], batch size: 64, lr: 3.26e-04
+2022-06-19 04:31:39,212 INFO [train.py:874] (2/4) Epoch 25, batch 2800, aishell_loss[loss=0.1667, simple_loss=0.244, pruned_loss=0.04463, over 4970.00 frames.], tot_loss[loss=0.1425, simple_loss=0.2248, pruned_loss=0.03013, over 985050.53 frames.], batch size: 61, aishell_tot_loss[loss=0.1453, simple_loss=0.2319, pruned_loss=0.02937, over 983725.05 frames.], datatang_tot_loss[loss=0.1406, simple_loss=0.2184, pruned_loss=0.03143, over 984871.97 frames.], batch size: 61, lr: 3.26e-04
+2022-06-19 04:32:09,648 INFO [train.py:874] (2/4) Epoch 25, batch 2850, aishell_loss[loss=0.1752, simple_loss=0.2621, pruned_loss=0.04417, over 4938.00 frames.], tot_loss[loss=0.1429, simple_loss=0.2252, pruned_loss=0.03024, over 985413.12 frames.], batch size: 49, aishell_tot_loss[loss=0.146, simple_loss=0.2327, pruned_loss=0.02971, over 983999.59 frames.], datatang_tot_loss[loss=0.1402, simple_loss=0.2181, pruned_loss=0.03112, over 985140.83 frames.], batch size: 49, lr: 3.26e-04
+2022-06-19 04:32:36,506 INFO [train.py:874] (2/4) Epoch 25, batch 2900, datatang_loss[loss=0.1501, simple_loss=0.2213, pruned_loss=0.03947, over 4930.00 frames.], tot_loss[loss=0.1428, simple_loss=0.2251, pruned_loss=0.03026, over 985481.83 frames.], batch size: 79, aishell_tot_loss[loss=0.1454, simple_loss=0.232, pruned_loss=0.02942, over 984260.85 frames.], datatang_tot_loss[loss=0.1406, simple_loss=0.2183, pruned_loss=0.0314, over 985171.80 frames.], batch size: 79, lr: 3.26e-04
+2022-06-19 04:33:06,598 INFO [train.py:874] (2/4) Epoch 25, batch 2950, datatang_loss[loss=0.2093, simple_loss=0.2814, pruned_loss=0.06863, over 4927.00 frames.], tot_loss[loss=0.1424, simple_loss=0.2247, pruned_loss=0.03006, over 985623.06 frames.], batch size: 108, aishell_tot_loss[loss=0.1451, simple_loss=0.2317, pruned_loss=0.02919, over 984410.02 frames.], datatang_tot_loss[loss=0.1405, simple_loss=0.2183, pruned_loss=0.03135, over 985336.59 frames.], batch size: 108, lr: 3.26e-04
+2022-06-19 04:33:34,207 INFO [train.py:874] (2/4) Epoch 25, batch 3000, aishell_loss[loss=0.1584, simple_loss=0.2423, pruned_loss=0.03724, over 4951.00 frames.], tot_loss[loss=0.1434, simple_loss=0.2264, pruned_loss=0.03018, over 985572.57 frames.], batch size: 45, aishell_tot_loss[loss=0.1459, simple_loss=0.2329, pruned_loss=0.02946, over 984778.52 frames.], datatang_tot_loss[loss=0.1404, simple_loss=0.2185, pruned_loss=0.03121, over 985091.68 frames.], batch size: 45, lr: 3.26e-04
+2022-06-19 04:33:34,207 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 04:33:49,968 INFO [train.py:914] (2/4) Epoch 25, validation: loss=0.1645, simple_loss=0.2487, pruned_loss=0.04017, over 1622729.00 frames.
+2022-06-19 04:34:20,838 INFO [train.py:874] (2/4) Epoch 25, batch 3050, aishell_loss[loss=0.1543, simple_loss=0.2399, pruned_loss=0.03434, over 4935.00 frames.], tot_loss[loss=0.1444, simple_loss=0.2272, pruned_loss=0.03078, over 985811.95 frames.], batch size: 32, aishell_tot_loss[loss=0.146, simple_loss=0.2329, pruned_loss=0.02952, over 985056.31 frames.], datatang_tot_loss[loss=0.1415, simple_loss=0.2193, pruned_loss=0.0318, over 985219.32 frames.], batch size: 32, lr: 3.26e-04
+2022-06-19 04:34:49,379 INFO [train.py:874] (2/4) Epoch 25, batch 3100, aishell_loss[loss=0.1599, simple_loss=0.2478, pruned_loss=0.03597, over 4975.00 frames.], tot_loss[loss=0.144, simple_loss=0.227, pruned_loss=0.03054, over 985301.75 frames.], batch size: 51, aishell_tot_loss[loss=0.1454, simple_loss=0.2322, pruned_loss=0.02929, over 984788.75 frames.], datatang_tot_loss[loss=0.1417, simple_loss=0.2196, pruned_loss=0.03189, over 985122.38 frames.], batch size: 51, lr: 3.26e-04
+2022-06-19 04:35:19,308 INFO [train.py:874] (2/4) Epoch 25, batch 3150, datatang_loss[loss=0.1308, simple_loss=0.2187, pruned_loss=0.02152, over 4917.00 frames.], tot_loss[loss=0.1438, simple_loss=0.2263, pruned_loss=0.03059, over 985447.28 frames.], batch size: 42, aishell_tot_loss[loss=0.1455, simple_loss=0.2321, pruned_loss=0.02942, over 984960.43 frames.], datatang_tot_loss[loss=0.1414, simple_loss=0.2193, pruned_loss=0.03179, over 985188.84 frames.], batch size: 42, lr: 3.26e-04
+2022-06-19 04:35:49,025 INFO [train.py:874] (2/4) Epoch 25, batch 3200, datatang_loss[loss=0.1301, simple_loss=0.2095, pruned_loss=0.02535, over 4960.00 frames.], tot_loss[loss=0.1428, simple_loss=0.2256, pruned_loss=0.03005, over 985174.66 frames.], batch size: 60, aishell_tot_loss[loss=0.1446, simple_loss=0.2313, pruned_loss=0.02895, over 984781.49 frames.], datatang_tot_loss[loss=0.1413, simple_loss=0.2192, pruned_loss=0.03171, over 985188.84 frames.], batch size: 60, lr: 3.26e-04
+2022-06-19 04:36:17,745 INFO [train.py:874] (2/4) Epoch 25, batch 3250, aishell_loss[loss=0.1352, simple_loss=0.2174, pruned_loss=0.02651, over 4859.00 frames.], tot_loss[loss=0.1432, simple_loss=0.226, pruned_loss=0.0302, over 985232.18 frames.], batch size: 37, aishell_tot_loss[loss=0.1446, simple_loss=0.2314, pruned_loss=0.02884, over 984714.65 frames.], datatang_tot_loss[loss=0.1417, simple_loss=0.2195, pruned_loss=0.03196, over 985385.43 frames.], batch size: 37, lr: 3.26e-04
+2022-06-19 04:36:47,044 INFO [train.py:874] (2/4) Epoch 25, batch 3300, datatang_loss[loss=0.2024, simple_loss=0.2722, pruned_loss=0.06626, over 4946.00 frames.], tot_loss[loss=0.1438, simple_loss=0.2263, pruned_loss=0.03071, over 985322.50 frames.], batch size: 109, aishell_tot_loss[loss=0.1444, simple_loss=0.2313, pruned_loss=0.02878, over 984875.91 frames.], datatang_tot_loss[loss=0.1425, simple_loss=0.2202, pruned_loss=0.03243, over 985359.48 frames.], batch size: 109, lr: 3.26e-04
+2022-06-19 04:37:14,506 INFO [train.py:874] (2/4) Epoch 25, batch 3350, datatang_loss[loss=0.1382, simple_loss=0.2198, pruned_loss=0.02832, over 4933.00 frames.], tot_loss[loss=0.1445, simple_loss=0.2265, pruned_loss=0.03122, over 985313.76 frames.], batch size: 79, aishell_tot_loss[loss=0.1449, simple_loss=0.2316, pruned_loss=0.02916, over 985046.88 frames.], datatang_tot_loss[loss=0.1428, simple_loss=0.2203, pruned_loss=0.03265, over 985231.69 frames.], batch size: 79, lr: 3.26e-04
+2022-06-19 04:37:42,871 INFO [train.py:874] (2/4) Epoch 25, batch 3400, aishell_loss[loss=0.1539, simple_loss=0.2508, pruned_loss=0.02857, over 4859.00 frames.], tot_loss[loss=0.144, simple_loss=0.2265, pruned_loss=0.03071, over 985730.98 frames.], batch size: 36, aishell_tot_loss[loss=0.1452, simple_loss=0.232, pruned_loss=0.02925, over 985447.78 frames.], datatang_tot_loss[loss=0.1421, simple_loss=0.2198, pruned_loss=0.03216, over 985331.31 frames.], batch size: 36, lr: 3.25e-04
+2022-06-19 04:38:11,574 INFO [train.py:874] (2/4) Epoch 25, batch 3450, datatang_loss[loss=0.1232, simple_loss=0.2128, pruned_loss=0.01683, over 4954.00 frames.], tot_loss[loss=0.1426, simple_loss=0.225, pruned_loss=0.03009, over 985447.95 frames.], batch size: 67, aishell_tot_loss[loss=0.1443, simple_loss=0.2307, pruned_loss=0.02894, over 985246.94 frames.], datatang_tot_loss[loss=0.1416, simple_loss=0.2194, pruned_loss=0.03186, over 985313.53 frames.], batch size: 67, lr: 3.25e-04
+2022-06-19 04:38:39,597 INFO [train.py:874] (2/4) Epoch 25, batch 3500, aishell_loss[loss=0.1423, simple_loss=0.2243, pruned_loss=0.03016, over 4822.00 frames.], tot_loss[loss=0.1425, simple_loss=0.2251, pruned_loss=0.03, over 985201.80 frames.], batch size: 29, aishell_tot_loss[loss=0.1442, simple_loss=0.2306, pruned_loss=0.02892, over 985055.66 frames.], datatang_tot_loss[loss=0.1414, simple_loss=0.2191, pruned_loss=0.03182, over 985285.12 frames.], batch size: 29, lr: 3.25e-04
+2022-06-19 04:39:08,256 INFO [train.py:874] (2/4) Epoch 25, batch 3550, datatang_loss[loss=0.1407, simple_loss=0.2217, pruned_loss=0.02988, over 4941.00 frames.], tot_loss[loss=0.1416, simple_loss=0.224, pruned_loss=0.02959, over 985437.98 frames.], batch size: 50, aishell_tot_loss[loss=0.144, simple_loss=0.2302, pruned_loss=0.02887, over 985127.40 frames.], datatang_tot_loss[loss=0.1406, simple_loss=0.2185, pruned_loss=0.03133, over 985475.14 frames.], batch size: 50, lr: 3.25e-04
+2022-06-19 04:39:38,074 INFO [train.py:874] (2/4) Epoch 25, batch 3600, aishell_loss[loss=0.1523, simple_loss=0.2348, pruned_loss=0.03493, over 4970.00 frames.], tot_loss[loss=0.143, simple_loss=0.2257, pruned_loss=0.03019, over 985330.04 frames.], batch size: 31, aishell_tot_loss[loss=0.1449, simple_loss=0.2314, pruned_loss=0.02921, over 985138.34 frames.], datatang_tot_loss[loss=0.1409, simple_loss=0.2186, pruned_loss=0.03162, over 985399.98 frames.], batch size: 31, lr: 3.25e-04
+2022-06-19 04:40:04,734 INFO [train.py:874] (2/4) Epoch 25, batch 3650, aishell_loss[loss=0.1513, simple_loss=0.2352, pruned_loss=0.03371, over 4969.00 frames.], tot_loss[loss=0.1435, simple_loss=0.2261, pruned_loss=0.03043, over 985210.79 frames.], batch size: 48, aishell_tot_loss[loss=0.1452, simple_loss=0.2316, pruned_loss=0.02935, over 985097.82 frames.], datatang_tot_loss[loss=0.1412, simple_loss=0.2188, pruned_loss=0.03173, over 985320.00 frames.], batch size: 48, lr: 3.25e-04
+2022-06-19 04:40:33,843 INFO [train.py:874] (2/4) Epoch 25, batch 3700, datatang_loss[loss=0.1124, simple_loss=0.1804, pruned_loss=0.02226, over 4959.00 frames.], tot_loss[loss=0.143, simple_loss=0.2256, pruned_loss=0.0302, over 985738.49 frames.], batch size: 55, aishell_tot_loss[loss=0.1451, simple_loss=0.2317, pruned_loss=0.02923, over 985133.32 frames.], datatang_tot_loss[loss=0.1408, simple_loss=0.2186, pruned_loss=0.03154, over 985832.62 frames.], batch size: 55, lr: 3.25e-04
+2022-06-19 04:41:00,274 INFO [train.py:874] (2/4) Epoch 25, batch 3750, datatang_loss[loss=0.156, simple_loss=0.229, pruned_loss=0.04147, over 4932.00 frames.], tot_loss[loss=0.1431, simple_loss=0.2256, pruned_loss=0.03033, over 985925.52 frames.], batch size: 62, aishell_tot_loss[loss=0.145, simple_loss=0.2316, pruned_loss=0.02919, over 985517.23 frames.], datatang_tot_loss[loss=0.1411, simple_loss=0.2188, pruned_loss=0.0317, over 985702.24 frames.], batch size: 62, lr: 3.25e-04
+2022-06-19 04:41:29,107 INFO [train.py:874] (2/4) Epoch 25, batch 3800, datatang_loss[loss=0.1191, simple_loss=0.2028, pruned_loss=0.0177, over 4925.00 frames.], tot_loss[loss=0.1421, simple_loss=0.2245, pruned_loss=0.02981, over 985723.55 frames.], batch size: 73, aishell_tot_loss[loss=0.1449, simple_loss=0.2315, pruned_loss=0.0291, over 985257.95 frames.], datatang_tot_loss[loss=0.1401, simple_loss=0.2178, pruned_loss=0.0312, over 985810.28 frames.], batch size: 73, lr: 3.25e-04
+2022-06-19 04:41:57,149 INFO [train.py:874] (2/4) Epoch 25, batch 3850, datatang_loss[loss=0.1535, simple_loss=0.2359, pruned_loss=0.03556, over 4935.00 frames.], tot_loss[loss=0.1423, simple_loss=0.2247, pruned_loss=0.02992, over 985472.43 frames.], batch size: 88, aishell_tot_loss[loss=0.1452, simple_loss=0.2319, pruned_loss=0.02924, over 985116.05 frames.], datatang_tot_loss[loss=0.14, simple_loss=0.2177, pruned_loss=0.0311, over 985728.64 frames.], batch size: 88, lr: 3.25e-04
+2022-06-19 04:42:25,226 INFO [train.py:874] (2/4) Epoch 25, batch 3900, aishell_loss[loss=0.1423, simple_loss=0.2298, pruned_loss=0.02741, over 4878.00 frames.], tot_loss[loss=0.1417, simple_loss=0.2245, pruned_loss=0.02939, over 985266.97 frames.], batch size: 47, aishell_tot_loss[loss=0.1446, simple_loss=0.2315, pruned_loss=0.0289, over 985039.95 frames.], datatang_tot_loss[loss=0.1397, simple_loss=0.2178, pruned_loss=0.03085, over 985601.87 frames.], batch size: 47, lr: 3.25e-04
+2022-06-19 04:42:52,890 INFO [train.py:874] (2/4) Epoch 25, batch 3950, aishell_loss[loss=0.1476, simple_loss=0.231, pruned_loss=0.03206, over 4917.00 frames.], tot_loss[loss=0.141, simple_loss=0.224, pruned_loss=0.02901, over 985301.15 frames.], batch size: 41, aishell_tot_loss[loss=0.1441, simple_loss=0.231, pruned_loss=0.0286, over 984979.93 frames.], datatang_tot_loss[loss=0.1394, simple_loss=0.2176, pruned_loss=0.03063, over 985688.51 frames.], batch size: 41, lr: 3.25e-04
+2022-06-19 04:43:21,358 INFO [train.py:874] (2/4) Epoch 25, batch 4000, aishell_loss[loss=0.1456, simple_loss=0.2407, pruned_loss=0.02525, over 4942.00 frames.], tot_loss[loss=0.1416, simple_loss=0.2249, pruned_loss=0.02918, over 985528.98 frames.], batch size: 54, aishell_tot_loss[loss=0.1442, simple_loss=0.2314, pruned_loss=0.02849, over 985274.61 frames.], datatang_tot_loss[loss=0.1398, simple_loss=0.218, pruned_loss=0.0308, over 985635.58 frames.], batch size: 54, lr: 3.25e-04
+2022-06-19 04:43:21,359 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 04:43:37,168 INFO [train.py:914] (2/4) Epoch 25, validation: loss=0.1639, simple_loss=0.2478, pruned_loss=0.03995, over 1622729.00 frames.
+2022-06-19 04:44:42,286 INFO [train.py:874] (2/4) Epoch 26, batch 50, aishell_loss[loss=0.128, simple_loss=0.2086, pruned_loss=0.02375, over 4861.00 frames.], tot_loss[loss=0.1347, simple_loss=0.2163, pruned_loss=0.02658, over 218464.05 frames.], batch size: 36, aishell_tot_loss[loss=0.1359, simple_loss=0.2191, pruned_loss=0.0263, over 98277.42 frames.], datatang_tot_loss[loss=0.1339, simple_loss=0.2141, pruned_loss=0.02685, over 133509.49 frames.], batch size: 36, lr: 3.18e-04
+2022-06-19 04:45:12,079 INFO [train.py:874] (2/4) Epoch 26, batch 100, datatang_loss[loss=0.1424, simple_loss=0.2266, pruned_loss=0.02912, over 4932.00 frames.], tot_loss[loss=0.1403, simple_loss=0.2226, pruned_loss=0.02903, over 388181.91 frames.], batch size: 94, aishell_tot_loss[loss=0.1436, simple_loss=0.2288, pruned_loss=0.02921, over 206406.13 frames.], datatang_tot_loss[loss=0.1369, simple_loss=0.2164, pruned_loss=0.02873, over 230010.87 frames.], batch size: 94, lr: 3.18e-04
+2022-06-19 04:45:40,324 INFO [train.py:874] (2/4) Epoch 26, batch 150, datatang_loss[loss=0.132, simple_loss=0.2102, pruned_loss=0.02691, over 4899.00 frames.], tot_loss[loss=0.1409, simple_loss=0.2235, pruned_loss=0.02917, over 520633.38 frames.], batch size: 52, aishell_tot_loss[loss=0.1454, simple_loss=0.2322, pruned_loss=0.02928, over 298203.02 frames.], datatang_tot_loss[loss=0.1364, simple_loss=0.2149, pruned_loss=0.02898, over 318984.45 frames.], batch size: 52, lr: 3.18e-04
+2022-06-19 04:46:10,014 INFO [train.py:874] (2/4) Epoch 26, batch 200, datatang_loss[loss=0.1805, simple_loss=0.254, pruned_loss=0.05353, over 4960.00 frames.], tot_loss[loss=0.1406, simple_loss=0.2229, pruned_loss=0.02916, over 624349.70 frames.], batch size: 99, aishell_tot_loss[loss=0.1446, simple_loss=0.2314, pruned_loss=0.02886, over 370348.54 frames.], datatang_tot_loss[loss=0.1369, simple_loss=0.215, pruned_loss=0.02936, over 406585.97 frames.], batch size: 99, lr: 3.18e-04
+2022-06-19 04:46:40,011 INFO [train.py:874] (2/4) Epoch 26, batch 250, datatang_loss[loss=0.1422, simple_loss=0.2239, pruned_loss=0.0303, over 4904.00 frames.], tot_loss[loss=0.1398, simple_loss=0.2215, pruned_loss=0.02907, over 704287.16 frames.], batch size: 47, aishell_tot_loss[loss=0.1428, simple_loss=0.2289, pruned_loss=0.02829, over 423492.74 frames.], datatang_tot_loss[loss=0.1376, simple_loss=0.2158, pruned_loss=0.02973, over 492179.86 frames.], batch size: 47, lr: 3.18e-04
+2022-06-19 04:47:07,485 INFO [train.py:874] (2/4) Epoch 26, batch 300, aishell_loss[loss=0.1172, simple_loss=0.2048, pruned_loss=0.01479, over 4822.00 frames.], tot_loss[loss=0.1395, simple_loss=0.2216, pruned_loss=0.0287, over 766352.66 frames.], batch size: 29, aishell_tot_loss[loss=0.1426, simple_loss=0.2294, pruned_loss=0.02789, over 481911.99 frames.], datatang_tot_loss[loss=0.1373, simple_loss=0.2155, pruned_loss=0.02952, over 556762.33 frames.], batch size: 29, lr: 3.18e-04
+2022-06-19 04:47:36,377 INFO [train.py:874] (2/4) Epoch 26, batch 350, aishell_loss[loss=0.1444, simple_loss=0.2323, pruned_loss=0.02827, over 4881.00 frames.], tot_loss[loss=0.14, simple_loss=0.2218, pruned_loss=0.0291, over 814737.15 frames.], batch size: 42, aishell_tot_loss[loss=0.144, simple_loss=0.2306, pruned_loss=0.02864, over 536643.64 frames.], datatang_tot_loss[loss=0.1367, simple_loss=0.2146, pruned_loss=0.02943, over 610925.53 frames.], batch size: 42, lr: 3.18e-04
+2022-06-19 04:48:05,362 INFO [train.py:874] (2/4) Epoch 26, batch 400, aishell_loss[loss=0.1273, simple_loss=0.2183, pruned_loss=0.01811, over 4927.00 frames.], tot_loss[loss=0.1398, simple_loss=0.2215, pruned_loss=0.02907, over 852548.91 frames.], batch size: 52, aishell_tot_loss[loss=0.1439, simple_loss=0.2303, pruned_loss=0.0288, over 595482.54 frames.], datatang_tot_loss[loss=0.1363, simple_loss=0.2139, pruned_loss=0.0293, over 649935.82 frames.], batch size: 52, lr: 3.18e-04
+2022-06-19 04:48:32,765 INFO [train.py:874] (2/4) Epoch 26, batch 450, aishell_loss[loss=0.1451, simple_loss=0.2305, pruned_loss=0.02984, over 4883.00 frames.], tot_loss[loss=0.1404, simple_loss=0.2222, pruned_loss=0.02928, over 882126.18 frames.], batch size: 47, aishell_tot_loss[loss=0.1435, simple_loss=0.2299, pruned_loss=0.0286, over 644730.16 frames.], datatang_tot_loss[loss=0.1372, simple_loss=0.2149, pruned_loss=0.02978, over 686693.49 frames.], batch size: 47, lr: 3.18e-04
+2022-06-19 04:49:06,564 INFO [train.py:874] (2/4) Epoch 26, batch 500, datatang_loss[loss=0.1314, simple_loss=0.2123, pruned_loss=0.0253, over 4945.00 frames.], tot_loss[loss=0.1403, simple_loss=0.2218, pruned_loss=0.02938, over 905066.30 frames.], batch size: 50, aishell_tot_loss[loss=0.1435, simple_loss=0.2297, pruned_loss=0.02862, over 675547.30 frames.], datatang_tot_loss[loss=0.1374, simple_loss=0.2151, pruned_loss=0.02987, over 729879.55 frames.], batch size: 50, lr: 3.18e-04
+2022-06-19 04:49:36,224 INFO [train.py:874] (2/4) Epoch 26, batch 550, aishell_loss[loss=0.1394, simple_loss=0.2286, pruned_loss=0.02517, over 4947.00 frames.], tot_loss[loss=0.1393, simple_loss=0.2216, pruned_loss=0.02852, over 922860.26 frames.], batch size: 49, aishell_tot_loss[loss=0.1429, simple_loss=0.2296, pruned_loss=0.02807, over 718885.21 frames.], datatang_tot_loss[loss=0.1365, simple_loss=0.2142, pruned_loss=0.02936, over 754148.00 frames.], batch size: 49, lr: 3.17e-04
+2022-06-19 04:50:04,779 INFO [train.py:874] (2/4) Epoch 26, batch 600, datatang_loss[loss=0.1203, simple_loss=0.1956, pruned_loss=0.02244, over 4957.00 frames.], tot_loss[loss=0.14, simple_loss=0.2219, pruned_loss=0.02907, over 936845.96 frames.], batch size: 55, aishell_tot_loss[loss=0.1429, simple_loss=0.2291, pruned_loss=0.02832, over 748967.24 frames.], datatang_tot_loss[loss=0.1373, simple_loss=0.215, pruned_loss=0.02976, over 782647.77 frames.], batch size: 55, lr: 3.17e-04
+2022-06-19 04:50:33,420 INFO [train.py:874] (2/4) Epoch 26, batch 650, datatang_loss[loss=0.1328, simple_loss=0.2092, pruned_loss=0.02816, over 4965.00 frames.], tot_loss[loss=0.1396, simple_loss=0.2218, pruned_loss=0.02872, over 947823.82 frames.], batch size: 55, aishell_tot_loss[loss=0.1426, simple_loss=0.229, pruned_loss=0.0281, over 777809.54 frames.], datatang_tot_loss[loss=0.137, simple_loss=0.2149, pruned_loss=0.02955, over 805917.92 frames.], batch size: 55, lr: 3.17e-04
+2022-06-19 04:51:04,491 INFO [train.py:874] (2/4) Epoch 26, batch 700, datatang_loss[loss=0.1485, simple_loss=0.2188, pruned_loss=0.0391, over 4971.00 frames.], tot_loss[loss=0.1403, simple_loss=0.2221, pruned_loss=0.02928, over 956319.76 frames.], batch size: 34, aishell_tot_loss[loss=0.1429, simple_loss=0.2291, pruned_loss=0.02838, over 804107.95 frames.], datatang_tot_loss[loss=0.1374, simple_loss=0.2149, pruned_loss=0.02997, over 825619.94 frames.], batch size: 34, lr: 3.17e-04
+2022-06-19 04:51:32,313 INFO [train.py:874] (2/4) Epoch 26, batch 750, aishell_loss[loss=0.1771, simple_loss=0.2705, pruned_loss=0.04187, over 4964.00 frames.], tot_loss[loss=0.1407, simple_loss=0.2224, pruned_loss=0.02948, over 963120.93 frames.], batch size: 79, aishell_tot_loss[loss=0.1431, simple_loss=0.2295, pruned_loss=0.02841, over 825617.33 frames.], datatang_tot_loss[loss=0.1377, simple_loss=0.215, pruned_loss=0.03021, over 844663.98 frames.], batch size: 79, lr: 3.17e-04
+2022-06-19 04:52:00,214 INFO [train.py:874] (2/4) Epoch 26, batch 800, aishell_loss[loss=0.1419, simple_loss=0.2293, pruned_loss=0.02723, over 4970.00 frames.], tot_loss[loss=0.1409, simple_loss=0.2226, pruned_loss=0.02964, over 968145.33 frames.], batch size: 48, aishell_tot_loss[loss=0.1432, simple_loss=0.2297, pruned_loss=0.02837, over 840851.06 frames.], datatang_tot_loss[loss=0.1381, simple_loss=0.2153, pruned_loss=0.03043, over 864393.18 frames.], batch size: 48, lr: 3.17e-04
+2022-06-19 04:52:30,612 INFO [train.py:874] (2/4) Epoch 26, batch 850, aishell_loss[loss=0.1469, simple_loss=0.2322, pruned_loss=0.0308, over 4973.00 frames.], tot_loss[loss=0.1418, simple_loss=0.2234, pruned_loss=0.0301, over 972638.80 frames.], batch size: 44, aishell_tot_loss[loss=0.1439, simple_loss=0.2301, pruned_loss=0.02884, over 861575.73 frames.], datatang_tot_loss[loss=0.1384, simple_loss=0.2156, pruned_loss=0.03063, over 876097.04 frames.], batch size: 44, lr: 3.17e-04
+2022-06-19 04:52:59,208 INFO [train.py:874] (2/4) Epoch 26, batch 900, aishell_loss[loss=0.1429, simple_loss=0.2329, pruned_loss=0.02646, over 4941.00 frames.], tot_loss[loss=0.1421, simple_loss=0.2235, pruned_loss=0.03036, over 975536.80 frames.], batch size: 79, aishell_tot_loss[loss=0.1437, simple_loss=0.23, pruned_loss=0.02874, over 872818.69 frames.], datatang_tot_loss[loss=0.1392, simple_loss=0.2163, pruned_loss=0.03106, over 891906.83 frames.], batch size: 79, lr: 3.17e-04
+2022-06-19 04:53:26,721 INFO [train.py:874] (2/4) Epoch 26, batch 950, aishell_loss[loss=0.1117, simple_loss=0.1919, pruned_loss=0.01578, over 4954.00 frames.], tot_loss[loss=0.1423, simple_loss=0.2238, pruned_loss=0.03045, over 977793.68 frames.], batch size: 27, aishell_tot_loss[loss=0.1437, simple_loss=0.2299, pruned_loss=0.02876, over 885059.90 frames.], datatang_tot_loss[loss=0.1397, simple_loss=0.217, pruned_loss=0.03126, over 903838.80 frames.], batch size: 27, lr: 3.17e-04
+2022-06-19 04:53:57,677 INFO [train.py:874] (2/4) Epoch 26, batch 1000, datatang_loss[loss=0.1291, simple_loss=0.2034, pruned_loss=0.02747, over 4942.00 frames.], tot_loss[loss=0.1423, simple_loss=0.224, pruned_loss=0.03034, over 979819.06 frames.], batch size: 34, aishell_tot_loss[loss=0.1442, simple_loss=0.2307, pruned_loss=0.02885, over 895104.73 frames.], datatang_tot_loss[loss=0.1395, simple_loss=0.2168, pruned_loss=0.03112, over 915208.09 frames.], batch size: 34, lr: 3.17e-04
+2022-06-19 04:53:57,678 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 04:54:14,574 INFO [train.py:914] (2/4) Epoch 26, validation: loss=0.165, simple_loss=0.2487, pruned_loss=0.04059, over 1622729.00 frames.
+2022-06-19 04:54:43,207 INFO [train.py:874] (2/4) Epoch 26, batch 1050, aishell_loss[loss=0.1525, simple_loss=0.2372, pruned_loss=0.03395, over 4953.00 frames.], tot_loss[loss=0.1422, simple_loss=0.2238, pruned_loss=0.03026, over 981297.67 frames.], batch size: 64, aishell_tot_loss[loss=0.1441, simple_loss=0.2307, pruned_loss=0.02873, over 905220.76 frames.], datatang_tot_loss[loss=0.1396, simple_loss=0.2167, pruned_loss=0.03119, over 924137.59 frames.], batch size: 64, lr: 3.17e-04
+2022-06-19 04:55:11,949 INFO [train.py:874] (2/4) Epoch 26, batch 1100, datatang_loss[loss=0.1197, simple_loss=0.1919, pruned_loss=0.02375, over 4950.00 frames.], tot_loss[loss=0.1417, simple_loss=0.2229, pruned_loss=0.03022, over 982393.42 frames.], batch size: 50, aishell_tot_loss[loss=0.1438, simple_loss=0.2303, pruned_loss=0.02865, over 912272.78 frames.], datatang_tot_loss[loss=0.1395, simple_loss=0.2166, pruned_loss=0.03122, over 933275.96 frames.], batch size: 50, lr: 3.17e-04
+2022-06-19 04:55:39,894 INFO [train.py:874] (2/4) Epoch 26, batch 1150, aishell_loss[loss=0.1436, simple_loss=0.2326, pruned_loss=0.02733, over 4889.00 frames.], tot_loss[loss=0.1421, simple_loss=0.2239, pruned_loss=0.0302, over 983117.66 frames.], batch size: 34, aishell_tot_loss[loss=0.1445, simple_loss=0.2313, pruned_loss=0.02889, over 920569.61 frames.], datatang_tot_loss[loss=0.1393, simple_loss=0.2166, pruned_loss=0.03101, over 939701.46 frames.], batch size: 34, lr: 3.17e-04
+2022-06-19 04:56:09,241 INFO [train.py:874] (2/4) Epoch 26, batch 1200, datatang_loss[loss=0.1579, simple_loss=0.2321, pruned_loss=0.04191, over 4962.00 frames.], tot_loss[loss=0.1416, simple_loss=0.2235, pruned_loss=0.02981, over 983697.98 frames.], batch size: 91, aishell_tot_loss[loss=0.1441, simple_loss=0.2305, pruned_loss=0.02882, over 929268.59 frames.], datatang_tot_loss[loss=0.1391, simple_loss=0.2167, pruned_loss=0.03075, over 944422.72 frames.], batch size: 91, lr: 3.16e-04
+2022-06-19 04:56:37,387 INFO [train.py:874] (2/4) Epoch 26, batch 1250, datatang_loss[loss=0.1401, simple_loss=0.2132, pruned_loss=0.03352, over 4931.00 frames.], tot_loss[loss=0.1413, simple_loss=0.2234, pruned_loss=0.02956, over 984369.38 frames.], batch size: 73, aishell_tot_loss[loss=0.1437, simple_loss=0.23, pruned_loss=0.02865, over 938057.78 frames.], datatang_tot_loss[loss=0.139, simple_loss=0.2165, pruned_loss=0.03072, over 947844.22 frames.], batch size: 73, lr: 3.16e-04
+2022-06-19 04:57:05,520 INFO [train.py:874] (2/4) Epoch 26, batch 1300, aishell_loss[loss=0.1362, simple_loss=0.2298, pruned_loss=0.02134, over 4972.00 frames.], tot_loss[loss=0.141, simple_loss=0.2236, pruned_loss=0.02927, over 984944.38 frames.], batch size: 51, aishell_tot_loss[loss=0.1437, simple_loss=0.2305, pruned_loss=0.02847, over 943028.28 frames.], datatang_tot_loss[loss=0.1388, simple_loss=0.2165, pruned_loss=0.03051, over 953032.09 frames.], batch size: 51, lr: 3.16e-04
+2022-06-19 04:57:34,760 INFO [train.py:874] (2/4) Epoch 26, batch 1350, datatang_loss[loss=0.1426, simple_loss=0.2211, pruned_loss=0.03207, over 4933.00 frames.], tot_loss[loss=0.1419, simple_loss=0.2245, pruned_loss=0.02969, over 985084.82 frames.], batch size: 83, aishell_tot_loss[loss=0.1438, simple_loss=0.2305, pruned_loss=0.0285, over 948246.83 frames.], datatang_tot_loss[loss=0.1396, simple_loss=0.2175, pruned_loss=0.03086, over 956654.29 frames.], batch size: 83, lr: 3.16e-04
+2022-06-19 04:58:03,278 INFO [train.py:874] (2/4) Epoch 26, batch 1400, datatang_loss[loss=0.1274, simple_loss=0.2132, pruned_loss=0.02079, over 4917.00 frames.], tot_loss[loss=0.1418, simple_loss=0.224, pruned_loss=0.02976, over 985138.22 frames.], batch size: 81, aishell_tot_loss[loss=0.1438, simple_loss=0.2305, pruned_loss=0.02856, over 952376.34 frames.], datatang_tot_loss[loss=0.1395, simple_loss=0.2172, pruned_loss=0.03086, over 960179.56 frames.], batch size: 81, lr: 3.16e-04
+2022-06-19 04:58:31,202 INFO [train.py:874] (2/4) Epoch 26, batch 1450, aishell_loss[loss=0.1439, simple_loss=0.2288, pruned_loss=0.02951, over 4914.00 frames.], tot_loss[loss=0.1421, simple_loss=0.224, pruned_loss=0.03007, over 985739.70 frames.], batch size: 41, aishell_tot_loss[loss=0.144, simple_loss=0.2306, pruned_loss=0.0287, over 956246.46 frames.], datatang_tot_loss[loss=0.1397, simple_loss=0.2173, pruned_loss=0.03105, over 963678.16 frames.], batch size: 41, lr: 3.16e-04
+2022-06-19 04:59:01,410 INFO [train.py:874] (2/4) Epoch 26, batch 1500, aishell_loss[loss=0.146, simple_loss=0.2416, pruned_loss=0.02521, over 4915.00 frames.], tot_loss[loss=0.1424, simple_loss=0.2247, pruned_loss=0.03006, over 985545.78 frames.], batch size: 41, aishell_tot_loss[loss=0.144, simple_loss=0.2307, pruned_loss=0.02865, over 959963.25 frames.], datatang_tot_loss[loss=0.14, simple_loss=0.2177, pruned_loss=0.03117, over 965938.00 frames.], batch size: 41, lr: 3.16e-04
+2022-06-19 04:59:30,366 INFO [train.py:874] (2/4) Epoch 26, batch 1550, aishell_loss[loss=0.1401, simple_loss=0.225, pruned_loss=0.02759, over 4900.00 frames.], tot_loss[loss=0.142, simple_loss=0.2245, pruned_loss=0.02978, over 985566.15 frames.], batch size: 52, aishell_tot_loss[loss=0.1439, simple_loss=0.2307, pruned_loss=0.02862, over 963327.83 frames.], datatang_tot_loss[loss=0.1397, simple_loss=0.2175, pruned_loss=0.03097, over 967981.16 frames.], batch size: 52, lr: 3.16e-04
+2022-06-19 04:59:58,544 INFO [train.py:874] (2/4) Epoch 26, batch 1600, datatang_loss[loss=0.1311, simple_loss=0.2135, pruned_loss=0.02434, over 4918.00 frames.], tot_loss[loss=0.1424, simple_loss=0.2248, pruned_loss=0.03001, over 985459.70 frames.], batch size: 81, aishell_tot_loss[loss=0.1437, simple_loss=0.2306, pruned_loss=0.02844, over 965886.82 frames.], datatang_tot_loss[loss=0.1405, simple_loss=0.2181, pruned_loss=0.03142, over 969998.73 frames.], batch size: 81, lr: 3.16e-04
+2022-06-19 05:00:28,445 INFO [train.py:874] (2/4) Epoch 26, batch 1650, aishell_loss[loss=0.1066, simple_loss=0.167, pruned_loss=0.02307, over 4838.00 frames.], tot_loss[loss=0.1424, simple_loss=0.2244, pruned_loss=0.0302, over 985844.97 frames.], batch size: 21, aishell_tot_loss[loss=0.1438, simple_loss=0.2304, pruned_loss=0.0286, over 967698.36 frames.], datatang_tot_loss[loss=0.1405, simple_loss=0.2183, pruned_loss=0.03141, over 972614.32 frames.], batch size: 21, lr: 3.16e-04
+2022-06-19 05:00:57,298 INFO [train.py:874] (2/4) Epoch 26, batch 1700, datatang_loss[loss=0.1356, simple_loss=0.2129, pruned_loss=0.02917, over 4935.00 frames.], tot_loss[loss=0.1419, simple_loss=0.2239, pruned_loss=0.02996, over 985794.02 frames.], batch size: 79, aishell_tot_loss[loss=0.1438, simple_loss=0.2303, pruned_loss=0.02865, over 969627.97 frames.], datatang_tot_loss[loss=0.1401, simple_loss=0.2179, pruned_loss=0.03114, over 974330.60 frames.], batch size: 79, lr: 3.16e-04
+2022-06-19 05:01:26,594 INFO [train.py:874] (2/4) Epoch 26, batch 1750, datatang_loss[loss=0.1332, simple_loss=0.2244, pruned_loss=0.02094, over 4941.00 frames.], tot_loss[loss=0.1416, simple_loss=0.2233, pruned_loss=0.02991, over 985688.35 frames.], batch size: 88, aishell_tot_loss[loss=0.1439, simple_loss=0.2301, pruned_loss=0.02882, over 971313.38 frames.], datatang_tot_loss[loss=0.1396, simple_loss=0.2174, pruned_loss=0.03091, over 975763.73 frames.], batch size: 88, lr: 3.16e-04
+2022-06-19 05:01:57,142 INFO [train.py:874] (2/4) Epoch 26, batch 1800, aishell_loss[loss=0.1461, simple_loss=0.2376, pruned_loss=0.02732, over 4944.00 frames.], tot_loss[loss=0.141, simple_loss=0.2224, pruned_loss=0.02983, over 985576.54 frames.], batch size: 40, aishell_tot_loss[loss=0.1438, simple_loss=0.2299, pruned_loss=0.02891, over 972659.77 frames.], datatang_tot_loss[loss=0.1391, simple_loss=0.2168, pruned_loss=0.0307, over 977080.26 frames.], batch size: 40, lr: 3.16e-04
+2022-06-19 05:02:25,515 INFO [train.py:874] (2/4) Epoch 26, batch 1850, datatang_loss[loss=0.1505, simple_loss=0.2265, pruned_loss=0.03728, over 4928.00 frames.], tot_loss[loss=0.1418, simple_loss=0.2231, pruned_loss=0.03021, over 985457.25 frames.], batch size: 83, aishell_tot_loss[loss=0.1444, simple_loss=0.2305, pruned_loss=0.02917, over 973570.80 frames.], datatang_tot_loss[loss=0.1394, simple_loss=0.2171, pruned_loss=0.0308, over 978417.66 frames.], batch size: 83, lr: 3.16e-04
+2022-06-19 05:02:55,373 INFO [train.py:874] (2/4) Epoch 26, batch 1900, datatang_loss[loss=0.1356, simple_loss=0.2089, pruned_loss=0.0312, over 4918.00 frames.], tot_loss[loss=0.1415, simple_loss=0.2233, pruned_loss=0.02983, over 985889.30 frames.], batch size: 42, aishell_tot_loss[loss=0.1445, simple_loss=0.2306, pruned_loss=0.0292, over 975250.01 frames.], datatang_tot_loss[loss=0.1389, simple_loss=0.2169, pruned_loss=0.03043, over 979477.82 frames.], batch size: 42, lr: 3.15e-04
+2022-06-19 05:03:25,568 INFO [train.py:874] (2/4) Epoch 26, batch 1950, datatang_loss[loss=0.1378, simple_loss=0.2159, pruned_loss=0.02988, over 4973.00 frames.], tot_loss[loss=0.1417, simple_loss=0.2238, pruned_loss=0.02979, over 986590.53 frames.], batch size: 45, aishell_tot_loss[loss=0.145, simple_loss=0.2313, pruned_loss=0.02941, over 977129.00 frames.], datatang_tot_loss[loss=0.1385, simple_loss=0.2166, pruned_loss=0.03019, over 980371.71 frames.], batch size: 45, lr: 3.15e-04
+2022-06-19 05:03:53,621 INFO [train.py:874] (2/4) Epoch 26, batch 2000, datatang_loss[loss=0.1687, simple_loss=0.2442, pruned_loss=0.04659, over 4919.00 frames.], tot_loss[loss=0.1423, simple_loss=0.2245, pruned_loss=0.03001, over 986244.74 frames.], batch size: 42, aishell_tot_loss[loss=0.1454, simple_loss=0.2318, pruned_loss=0.02955, over 977938.31 frames.], datatang_tot_loss[loss=0.1386, simple_loss=0.2166, pruned_loss=0.03027, over 981080.62 frames.], batch size: 42, lr: 3.15e-04
+2022-06-19 05:03:53,621 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 05:04:09,286 INFO [train.py:914] (2/4) Epoch 26, validation: loss=0.1659, simple_loss=0.2484, pruned_loss=0.04175, over 1622729.00 frames.
+2022-06-19 05:04:38,949 INFO [train.py:874] (2/4) Epoch 26, batch 2050, aishell_loss[loss=0.1854, simple_loss=0.2747, pruned_loss=0.04808, over 4949.00 frames.], tot_loss[loss=0.142, simple_loss=0.2243, pruned_loss=0.02988, over 985963.69 frames.], batch size: 80, aishell_tot_loss[loss=0.1452, simple_loss=0.2316, pruned_loss=0.02942, over 978467.65 frames.], datatang_tot_loss[loss=0.1386, simple_loss=0.2167, pruned_loss=0.03029, over 981836.05 frames.], batch size: 80, lr: 3.15e-04
+2022-06-19 05:05:07,638 INFO [train.py:874] (2/4) Epoch 26, batch 2100, datatang_loss[loss=0.1826, simple_loss=0.247, pruned_loss=0.05911, over 4926.00 frames.], tot_loss[loss=0.1423, simple_loss=0.2244, pruned_loss=0.03012, over 985619.49 frames.], batch size: 81, aishell_tot_loss[loss=0.1449, simple_loss=0.2313, pruned_loss=0.02925, over 979019.52 frames.], datatang_tot_loss[loss=0.1393, simple_loss=0.2172, pruned_loss=0.03069, over 982270.56 frames.], batch size: 81, lr: 3.15e-04
+2022-06-19 05:05:37,215 INFO [train.py:874] (2/4) Epoch 26, batch 2150, aishell_loss[loss=0.116, simple_loss=0.197, pruned_loss=0.01748, over 4940.00 frames.], tot_loss[loss=0.1422, simple_loss=0.2244, pruned_loss=0.02996, over 985783.51 frames.], batch size: 27, aishell_tot_loss[loss=0.1447, simple_loss=0.231, pruned_loss=0.02918, over 980083.57 frames.], datatang_tot_loss[loss=0.1393, simple_loss=0.2173, pruned_loss=0.03065, over 982593.52 frames.], batch size: 27, lr: 3.15e-04
+2022-06-19 05:06:06,679 INFO [train.py:874] (2/4) Epoch 26, batch 2200, datatang_loss[loss=0.1359, simple_loss=0.2089, pruned_loss=0.03143, over 4936.00 frames.], tot_loss[loss=0.1428, simple_loss=0.2248, pruned_loss=0.0304, over 985952.84 frames.], batch size: 37, aishell_tot_loss[loss=0.1448, simple_loss=0.2314, pruned_loss=0.02914, over 980834.25 frames.], datatang_tot_loss[loss=0.14, simple_loss=0.2177, pruned_loss=0.03112, over 983026.42 frames.], batch size: 37, lr: 3.15e-04
+2022-06-19 05:06:34,930 INFO [train.py:874] (2/4) Epoch 26, batch 2250, datatang_loss[loss=0.1368, simple_loss=0.2151, pruned_loss=0.02926, over 4947.00 frames.], tot_loss[loss=0.1426, simple_loss=0.2245, pruned_loss=0.03029, over 985635.42 frames.], batch size: 62, aishell_tot_loss[loss=0.1448, simple_loss=0.2313, pruned_loss=0.02917, over 981270.06 frames.], datatang_tot_loss[loss=0.1397, simple_loss=0.2172, pruned_loss=0.03108, over 983239.42 frames.], batch size: 62, lr: 3.15e-04
+2022-06-19 05:07:05,041 INFO [train.py:874] (2/4) Epoch 26, batch 2300, datatang_loss[loss=0.1429, simple_loss=0.226, pruned_loss=0.02991, over 4925.00 frames.], tot_loss[loss=0.1432, simple_loss=0.2252, pruned_loss=0.03061, over 985601.81 frames.], batch size: 71, aishell_tot_loss[loss=0.1449, simple_loss=0.2315, pruned_loss=0.02918, over 981675.27 frames.], datatang_tot_loss[loss=0.1404, simple_loss=0.2179, pruned_loss=0.03144, over 983581.98 frames.], batch size: 71, lr: 3.15e-04
+2022-06-19 05:07:33,214 INFO [train.py:874] (2/4) Epoch 26, batch 2350, aishell_loss[loss=0.1362, simple_loss=0.2336, pruned_loss=0.01944, over 4885.00 frames.], tot_loss[loss=0.1418, simple_loss=0.2239, pruned_loss=0.02992, over 985548.92 frames.], batch size: 42, aishell_tot_loss[loss=0.1441, simple_loss=0.2308, pruned_loss=0.02869, over 981965.50 frames.], datatang_tot_loss[loss=0.1399, simple_loss=0.2173, pruned_loss=0.03126, over 983937.26 frames.], batch size: 42, lr: 3.15e-04
+2022-06-19 05:08:00,961 INFO [train.py:874] (2/4) Epoch 26, batch 2400, datatang_loss[loss=0.1219, simple_loss=0.2072, pruned_loss=0.01834, over 4921.00 frames.], tot_loss[loss=0.1417, simple_loss=0.2237, pruned_loss=0.02982, over 985472.01 frames.], batch size: 73, aishell_tot_loss[loss=0.1441, simple_loss=0.231, pruned_loss=0.02865, over 982240.44 frames.], datatang_tot_loss[loss=0.1397, simple_loss=0.217, pruned_loss=0.03119, over 984197.56 frames.], batch size: 73, lr: 3.15e-04
+2022-06-19 05:08:30,374 INFO [train.py:874] (2/4) Epoch 26, batch 2450, aishell_loss[loss=0.117, simple_loss=0.2029, pruned_loss=0.01557, over 4980.00 frames.], tot_loss[loss=0.1406, simple_loss=0.2228, pruned_loss=0.02916, over 985337.67 frames.], batch size: 27, aishell_tot_loss[loss=0.1438, simple_loss=0.2307, pruned_loss=0.02838, over 982349.91 frames.], datatang_tot_loss[loss=0.1389, simple_loss=0.2164, pruned_loss=0.03071, over 984460.31 frames.], batch size: 27, lr: 3.15e-04
+2022-06-19 05:08:59,235 INFO [train.py:874] (2/4) Epoch 26, batch 2500, aishell_loss[loss=0.17, simple_loss=0.2457, pruned_loss=0.04714, over 4922.00 frames.], tot_loss[loss=0.1414, simple_loss=0.2237, pruned_loss=0.02953, over 985227.77 frames.], batch size: 33, aishell_tot_loss[loss=0.1443, simple_loss=0.2312, pruned_loss=0.02871, over 982531.43 frames.], datatang_tot_loss[loss=0.139, simple_loss=0.2166, pruned_loss=0.0307, over 984621.91 frames.], batch size: 33, lr: 3.15e-04
+2022-06-19 05:09:26,860 INFO [train.py:874] (2/4) Epoch 26, batch 2550, aishell_loss[loss=0.1329, simple_loss=0.2184, pruned_loss=0.02371, over 4959.00 frames.], tot_loss[loss=0.141, simple_loss=0.2237, pruned_loss=0.02912, over 985425.77 frames.], batch size: 56, aishell_tot_loss[loss=0.1439, simple_loss=0.231, pruned_loss=0.02841, over 982885.09 frames.], datatang_tot_loss[loss=0.1388, simple_loss=0.2165, pruned_loss=0.03058, over 984912.40 frames.], batch size: 56, lr: 3.14e-04
+2022-06-19 05:09:56,386 INFO [train.py:874] (2/4) Epoch 26, batch 2600, datatang_loss[loss=0.1182, simple_loss=0.1868, pruned_loss=0.02479, over 4964.00 frames.], tot_loss[loss=0.1403, simple_loss=0.2226, pruned_loss=0.02899, over 985268.09 frames.], batch size: 34, aishell_tot_loss[loss=0.1438, simple_loss=0.2307, pruned_loss=0.0285, over 983084.46 frames.], datatang_tot_loss[loss=0.138, simple_loss=0.2154, pruned_loss=0.03028, over 984928.56 frames.], batch size: 34, lr: 3.14e-04
+2022-06-19 05:10:26,625 INFO [train.py:874] (2/4) Epoch 26, batch 2650, datatang_loss[loss=0.153, simple_loss=0.2284, pruned_loss=0.03884, over 4901.00 frames.], tot_loss[loss=0.1403, simple_loss=0.2224, pruned_loss=0.0291, over 985066.59 frames.], batch size: 52, aishell_tot_loss[loss=0.1436, simple_loss=0.2303, pruned_loss=0.02847, over 983049.76 frames.], datatang_tot_loss[loss=0.1382, simple_loss=0.2157, pruned_loss=0.0303, over 985029.74 frames.], batch size: 52, lr: 3.14e-04
+2022-06-19 05:10:56,066 INFO [train.py:874] (2/4) Epoch 26, batch 2700, datatang_loss[loss=0.178, simple_loss=0.2565, pruned_loss=0.0497, over 4955.00 frames.], tot_loss[loss=0.1411, simple_loss=0.2234, pruned_loss=0.02946, over 985453.42 frames.], batch size: 99, aishell_tot_loss[loss=0.1441, simple_loss=0.2307, pruned_loss=0.02869, over 983555.31 frames.], datatang_tot_loss[loss=0.1385, simple_loss=0.2162, pruned_loss=0.03038, over 985171.07 frames.], batch size: 99, lr: 3.14e-04
+2022-06-19 05:11:24,555 INFO [train.py:874] (2/4) Epoch 26, batch 2750, aishell_loss[loss=0.1579, simple_loss=0.238, pruned_loss=0.03888, over 4883.00 frames.], tot_loss[loss=0.1426, simple_loss=0.2248, pruned_loss=0.03017, over 985741.82 frames.], batch size: 42, aishell_tot_loss[loss=0.1448, simple_loss=0.2315, pruned_loss=0.02899, over 984080.98 frames.], datatang_tot_loss[loss=0.1391, simple_loss=0.2165, pruned_loss=0.03082, over 985239.02 frames.], batch size: 42, lr: 3.14e-04
+2022-06-19 05:11:53,309 INFO [train.py:874] (2/4) Epoch 26, batch 2800, aishell_loss[loss=0.1308, simple_loss=0.218, pruned_loss=0.02179, over 4909.00 frames.], tot_loss[loss=0.1426, simple_loss=0.2252, pruned_loss=0.03003, over 985476.79 frames.], batch size: 52, aishell_tot_loss[loss=0.1443, simple_loss=0.2312, pruned_loss=0.02872, over 983986.50 frames.], datatang_tot_loss[loss=0.1395, simple_loss=0.2167, pruned_loss=0.03112, over 985366.01 frames.], batch size: 52, lr: 3.14e-04
+2022-06-19 05:12:22,079 INFO [train.py:874] (2/4) Epoch 26, batch 2850, aishell_loss[loss=0.1756, simple_loss=0.2684, pruned_loss=0.04146, over 4948.00 frames.], tot_loss[loss=0.1427, simple_loss=0.225, pruned_loss=0.03013, over 985724.98 frames.], batch size: 32, aishell_tot_loss[loss=0.1437, simple_loss=0.2305, pruned_loss=0.0285, over 984589.24 frames.], datatang_tot_loss[loss=0.1402, simple_loss=0.2173, pruned_loss=0.03153, over 985229.24 frames.], batch size: 32, lr: 3.14e-04
+2022-06-19 05:12:50,267 INFO [train.py:874] (2/4) Epoch 26, batch 2900, datatang_loss[loss=0.1494, simple_loss=0.2098, pruned_loss=0.0445, over 4940.00 frames.], tot_loss[loss=0.1422, simple_loss=0.225, pruned_loss=0.02971, over 985718.48 frames.], batch size: 50, aishell_tot_loss[loss=0.1436, simple_loss=0.2305, pruned_loss=0.02834, over 984611.83 frames.], datatang_tot_loss[loss=0.14, simple_loss=0.2173, pruned_loss=0.03133, over 985410.72 frames.], batch size: 50, lr: 3.14e-04
+2022-06-19 05:13:18,641 INFO [train.py:874] (2/4) Epoch 26, batch 2950, aishell_loss[loss=0.154, simple_loss=0.2442, pruned_loss=0.03185, over 4966.00 frames.], tot_loss[loss=0.1419, simple_loss=0.2247, pruned_loss=0.02954, over 985920.27 frames.], batch size: 39, aishell_tot_loss[loss=0.1437, simple_loss=0.2306, pruned_loss=0.02835, over 985139.70 frames.], datatang_tot_loss[loss=0.1396, simple_loss=0.2169, pruned_loss=0.03117, over 985266.72 frames.], batch size: 39, lr: 3.14e-04
+2022-06-19 05:13:48,852 INFO [train.py:874] (2/4) Epoch 26, batch 3000, aishell_loss[loss=0.1503, simple_loss=0.2432, pruned_loss=0.02872, over 4930.00 frames.], tot_loss[loss=0.1427, simple_loss=0.2254, pruned_loss=0.02998, over 985921.22 frames.], batch size: 68, aishell_tot_loss[loss=0.1442, simple_loss=0.2309, pruned_loss=0.0287, over 985463.89 frames.], datatang_tot_loss[loss=0.1399, simple_loss=0.2172, pruned_loss=0.03132, over 985126.60 frames.], batch size: 68, lr: 3.14e-04
+2022-06-19 05:13:48,853 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 05:14:05,912 INFO [train.py:914] (2/4) Epoch 26, validation: loss=0.1644, simple_loss=0.2479, pruned_loss=0.04043, over 1622729.00 frames.
+2022-06-19 05:14:34,724 INFO [train.py:874] (2/4) Epoch 26, batch 3050, aishell_loss[loss=0.1514, simple_loss=0.2455, pruned_loss=0.02865, over 4904.00 frames.], tot_loss[loss=0.1424, simple_loss=0.2248, pruned_loss=0.03004, over 986001.71 frames.], batch size: 60, aishell_tot_loss[loss=0.1446, simple_loss=0.2314, pruned_loss=0.02893, over 985615.16 frames.], datatang_tot_loss[loss=0.1395, simple_loss=0.2169, pruned_loss=0.03107, over 985231.69 frames.], batch size: 60, lr: 3.14e-04
+2022-06-19 05:15:02,362 INFO [train.py:874] (2/4) Epoch 26, batch 3100, datatang_loss[loss=0.1287, simple_loss=0.2067, pruned_loss=0.02536, over 4920.00 frames.], tot_loss[loss=0.1423, simple_loss=0.2244, pruned_loss=0.03009, over 986033.11 frames.], batch size: 83, aishell_tot_loss[loss=0.1447, simple_loss=0.2316, pruned_loss=0.02894, over 985616.98 frames.], datatang_tot_loss[loss=0.1395, simple_loss=0.2168, pruned_loss=0.03105, over 985413.06 frames.], batch size: 83, lr: 3.14e-04
+2022-06-19 05:15:30,147 INFO [train.py:874] (2/4) Epoch 26, batch 3150, datatang_loss[loss=0.1407, simple_loss=0.2242, pruned_loss=0.02857, over 4913.00 frames.], tot_loss[loss=0.1426, simple_loss=0.2246, pruned_loss=0.03024, over 986002.91 frames.], batch size: 75, aishell_tot_loss[loss=0.145, simple_loss=0.2319, pruned_loss=0.02905, over 985607.65 frames.], datatang_tot_loss[loss=0.1395, simple_loss=0.2168, pruned_loss=0.03114, over 985518.04 frames.], batch size: 75, lr: 3.14e-04
+2022-06-19 05:16:00,419 INFO [train.py:874] (2/4) Epoch 26, batch 3200, datatang_loss[loss=0.1377, simple_loss=0.2204, pruned_loss=0.02746, over 4953.00 frames.], tot_loss[loss=0.143, simple_loss=0.2249, pruned_loss=0.03054, over 986067.29 frames.], batch size: 86, aishell_tot_loss[loss=0.1448, simple_loss=0.2318, pruned_loss=0.02892, over 985723.49 frames.], datatang_tot_loss[loss=0.1402, simple_loss=0.2172, pruned_loss=0.0316, over 985578.95 frames.], batch size: 86, lr: 3.14e-04
+2022-06-19 05:16:28,118 INFO [train.py:874] (2/4) Epoch 26, batch 3250, aishell_loss[loss=0.1322, simple_loss=0.2334, pruned_loss=0.01557, over 4978.00 frames.], tot_loss[loss=0.1426, simple_loss=0.2249, pruned_loss=0.03012, over 985923.32 frames.], batch size: 48, aishell_tot_loss[loss=0.1447, simple_loss=0.2317, pruned_loss=0.02892, over 985521.18 frames.], datatang_tot_loss[loss=0.1398, simple_loss=0.2171, pruned_loss=0.0313, over 985709.42 frames.], batch size: 48, lr: 3.13e-04
+2022-06-19 05:16:56,613 INFO [train.py:874] (2/4) Epoch 26, batch 3300, datatang_loss[loss=0.1229, simple_loss=0.2038, pruned_loss=0.02099, over 4985.00 frames.], tot_loss[loss=0.1424, simple_loss=0.2243, pruned_loss=0.03027, over 985804.91 frames.], batch size: 31, aishell_tot_loss[loss=0.145, simple_loss=0.232, pruned_loss=0.02903, over 985265.05 frames.], datatang_tot_loss[loss=0.1396, simple_loss=0.2166, pruned_loss=0.03126, over 985887.13 frames.], batch size: 31, lr: 3.13e-04
+2022-06-19 05:17:26,233 INFO [train.py:874] (2/4) Epoch 26, batch 3350, datatang_loss[loss=0.1541, simple_loss=0.224, pruned_loss=0.0421, over 4926.00 frames.], tot_loss[loss=0.142, simple_loss=0.2243, pruned_loss=0.02983, over 985973.32 frames.], batch size: 71, aishell_tot_loss[loss=0.1449, simple_loss=0.2322, pruned_loss=0.02881, over 985566.87 frames.], datatang_tot_loss[loss=0.1392, simple_loss=0.2163, pruned_loss=0.03106, over 985821.18 frames.], batch size: 71, lr: 3.13e-04
+2022-06-19 05:17:54,020 INFO [train.py:874] (2/4) Epoch 26, batch 3400, datatang_loss[loss=0.1243, simple_loss=0.1925, pruned_loss=0.02805, over 4969.00 frames.], tot_loss[loss=0.1422, simple_loss=0.2247, pruned_loss=0.0298, over 985887.61 frames.], batch size: 37, aishell_tot_loss[loss=0.1451, simple_loss=0.2325, pruned_loss=0.02882, over 985476.28 frames.], datatang_tot_loss[loss=0.1391, simple_loss=0.2162, pruned_loss=0.03104, over 985891.31 frames.], batch size: 37, lr: 3.13e-04
+2022-06-19 05:18:21,623 INFO [train.py:874] (2/4) Epoch 26, batch 3450, aishell_loss[loss=0.1122, simple_loss=0.1981, pruned_loss=0.01318, over 4803.00 frames.], tot_loss[loss=0.1424, simple_loss=0.2251, pruned_loss=0.02982, over 986116.92 frames.], batch size: 26, aishell_tot_loss[loss=0.1455, simple_loss=0.2329, pruned_loss=0.02901, over 985557.25 frames.], datatang_tot_loss[loss=0.1391, simple_loss=0.2165, pruned_loss=0.03083, over 986093.67 frames.], batch size: 26, lr: 3.13e-04
+2022-06-19 05:18:52,411 INFO [train.py:874] (2/4) Epoch 26, batch 3500, datatang_loss[loss=0.1169, simple_loss=0.2014, pruned_loss=0.01616, over 4921.00 frames.], tot_loss[loss=0.1412, simple_loss=0.2238, pruned_loss=0.02934, over 986310.21 frames.], batch size: 81, aishell_tot_loss[loss=0.1446, simple_loss=0.2318, pruned_loss=0.02868, over 985534.92 frames.], datatang_tot_loss[loss=0.1389, simple_loss=0.2166, pruned_loss=0.03057, over 986394.11 frames.], batch size: 81, lr: 3.13e-04
+2022-06-19 05:19:22,037 INFO [train.py:874] (2/4) Epoch 26, batch 3550, datatang_loss[loss=0.1265, simple_loss=0.2064, pruned_loss=0.02326, over 4913.00 frames.], tot_loss[loss=0.1402, simple_loss=0.2226, pruned_loss=0.0289, over 985956.72 frames.], batch size: 64, aishell_tot_loss[loss=0.1443, simple_loss=0.2313, pruned_loss=0.0287, over 985336.24 frames.], datatang_tot_loss[loss=0.1379, simple_loss=0.2158, pruned_loss=0.03001, over 986283.89 frames.], batch size: 64, lr: 3.13e-04
+2022-06-19 05:19:50,837 INFO [train.py:874] (2/4) Epoch 26, batch 3600, aishell_loss[loss=0.1346, simple_loss=0.2254, pruned_loss=0.02193, over 4968.00 frames.], tot_loss[loss=0.1402, simple_loss=0.2229, pruned_loss=0.02882, over 985471.88 frames.], batch size: 51, aishell_tot_loss[loss=0.1438, simple_loss=0.2307, pruned_loss=0.02843, over 985047.21 frames.], datatang_tot_loss[loss=0.1383, simple_loss=0.2163, pruned_loss=0.03011, over 986113.45 frames.], batch size: 51, lr: 3.13e-04
+2022-06-19 05:20:19,575 INFO [train.py:874] (2/4) Epoch 26, batch 3650, aishell_loss[loss=0.16, simple_loss=0.2441, pruned_loss=0.03799, over 4925.00 frames.], tot_loss[loss=0.1409, simple_loss=0.2238, pruned_loss=0.02901, over 985665.46 frames.], batch size: 68, aishell_tot_loss[loss=0.1443, simple_loss=0.2312, pruned_loss=0.02873, over 985281.75 frames.], datatang_tot_loss[loss=0.1381, simple_loss=0.2163, pruned_loss=0.02992, over 986076.13 frames.], batch size: 68, lr: 3.13e-04
+2022-06-19 05:20:49,061 INFO [train.py:874] (2/4) Epoch 26, batch 3700, datatang_loss[loss=0.1201, simple_loss=0.1981, pruned_loss=0.02101, over 4939.00 frames.], tot_loss[loss=0.1411, simple_loss=0.2237, pruned_loss=0.02926, over 985403.73 frames.], batch size: 62, aishell_tot_loss[loss=0.1444, simple_loss=0.2312, pruned_loss=0.02874, over 984916.63 frames.], datatang_tot_loss[loss=0.1383, simple_loss=0.2165, pruned_loss=0.03008, over 986125.93 frames.], batch size: 62, lr: 3.13e-04
+2022-06-19 05:21:16,719 INFO [train.py:874] (2/4) Epoch 26, batch 3750, aishell_loss[loss=0.1103, simple_loss=0.1757, pruned_loss=0.0225, over 4740.00 frames.], tot_loss[loss=0.1412, simple_loss=0.2238, pruned_loss=0.02928, over 985152.31 frames.], batch size: 20, aishell_tot_loss[loss=0.1444, simple_loss=0.2313, pruned_loss=0.02882, over 984785.00 frames.], datatang_tot_loss[loss=0.1382, simple_loss=0.2165, pruned_loss=0.02998, over 985973.45 frames.], batch size: 20, lr: 3.13e-04
+2022-06-19 05:21:44,665 INFO [train.py:874] (2/4) Epoch 26, batch 3800, aishell_loss[loss=0.1501, simple_loss=0.2428, pruned_loss=0.02869, over 4977.00 frames.], tot_loss[loss=0.1418, simple_loss=0.2244, pruned_loss=0.02954, over 985435.86 frames.], batch size: 44, aishell_tot_loss[loss=0.1447, simple_loss=0.2316, pruned_loss=0.02889, over 984938.46 frames.], datatang_tot_loss[loss=0.1385, simple_loss=0.2168, pruned_loss=0.03017, over 986091.76 frames.], batch size: 44, lr: 3.13e-04
+2022-06-19 05:22:11,979 INFO [train.py:874] (2/4) Epoch 26, batch 3850, aishell_loss[loss=0.1416, simple_loss=0.242, pruned_loss=0.0206, over 4978.00 frames.], tot_loss[loss=0.1413, simple_loss=0.2238, pruned_loss=0.02941, over 985287.62 frames.], batch size: 51, aishell_tot_loss[loss=0.144, simple_loss=0.2307, pruned_loss=0.02869, over 984959.45 frames.], datatang_tot_loss[loss=0.1387, simple_loss=0.2168, pruned_loss=0.03027, over 985915.49 frames.], batch size: 51, lr: 3.13e-04
+2022-06-19 05:22:40,503 INFO [train.py:874] (2/4) Epoch 26, batch 3900, datatang_loss[loss=0.1256, simple_loss=0.1992, pruned_loss=0.026, over 4888.00 frames.], tot_loss[loss=0.1412, simple_loss=0.2234, pruned_loss=0.02945, over 985234.07 frames.], batch size: 52, aishell_tot_loss[loss=0.1443, simple_loss=0.2309, pruned_loss=0.02881, over 984776.55 frames.], datatang_tot_loss[loss=0.1384, simple_loss=0.2165, pruned_loss=0.03014, over 985974.30 frames.], batch size: 52, lr: 3.12e-04
+2022-06-19 05:23:06,639 INFO [train.py:874] (2/4) Epoch 26, batch 3950, aishell_loss[loss=0.1542, simple_loss=0.2476, pruned_loss=0.03039, over 4873.00 frames.], tot_loss[loss=0.1412, simple_loss=0.2237, pruned_loss=0.02935, over 985163.51 frames.], batch size: 36, aishell_tot_loss[loss=0.1442, simple_loss=0.2308, pruned_loss=0.02876, over 984573.09 frames.], datatang_tot_loss[loss=0.1384, simple_loss=0.2166, pruned_loss=0.03009, over 986104.42 frames.], batch size: 36, lr: 3.12e-04
+2022-06-19 05:23:34,140 INFO [train.py:874] (2/4) Epoch 26, batch 4000, aishell_loss[loss=0.162, simple_loss=0.254, pruned_loss=0.03496, over 4946.00 frames.], tot_loss[loss=0.1407, simple_loss=0.2232, pruned_loss=0.02908, over 985399.69 frames.], batch size: 78, aishell_tot_loss[loss=0.1442, simple_loss=0.2311, pruned_loss=0.02867, over 985064.68 frames.], datatang_tot_loss[loss=0.1379, simple_loss=0.216, pruned_loss=0.02988, over 985818.63 frames.], batch size: 78, lr: 3.12e-04
+2022-06-19 05:23:34,141 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 05:23:50,808 INFO [train.py:914] (2/4) Epoch 26, validation: loss=0.1644, simple_loss=0.2484, pruned_loss=0.04015, over 1622729.00 frames.
+2022-06-19 05:24:19,277 INFO [train.py:874] (2/4) Epoch 26, batch 4050, aishell_loss[loss=0.153, simple_loss=0.236, pruned_loss=0.03498, over 4948.00 frames.], tot_loss[loss=0.1407, simple_loss=0.2234, pruned_loss=0.02902, over 985853.86 frames.], batch size: 31, aishell_tot_loss[loss=0.144, simple_loss=0.2309, pruned_loss=0.02857, over 985561.95 frames.], datatang_tot_loss[loss=0.138, simple_loss=0.2162, pruned_loss=0.02989, over 985808.11 frames.], batch size: 31, lr: 3.12e-04
+2022-06-19 05:24:46,161 INFO [train.py:874] (2/4) Epoch 26, batch 4100, datatang_loss[loss=0.1229, simple_loss=0.21, pruned_loss=0.01786, over 4961.00 frames.], tot_loss[loss=0.141, simple_loss=0.2239, pruned_loss=0.02908, over 986193.75 frames.], batch size: 67, aishell_tot_loss[loss=0.1447, simple_loss=0.2318, pruned_loss=0.02882, over 985680.13 frames.], datatang_tot_loss[loss=0.1375, simple_loss=0.2157, pruned_loss=0.02966, over 986092.21 frames.], batch size: 67, lr: 3.12e-04
+2022-06-19 05:26:04,678 INFO [train.py:874] (2/4) Epoch 27, batch 50, datatang_loss[loss=0.1321, simple_loss=0.2217, pruned_loss=0.02127, over 4911.00 frames.], tot_loss[loss=0.136, simple_loss=0.2177, pruned_loss=0.02715, over 218916.67 frames.], batch size: 98, aishell_tot_loss[loss=0.143, simple_loss=0.2313, pruned_loss=0.02737, over 111983.22 frames.], datatang_tot_loss[loss=0.1297, simple_loss=0.2053, pruned_loss=0.02702, over 120606.12 frames.], batch size: 98, lr: 3.06e-04
+2022-06-19 05:26:34,281 INFO [train.py:874] (2/4) Epoch 27, batch 100, datatang_loss[loss=0.1308, simple_loss=0.2128, pruned_loss=0.0244, over 4925.00 frames.], tot_loss[loss=0.1369, simple_loss=0.219, pruned_loss=0.02737, over 388456.08 frames.], batch size: 83, aishell_tot_loss[loss=0.1435, simple_loss=0.2313, pruned_loss=0.02786, over 206753.22 frames.], datatang_tot_loss[loss=0.131, simple_loss=0.2081, pruned_loss=0.02698, over 229989.39 frames.], batch size: 83, lr: 3.06e-04
+2022-06-19 05:27:02,754 INFO [train.py:874] (2/4) Epoch 27, batch 150, aishell_loss[loss=0.1325, simple_loss=0.2076, pruned_loss=0.02867, over 4964.00 frames.], tot_loss[loss=0.139, simple_loss=0.2219, pruned_loss=0.02809, over 521044.22 frames.], batch size: 25, aishell_tot_loss[loss=0.1435, simple_loss=0.2306, pruned_loss=0.02823, over 335419.41 frames.], datatang_tot_loss[loss=0.1327, simple_loss=0.2099, pruned_loss=0.02777, over 281349.21 frames.], batch size: 25, lr: 3.06e-04
+2022-06-19 05:27:29,181 INFO [train.py:874] (2/4) Epoch 27, batch 200, aishell_loss[loss=0.1585, simple_loss=0.2458, pruned_loss=0.03565, over 4917.00 frames.], tot_loss[loss=0.14, simple_loss=0.2226, pruned_loss=0.02875, over 624453.24 frames.], batch size: 33, aishell_tot_loss[loss=0.1445, simple_loss=0.2315, pruned_loss=0.02875, over 426558.97 frames.], datatang_tot_loss[loss=0.1333, simple_loss=0.2098, pruned_loss=0.02842, over 348594.06 frames.], batch size: 33, lr: 3.06e-04
+2022-06-19 05:27:59,031 INFO [train.py:874] (2/4) Epoch 27, batch 250, datatang_loss[loss=0.1797, simple_loss=0.256, pruned_loss=0.05168, over 4928.00 frames.], tot_loss[loss=0.1393, simple_loss=0.2221, pruned_loss=0.02829, over 704505.61 frames.], batch size: 108, aishell_tot_loss[loss=0.1441, simple_loss=0.2315, pruned_loss=0.02829, over 489912.23 frames.], datatang_tot_loss[loss=0.1333, simple_loss=0.2101, pruned_loss=0.02822, over 426435.24 frames.], batch size: 108, lr: 3.06e-04
+2022-06-19 05:28:28,415 INFO [train.py:874] (2/4) Epoch 27, batch 300, datatang_loss[loss=0.1343, simple_loss=0.206, pruned_loss=0.03128, over 4888.00 frames.], tot_loss[loss=0.1392, simple_loss=0.2222, pruned_loss=0.02806, over 766879.79 frames.], batch size: 39, aishell_tot_loss[loss=0.1439, simple_loss=0.2317, pruned_loss=0.02809, over 550582.95 frames.], datatang_tot_loss[loss=0.1334, simple_loss=0.2106, pruned_loss=0.02809, over 489776.45 frames.], batch size: 39, lr: 3.06e-04
+2022-06-19 05:29:00,122 INFO [train.py:874] (2/4) Epoch 27, batch 350, datatang_loss[loss=0.1426, simple_loss=0.2108, pruned_loss=0.03719, over 4923.00 frames.], tot_loss[loss=0.1401, simple_loss=0.223, pruned_loss=0.02865, over 815341.68 frames.], batch size: 77, aishell_tot_loss[loss=0.1444, simple_loss=0.2319, pruned_loss=0.02845, over 595817.39 frames.], datatang_tot_loss[loss=0.1348, simple_loss=0.2124, pruned_loss=0.0286, over 554858.15 frames.], batch size: 77, lr: 3.06e-04
+2022-06-19 05:29:29,884 INFO [train.py:874] (2/4) Epoch 27, batch 400, datatang_loss[loss=0.1551, simple_loss=0.2268, pruned_loss=0.04167, over 4942.00 frames.], tot_loss[loss=0.1406, simple_loss=0.2233, pruned_loss=0.02889, over 853344.75 frames.], batch size: 37, aishell_tot_loss[loss=0.1449, simple_loss=0.2324, pruned_loss=0.02868, over 636704.56 frames.], datatang_tot_loss[loss=0.1353, simple_loss=0.2131, pruned_loss=0.02873, over 611356.33 frames.], batch size: 37, lr: 3.06e-04
+2022-06-19 05:29:59,161 INFO [train.py:874] (2/4) Epoch 27, batch 450, datatang_loss[loss=0.146, simple_loss=0.2372, pruned_loss=0.02743, over 4863.00 frames.], tot_loss[loss=0.1397, simple_loss=0.2223, pruned_loss=0.02856, over 882575.30 frames.], batch size: 36, aishell_tot_loss[loss=0.1448, simple_loss=0.232, pruned_loss=0.02876, over 665004.72 frames.], datatang_tot_loss[loss=0.1349, simple_loss=0.2133, pruned_loss=0.02826, over 668505.95 frames.], batch size: 36, lr: 3.06e-04
+2022-06-19 05:30:26,671 INFO [train.py:874] (2/4) Epoch 27, batch 500, datatang_loss[loss=0.1448, simple_loss=0.2136, pruned_loss=0.03804, over 4959.00 frames.], tot_loss[loss=0.1398, simple_loss=0.2218, pruned_loss=0.02894, over 905367.20 frames.], batch size: 45, aishell_tot_loss[loss=0.1446, simple_loss=0.2312, pruned_loss=0.02896, over 701280.26 frames.], datatang_tot_loss[loss=0.1352, simple_loss=0.2134, pruned_loss=0.02856, over 707294.12 frames.], batch size: 45, lr: 3.06e-04
+2022-06-19 05:30:56,945 INFO [train.py:874] (2/4) Epoch 27, batch 550, datatang_loss[loss=0.1586, simple_loss=0.2337, pruned_loss=0.04174, over 4974.00 frames.], tot_loss[loss=0.1406, simple_loss=0.2223, pruned_loss=0.02944, over 923155.91 frames.], batch size: 34, aishell_tot_loss[loss=0.1447, simple_loss=0.231, pruned_loss=0.0292, over 730835.83 frames.], datatang_tot_loss[loss=0.1363, simple_loss=0.2145, pruned_loss=0.02904, over 743867.06 frames.], batch size: 34, lr: 3.06e-04
+2022-06-19 05:31:26,667 INFO [train.py:874] (2/4) Epoch 27, batch 600, datatang_loss[loss=0.1627, simple_loss=0.2271, pruned_loss=0.04916, over 4931.00 frames.], tot_loss[loss=0.1405, simple_loss=0.2226, pruned_loss=0.02917, over 936726.97 frames.], batch size: 79, aishell_tot_loss[loss=0.1444, simple_loss=0.231, pruned_loss=0.02895, over 764969.08 frames.], datatang_tot_loss[loss=0.1362, simple_loss=0.2144, pruned_loss=0.02902, over 768074.41 frames.], batch size: 79, lr: 3.06e-04
+2022-06-19 05:31:54,492 INFO [train.py:874] (2/4) Epoch 27, batch 650, aishell_loss[loss=0.1272, simple_loss=0.2176, pruned_loss=0.01838, over 4955.00 frames.], tot_loss[loss=0.1407, simple_loss=0.2225, pruned_loss=0.02943, over 947135.03 frames.], batch size: 64, aishell_tot_loss[loss=0.1443, simple_loss=0.2305, pruned_loss=0.02907, over 790673.36 frames.], datatang_tot_loss[loss=0.1366, simple_loss=0.2147, pruned_loss=0.02927, over 793512.97 frames.], batch size: 64, lr: 3.05e-04
+2022-06-19 05:32:23,240 INFO [train.py:874] (2/4) Epoch 27, batch 700, datatang_loss[loss=0.1751, simple_loss=0.2452, pruned_loss=0.0525, over 4951.00 frames.], tot_loss[loss=0.1407, simple_loss=0.2227, pruned_loss=0.0293, over 955829.45 frames.], batch size: 86, aishell_tot_loss[loss=0.144, simple_loss=0.2301, pruned_loss=0.02889, over 819632.76 frames.], datatang_tot_loss[loss=0.1367, simple_loss=0.2146, pruned_loss=0.02936, over 810194.98 frames.], batch size: 86, lr: 3.05e-04
+2022-06-19 05:32:53,866 INFO [train.py:874] (2/4) Epoch 27, batch 750, aishell_loss[loss=0.1456, simple_loss=0.2301, pruned_loss=0.0306, over 4907.00 frames.], tot_loss[loss=0.1412, simple_loss=0.2233, pruned_loss=0.02952, over 962310.60 frames.], batch size: 33, aishell_tot_loss[loss=0.1444, simple_loss=0.2305, pruned_loss=0.02918, over 838310.37 frames.], datatang_tot_loss[loss=0.137, simple_loss=0.2152, pruned_loss=0.02941, over 831638.64 frames.], batch size: 33, lr: 3.05e-04
+2022-06-19 05:33:22,657 INFO [train.py:874] (2/4) Epoch 27, batch 800, aishell_loss[loss=0.1646, simple_loss=0.2449, pruned_loss=0.0422, over 4960.00 frames.], tot_loss[loss=0.1413, simple_loss=0.2233, pruned_loss=0.02969, over 967401.80 frames.], batch size: 40, aishell_tot_loss[loss=0.1442, simple_loss=0.2301, pruned_loss=0.02911, over 854977.76 frames.], datatang_tot_loss[loss=0.1376, simple_loss=0.2157, pruned_loss=0.02971, over 850398.41 frames.], batch size: 40, lr: 3.05e-04
+2022-06-19 05:33:51,880 INFO [train.py:874] (2/4) Epoch 27, batch 850, datatang_loss[loss=0.1361, simple_loss=0.2118, pruned_loss=0.03023, over 4914.00 frames.], tot_loss[loss=0.1415, simple_loss=0.2235, pruned_loss=0.02976, over 971236.98 frames.], batch size: 57, aishell_tot_loss[loss=0.1443, simple_loss=0.2302, pruned_loss=0.0292, over 870707.86 frames.], datatang_tot_loss[loss=0.1377, simple_loss=0.2159, pruned_loss=0.02978, over 865749.98 frames.], batch size: 57, lr: 3.05e-04
+2022-06-19 05:34:20,054 INFO [train.py:874] (2/4) Epoch 27, batch 900, datatang_loss[loss=0.1555, simple_loss=0.24, pruned_loss=0.03547, over 4936.00 frames.], tot_loss[loss=0.1414, simple_loss=0.2237, pruned_loss=0.02959, over 974569.54 frames.], batch size: 94, aishell_tot_loss[loss=0.1442, simple_loss=0.2304, pruned_loss=0.02896, over 883189.77 frames.], datatang_tot_loss[loss=0.138, simple_loss=0.2162, pruned_loss=0.02988, over 881115.25 frames.], batch size: 94, lr: 3.05e-04
+2022-06-19 05:34:50,691 INFO [train.py:874] (2/4) Epoch 27, batch 950, datatang_loss[loss=0.1997, simple_loss=0.276, pruned_loss=0.06174, over 4956.00 frames.], tot_loss[loss=0.1415, simple_loss=0.2234, pruned_loss=0.02979, over 976989.23 frames.], batch size: 109, aishell_tot_loss[loss=0.1441, simple_loss=0.2301, pruned_loss=0.02908, over 893629.27 frames.], datatang_tot_loss[loss=0.1382, simple_loss=0.2164, pruned_loss=0.03002, over 895016.52 frames.], batch size: 109, lr: 3.05e-04
+2022-06-19 05:35:20,656 INFO [train.py:874] (2/4) Epoch 27, batch 1000, datatang_loss[loss=0.1545, simple_loss=0.2371, pruned_loss=0.03599, over 4963.00 frames.], tot_loss[loss=0.1418, simple_loss=0.2243, pruned_loss=0.02964, over 978828.63 frames.], batch size: 34, aishell_tot_loss[loss=0.1437, simple_loss=0.23, pruned_loss=0.02868, over 907432.04 frames.], datatang_tot_loss[loss=0.1389, simple_loss=0.2171, pruned_loss=0.03036, over 902569.31 frames.], batch size: 34, lr: 3.05e-04
+2022-06-19 05:35:20,656 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 05:35:36,379 INFO [train.py:914] (2/4) Epoch 27, validation: loss=0.1644, simple_loss=0.2483, pruned_loss=0.04028, over 1622729.00 frames.
+2022-06-19 05:36:07,060 INFO [train.py:874] (2/4) Epoch 27, batch 1050, aishell_loss[loss=0.1344, simple_loss=0.2268, pruned_loss=0.02101, over 4864.00 frames.], tot_loss[loss=0.1415, simple_loss=0.2239, pruned_loss=0.02951, over 980241.12 frames.], batch size: 37, aishell_tot_loss[loss=0.1438, simple_loss=0.2305, pruned_loss=0.02853, over 916090.58 frames.], datatang_tot_loss[loss=0.1386, simple_loss=0.2164, pruned_loss=0.03038, over 912830.66 frames.], batch size: 37, lr: 3.05e-04
+2022-06-19 05:36:36,808 INFO [train.py:874] (2/4) Epoch 27, batch 1100, aishell_loss[loss=0.1324, simple_loss=0.2165, pruned_loss=0.02417, over 4921.00 frames.], tot_loss[loss=0.1416, simple_loss=0.2242, pruned_loss=0.02945, over 981479.88 frames.], batch size: 33, aishell_tot_loss[loss=0.1436, simple_loss=0.2303, pruned_loss=0.02842, over 925144.14 frames.], datatang_tot_loss[loss=0.1389, simple_loss=0.2169, pruned_loss=0.03048, over 920536.89 frames.], batch size: 33, lr: 3.05e-04
+2022-06-19 05:37:06,381 INFO [train.py:874] (2/4) Epoch 27, batch 1150, datatang_loss[loss=0.1563, simple_loss=0.2368, pruned_loss=0.03793, over 4880.00 frames.], tot_loss[loss=0.1411, simple_loss=0.2238, pruned_loss=0.02917, over 982344.18 frames.], batch size: 39, aishell_tot_loss[loss=0.1434, simple_loss=0.2301, pruned_loss=0.02837, over 932527.40 frames.], datatang_tot_loss[loss=0.1386, simple_loss=0.2168, pruned_loss=0.03023, over 927862.19 frames.], batch size: 39, lr: 3.05e-04
+2022-06-19 05:37:36,215 INFO [train.py:874] (2/4) Epoch 27, batch 1200, aishell_loss[loss=0.1294, simple_loss=0.2247, pruned_loss=0.0171, over 4954.00 frames.], tot_loss[loss=0.1414, simple_loss=0.224, pruned_loss=0.02939, over 983260.34 frames.], batch size: 31, aishell_tot_loss[loss=0.1436, simple_loss=0.2304, pruned_loss=0.02841, over 938302.69 frames.], datatang_tot_loss[loss=0.1388, simple_loss=0.2168, pruned_loss=0.03038, over 935373.88 frames.], batch size: 31, lr: 3.05e-04
+2022-06-19 05:38:04,625 INFO [train.py:874] (2/4) Epoch 27, batch 1250, aishell_loss[loss=0.1477, simple_loss=0.246, pruned_loss=0.02468, over 4946.00 frames.], tot_loss[loss=0.1415, simple_loss=0.2244, pruned_loss=0.02935, over 983792.65 frames.], batch size: 64, aishell_tot_loss[loss=0.1441, simple_loss=0.2311, pruned_loss=0.0286, over 943505.48 frames.], datatang_tot_loss[loss=0.1385, simple_loss=0.2168, pruned_loss=0.03013, over 941688.29 frames.], batch size: 64, lr: 3.05e-04
+2022-06-19 05:38:34,998 INFO [train.py:874] (2/4) Epoch 27, batch 1300, aishell_loss[loss=0.1625, simple_loss=0.2454, pruned_loss=0.03979, over 4908.00 frames.], tot_loss[loss=0.1408, simple_loss=0.2237, pruned_loss=0.02889, over 984264.67 frames.], batch size: 41, aishell_tot_loss[loss=0.1442, simple_loss=0.2311, pruned_loss=0.0286, over 948266.19 frames.], datatang_tot_loss[loss=0.1377, simple_loss=0.2163, pruned_loss=0.02961, over 947143.94 frames.], batch size: 41, lr: 3.05e-04
+2022-06-19 05:39:05,110 INFO [train.py:874] (2/4) Epoch 27, batch 1350, datatang_loss[loss=0.1332, simple_loss=0.2057, pruned_loss=0.03035, over 4947.00 frames.], tot_loss[loss=0.1397, simple_loss=0.2221, pruned_loss=0.02863, over 984882.17 frames.], batch size: 67, aishell_tot_loss[loss=0.1442, simple_loss=0.2312, pruned_loss=0.02861, over 951148.18 frames.], datatang_tot_loss[loss=0.1368, simple_loss=0.2151, pruned_loss=0.02922, over 953493.46 frames.], batch size: 67, lr: 3.04e-04
+2022-06-19 05:39:34,200 INFO [train.py:874] (2/4) Epoch 27, batch 1400, datatang_loss[loss=0.1343, simple_loss=0.2054, pruned_loss=0.03162, over 4907.00 frames.], tot_loss[loss=0.1402, simple_loss=0.2231, pruned_loss=0.02869, over 984982.56 frames.], batch size: 25, aishell_tot_loss[loss=0.1441, simple_loss=0.2314, pruned_loss=0.02837, over 954997.57 frames.], datatang_tot_loss[loss=0.1373, simple_loss=0.2157, pruned_loss=0.02945, over 957422.42 frames.], batch size: 25, lr: 3.04e-04
+2022-06-19 05:40:02,655 INFO [train.py:874] (2/4) Epoch 27, batch 1450, datatang_loss[loss=0.1199, simple_loss=0.1951, pruned_loss=0.02238, over 4934.00 frames.], tot_loss[loss=0.1411, simple_loss=0.2236, pruned_loss=0.02925, over 985263.61 frames.], batch size: 50, aishell_tot_loss[loss=0.1448, simple_loss=0.2321, pruned_loss=0.02876, over 958472.66 frames.], datatang_tot_loss[loss=0.1374, simple_loss=0.2156, pruned_loss=0.02963, over 960986.98 frames.], batch size: 50, lr: 3.04e-04
+2022-06-19 05:40:32,270 INFO [train.py:874] (2/4) Epoch 27, batch 1500, aishell_loss[loss=0.1227, simple_loss=0.1943, pruned_loss=0.02556, over 4945.00 frames.], tot_loss[loss=0.1415, simple_loss=0.224, pruned_loss=0.02955, over 985810.83 frames.], batch size: 25, aishell_tot_loss[loss=0.1454, simple_loss=0.2326, pruned_loss=0.02912, over 961762.81 frames.], datatang_tot_loss[loss=0.1374, simple_loss=0.2155, pruned_loss=0.02963, over 964288.20 frames.], batch size: 25, lr: 3.04e-04
+2022-06-19 05:41:00,509 INFO [train.py:874] (2/4) Epoch 27, batch 1550, aishell_loss[loss=0.1594, simple_loss=0.2454, pruned_loss=0.03665, over 4944.00 frames.], tot_loss[loss=0.141, simple_loss=0.2232, pruned_loss=0.02934, over 985944.95 frames.], batch size: 56, aishell_tot_loss[loss=0.1451, simple_loss=0.2322, pruned_loss=0.029, over 964423.27 frames.], datatang_tot_loss[loss=0.1372, simple_loss=0.2152, pruned_loss=0.02956, over 967105.88 frames.], batch size: 56, lr: 3.04e-04
+2022-06-19 05:41:29,483 INFO [train.py:874] (2/4) Epoch 27, batch 1600, aishell_loss[loss=0.1442, simple_loss=0.2316, pruned_loss=0.02844, over 4885.00 frames.], tot_loss[loss=0.1406, simple_loss=0.2232, pruned_loss=0.02899, over 986126.80 frames.], batch size: 47, aishell_tot_loss[loss=0.1442, simple_loss=0.2313, pruned_loss=0.02857, over 967690.76 frames.], datatang_tot_loss[loss=0.1374, simple_loss=0.2155, pruned_loss=0.02965, over 968841.01 frames.], batch size: 47, lr: 3.04e-04
+2022-06-19 05:41:58,666 INFO [train.py:874] (2/4) Epoch 27, batch 1650, aishell_loss[loss=0.1588, simple_loss=0.2408, pruned_loss=0.03843, over 4910.00 frames.], tot_loss[loss=0.1409, simple_loss=0.2232, pruned_loss=0.02935, over 985692.51 frames.], batch size: 33, aishell_tot_loss[loss=0.1441, simple_loss=0.2312, pruned_loss=0.02851, over 969178.51 frames.], datatang_tot_loss[loss=0.1378, simple_loss=0.2155, pruned_loss=0.03007, over 971098.42 frames.], batch size: 33, lr: 3.04e-04
+2022-06-19 05:42:26,851 INFO [train.py:874] (2/4) Epoch 27, batch 1700, aishell_loss[loss=0.1559, simple_loss=0.2349, pruned_loss=0.03844, over 4940.00 frames.], tot_loss[loss=0.1412, simple_loss=0.2234, pruned_loss=0.02954, over 985829.72 frames.], batch size: 32, aishell_tot_loss[loss=0.1441, simple_loss=0.231, pruned_loss=0.02858, over 970998.98 frames.], datatang_tot_loss[loss=0.1381, simple_loss=0.2159, pruned_loss=0.0302, over 973077.95 frames.], batch size: 32, lr: 3.04e-04
+2022-06-19 05:42:56,261 INFO [train.py:874] (2/4) Epoch 27, batch 1750, datatang_loss[loss=0.1616, simple_loss=0.2363, pruned_loss=0.04352, over 4978.00 frames.], tot_loss[loss=0.1414, simple_loss=0.2239, pruned_loss=0.02938, over 985875.06 frames.], batch size: 31, aishell_tot_loss[loss=0.1435, simple_loss=0.2305, pruned_loss=0.02826, over 973245.36 frames.], datatang_tot_loss[loss=0.1388, simple_loss=0.2167, pruned_loss=0.03042, over 974159.52 frames.], batch size: 31, lr: 3.04e-04
+2022-06-19 05:43:25,973 INFO [train.py:874] (2/4) Epoch 27, batch 1800, aishell_loss[loss=0.1508, simple_loss=0.2446, pruned_loss=0.02853, over 4979.00 frames.], tot_loss[loss=0.1401, simple_loss=0.2226, pruned_loss=0.02878, over 985633.37 frames.], batch size: 44, aishell_tot_loss[loss=0.1429, simple_loss=0.23, pruned_loss=0.02794, over 974607.36 frames.], datatang_tot_loss[loss=0.1381, simple_loss=0.2159, pruned_loss=0.0301, over 975411.67 frames.], batch size: 44, lr: 3.04e-04
+2022-06-19 05:43:54,024 INFO [train.py:874] (2/4) Epoch 27, batch 1850, aishell_loss[loss=0.1393, simple_loss=0.2282, pruned_loss=0.02517, over 4886.00 frames.], tot_loss[loss=0.1399, simple_loss=0.2226, pruned_loss=0.02859, over 985668.66 frames.], batch size: 47, aishell_tot_loss[loss=0.1423, simple_loss=0.2293, pruned_loss=0.02768, over 975998.33 frames.], datatang_tot_loss[loss=0.1381, simple_loss=0.216, pruned_loss=0.03017, over 976583.42 frames.], batch size: 47, lr: 3.04e-04
+2022-06-19 05:44:24,529 INFO [train.py:874] (2/4) Epoch 27, batch 1900, aishell_loss[loss=0.1599, simple_loss=0.2487, pruned_loss=0.03561, over 4947.00 frames.], tot_loss[loss=0.1406, simple_loss=0.2233, pruned_loss=0.02897, over 985348.56 frames.], batch size: 45, aishell_tot_loss[loss=0.1427, simple_loss=0.2297, pruned_loss=0.02782, over 976784.76 frames.], datatang_tot_loss[loss=0.1385, simple_loss=0.2163, pruned_loss=0.03037, over 977653.39 frames.], batch size: 45, lr: 3.04e-04
+2022-06-19 05:44:54,842 INFO [train.py:874] (2/4) Epoch 27, batch 1950, aishell_loss[loss=0.1559, simple_loss=0.2363, pruned_loss=0.03772, over 4898.00 frames.], tot_loss[loss=0.141, simple_loss=0.2234, pruned_loss=0.02932, over 985136.67 frames.], batch size: 34, aishell_tot_loss[loss=0.1431, simple_loss=0.2301, pruned_loss=0.02803, over 977758.07 frames.], datatang_tot_loss[loss=0.1385, simple_loss=0.216, pruned_loss=0.03052, over 978373.48 frames.], batch size: 34, lr: 3.04e-04
+2022-06-19 05:45:24,370 INFO [train.py:874] (2/4) Epoch 27, batch 2000, datatang_loss[loss=0.1269, simple_loss=0.206, pruned_loss=0.02385, over 4915.00 frames.], tot_loss[loss=0.1415, simple_loss=0.2239, pruned_loss=0.02954, over 985581.43 frames.], batch size: 57, aishell_tot_loss[loss=0.1432, simple_loss=0.2301, pruned_loss=0.02815, over 978736.25 frames.], datatang_tot_loss[loss=0.1389, simple_loss=0.2165, pruned_loss=0.03066, over 979538.67 frames.], batch size: 57, lr: 3.04e-04
+2022-06-19 05:45:24,370 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 05:45:41,439 INFO [train.py:914] (2/4) Epoch 27, validation: loss=0.1643, simple_loss=0.2484, pruned_loss=0.04011, over 1622729.00 frames.
+2022-06-19 05:46:10,064 INFO [train.py:874] (2/4) Epoch 27, batch 2050, aishell_loss[loss=0.128, simple_loss=0.2235, pruned_loss=0.01622, over 4966.00 frames.], tot_loss[loss=0.1418, simple_loss=0.224, pruned_loss=0.02975, over 985780.15 frames.], batch size: 44, aishell_tot_loss[loss=0.1439, simple_loss=0.2307, pruned_loss=0.02851, over 979684.08 frames.], datatang_tot_loss[loss=0.1387, simple_loss=0.2164, pruned_loss=0.0305, over 980282.01 frames.], batch size: 44, lr: 3.04e-04
+2022-06-19 05:46:39,363 INFO [train.py:874] (2/4) Epoch 27, batch 2100, datatang_loss[loss=0.155, simple_loss=0.2331, pruned_loss=0.03845, over 4903.00 frames.], tot_loss[loss=0.1413, simple_loss=0.2239, pruned_loss=0.0293, over 985910.71 frames.], batch size: 64, aishell_tot_loss[loss=0.1433, simple_loss=0.2302, pruned_loss=0.02817, over 980763.41 frames.], datatang_tot_loss[loss=0.1387, simple_loss=0.2165, pruned_loss=0.03046, over 980726.84 frames.], batch size: 64, lr: 3.03e-04
+2022-06-19 05:47:08,908 INFO [train.py:874] (2/4) Epoch 27, batch 2150, aishell_loss[loss=0.1378, simple_loss=0.2283, pruned_loss=0.02367, over 4955.00 frames.], tot_loss[loss=0.1417, simple_loss=0.2245, pruned_loss=0.02945, over 986281.56 frames.], batch size: 56, aishell_tot_loss[loss=0.1438, simple_loss=0.2308, pruned_loss=0.02835, over 981640.55 frames.], datatang_tot_loss[loss=0.1388, simple_loss=0.2166, pruned_loss=0.03043, over 981463.50 frames.], batch size: 56, lr: 3.03e-04
+2022-06-19 05:47:37,160 INFO [train.py:874] (2/4) Epoch 27, batch 2200, aishell_loss[loss=0.1535, simple_loss=0.2442, pruned_loss=0.03137, over 4952.00 frames.], tot_loss[loss=0.1422, simple_loss=0.2245, pruned_loss=0.02992, over 985794.12 frames.], batch size: 64, aishell_tot_loss[loss=0.144, simple_loss=0.231, pruned_loss=0.0285, over 981910.99 frames.], datatang_tot_loss[loss=0.1391, simple_loss=0.2165, pruned_loss=0.03082, over 981793.56 frames.], batch size: 64, lr: 3.03e-04
+2022-06-19 05:48:07,843 INFO [train.py:874] (2/4) Epoch 27, batch 2250, aishell_loss[loss=0.1431, simple_loss=0.2277, pruned_loss=0.02921, over 4954.00 frames.], tot_loss[loss=0.1418, simple_loss=0.2238, pruned_loss=0.02988, over 985987.80 frames.], batch size: 27, aishell_tot_loss[loss=0.1437, simple_loss=0.2306, pruned_loss=0.02842, over 982719.23 frames.], datatang_tot_loss[loss=0.1391, simple_loss=0.2164, pruned_loss=0.0309, over 982111.83 frames.], batch size: 27, lr: 3.03e-04
+2022-06-19 05:48:35,824 INFO [train.py:874] (2/4) Epoch 27, batch 2300, aishell_loss[loss=0.1466, simple_loss=0.2326, pruned_loss=0.03028, over 4917.00 frames.], tot_loss[loss=0.1421, simple_loss=0.2241, pruned_loss=0.03002, over 985730.41 frames.], batch size: 33, aishell_tot_loss[loss=0.1436, simple_loss=0.2304, pruned_loss=0.0284, over 982736.56 frames.], datatang_tot_loss[loss=0.1396, simple_loss=0.2168, pruned_loss=0.03118, over 982665.69 frames.], batch size: 33, lr: 3.03e-04
+2022-06-19 05:49:05,340 INFO [train.py:874] (2/4) Epoch 27, batch 2350, aishell_loss[loss=0.1402, simple_loss=0.2249, pruned_loss=0.02775, over 4860.00 frames.], tot_loss[loss=0.1425, simple_loss=0.2244, pruned_loss=0.0303, over 985883.41 frames.], batch size: 36, aishell_tot_loss[loss=0.1434, simple_loss=0.2301, pruned_loss=0.02838, over 983139.16 frames.], datatang_tot_loss[loss=0.1403, simple_loss=0.2176, pruned_loss=0.03152, over 983138.26 frames.], batch size: 36, lr: 3.03e-04
+2022-06-19 05:49:34,600 INFO [train.py:874] (2/4) Epoch 27, batch 2400, aishell_loss[loss=0.1471, simple_loss=0.2488, pruned_loss=0.02269, over 4940.00 frames.], tot_loss[loss=0.1428, simple_loss=0.2251, pruned_loss=0.03025, over 985848.06 frames.], batch size: 54, aishell_tot_loss[loss=0.1441, simple_loss=0.231, pruned_loss=0.0286, over 983389.84 frames.], datatang_tot_loss[loss=0.1401, simple_loss=0.2175, pruned_loss=0.03136, over 983504.93 frames.], batch size: 54, lr: 3.03e-04
+2022-06-19 05:50:01,632 INFO [train.py:874] (2/4) Epoch 27, batch 2450, datatang_loss[loss=0.1354, simple_loss=0.2214, pruned_loss=0.02474, over 4912.00 frames.], tot_loss[loss=0.1422, simple_loss=0.2247, pruned_loss=0.0299, over 985646.63 frames.], batch size: 77, aishell_tot_loss[loss=0.1439, simple_loss=0.2305, pruned_loss=0.02861, over 983689.44 frames.], datatang_tot_loss[loss=0.1398, simple_loss=0.2176, pruned_loss=0.03105, over 983565.46 frames.], batch size: 77, lr: 3.03e-04
+2022-06-19 05:50:30,997 INFO [train.py:874] (2/4) Epoch 27, batch 2500, aishell_loss[loss=0.1488, simple_loss=0.2425, pruned_loss=0.0276, over 4928.00 frames.], tot_loss[loss=0.143, simple_loss=0.2255, pruned_loss=0.03028, over 986002.02 frames.], batch size: 41, aishell_tot_loss[loss=0.1447, simple_loss=0.2313, pruned_loss=0.02899, over 984008.93 frames.], datatang_tot_loss[loss=0.1399, simple_loss=0.2176, pruned_loss=0.03108, over 984104.87 frames.], batch size: 41, lr: 3.03e-04
+2022-06-19 05:51:00,250 INFO [train.py:874] (2/4) Epoch 27, batch 2550, datatang_loss[loss=0.1156, simple_loss=0.2021, pruned_loss=0.01461, over 4916.00 frames.], tot_loss[loss=0.1426, simple_loss=0.2246, pruned_loss=0.03027, over 985855.59 frames.], batch size: 75, aishell_tot_loss[loss=0.144, simple_loss=0.2306, pruned_loss=0.02871, over 984042.27 frames.], datatang_tot_loss[loss=0.1402, simple_loss=0.2177, pruned_loss=0.03139, over 984370.63 frames.], batch size: 75, lr: 3.03e-04
+2022-06-19 05:51:28,166 INFO [train.py:874] (2/4) Epoch 27, batch 2600, aishell_loss[loss=0.1383, simple_loss=0.2369, pruned_loss=0.01991, over 4873.00 frames.], tot_loss[loss=0.1418, simple_loss=0.224, pruned_loss=0.0298, over 985669.74 frames.], batch size: 34, aishell_tot_loss[loss=0.1432, simple_loss=0.2298, pruned_loss=0.02832, over 984172.10 frames.], datatang_tot_loss[loss=0.1402, simple_loss=0.2176, pruned_loss=0.03142, over 984462.25 frames.], batch size: 34, lr: 3.03e-04
+2022-06-19 05:51:58,806 INFO [train.py:874] (2/4) Epoch 27, batch 2650, aishell_loss[loss=0.1544, simple_loss=0.2348, pruned_loss=0.03701, over 4982.00 frames.], tot_loss[loss=0.1416, simple_loss=0.2238, pruned_loss=0.02971, over 985969.86 frames.], batch size: 51, aishell_tot_loss[loss=0.1435, simple_loss=0.2302, pruned_loss=0.02842, over 984410.73 frames.], datatang_tot_loss[loss=0.1398, simple_loss=0.2173, pruned_loss=0.03113, over 984851.65 frames.], batch size: 51, lr: 3.03e-04
+2022-06-19 05:52:27,275 INFO [train.py:874] (2/4) Epoch 27, batch 2700, datatang_loss[loss=0.138, simple_loss=0.2296, pruned_loss=0.02323, over 4893.00 frames.], tot_loss[loss=0.141, simple_loss=0.2234, pruned_loss=0.02926, over 985823.66 frames.], batch size: 52, aishell_tot_loss[loss=0.1436, simple_loss=0.2302, pruned_loss=0.02851, over 984506.08 frames.], datatang_tot_loss[loss=0.1389, simple_loss=0.2167, pruned_loss=0.0306, over 984922.52 frames.], batch size: 52, lr: 3.03e-04
+2022-06-19 05:52:55,905 INFO [train.py:874] (2/4) Epoch 27, batch 2750, aishell_loss[loss=0.1551, simple_loss=0.242, pruned_loss=0.03415, over 4971.00 frames.], tot_loss[loss=0.1408, simple_loss=0.2231, pruned_loss=0.02927, over 986031.98 frames.], batch size: 44, aishell_tot_loss[loss=0.1437, simple_loss=0.2304, pruned_loss=0.02849, over 984662.65 frames.], datatang_tot_loss[loss=0.1387, simple_loss=0.2163, pruned_loss=0.03053, over 985241.25 frames.], batch size: 44, lr: 3.03e-04
+2022-06-19 05:53:25,470 INFO [train.py:874] (2/4) Epoch 27, batch 2800, aishell_loss[loss=0.1235, simple_loss=0.2084, pruned_loss=0.01927, over 4956.00 frames.], tot_loss[loss=0.141, simple_loss=0.2237, pruned_loss=0.0292, over 986223.32 frames.], batch size: 31, aishell_tot_loss[loss=0.1439, simple_loss=0.2309, pruned_loss=0.02843, over 984999.41 frames.], datatang_tot_loss[loss=0.1387, simple_loss=0.2164, pruned_loss=0.03046, over 985358.73 frames.], batch size: 31, lr: 3.02e-04
+2022-06-19 05:53:55,503 INFO [train.py:874] (2/4) Epoch 27, batch 2850, datatang_loss[loss=0.1346, simple_loss=0.2062, pruned_loss=0.03148, over 4965.00 frames.], tot_loss[loss=0.1411, simple_loss=0.2236, pruned_loss=0.02926, over 985916.32 frames.], batch size: 37, aishell_tot_loss[loss=0.144, simple_loss=0.2311, pruned_loss=0.02839, over 985079.97 frames.], datatang_tot_loss[loss=0.1385, simple_loss=0.2161, pruned_loss=0.03049, over 985207.61 frames.], batch size: 37, lr: 3.02e-04
+2022-06-19 05:54:23,634 INFO [train.py:874] (2/4) Epoch 27, batch 2900, datatang_loss[loss=0.1773, simple_loss=0.2526, pruned_loss=0.05103, over 4928.00 frames.], tot_loss[loss=0.1416, simple_loss=0.2242, pruned_loss=0.02955, over 986062.48 frames.], batch size: 108, aishell_tot_loss[loss=0.1442, simple_loss=0.2313, pruned_loss=0.02853, over 985306.52 frames.], datatang_tot_loss[loss=0.1389, simple_loss=0.2165, pruned_loss=0.03062, over 985320.77 frames.], batch size: 108, lr: 3.02e-04
+2022-06-19 05:54:53,399 INFO [train.py:874] (2/4) Epoch 27, batch 2950, datatang_loss[loss=0.1475, simple_loss=0.2325, pruned_loss=0.03126, over 4927.00 frames.], tot_loss[loss=0.1408, simple_loss=0.2236, pruned_loss=0.02905, over 985958.01 frames.], batch size: 94, aishell_tot_loss[loss=0.1441, simple_loss=0.2312, pruned_loss=0.0285, over 985508.89 frames.], datatang_tot_loss[loss=0.1382, simple_loss=0.2162, pruned_loss=0.03009, over 985193.72 frames.], batch size: 94, lr: 3.02e-04
+2022-06-19 05:55:22,539 INFO [train.py:874] (2/4) Epoch 27, batch 3000, aishell_loss[loss=0.1777, simple_loss=0.2636, pruned_loss=0.04585, over 4976.00 frames.], tot_loss[loss=0.1407, simple_loss=0.2233, pruned_loss=0.02904, over 986135.83 frames.], batch size: 69, aishell_tot_loss[loss=0.144, simple_loss=0.2312, pruned_loss=0.02844, over 985650.04 frames.], datatang_tot_loss[loss=0.1381, simple_loss=0.2162, pruned_loss=0.03004, over 985395.01 frames.], batch size: 69, lr: 3.02e-04
+2022-06-19 05:55:22,539 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 05:55:39,143 INFO [train.py:914] (2/4) Epoch 27, validation: loss=0.1642, simple_loss=0.2486, pruned_loss=0.03985, over 1622729.00 frames.
+2022-06-19 05:56:10,023 INFO [train.py:874] (2/4) Epoch 27, batch 3050, datatang_loss[loss=0.1226, simple_loss=0.2049, pruned_loss=0.02014, over 4931.00 frames.], tot_loss[loss=0.1406, simple_loss=0.2228, pruned_loss=0.02921, over 986135.33 frames.], batch size: 79, aishell_tot_loss[loss=0.1441, simple_loss=0.2309, pruned_loss=0.02859, over 985867.21 frames.], datatang_tot_loss[loss=0.138, simple_loss=0.2159, pruned_loss=0.03001, over 985331.15 frames.], batch size: 79, lr: 3.02e-04
+2022-06-19 05:56:38,573 INFO [train.py:874] (2/4) Epoch 27, batch 3100, datatang_loss[loss=0.1634, simple_loss=0.2344, pruned_loss=0.04617, over 4949.00 frames.], tot_loss[loss=0.1414, simple_loss=0.2239, pruned_loss=0.02944, over 986096.16 frames.], batch size: 91, aishell_tot_loss[loss=0.1443, simple_loss=0.2313, pruned_loss=0.0286, over 985956.87 frames.], datatang_tot_loss[loss=0.1385, simple_loss=0.2165, pruned_loss=0.03022, over 985331.54 frames.], batch size: 91, lr: 3.02e-04
+2022-06-19 05:57:09,356 INFO [train.py:874] (2/4) Epoch 27, batch 3150, datatang_loss[loss=0.1434, simple_loss=0.2207, pruned_loss=0.03301, over 4918.00 frames.], tot_loss[loss=0.1418, simple_loss=0.224, pruned_loss=0.02976, over 986108.37 frames.], batch size: 83, aishell_tot_loss[loss=0.1444, simple_loss=0.2315, pruned_loss=0.02864, over 986007.61 frames.], datatang_tot_loss[loss=0.1389, simple_loss=0.2168, pruned_loss=0.0305, over 985411.28 frames.], batch size: 83, lr: 3.02e-04
+2022-06-19 05:57:40,520 INFO [train.py:874] (2/4) Epoch 27, batch 3200, datatang_loss[loss=0.1411, simple_loss=0.2171, pruned_loss=0.0326, over 4929.00 frames.], tot_loss[loss=0.1426, simple_loss=0.2248, pruned_loss=0.03024, over 986035.81 frames.], batch size: 71, aishell_tot_loss[loss=0.1448, simple_loss=0.2318, pruned_loss=0.02894, over 986051.72 frames.], datatang_tot_loss[loss=0.1393, simple_loss=0.2171, pruned_loss=0.03076, over 985365.16 frames.], batch size: 71, lr: 3.02e-04
+2022-06-19 05:58:09,194 INFO [train.py:874] (2/4) Epoch 27, batch 3250, aishell_loss[loss=0.1416, simple_loss=0.2285, pruned_loss=0.02738, over 4862.00 frames.], tot_loss[loss=0.1426, simple_loss=0.2247, pruned_loss=0.03024, over 985688.25 frames.], batch size: 37, aishell_tot_loss[loss=0.1448, simple_loss=0.232, pruned_loss=0.02881, over 985701.53 frames.], datatang_tot_loss[loss=0.1394, simple_loss=0.2168, pruned_loss=0.031, over 985426.25 frames.], batch size: 37, lr: 3.02e-04
+2022-06-19 05:58:39,173 INFO [train.py:874] (2/4) Epoch 27, batch 3300, aishell_loss[loss=0.1528, simple_loss=0.2354, pruned_loss=0.0351, over 4920.00 frames.], tot_loss[loss=0.1421, simple_loss=0.2242, pruned_loss=0.03002, over 985386.31 frames.], batch size: 41, aishell_tot_loss[loss=0.1447, simple_loss=0.2317, pruned_loss=0.02888, over 985516.54 frames.], datatang_tot_loss[loss=0.1391, simple_loss=0.2166, pruned_loss=0.03077, over 985322.16 frames.], batch size: 41, lr: 3.02e-04
+2022-06-19 05:59:09,428 INFO [train.py:874] (2/4) Epoch 27, batch 3350, aishell_loss[loss=0.1409, simple_loss=0.2296, pruned_loss=0.02608, over 4965.00 frames.], tot_loss[loss=0.142, simple_loss=0.2237, pruned_loss=0.03017, over 985460.27 frames.], batch size: 56, aishell_tot_loss[loss=0.1447, simple_loss=0.2315, pruned_loss=0.02892, over 985408.98 frames.], datatang_tot_loss[loss=0.1392, simple_loss=0.2167, pruned_loss=0.03088, over 985510.79 frames.], batch size: 56, lr: 3.02e-04
+2022-06-19 05:59:36,781 INFO [train.py:874] (2/4) Epoch 27, batch 3400, aishell_loss[loss=0.142, simple_loss=0.2364, pruned_loss=0.02377, over 4943.00 frames.], tot_loss[loss=0.1413, simple_loss=0.2229, pruned_loss=0.02984, over 985565.60 frames.], batch size: 64, aishell_tot_loss[loss=0.1447, simple_loss=0.2316, pruned_loss=0.02889, over 985441.97 frames.], datatang_tot_loss[loss=0.1386, simple_loss=0.216, pruned_loss=0.03058, over 985588.47 frames.], batch size: 64, lr: 3.02e-04
+2022-06-19 06:00:05,980 INFO [train.py:874] (2/4) Epoch 27, batch 3450, aishell_loss[loss=0.1553, simple_loss=0.2329, pruned_loss=0.03879, over 4879.00 frames.], tot_loss[loss=0.1406, simple_loss=0.2226, pruned_loss=0.0293, over 985532.33 frames.], batch size: 34, aishell_tot_loss[loss=0.1441, simple_loss=0.2311, pruned_loss=0.02852, over 985259.61 frames.], datatang_tot_loss[loss=0.1383, simple_loss=0.2159, pruned_loss=0.03039, over 985752.00 frames.], batch size: 34, lr: 3.02e-04
+2022-06-19 06:00:35,350 INFO [train.py:874] (2/4) Epoch 27, batch 3500, aishell_loss[loss=0.1633, simple_loss=0.2482, pruned_loss=0.03923, over 4972.00 frames.], tot_loss[loss=0.1408, simple_loss=0.2227, pruned_loss=0.02945, over 985608.42 frames.], batch size: 44, aishell_tot_loss[loss=0.1443, simple_loss=0.2311, pruned_loss=0.02872, over 985280.90 frames.], datatang_tot_loss[loss=0.1383, simple_loss=0.2159, pruned_loss=0.0303, over 985803.29 frames.], batch size: 44, lr: 3.02e-04
+2022-06-19 06:01:04,243 INFO [train.py:874] (2/4) Epoch 27, batch 3550, datatang_loss[loss=0.1383, simple_loss=0.2056, pruned_loss=0.03554, over 4908.00 frames.], tot_loss[loss=0.1407, simple_loss=0.2223, pruned_loss=0.0295, over 986067.01 frames.], batch size: 64, aishell_tot_loss[loss=0.1443, simple_loss=0.2311, pruned_loss=0.02879, over 985479.16 frames.], datatang_tot_loss[loss=0.1381, simple_loss=0.2158, pruned_loss=0.03023, over 986089.83 frames.], batch size: 64, lr: 3.01e-04
+2022-06-19 06:01:33,725 INFO [train.py:874] (2/4) Epoch 27, batch 3600, aishell_loss[loss=0.1353, simple_loss=0.2203, pruned_loss=0.02515, over 4835.00 frames.], tot_loss[loss=0.1411, simple_loss=0.223, pruned_loss=0.02963, over 986313.04 frames.], batch size: 29, aishell_tot_loss[loss=0.1445, simple_loss=0.2312, pruned_loss=0.02889, over 985817.80 frames.], datatang_tot_loss[loss=0.1384, simple_loss=0.2162, pruned_loss=0.03028, over 986087.83 frames.], batch size: 29, lr: 3.01e-04
+2022-06-19 06:02:05,262 INFO [train.py:874] (2/4) Epoch 27, batch 3650, aishell_loss[loss=0.1444, simple_loss=0.2167, pruned_loss=0.03604, over 4791.00 frames.], tot_loss[loss=0.1408, simple_loss=0.2227, pruned_loss=0.0295, over 986234.83 frames.], batch size: 24, aishell_tot_loss[loss=0.1441, simple_loss=0.2308, pruned_loss=0.02872, over 985691.94 frames.], datatang_tot_loss[loss=0.1386, simple_loss=0.2167, pruned_loss=0.03026, over 986214.39 frames.], batch size: 24, lr: 3.01e-04
+2022-06-19 06:02:32,516 INFO [train.py:874] (2/4) Epoch 27, batch 3700, datatang_loss[loss=0.1392, simple_loss=0.224, pruned_loss=0.02724, over 4957.00 frames.], tot_loss[loss=0.1412, simple_loss=0.2231, pruned_loss=0.02963, over 985971.79 frames.], batch size: 86, aishell_tot_loss[loss=0.1442, simple_loss=0.2308, pruned_loss=0.02882, over 985540.02 frames.], datatang_tot_loss[loss=0.1387, simple_loss=0.2167, pruned_loss=0.03032, over 986158.05 frames.], batch size: 86, lr: 3.01e-04
+2022-06-19 06:03:03,248 INFO [train.py:874] (2/4) Epoch 27, batch 3750, datatang_loss[loss=0.1023, simple_loss=0.1832, pruned_loss=0.01071, over 4922.00 frames.], tot_loss[loss=0.1405, simple_loss=0.2223, pruned_loss=0.02937, over 985576.15 frames.], batch size: 71, aishell_tot_loss[loss=0.1433, simple_loss=0.2299, pruned_loss=0.02836, over 985474.59 frames.], datatang_tot_loss[loss=0.1388, simple_loss=0.2167, pruned_loss=0.03047, over 985833.10 frames.], batch size: 71, lr: 3.01e-04
+2022-06-19 06:03:32,651 INFO [train.py:874] (2/4) Epoch 27, batch 3800, aishell_loss[loss=0.1343, simple_loss=0.2233, pruned_loss=0.02264, over 4916.00 frames.], tot_loss[loss=0.1404, simple_loss=0.2224, pruned_loss=0.02922, over 985711.80 frames.], batch size: 41, aishell_tot_loss[loss=0.1438, simple_loss=0.2303, pruned_loss=0.02868, over 985491.50 frames.], datatang_tot_loss[loss=0.138, simple_loss=0.2161, pruned_loss=0.03, over 985951.76 frames.], batch size: 41, lr: 3.01e-04
+2022-06-19 06:04:01,643 INFO [train.py:874] (2/4) Epoch 27, batch 3850, datatang_loss[loss=0.1633, simple_loss=0.2464, pruned_loss=0.04011, over 4952.00 frames.], tot_loss[loss=0.1408, simple_loss=0.2228, pruned_loss=0.02943, over 985574.16 frames.], batch size: 91, aishell_tot_loss[loss=0.1443, simple_loss=0.2308, pruned_loss=0.02887, over 985276.57 frames.], datatang_tot_loss[loss=0.138, simple_loss=0.2161, pruned_loss=0.02998, over 986008.10 frames.], batch size: 91, lr: 3.01e-04
+2022-06-19 06:04:29,027 INFO [train.py:874] (2/4) Epoch 27, batch 3900, aishell_loss[loss=0.1509, simple_loss=0.2421, pruned_loss=0.02983, over 4852.00 frames.], tot_loss[loss=0.1411, simple_loss=0.2236, pruned_loss=0.02929, over 985244.10 frames.], batch size: 35, aishell_tot_loss[loss=0.1443, simple_loss=0.2311, pruned_loss=0.02877, over 985003.94 frames.], datatang_tot_loss[loss=0.1379, simple_loss=0.2159, pruned_loss=0.02997, over 985953.10 frames.], batch size: 35, lr: 3.01e-04
+2022-06-19 06:04:58,592 INFO [train.py:874] (2/4) Epoch 27, batch 3950, aishell_loss[loss=0.1356, simple_loss=0.22, pruned_loss=0.0256, over 4891.00 frames.], tot_loss[loss=0.1399, simple_loss=0.2223, pruned_loss=0.02879, over 985285.75 frames.], batch size: 47, aishell_tot_loss[loss=0.1438, simple_loss=0.2306, pruned_loss=0.02848, over 984979.09 frames.], datatang_tot_loss[loss=0.1374, simple_loss=0.2154, pruned_loss=0.02969, over 985955.87 frames.], batch size: 47, lr: 3.01e-04
+2022-06-19 06:05:27,339 INFO [train.py:874] (2/4) Epoch 27, batch 4000, datatang_loss[loss=0.1202, simple_loss=0.1974, pruned_loss=0.02149, over 4923.00 frames.], tot_loss[loss=0.14, simple_loss=0.2224, pruned_loss=0.02875, over 985938.53 frames.], batch size: 79, aishell_tot_loss[loss=0.1436, simple_loss=0.2305, pruned_loss=0.02838, over 985533.84 frames.], datatang_tot_loss[loss=0.1374, simple_loss=0.2153, pruned_loss=0.02969, over 986051.71 frames.], batch size: 79, lr: 3.01e-04
+2022-06-19 06:05:27,340 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 06:05:43,549 INFO [train.py:914] (2/4) Epoch 27, validation: loss=0.1646, simple_loss=0.249, pruned_loss=0.0401, over 1622729.00 frames.
+2022-06-19 06:06:11,903 INFO [train.py:874] (2/4) Epoch 27, batch 4050, aishell_loss[loss=0.1206, simple_loss=0.2087, pruned_loss=0.01627, over 4970.00 frames.], tot_loss[loss=0.1396, simple_loss=0.2221, pruned_loss=0.0285, over 985949.98 frames.], batch size: 39, aishell_tot_loss[loss=0.144, simple_loss=0.231, pruned_loss=0.02853, over 985449.43 frames.], datatang_tot_loss[loss=0.1365, simple_loss=0.2146, pruned_loss=0.02921, over 986163.12 frames.], batch size: 39, lr: 3.01e-04
+2022-06-19 06:06:38,048 INFO [train.py:874] (2/4) Epoch 27, batch 4100, datatang_loss[loss=0.1199, simple_loss=0.2014, pruned_loss=0.01919, over 4925.00 frames.], tot_loss[loss=0.1395, simple_loss=0.2224, pruned_loss=0.02825, over 986130.02 frames.], batch size: 64, aishell_tot_loss[loss=0.1439, simple_loss=0.231, pruned_loss=0.02837, over 985482.53 frames.], datatang_tot_loss[loss=0.1362, simple_loss=0.2144, pruned_loss=0.02901, over 986355.11 frames.], batch size: 64, lr: 3.01e-04
+2022-06-19 06:07:06,627 INFO [train.py:874] (2/4) Epoch 27, batch 4150, aishell_loss[loss=0.1594, simple_loss=0.2528, pruned_loss=0.03297, over 4978.00 frames.], tot_loss[loss=0.1406, simple_loss=0.2232, pruned_loss=0.02899, over 985835.20 frames.], batch size: 61, aishell_tot_loss[loss=0.1446, simple_loss=0.2317, pruned_loss=0.02879, over 985366.68 frames.], datatang_tot_loss[loss=0.1366, simple_loss=0.2147, pruned_loss=0.02925, over 986176.18 frames.], batch size: 61, lr: 3.01e-04
+2022-06-19 06:08:12,728 INFO [train.py:874] (2/4) Epoch 28, batch 50, datatang_loss[loss=0.1203, simple_loss=0.1911, pruned_loss=0.02479, over 4964.00 frames.], tot_loss[loss=0.1368, simple_loss=0.2175, pruned_loss=0.02806, over 218535.15 frames.], batch size: 55, aishell_tot_loss[loss=0.1459, simple_loss=0.2303, pruned_loss=0.03071, over 120280.99 frames.], datatang_tot_loss[loss=0.1272, simple_loss=0.2039, pruned_loss=0.02521, over 111893.86 frames.], batch size: 55, lr: 2.95e-04
+2022-06-19 06:08:39,691 INFO [train.py:874] (2/4) Epoch 28, batch 100, aishell_loss[loss=0.139, simple_loss=0.2283, pruned_loss=0.02483, over 4965.00 frames.], tot_loss[loss=0.1375, simple_loss=0.2198, pruned_loss=0.02759, over 388750.19 frames.], batch size: 39, aishell_tot_loss[loss=0.1454, simple_loss=0.2322, pruned_loss=0.02934, over 229836.91 frames.], datatang_tot_loss[loss=0.1285, simple_loss=0.2056, pruned_loss=0.02569, over 207179.25 frames.], batch size: 39, lr: 2.95e-04
+2022-06-19 06:09:09,738 INFO [train.py:874] (2/4) Epoch 28, batch 150, aishell_loss[loss=0.1363, simple_loss=0.2191, pruned_loss=0.02675, over 4933.00 frames.], tot_loss[loss=0.1377, simple_loss=0.2207, pruned_loss=0.0274, over 520780.90 frames.], batch size: 33, aishell_tot_loss[loss=0.1452, simple_loss=0.2318, pruned_loss=0.02931, over 335134.53 frames.], datatang_tot_loss[loss=0.1283, simple_loss=0.2064, pruned_loss=0.02512, over 281349.74 frames.], batch size: 33, lr: 2.95e-04
+2022-06-19 06:09:42,577 INFO [train.py:874] (2/4) Epoch 28, batch 200, datatang_loss[loss=0.124, simple_loss=0.2034, pruned_loss=0.02227, over 4924.00 frames.], tot_loss[loss=0.1375, simple_loss=0.2208, pruned_loss=0.02714, over 623987.49 frames.], batch size: 77, aishell_tot_loss[loss=0.1458, simple_loss=0.2327, pruned_loss=0.02946, over 423149.75 frames.], datatang_tot_loss[loss=0.1271, simple_loss=0.2054, pruned_loss=0.02439, over 351870.53 frames.], batch size: 77, lr: 2.95e-04
+2022-06-19 06:10:12,329 INFO [train.py:874] (2/4) Epoch 28, batch 250, datatang_loss[loss=0.1339, simple_loss=0.2175, pruned_loss=0.02517, over 4878.00 frames.], tot_loss[loss=0.1374, simple_loss=0.2195, pruned_loss=0.02766, over 704017.14 frames.], batch size: 44, aishell_tot_loss[loss=0.1454, simple_loss=0.2318, pruned_loss=0.02943, over 484332.20 frames.], datatang_tot_loss[loss=0.1283, simple_loss=0.2056, pruned_loss=0.02554, over 432017.92 frames.], batch size: 44, lr: 2.95e-04
+2022-06-19 06:10:42,401 INFO [train.py:874] (2/4) Epoch 28, batch 300, aishell_loss[loss=0.1364, simple_loss=0.2287, pruned_loss=0.02205, over 4948.00 frames.], tot_loss[loss=0.1386, simple_loss=0.2208, pruned_loss=0.02819, over 766460.80 frames.], batch size: 49, aishell_tot_loss[loss=0.1459, simple_loss=0.2325, pruned_loss=0.02963, over 545480.94 frames.], datatang_tot_loss[loss=0.1298, simple_loss=0.2072, pruned_loss=0.02623, over 494881.80 frames.], batch size: 49, lr: 2.95e-04
+2022-06-19 06:11:09,564 INFO [train.py:874] (2/4) Epoch 28, batch 350, datatang_loss[loss=0.1774, simple_loss=0.2527, pruned_loss=0.05104, over 4911.00 frames.], tot_loss[loss=0.1378, simple_loss=0.22, pruned_loss=0.02781, over 815272.74 frames.], batch size: 108, aishell_tot_loss[loss=0.1448, simple_loss=0.2318, pruned_loss=0.02891, over 587677.64 frames.], datatang_tot_loss[loss=0.1307, simple_loss=0.2082, pruned_loss=0.02661, over 563430.99 frames.], batch size: 108, lr: 2.95e-04
+2022-06-19 06:11:39,749 INFO [train.py:874] (2/4) Epoch 28, batch 400, aishell_loss[loss=0.1329, simple_loss=0.2205, pruned_loss=0.02267, over 4881.00 frames.], tot_loss[loss=0.1384, simple_loss=0.2205, pruned_loss=0.02813, over 852992.34 frames.], batch size: 28, aishell_tot_loss[loss=0.1446, simple_loss=0.2315, pruned_loss=0.02891, over 629188.60 frames.], datatang_tot_loss[loss=0.132, simple_loss=0.2097, pruned_loss=0.02714, over 618712.52 frames.], batch size: 28, lr: 2.95e-04
+2022-06-19 06:12:10,415 INFO [train.py:874] (2/4) Epoch 28, batch 450, datatang_loss[loss=0.1271, simple_loss=0.2109, pruned_loss=0.02165, over 4919.00 frames.], tot_loss[loss=0.1384, simple_loss=0.2204, pruned_loss=0.02815, over 882394.05 frames.], batch size: 81, aishell_tot_loss[loss=0.1444, simple_loss=0.2313, pruned_loss=0.02878, over 666476.20 frames.], datatang_tot_loss[loss=0.1325, simple_loss=0.2103, pruned_loss=0.02737, over 666690.08 frames.], batch size: 81, lr: 2.95e-04
+2022-06-19 06:12:37,878 INFO [train.py:874] (2/4) Epoch 28, batch 500, aishell_loss[loss=0.1457, simple_loss=0.2173, pruned_loss=0.03702, over 4963.00 frames.], tot_loss[loss=0.1391, simple_loss=0.2209, pruned_loss=0.0287, over 905195.27 frames.], batch size: 31, aishell_tot_loss[loss=0.1445, simple_loss=0.2309, pruned_loss=0.02907, over 708111.01 frames.], datatang_tot_loss[loss=0.1332, simple_loss=0.2108, pruned_loss=0.02782, over 700052.83 frames.], batch size: 31, lr: 2.95e-04
+2022-06-19 06:13:06,539 INFO [train.py:874] (2/4) Epoch 28, batch 550, datatang_loss[loss=0.1326, simple_loss=0.214, pruned_loss=0.02559, over 4956.00 frames.], tot_loss[loss=0.1395, simple_loss=0.2214, pruned_loss=0.02875, over 923252.61 frames.], batch size: 86, aishell_tot_loss[loss=0.1447, simple_loss=0.2311, pruned_loss=0.0292, over 743350.00 frames.], datatang_tot_loss[loss=0.1333, simple_loss=0.211, pruned_loss=0.02782, over 731282.24 frames.], batch size: 86, lr: 2.95e-04
+2022-06-19 06:13:35,837 INFO [train.py:874] (2/4) Epoch 28, batch 600, aishell_loss[loss=0.1539, simple_loss=0.2424, pruned_loss=0.03274, over 4953.00 frames.], tot_loss[loss=0.1393, simple_loss=0.2216, pruned_loss=0.02847, over 937078.85 frames.], batch size: 54, aishell_tot_loss[loss=0.1443, simple_loss=0.2306, pruned_loss=0.02903, over 778392.33 frames.], datatang_tot_loss[loss=0.1332, simple_loss=0.2111, pruned_loss=0.02765, over 754206.81 frames.], batch size: 54, lr: 2.95e-04
+2022-06-19 06:14:02,736 INFO [train.py:874] (2/4) Epoch 28, batch 650, aishell_loss[loss=0.17, simple_loss=0.2485, pruned_loss=0.04578, over 4960.00 frames.], tot_loss[loss=0.1387, simple_loss=0.221, pruned_loss=0.02813, over 947925.23 frames.], batch size: 31, aishell_tot_loss[loss=0.144, simple_loss=0.2304, pruned_loss=0.02878, over 802965.74 frames.], datatang_tot_loss[loss=0.1329, simple_loss=0.2108, pruned_loss=0.02752, over 781392.60 frames.], batch size: 31, lr: 2.94e-04
+2022-06-19 06:14:32,403 INFO [train.py:874] (2/4) Epoch 28, batch 700, aishell_loss[loss=0.1282, simple_loss=0.2209, pruned_loss=0.01774, over 4955.00 frames.], tot_loss[loss=0.1384, simple_loss=0.2205, pruned_loss=0.02814, over 956285.42 frames.], batch size: 56, aishell_tot_loss[loss=0.1435, simple_loss=0.23, pruned_loss=0.02853, over 822808.55 frames.], datatang_tot_loss[loss=0.1332, simple_loss=0.2109, pruned_loss=0.02781, over 807282.90 frames.], batch size: 56, lr: 2.94e-04
+2022-06-19 06:15:01,801 INFO [train.py:874] (2/4) Epoch 28, batch 750, aishell_loss[loss=0.1371, simple_loss=0.2253, pruned_loss=0.0245, over 4912.00 frames.], tot_loss[loss=0.1382, simple_loss=0.2206, pruned_loss=0.02795, over 963092.02 frames.], batch size: 52, aishell_tot_loss[loss=0.1427, simple_loss=0.2293, pruned_loss=0.02808, over 842015.33 frames.], datatang_tot_loss[loss=0.1339, simple_loss=0.2118, pruned_loss=0.02801, over 828571.89 frames.], batch size: 52, lr: 2.94e-04
+2022-06-19 06:15:30,190 INFO [train.py:874] (2/4) Epoch 28, batch 800, datatang_loss[loss=0.1469, simple_loss=0.2235, pruned_loss=0.0352, over 4937.00 frames.], tot_loss[loss=0.1391, simple_loss=0.2212, pruned_loss=0.02845, over 967998.45 frames.], batch size: 69, aishell_tot_loss[loss=0.1431, simple_loss=0.2299, pruned_loss=0.0281, over 854402.91 frames.], datatang_tot_loss[loss=0.1349, simple_loss=0.2126, pruned_loss=0.02858, over 851764.58 frames.], batch size: 69, lr: 2.94e-04
+2022-06-19 06:16:00,393 INFO [train.py:874] (2/4) Epoch 28, batch 850, aishell_loss[loss=0.1562, simple_loss=0.2563, pruned_loss=0.02809, over 4929.00 frames.], tot_loss[loss=0.1398, simple_loss=0.2218, pruned_loss=0.02886, over 971994.78 frames.], batch size: 68, aishell_tot_loss[loss=0.1423, simple_loss=0.229, pruned_loss=0.02784, over 870513.14 frames.], datatang_tot_loss[loss=0.1365, simple_loss=0.2142, pruned_loss=0.02938, over 866958.10 frames.], batch size: 68, lr: 2.94e-04
+2022-06-19 06:16:30,596 INFO [train.py:874] (2/4) Epoch 28, batch 900, datatang_loss[loss=0.1292, simple_loss=0.2047, pruned_loss=0.02685, over 4931.00 frames.], tot_loss[loss=0.1394, simple_loss=0.2214, pruned_loss=0.02872, over 975141.15 frames.], batch size: 57, aishell_tot_loss[loss=0.1421, simple_loss=0.2287, pruned_loss=0.02777, over 884106.80 frames.], datatang_tot_loss[loss=0.1364, simple_loss=0.2141, pruned_loss=0.02933, over 881028.39 frames.], batch size: 57, lr: 2.94e-04
+2022-06-19 06:16:58,022 INFO [train.py:874] (2/4) Epoch 28, batch 950, datatang_loss[loss=0.1296, simple_loss=0.2045, pruned_loss=0.02735, over 4924.00 frames.], tot_loss[loss=0.1393, simple_loss=0.2214, pruned_loss=0.0286, over 977700.36 frames.], batch size: 77, aishell_tot_loss[loss=0.1419, simple_loss=0.2283, pruned_loss=0.02771, over 895790.28 frames.], datatang_tot_loss[loss=0.1365, simple_loss=0.2144, pruned_loss=0.02928, over 893879.39 frames.], batch size: 77, lr: 2.94e-04
+2022-06-19 06:17:29,829 INFO [train.py:874] (2/4) Epoch 28, batch 1000, datatang_loss[loss=0.122, simple_loss=0.212, pruned_loss=0.01601, over 4921.00 frames.], tot_loss[loss=0.1395, simple_loss=0.2221, pruned_loss=0.02841, over 979273.80 frames.], batch size: 77, aishell_tot_loss[loss=0.1421, simple_loss=0.2289, pruned_loss=0.0277, over 906823.45 frames.], datatang_tot_loss[loss=0.1364, simple_loss=0.2146, pruned_loss=0.02911, over 904023.34 frames.], batch size: 77, lr: 2.94e-04
+2022-06-19 06:17:29,830 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 06:17:46,265 INFO [train.py:914] (2/4) Epoch 28, validation: loss=0.1641, simple_loss=0.2478, pruned_loss=0.0402, over 1622729.00 frames.
+2022-06-19 06:18:14,331 INFO [train.py:874] (2/4) Epoch 28, batch 1050, aishell_loss[loss=0.1367, simple_loss=0.2215, pruned_loss=0.02588, over 4921.00 frames.], tot_loss[loss=0.1391, simple_loss=0.2217, pruned_loss=0.02828, over 980419.65 frames.], batch size: 41, aishell_tot_loss[loss=0.1421, simple_loss=0.229, pruned_loss=0.02762, over 915874.82 frames.], datatang_tot_loss[loss=0.1361, simple_loss=0.2142, pruned_loss=0.02905, over 913597.74 frames.], batch size: 41, lr: 2.94e-04
+2022-06-19 06:18:45,861 INFO [train.py:874] (2/4) Epoch 28, batch 1100, aishell_loss[loss=0.1585, simple_loss=0.2364, pruned_loss=0.04031, over 4887.00 frames.], tot_loss[loss=0.1396, simple_loss=0.2223, pruned_loss=0.02849, over 981478.15 frames.], batch size: 42, aishell_tot_loss[loss=0.1427, simple_loss=0.2295, pruned_loss=0.02793, over 923348.33 frames.], datatang_tot_loss[loss=0.1362, simple_loss=0.2145, pruned_loss=0.02894, over 922731.48 frames.], batch size: 42, lr: 2.94e-04
+2022-06-19 06:19:12,533 INFO [train.py:874] (2/4) Epoch 28, batch 1150, aishell_loss[loss=0.1289, simple_loss=0.2174, pruned_loss=0.02027, over 4972.00 frames.], tot_loss[loss=0.1391, simple_loss=0.2217, pruned_loss=0.02825, over 981897.78 frames.], batch size: 44, aishell_tot_loss[loss=0.1426, simple_loss=0.2293, pruned_loss=0.02795, over 930013.85 frames.], datatang_tot_loss[loss=0.1358, simple_loss=0.2143, pruned_loss=0.02866, over 930287.67 frames.], batch size: 44, lr: 2.94e-04
+2022-06-19 06:19:43,165 INFO [train.py:874] (2/4) Epoch 28, batch 1200, datatang_loss[loss=0.1415, simple_loss=0.1998, pruned_loss=0.04162, over 4961.00 frames.], tot_loss[loss=0.1399, simple_loss=0.2227, pruned_loss=0.0286, over 982718.65 frames.], batch size: 45, aishell_tot_loss[loss=0.1428, simple_loss=0.2298, pruned_loss=0.02789, over 937126.02 frames.], datatang_tot_loss[loss=0.1364, simple_loss=0.2146, pruned_loss=0.02909, over 936228.70 frames.], batch size: 45, lr: 2.94e-04
+2022-06-19 06:20:13,505 INFO [train.py:874] (2/4) Epoch 28, batch 1250, aishell_loss[loss=0.1384, simple_loss=0.2309, pruned_loss=0.02294, over 4959.00 frames.], tot_loss[loss=0.1409, simple_loss=0.2235, pruned_loss=0.0291, over 983014.34 frames.], batch size: 64, aishell_tot_loss[loss=0.1432, simple_loss=0.2301, pruned_loss=0.0282, over 942915.24 frames.], datatang_tot_loss[loss=0.137, simple_loss=0.2153, pruned_loss=0.02937, over 941589.35 frames.], batch size: 64, lr: 2.94e-04
+2022-06-19 06:20:39,905 INFO [train.py:874] (2/4) Epoch 28, batch 1300, datatang_loss[loss=0.1268, simple_loss=0.2121, pruned_loss=0.02077, over 4922.00 frames.], tot_loss[loss=0.1402, simple_loss=0.2232, pruned_loss=0.02862, over 983382.73 frames.], batch size: 83, aishell_tot_loss[loss=0.1428, simple_loss=0.2299, pruned_loss=0.02787, over 948633.01 frames.], datatang_tot_loss[loss=0.1368, simple_loss=0.2151, pruned_loss=0.02928, over 945832.37 frames.], batch size: 83, lr: 2.94e-04
+2022-06-19 06:21:10,167 INFO [train.py:874] (2/4) Epoch 28, batch 1350, datatang_loss[loss=0.1577, simple_loss=0.2452, pruned_loss=0.03508, over 4957.00 frames.], tot_loss[loss=0.1409, simple_loss=0.2239, pruned_loss=0.02891, over 984204.90 frames.], batch size: 91, aishell_tot_loss[loss=0.1428, simple_loss=0.2301, pruned_loss=0.02779, over 952968.35 frames.], datatang_tot_loss[loss=0.1376, simple_loss=0.2159, pruned_loss=0.02968, over 950894.05 frames.], batch size: 91, lr: 2.94e-04
+2022-06-19 06:21:39,766 INFO [train.py:874] (2/4) Epoch 28, batch 1400, datatang_loss[loss=0.1359, simple_loss=0.1995, pruned_loss=0.03613, over 4971.00 frames.], tot_loss[loss=0.1413, simple_loss=0.224, pruned_loss=0.02933, over 984496.13 frames.], batch size: 67, aishell_tot_loss[loss=0.1432, simple_loss=0.2304, pruned_loss=0.02799, over 956448.81 frames.], datatang_tot_loss[loss=0.1379, simple_loss=0.2159, pruned_loss=0.02996, over 955326.75 frames.], batch size: 67, lr: 2.93e-04
+2022-06-19 06:22:07,140 INFO [train.py:874] (2/4) Epoch 28, batch 1450, aishell_loss[loss=0.1322, simple_loss=0.2142, pruned_loss=0.02507, over 4925.00 frames.], tot_loss[loss=0.1409, simple_loss=0.2234, pruned_loss=0.02922, over 984797.99 frames.], batch size: 41, aishell_tot_loss[loss=0.1434, simple_loss=0.2305, pruned_loss=0.02819, over 959687.21 frames.], datatang_tot_loss[loss=0.1374, simple_loss=0.2155, pruned_loss=0.02968, over 959147.93 frames.], batch size: 41, lr: 2.93e-04
+2022-06-19 06:22:39,117 INFO [train.py:874] (2/4) Epoch 28, batch 1500, datatang_loss[loss=0.1496, simple_loss=0.2199, pruned_loss=0.0396, over 4976.00 frames.], tot_loss[loss=0.1406, simple_loss=0.2231, pruned_loss=0.02901, over 985006.25 frames.], batch size: 40, aishell_tot_loss[loss=0.1432, simple_loss=0.2302, pruned_loss=0.02812, over 962389.43 frames.], datatang_tot_loss[loss=0.1375, simple_loss=0.2158, pruned_loss=0.02958, over 962638.56 frames.], batch size: 40, lr: 2.93e-04
+2022-06-19 06:23:09,509 INFO [train.py:874] (2/4) Epoch 28, batch 1550, aishell_loss[loss=0.1539, simple_loss=0.2418, pruned_loss=0.03299, over 4945.00 frames.], tot_loss[loss=0.1402, simple_loss=0.2226, pruned_loss=0.02891, over 985855.00 frames.], batch size: 54, aishell_tot_loss[loss=0.143, simple_loss=0.23, pruned_loss=0.02802, over 965382.35 frames.], datatang_tot_loss[loss=0.1373, simple_loss=0.2155, pruned_loss=0.02959, over 965845.93 frames.], batch size: 54, lr: 2.93e-04
+2022-06-19 06:23:37,900 INFO [train.py:874] (2/4) Epoch 28, batch 1600, datatang_loss[loss=0.1166, simple_loss=0.1825, pruned_loss=0.02535, over 4951.00 frames.], tot_loss[loss=0.1394, simple_loss=0.222, pruned_loss=0.02844, over 985710.97 frames.], batch size: 62, aishell_tot_loss[loss=0.1423, simple_loss=0.2294, pruned_loss=0.02765, over 967373.56 frames.], datatang_tot_loss[loss=0.1372, simple_loss=0.2156, pruned_loss=0.02945, over 968451.52 frames.], batch size: 62, lr: 2.93e-04
+2022-06-19 06:24:07,776 INFO [train.py:874] (2/4) Epoch 28, batch 1650, datatang_loss[loss=0.1286, simple_loss=0.2066, pruned_loss=0.02527, over 4920.00 frames.], tot_loss[loss=0.1393, simple_loss=0.222, pruned_loss=0.0283, over 985694.85 frames.], batch size: 83, aishell_tot_loss[loss=0.1425, simple_loss=0.2297, pruned_loss=0.02768, over 969467.83 frames.], datatang_tot_loss[loss=0.1368, simple_loss=0.2151, pruned_loss=0.02925, over 970535.47 frames.], batch size: 83, lr: 2.93e-04
+2022-06-19 06:24:36,381 INFO [train.py:874] (2/4) Epoch 28, batch 1700, aishell_loss[loss=0.1556, simple_loss=0.2485, pruned_loss=0.03136, over 4958.00 frames.], tot_loss[loss=0.1396, simple_loss=0.2228, pruned_loss=0.02819, over 985430.98 frames.], batch size: 40, aishell_tot_loss[loss=0.1423, simple_loss=0.2296, pruned_loss=0.02752, over 971479.97 frames.], datatang_tot_loss[loss=0.1371, simple_loss=0.2155, pruned_loss=0.02929, over 971961.27 frames.], batch size: 40, lr: 2.93e-04
+2022-06-19 06:25:04,621 INFO [train.py:874] (2/4) Epoch 28, batch 1750, datatang_loss[loss=0.1307, simple_loss=0.2001, pruned_loss=0.03069, over 4968.00 frames.], tot_loss[loss=0.1393, simple_loss=0.2226, pruned_loss=0.02804, over 985367.25 frames.], batch size: 34, aishell_tot_loss[loss=0.1421, simple_loss=0.2297, pruned_loss=0.0273, over 973007.07 frames.], datatang_tot_loss[loss=0.137, simple_loss=0.2154, pruned_loss=0.0293, over 973593.04 frames.], batch size: 34, lr: 2.93e-04
+2022-06-19 06:25:35,062 INFO [train.py:874] (2/4) Epoch 28, batch 1800, datatang_loss[loss=0.1428, simple_loss=0.2192, pruned_loss=0.03323, over 4962.00 frames.], tot_loss[loss=0.1393, simple_loss=0.2226, pruned_loss=0.02803, over 984720.98 frames.], batch size: 60, aishell_tot_loss[loss=0.1422, simple_loss=0.2298, pruned_loss=0.02731, over 973937.33 frames.], datatang_tot_loss[loss=0.1368, simple_loss=0.215, pruned_loss=0.02928, over 974819.20 frames.], batch size: 60, lr: 2.93e-04
+2022-06-19 06:26:02,300 INFO [train.py:874] (2/4) Epoch 28, batch 1850, datatang_loss[loss=0.1291, simple_loss=0.1992, pruned_loss=0.02949, over 4922.00 frames.], tot_loss[loss=0.1382, simple_loss=0.2212, pruned_loss=0.02764, over 984739.01 frames.], batch size: 73, aishell_tot_loss[loss=0.1415, simple_loss=0.2289, pruned_loss=0.02705, over 975027.13 frames.], datatang_tot_loss[loss=0.1363, simple_loss=0.2146, pruned_loss=0.02902, over 976171.85 frames.], batch size: 73, lr: 2.93e-04
+2022-06-19 06:26:30,986 INFO [train.py:874] (2/4) Epoch 28, batch 1900, aishell_loss[loss=0.1442, simple_loss=0.2323, pruned_loss=0.02802, over 4968.00 frames.], tot_loss[loss=0.1389, simple_loss=0.2218, pruned_loss=0.02796, over 985110.61 frames.], batch size: 39, aishell_tot_loss[loss=0.1416, simple_loss=0.2291, pruned_loss=0.02703, over 976268.77 frames.], datatang_tot_loss[loss=0.1368, simple_loss=0.215, pruned_loss=0.02931, over 977494.26 frames.], batch size: 39, lr: 2.93e-04
+2022-06-19 06:27:02,459 INFO [train.py:874] (2/4) Epoch 28, batch 1950, aishell_loss[loss=0.1463, simple_loss=0.2262, pruned_loss=0.03318, over 4930.00 frames.], tot_loss[loss=0.1387, simple_loss=0.2215, pruned_loss=0.02789, over 985014.86 frames.], batch size: 33, aishell_tot_loss[loss=0.1414, simple_loss=0.229, pruned_loss=0.02688, over 977132.50 frames.], datatang_tot_loss[loss=0.1367, simple_loss=0.2149, pruned_loss=0.0293, over 978462.28 frames.], batch size: 33, lr: 2.93e-04
+2022-06-19 06:27:30,176 INFO [train.py:874] (2/4) Epoch 28, batch 2000, datatang_loss[loss=0.1339, simple_loss=0.2199, pruned_loss=0.02397, over 4923.00 frames.], tot_loss[loss=0.1399, simple_loss=0.2231, pruned_loss=0.02837, over 985155.83 frames.], batch size: 64, aishell_tot_loss[loss=0.1424, simple_loss=0.2301, pruned_loss=0.02734, over 978328.62 frames.], datatang_tot_loss[loss=0.1369, simple_loss=0.2151, pruned_loss=0.02933, over 979135.28 frames.], batch size: 64, lr: 2.93e-04
+2022-06-19 06:27:30,176 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 06:27:46,237 INFO [train.py:914] (2/4) Epoch 28, validation: loss=0.1643, simple_loss=0.2484, pruned_loss=0.04007, over 1622729.00 frames.
+2022-06-19 06:28:15,603 INFO [train.py:874] (2/4) Epoch 28, batch 2050, datatang_loss[loss=0.1418, simple_loss=0.2127, pruned_loss=0.03547, over 4956.00 frames.], tot_loss[loss=0.1405, simple_loss=0.2236, pruned_loss=0.02876, over 985108.35 frames.], batch size: 34, aishell_tot_loss[loss=0.1429, simple_loss=0.2308, pruned_loss=0.02748, over 979059.47 frames.], datatang_tot_loss[loss=0.1371, simple_loss=0.215, pruned_loss=0.0296, over 979861.21 frames.], batch size: 34, lr: 2.93e-04
+2022-06-19 06:28:43,206 INFO [train.py:874] (2/4) Epoch 28, batch 2100, datatang_loss[loss=0.1224, simple_loss=0.1997, pruned_loss=0.02257, over 4925.00 frames.], tot_loss[loss=0.14, simple_loss=0.2228, pruned_loss=0.02861, over 985493.08 frames.], batch size: 79, aishell_tot_loss[loss=0.1429, simple_loss=0.2305, pruned_loss=0.02765, over 980048.71 frames.], datatang_tot_loss[loss=0.1366, simple_loss=0.2144, pruned_loss=0.02935, over 980614.55 frames.], batch size: 79, lr: 2.93e-04
+2022-06-19 06:29:13,534 INFO [train.py:874] (2/4) Epoch 28, batch 2150, datatang_loss[loss=0.1336, simple_loss=0.223, pruned_loss=0.02211, over 4926.00 frames.], tot_loss[loss=0.1398, simple_loss=0.2228, pruned_loss=0.02846, over 985640.56 frames.], batch size: 79, aishell_tot_loss[loss=0.1435, simple_loss=0.2311, pruned_loss=0.0279, over 980723.70 frames.], datatang_tot_loss[loss=0.1361, simple_loss=0.2143, pruned_loss=0.02894, over 981290.71 frames.], batch size: 79, lr: 2.93e-04
+2022-06-19 06:29:42,212 INFO [train.py:874] (2/4) Epoch 28, batch 2200, aishell_loss[loss=0.1538, simple_loss=0.2275, pruned_loss=0.04007, over 4944.00 frames.], tot_loss[loss=0.1395, simple_loss=0.2224, pruned_loss=0.02829, over 985823.62 frames.], batch size: 32, aishell_tot_loss[loss=0.1433, simple_loss=0.2309, pruned_loss=0.02781, over 981308.25 frames.], datatang_tot_loss[loss=0.136, simple_loss=0.2143, pruned_loss=0.02885, over 981982.88 frames.], batch size: 32, lr: 2.92e-04
+2022-06-19 06:30:10,883 INFO [train.py:874] (2/4) Epoch 28, batch 2250, aishell_loss[loss=0.1385, simple_loss=0.2265, pruned_loss=0.02519, over 4938.00 frames.], tot_loss[loss=0.1396, simple_loss=0.2223, pruned_loss=0.02847, over 985413.75 frames.], batch size: 58, aishell_tot_loss[loss=0.143, simple_loss=0.2304, pruned_loss=0.02777, over 981686.16 frames.], datatang_tot_loss[loss=0.1363, simple_loss=0.2144, pruned_loss=0.02909, over 982162.03 frames.], batch size: 58, lr: 2.92e-04
+2022-06-19 06:30:41,934 INFO [train.py:874] (2/4) Epoch 28, batch 2300, datatang_loss[loss=0.1182, simple_loss=0.1951, pruned_loss=0.02065, over 4982.00 frames.], tot_loss[loss=0.1392, simple_loss=0.2219, pruned_loss=0.02823, over 985321.33 frames.], batch size: 48, aishell_tot_loss[loss=0.1428, simple_loss=0.2301, pruned_loss=0.02774, over 982130.07 frames.], datatang_tot_loss[loss=0.136, simple_loss=0.2142, pruned_loss=0.02888, over 982423.93 frames.], batch size: 48, lr: 2.92e-04
+2022-06-19 06:31:09,123 INFO [train.py:874] (2/4) Epoch 28, batch 2350, datatang_loss[loss=0.1324, simple_loss=0.2069, pruned_loss=0.02897, over 4965.00 frames.], tot_loss[loss=0.1396, simple_loss=0.2222, pruned_loss=0.02845, over 985461.77 frames.], batch size: 55, aishell_tot_loss[loss=0.1427, simple_loss=0.2299, pruned_loss=0.02777, over 982681.27 frames.], datatang_tot_loss[loss=0.1364, simple_loss=0.2146, pruned_loss=0.02908, over 982730.74 frames.], batch size: 55, lr: 2.92e-04
+2022-06-19 06:31:38,570 INFO [train.py:874] (2/4) Epoch 28, batch 2400, aishell_loss[loss=0.1498, simple_loss=0.2332, pruned_loss=0.03324, over 4930.00 frames.], tot_loss[loss=0.1402, simple_loss=0.2229, pruned_loss=0.02872, over 985734.87 frames.], batch size: 58, aishell_tot_loss[loss=0.1428, simple_loss=0.23, pruned_loss=0.02775, over 983367.96 frames.], datatang_tot_loss[loss=0.1368, simple_loss=0.2148, pruned_loss=0.0294, over 982988.09 frames.], batch size: 58, lr: 2.92e-04
+2022-06-19 06:32:08,609 INFO [train.py:874] (2/4) Epoch 28, batch 2450, aishell_loss[loss=0.1382, simple_loss=0.238, pruned_loss=0.01923, over 4939.00 frames.], tot_loss[loss=0.1399, simple_loss=0.2231, pruned_loss=0.02831, over 986148.23 frames.], batch size: 64, aishell_tot_loss[loss=0.1426, simple_loss=0.2301, pruned_loss=0.02749, over 983937.36 frames.], datatang_tot_loss[loss=0.1367, simple_loss=0.2147, pruned_loss=0.02929, over 983455.71 frames.], batch size: 64, lr: 2.92e-04
+2022-06-19 06:32:35,290 INFO [train.py:874] (2/4) Epoch 28, batch 2500, datatang_loss[loss=0.121, simple_loss=0.1916, pruned_loss=0.02518, over 4967.00 frames.], tot_loss[loss=0.1399, simple_loss=0.2229, pruned_loss=0.02841, over 985971.59 frames.], batch size: 24, aishell_tot_loss[loss=0.1422, simple_loss=0.2297, pruned_loss=0.0273, over 984013.20 frames.], datatang_tot_loss[loss=0.1371, simple_loss=0.2151, pruned_loss=0.02955, over 983745.29 frames.], batch size: 24, lr: 2.92e-04
+2022-06-19 06:33:05,156 INFO [train.py:874] (2/4) Epoch 28, batch 2550, aishell_loss[loss=0.1316, simple_loss=0.2228, pruned_loss=0.02016, over 4974.00 frames.], tot_loss[loss=0.1394, simple_loss=0.2224, pruned_loss=0.02818, over 986003.79 frames.], batch size: 64, aishell_tot_loss[loss=0.1416, simple_loss=0.2291, pruned_loss=0.02703, over 984257.57 frames.], datatang_tot_loss[loss=0.1372, simple_loss=0.2153, pruned_loss=0.02957, over 984024.22 frames.], batch size: 64, lr: 2.92e-04
+2022-06-19 06:33:35,654 INFO [train.py:874] (2/4) Epoch 28, batch 2600, datatang_loss[loss=0.1321, simple_loss=0.2115, pruned_loss=0.02641, over 4954.00 frames.], tot_loss[loss=0.1402, simple_loss=0.2234, pruned_loss=0.02851, over 985982.78 frames.], batch size: 67, aishell_tot_loss[loss=0.1424, simple_loss=0.2298, pruned_loss=0.0275, over 984409.37 frames.], datatang_tot_loss[loss=0.1371, simple_loss=0.2153, pruned_loss=0.02947, over 984289.06 frames.], batch size: 67, lr: 2.92e-04
+2022-06-19 06:34:01,433 INFO [train.py:874] (2/4) Epoch 28, batch 2650, datatang_loss[loss=0.1222, simple_loss=0.1999, pruned_loss=0.02225, over 4910.00 frames.], tot_loss[loss=0.1405, simple_loss=0.2236, pruned_loss=0.02869, over 985645.98 frames.], batch size: 47, aishell_tot_loss[loss=0.1425, simple_loss=0.2297, pruned_loss=0.02768, over 984196.51 frames.], datatang_tot_loss[loss=0.1374, simple_loss=0.2157, pruned_loss=0.02952, over 984516.05 frames.], batch size: 47, lr: 2.92e-04
+2022-06-19 06:34:31,376 INFO [train.py:874] (2/4) Epoch 28, batch 2700, datatang_loss[loss=0.1235, simple_loss=0.2037, pruned_loss=0.02169, over 4941.00 frames.], tot_loss[loss=0.141, simple_loss=0.2242, pruned_loss=0.02888, over 985584.51 frames.], batch size: 62, aishell_tot_loss[loss=0.1429, simple_loss=0.2302, pruned_loss=0.0278, over 984273.41 frames.], datatang_tot_loss[loss=0.1376, simple_loss=0.2158, pruned_loss=0.02966, over 984674.64 frames.], batch size: 62, lr: 2.92e-04
+2022-06-19 06:34:59,551 INFO [train.py:874] (2/4) Epoch 28, batch 2750, aishell_loss[loss=0.1498, simple_loss=0.2365, pruned_loss=0.03153, over 4957.00 frames.], tot_loss[loss=0.141, simple_loss=0.2242, pruned_loss=0.02889, over 985187.93 frames.], batch size: 61, aishell_tot_loss[loss=0.1431, simple_loss=0.2304, pruned_loss=0.02796, over 984109.20 frames.], datatang_tot_loss[loss=0.1374, simple_loss=0.2157, pruned_loss=0.02958, over 984694.45 frames.], batch size: 61, lr: 2.92e-04
+2022-06-19 06:35:27,838 INFO [train.py:874] (2/4) Epoch 28, batch 2800, aishell_loss[loss=0.143, simple_loss=0.2399, pruned_loss=0.02309, over 4940.00 frames.], tot_loss[loss=0.1405, simple_loss=0.2235, pruned_loss=0.02874, over 985694.88 frames.], batch size: 64, aishell_tot_loss[loss=0.1429, simple_loss=0.23, pruned_loss=0.02785, over 984609.83 frames.], datatang_tot_loss[loss=0.1373, simple_loss=0.2155, pruned_loss=0.02958, over 984924.54 frames.], batch size: 64, lr: 2.92e-04
+2022-06-19 06:35:57,601 INFO [train.py:874] (2/4) Epoch 28, batch 2850, datatang_loss[loss=0.1493, simple_loss=0.2294, pruned_loss=0.03465, over 4956.00 frames.], tot_loss[loss=0.1401, simple_loss=0.2231, pruned_loss=0.02852, over 985692.44 frames.], batch size: 45, aishell_tot_loss[loss=0.1426, simple_loss=0.2299, pruned_loss=0.02762, over 984737.73 frames.], datatang_tot_loss[loss=0.1373, simple_loss=0.2155, pruned_loss=0.02958, over 985007.64 frames.], batch size: 45, lr: 2.92e-04
+2022-06-19 06:36:25,971 INFO [train.py:874] (2/4) Epoch 28, batch 2900, aishell_loss[loss=0.1594, simple_loss=0.2476, pruned_loss=0.03554, over 4872.00 frames.], tot_loss[loss=0.1401, simple_loss=0.2231, pruned_loss=0.0286, over 985475.69 frames.], batch size: 35, aishell_tot_loss[loss=0.1425, simple_loss=0.2298, pruned_loss=0.02758, over 984520.98 frames.], datatang_tot_loss[loss=0.1375, simple_loss=0.2158, pruned_loss=0.02964, over 985169.35 frames.], batch size: 35, lr: 2.92e-04
+2022-06-19 06:36:54,060 INFO [train.py:874] (2/4) Epoch 28, batch 2950, aishell_loss[loss=0.1758, simple_loss=0.2506, pruned_loss=0.05048, over 4872.00 frames.], tot_loss[loss=0.1397, simple_loss=0.2224, pruned_loss=0.02848, over 985261.09 frames.], batch size: 35, aishell_tot_loss[loss=0.1424, simple_loss=0.2297, pruned_loss=0.02753, over 984464.91 frames.], datatang_tot_loss[loss=0.1372, simple_loss=0.2153, pruned_loss=0.02954, over 985150.27 frames.], batch size: 35, lr: 2.91e-04
+2022-06-19 06:37:24,543 INFO [train.py:874] (2/4) Epoch 28, batch 3000, aishell_loss[loss=0.1294, simple_loss=0.2086, pruned_loss=0.02512, over 4945.00 frames.], tot_loss[loss=0.1408, simple_loss=0.2229, pruned_loss=0.02933, over 985631.41 frames.], batch size: 31, aishell_tot_loss[loss=0.1433, simple_loss=0.2304, pruned_loss=0.02804, over 984572.02 frames.], datatang_tot_loss[loss=0.1376, simple_loss=0.2155, pruned_loss=0.02987, over 985518.63 frames.], batch size: 31, lr: 2.91e-04
+2022-06-19 06:37:24,543 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 06:37:41,301 INFO [train.py:914] (2/4) Epoch 28, validation: loss=0.1644, simple_loss=0.2482, pruned_loss=0.04027, over 1622729.00 frames.
+2022-06-19 06:38:10,637 INFO [train.py:874] (2/4) Epoch 28, batch 3050, datatang_loss[loss=0.119, simple_loss=0.1881, pruned_loss=0.02494, over 4959.00 frames.], tot_loss[loss=0.1406, simple_loss=0.2228, pruned_loss=0.02922, over 985739.92 frames.], batch size: 55, aishell_tot_loss[loss=0.1432, simple_loss=0.2304, pruned_loss=0.02801, over 984819.28 frames.], datatang_tot_loss[loss=0.1376, simple_loss=0.2155, pruned_loss=0.02986, over 985511.79 frames.], batch size: 55, lr: 2.91e-04
+2022-06-19 06:38:41,007 INFO [train.py:874] (2/4) Epoch 28, batch 3100, datatang_loss[loss=0.1293, simple_loss=0.2131, pruned_loss=0.02282, over 4920.00 frames.], tot_loss[loss=0.1413, simple_loss=0.2236, pruned_loss=0.02948, over 985448.98 frames.], batch size: 73, aishell_tot_loss[loss=0.1434, simple_loss=0.2308, pruned_loss=0.02801, over 984693.12 frames.], datatang_tot_loss[loss=0.1383, simple_loss=0.2162, pruned_loss=0.03018, over 985453.98 frames.], batch size: 73, lr: 2.91e-04
+2022-06-19 06:39:07,311 INFO [train.py:874] (2/4) Epoch 28, batch 3150, aishell_loss[loss=0.1448, simple_loss=0.2304, pruned_loss=0.02957, over 4860.00 frames.], tot_loss[loss=0.1407, simple_loss=0.2231, pruned_loss=0.0291, over 985672.05 frames.], batch size: 36, aishell_tot_loss[loss=0.1436, simple_loss=0.2311, pruned_loss=0.02811, over 984952.12 frames.], datatang_tot_loss[loss=0.1374, simple_loss=0.2152, pruned_loss=0.02981, over 985547.47 frames.], batch size: 36, lr: 2.91e-04
+2022-06-19 06:39:36,943 INFO [train.py:874] (2/4) Epoch 28, batch 3200, datatang_loss[loss=0.1109, simple_loss=0.1745, pruned_loss=0.02361, over 4893.00 frames.], tot_loss[loss=0.1408, simple_loss=0.2233, pruned_loss=0.02912, over 985579.61 frames.], batch size: 47, aishell_tot_loss[loss=0.1438, simple_loss=0.2312, pruned_loss=0.02819, over 985092.32 frames.], datatang_tot_loss[loss=0.1374, simple_loss=0.2153, pruned_loss=0.02978, over 985414.16 frames.], batch size: 47, lr: 2.91e-04
+2022-06-19 06:40:07,280 INFO [train.py:874] (2/4) Epoch 28, batch 3250, aishell_loss[loss=0.1412, simple_loss=0.2329, pruned_loss=0.02474, over 4944.00 frames.], tot_loss[loss=0.1402, simple_loss=0.2228, pruned_loss=0.02876, over 985561.71 frames.], batch size: 49, aishell_tot_loss[loss=0.1442, simple_loss=0.2317, pruned_loss=0.02836, over 984978.73 frames.], datatang_tot_loss[loss=0.1364, simple_loss=0.2144, pruned_loss=0.02924, over 985587.73 frames.], batch size: 49, lr: 2.91e-04
+2022-06-19 06:40:34,280 INFO [train.py:874] (2/4) Epoch 28, batch 3300, aishell_loss[loss=0.1548, simple_loss=0.247, pruned_loss=0.03127, over 4969.00 frames.], tot_loss[loss=0.1405, simple_loss=0.2228, pruned_loss=0.02905, over 985418.50 frames.], batch size: 51, aishell_tot_loss[loss=0.1441, simple_loss=0.2315, pruned_loss=0.0284, over 984992.55 frames.], datatang_tot_loss[loss=0.1368, simple_loss=0.2146, pruned_loss=0.02949, over 985481.87 frames.], batch size: 51, lr: 2.91e-04
+2022-06-19 06:41:04,422 INFO [train.py:874] (2/4) Epoch 28, batch 3350, aishell_loss[loss=0.1485, simple_loss=0.2422, pruned_loss=0.02738, over 4895.00 frames.], tot_loss[loss=0.1415, simple_loss=0.2241, pruned_loss=0.02945, over 985779.39 frames.], batch size: 28, aishell_tot_loss[loss=0.1449, simple_loss=0.2324, pruned_loss=0.02869, over 985333.46 frames.], datatang_tot_loss[loss=0.1369, simple_loss=0.2144, pruned_loss=0.02968, over 985584.02 frames.], batch size: 28, lr: 2.91e-04
+2022-06-19 06:41:34,599 INFO [train.py:874] (2/4) Epoch 28, batch 3400, datatang_loss[loss=0.1535, simple_loss=0.2366, pruned_loss=0.03513, over 4933.00 frames.], tot_loss[loss=0.1414, simple_loss=0.2243, pruned_loss=0.02922, over 985832.43 frames.], batch size: 94, aishell_tot_loss[loss=0.1448, simple_loss=0.2326, pruned_loss=0.0285, over 985317.09 frames.], datatang_tot_loss[loss=0.137, simple_loss=0.2146, pruned_loss=0.02971, over 985731.11 frames.], batch size: 94, lr: 2.91e-04
+2022-06-19 06:42:02,106 INFO [train.py:874] (2/4) Epoch 28, batch 3450, datatang_loss[loss=0.1478, simple_loss=0.2172, pruned_loss=0.0392, over 4970.00 frames.], tot_loss[loss=0.1412, simple_loss=0.2237, pruned_loss=0.02935, over 985691.60 frames.], batch size: 60, aishell_tot_loss[loss=0.1449, simple_loss=0.2324, pruned_loss=0.02867, over 985102.62 frames.], datatang_tot_loss[loss=0.1369, simple_loss=0.2145, pruned_loss=0.02968, over 985856.20 frames.], batch size: 60, lr: 2.91e-04
+2022-06-19 06:42:33,173 INFO [train.py:874] (2/4) Epoch 28, batch 3500, datatang_loss[loss=0.1381, simple_loss=0.2265, pruned_loss=0.02484, over 4961.00 frames.], tot_loss[loss=0.1406, simple_loss=0.223, pruned_loss=0.02907, over 985540.62 frames.], batch size: 91, aishell_tot_loss[loss=0.1443, simple_loss=0.2316, pruned_loss=0.02856, over 984934.35 frames.], datatang_tot_loss[loss=0.1368, simple_loss=0.2145, pruned_loss=0.02953, over 985915.91 frames.], batch size: 91, lr: 2.91e-04
+2022-06-19 06:43:02,758 INFO [train.py:874] (2/4) Epoch 28, batch 3550, aishell_loss[loss=0.1699, simple_loss=0.2416, pruned_loss=0.04915, over 4919.00 frames.], tot_loss[loss=0.1413, simple_loss=0.2237, pruned_loss=0.02945, over 986187.45 frames.], batch size: 33, aishell_tot_loss[loss=0.1447, simple_loss=0.2319, pruned_loss=0.02871, over 985399.67 frames.], datatang_tot_loss[loss=0.1373, simple_loss=0.2151, pruned_loss=0.02977, over 986164.86 frames.], batch size: 33, lr: 2.91e-04
+2022-06-19 06:43:30,715 INFO [train.py:874] (2/4) Epoch 28, batch 3600, aishell_loss[loss=0.1474, simple_loss=0.239, pruned_loss=0.02789, over 4921.00 frames.], tot_loss[loss=0.1408, simple_loss=0.223, pruned_loss=0.02928, over 985851.74 frames.], batch size: 52, aishell_tot_loss[loss=0.1442, simple_loss=0.2313, pruned_loss=0.02857, over 985038.58 frames.], datatang_tot_loss[loss=0.1374, simple_loss=0.2153, pruned_loss=0.02975, over 986248.24 frames.], batch size: 52, lr: 2.91e-04
+2022-06-19 06:44:01,890 INFO [train.py:874] (2/4) Epoch 28, batch 3650, datatang_loss[loss=0.1323, simple_loss=0.2087, pruned_loss=0.02795, over 4921.00 frames.], tot_loss[loss=0.1403, simple_loss=0.2232, pruned_loss=0.0287, over 985678.52 frames.], batch size: 34, aishell_tot_loss[loss=0.1441, simple_loss=0.2315, pruned_loss=0.02836, over 985029.99 frames.], datatang_tot_loss[loss=0.1368, simple_loss=0.2149, pruned_loss=0.02937, over 986144.73 frames.], batch size: 34, lr: 2.91e-04
+2022-06-19 06:44:30,423 INFO [train.py:874] (2/4) Epoch 28, batch 3700, datatang_loss[loss=0.1146, simple_loss=0.1959, pruned_loss=0.01665, over 4928.00 frames.], tot_loss[loss=0.1404, simple_loss=0.2234, pruned_loss=0.02873, over 985554.57 frames.], batch size: 79, aishell_tot_loss[loss=0.1439, simple_loss=0.2314, pruned_loss=0.02825, over 984880.61 frames.], datatang_tot_loss[loss=0.1371, simple_loss=0.2151, pruned_loss=0.02948, over 986189.50 frames.], batch size: 79, lr: 2.91e-04
+2022-06-19 06:44:59,335 INFO [train.py:874] (2/4) Epoch 28, batch 3750, datatang_loss[loss=0.1299, simple_loss=0.1963, pruned_loss=0.03177, over 4963.00 frames.], tot_loss[loss=0.1404, simple_loss=0.2236, pruned_loss=0.02861, over 985531.81 frames.], batch size: 45, aishell_tot_loss[loss=0.1439, simple_loss=0.2315, pruned_loss=0.02815, over 984903.72 frames.], datatang_tot_loss[loss=0.137, simple_loss=0.2151, pruned_loss=0.02942, over 986162.36 frames.], batch size: 45, lr: 2.90e-04
+2022-06-19 06:45:28,856 INFO [train.py:874] (2/4) Epoch 28, batch 3800, aishell_loss[loss=0.1576, simple_loss=0.2439, pruned_loss=0.03569, over 4971.00 frames.], tot_loss[loss=0.1394, simple_loss=0.2225, pruned_loss=0.02819, over 985848.04 frames.], batch size: 39, aishell_tot_loss[loss=0.1436, simple_loss=0.2311, pruned_loss=0.02806, over 985099.53 frames.], datatang_tot_loss[loss=0.1362, simple_loss=0.2145, pruned_loss=0.02902, over 986287.95 frames.], batch size: 39, lr: 2.90e-04
+2022-06-19 06:45:56,690 INFO [train.py:874] (2/4) Epoch 28, batch 3850, datatang_loss[loss=0.1379, simple_loss=0.2126, pruned_loss=0.03165, over 4943.00 frames.], tot_loss[loss=0.1392, simple_loss=0.2223, pruned_loss=0.02807, over 985510.02 frames.], batch size: 62, aishell_tot_loss[loss=0.1432, simple_loss=0.2307, pruned_loss=0.02785, over 984870.48 frames.], datatang_tot_loss[loss=0.1363, simple_loss=0.2145, pruned_loss=0.02901, over 986181.43 frames.], batch size: 62, lr: 2.90e-04
+2022-06-19 06:46:25,719 INFO [train.py:874] (2/4) Epoch 28, batch 3900, aishell_loss[loss=0.1434, simple_loss=0.2312, pruned_loss=0.02779, over 4954.00 frames.], tot_loss[loss=0.1391, simple_loss=0.2222, pruned_loss=0.02804, over 985982.75 frames.], batch size: 56, aishell_tot_loss[loss=0.1433, simple_loss=0.2307, pruned_loss=0.0279, over 985200.92 frames.], datatang_tot_loss[loss=0.136, simple_loss=0.2141, pruned_loss=0.02888, over 986363.80 frames.], batch size: 56, lr: 2.90e-04
+2022-06-19 06:46:53,821 INFO [train.py:874] (2/4) Epoch 28, batch 3950, aishell_loss[loss=0.1233, simple_loss=0.2166, pruned_loss=0.01503, over 4944.00 frames.], tot_loss[loss=0.1395, simple_loss=0.2228, pruned_loss=0.02812, over 985876.47 frames.], batch size: 56, aishell_tot_loss[loss=0.1434, simple_loss=0.2309, pruned_loss=0.02795, over 985253.85 frames.], datatang_tot_loss[loss=0.136, simple_loss=0.2144, pruned_loss=0.02885, over 986243.65 frames.], batch size: 56, lr: 2.90e-04
+2022-06-19 06:47:20,488 INFO [train.py:874] (2/4) Epoch 28, batch 4000, aishell_loss[loss=0.1361, simple_loss=0.2321, pruned_loss=0.02011, over 4899.00 frames.], tot_loss[loss=0.1392, simple_loss=0.2228, pruned_loss=0.02784, over 985722.87 frames.], batch size: 52, aishell_tot_loss[loss=0.1435, simple_loss=0.2311, pruned_loss=0.02798, over 985390.43 frames.], datatang_tot_loss[loss=0.1354, simple_loss=0.2139, pruned_loss=0.02848, over 985985.05 frames.], batch size: 52, lr: 2.90e-04
+2022-06-19 06:47:20,489 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 06:47:36,385 INFO [train.py:914] (2/4) Epoch 28, validation: loss=0.1646, simple_loss=0.2484, pruned_loss=0.04036, over 1622729.00 frames.
+2022-06-19 06:48:02,200 INFO [train.py:874] (2/4) Epoch 28, batch 4050, aishell_loss[loss=0.1635, simple_loss=0.2509, pruned_loss=0.0381, over 4892.00 frames.], tot_loss[loss=0.1388, simple_loss=0.2224, pruned_loss=0.02762, over 985476.92 frames.], batch size: 34, aishell_tot_loss[loss=0.1431, simple_loss=0.2306, pruned_loss=0.02778, over 985182.82 frames.], datatang_tot_loss[loss=0.1351, simple_loss=0.2135, pruned_loss=0.02838, over 985967.78 frames.], batch size: 34, lr: 2.90e-04
+2022-06-19 06:49:08,007 INFO [train.py:874] (2/4) Epoch 29, batch 50, aishell_loss[loss=0.1448, simple_loss=0.2257, pruned_loss=0.03196, over 4922.00 frames.], tot_loss[loss=0.1316, simple_loss=0.2137, pruned_loss=0.02478, over 218777.07 frames.], batch size: 52, aishell_tot_loss[loss=0.1384, simple_loss=0.2251, pruned_loss=0.02583, over 120546.92 frames.], datatang_tot_loss[loss=0.1244, simple_loss=0.2016, pruned_loss=0.02356, over 111888.35 frames.], batch size: 52, lr: 2.85e-04
+2022-06-19 06:49:37,789 INFO [train.py:874] (2/4) Epoch 29, batch 100, datatang_loss[loss=0.1161, simple_loss=0.1998, pruned_loss=0.01617, over 4950.00 frames.], tot_loss[loss=0.1368, simple_loss=0.2189, pruned_loss=0.02739, over 388855.49 frames.], batch size: 86, aishell_tot_loss[loss=0.1439, simple_loss=0.2314, pruned_loss=0.02821, over 222373.31 frames.], datatang_tot_loss[loss=0.129, simple_loss=0.2054, pruned_loss=0.02624, over 214947.84 frames.], batch size: 86, lr: 2.85e-04
+2022-06-19 06:50:11,746 INFO [train.py:874] (2/4) Epoch 29, batch 150, aishell_loss[loss=0.1817, simple_loss=0.2621, pruned_loss=0.05059, over 4943.00 frames.], tot_loss[loss=0.1392, simple_loss=0.2208, pruned_loss=0.0288, over 521522.76 frames.], batch size: 45, aishell_tot_loss[loss=0.1455, simple_loss=0.2319, pruned_loss=0.02956, over 319096.67 frames.], datatang_tot_loss[loss=0.1315, simple_loss=0.2081, pruned_loss=0.02747, over 299158.63 frames.], batch size: 45, lr: 2.85e-04
+2022-06-19 06:50:40,057 INFO [train.py:874] (2/4) Epoch 29, batch 200, datatang_loss[loss=0.1389, simple_loss=0.2225, pruned_loss=0.02762, over 4910.00 frames.], tot_loss[loss=0.1383, simple_loss=0.2209, pruned_loss=0.02788, over 624491.03 frames.], batch size: 64, aishell_tot_loss[loss=0.1443, simple_loss=0.2316, pruned_loss=0.02847, over 406252.46 frames.], datatang_tot_loss[loss=0.1313, simple_loss=0.2082, pruned_loss=0.02717, over 371040.90 frames.], batch size: 64, lr: 2.85e-04
+2022-06-19 06:51:09,813 INFO [train.py:874] (2/4) Epoch 29, batch 250, datatang_loss[loss=0.1339, simple_loss=0.2085, pruned_loss=0.02966, over 4929.00 frames.], tot_loss[loss=0.137, simple_loss=0.2199, pruned_loss=0.02709, over 704493.71 frames.], batch size: 77, aishell_tot_loss[loss=0.1426, simple_loss=0.2298, pruned_loss=0.02768, over 474368.14 frames.], datatang_tot_loss[loss=0.1312, simple_loss=0.209, pruned_loss=0.02672, over 443488.49 frames.], batch size: 77, lr: 2.85e-04
+2022-06-19 06:51:37,965 INFO [train.py:874] (2/4) Epoch 29, batch 300, datatang_loss[loss=0.1698, simple_loss=0.247, pruned_loss=0.04633, over 4941.00 frames.], tot_loss[loss=0.1374, simple_loss=0.2206, pruned_loss=0.02713, over 766407.16 frames.], batch size: 108, aishell_tot_loss[loss=0.142, simple_loss=0.2291, pruned_loss=0.02751, over 549801.77 frames.], datatang_tot_loss[loss=0.1318, simple_loss=0.2099, pruned_loss=0.02685, over 490171.11 frames.], batch size: 108, lr: 2.85e-04
+2022-06-19 06:52:07,670 INFO [train.py:874] (2/4) Epoch 29, batch 350, aishell_loss[loss=0.1426, simple_loss=0.234, pruned_loss=0.02559, over 4977.00 frames.], tot_loss[loss=0.1392, simple_loss=0.2224, pruned_loss=0.02798, over 815537.57 frames.], batch size: 48, aishell_tot_loss[loss=0.143, simple_loss=0.2301, pruned_loss=0.02801, over 611057.70 frames.], datatang_tot_loss[loss=0.1333, simple_loss=0.2113, pruned_loss=0.02761, over 537576.17 frames.], batch size: 48, lr: 2.85e-04
+2022-06-19 06:52:37,144 INFO [train.py:874] (2/4) Epoch 29, batch 400, aishell_loss[loss=0.1497, simple_loss=0.2324, pruned_loss=0.03357, over 4924.00 frames.], tot_loss[loss=0.1396, simple_loss=0.2225, pruned_loss=0.0283, over 852819.96 frames.], batch size: 41, aishell_tot_loss[loss=0.1435, simple_loss=0.2305, pruned_loss=0.02824, over 656451.43 frames.], datatang_tot_loss[loss=0.1337, simple_loss=0.2116, pruned_loss=0.02789, over 588277.61 frames.], batch size: 41, lr: 2.85e-04
+2022-06-19 06:53:05,422 INFO [train.py:874] (2/4) Epoch 29, batch 450, datatang_loss[loss=0.1195, simple_loss=0.1975, pruned_loss=0.02071, over 4921.00 frames.], tot_loss[loss=0.1393, simple_loss=0.2221, pruned_loss=0.02822, over 881683.83 frames.], batch size: 71, aishell_tot_loss[loss=0.1426, simple_loss=0.2295, pruned_loss=0.02791, over 699238.49 frames.], datatang_tot_loss[loss=0.1343, simple_loss=0.2121, pruned_loss=0.02823, over 629478.05 frames.], batch size: 71, lr: 2.85e-04
+2022-06-19 06:53:35,500 INFO [train.py:874] (2/4) Epoch 29, batch 500, datatang_loss[loss=0.1471, simple_loss=0.2325, pruned_loss=0.03087, over 4949.00 frames.], tot_loss[loss=0.1398, simple_loss=0.2227, pruned_loss=0.0285, over 905079.98 frames.], batch size: 67, aishell_tot_loss[loss=0.1429, simple_loss=0.2299, pruned_loss=0.02797, over 733374.47 frames.], datatang_tot_loss[loss=0.1351, simple_loss=0.213, pruned_loss=0.02864, over 671388.07 frames.], batch size: 67, lr: 2.84e-04
+2022-06-19 06:54:05,084 INFO [train.py:874] (2/4) Epoch 29, batch 550, aishell_loss[loss=0.1458, simple_loss=0.2343, pruned_loss=0.02867, over 4934.00 frames.], tot_loss[loss=0.1402, simple_loss=0.2232, pruned_loss=0.02863, over 922522.36 frames.], batch size: 49, aishell_tot_loss[loss=0.1436, simple_loss=0.2307, pruned_loss=0.02824, over 757179.69 frames.], datatang_tot_loss[loss=0.1355, simple_loss=0.2139, pruned_loss=0.02858, over 715075.05 frames.], batch size: 49, lr: 2.84e-04
+2022-06-19 06:54:33,116 INFO [train.py:874] (2/4) Epoch 29, batch 600, datatang_loss[loss=0.131, simple_loss=0.2134, pruned_loss=0.02432, over 4969.00 frames.], tot_loss[loss=0.1406, simple_loss=0.2233, pruned_loss=0.0289, over 936670.75 frames.], batch size: 55, aishell_tot_loss[loss=0.1443, simple_loss=0.2314, pruned_loss=0.02863, over 786297.85 frames.], datatang_tot_loss[loss=0.1353, simple_loss=0.2133, pruned_loss=0.0286, over 744498.78 frames.], batch size: 55, lr: 2.84e-04
+2022-06-19 06:55:02,475 INFO [train.py:874] (2/4) Epoch 29, batch 650, aishell_loss[loss=0.1428, simple_loss=0.2326, pruned_loss=0.02651, over 4869.00 frames.], tot_loss[loss=0.1413, simple_loss=0.2241, pruned_loss=0.02928, over 947629.84 frames.], batch size: 36, aishell_tot_loss[loss=0.1445, simple_loss=0.2315, pruned_loss=0.0288, over 809766.21 frames.], datatang_tot_loss[loss=0.1362, simple_loss=0.2144, pruned_loss=0.02899, over 773026.42 frames.], batch size: 36, lr: 2.84e-04
+2022-06-19 06:55:31,654 INFO [train.py:874] (2/4) Epoch 29, batch 700, datatang_loss[loss=0.1318, simple_loss=0.2136, pruned_loss=0.02498, over 4941.00 frames.], tot_loss[loss=0.14, simple_loss=0.2227, pruned_loss=0.02866, over 956144.54 frames.], batch size: 69, aishell_tot_loss[loss=0.1439, simple_loss=0.2305, pruned_loss=0.02863, over 828037.00 frames.], datatang_tot_loss[loss=0.1357, simple_loss=0.2144, pruned_loss=0.02851, over 801036.40 frames.], batch size: 69, lr: 2.84e-04
+2022-06-19 06:56:00,640 INFO [train.py:874] (2/4) Epoch 29, batch 750, aishell_loss[loss=0.1215, simple_loss=0.2101, pruned_loss=0.01648, over 4841.00 frames.], tot_loss[loss=0.1408, simple_loss=0.2236, pruned_loss=0.02899, over 962442.47 frames.], batch size: 29, aishell_tot_loss[loss=0.1446, simple_loss=0.2314, pruned_loss=0.02885, over 846940.97 frames.], datatang_tot_loss[loss=0.136, simple_loss=0.2146, pruned_loss=0.02874, over 822073.90 frames.], batch size: 29, lr: 2.84e-04
+2022-06-19 06:56:31,422 INFO [train.py:874] (2/4) Epoch 29, batch 800, aishell_loss[loss=0.1343, simple_loss=0.2189, pruned_loss=0.0249, over 4913.00 frames.], tot_loss[loss=0.1396, simple_loss=0.2222, pruned_loss=0.02846, over 967148.86 frames.], batch size: 33, aishell_tot_loss[loss=0.1432, simple_loss=0.2296, pruned_loss=0.02839, over 864412.70 frames.], datatang_tot_loss[loss=0.1359, simple_loss=0.2146, pruned_loss=0.02862, over 839375.06 frames.], batch size: 33, lr: 2.84e-04
+2022-06-19 06:57:01,157 INFO [train.py:874] (2/4) Epoch 29, batch 850, aishell_loss[loss=0.1413, simple_loss=0.2339, pruned_loss=0.02439, over 4946.00 frames.], tot_loss[loss=0.1394, simple_loss=0.222, pruned_loss=0.02834, over 971271.87 frames.], batch size: 54, aishell_tot_loss[loss=0.1433, simple_loss=0.2299, pruned_loss=0.02833, over 879059.68 frames.], datatang_tot_loss[loss=0.1355, simple_loss=0.214, pruned_loss=0.0285, over 856134.87 frames.], batch size: 54, lr: 2.84e-04
+2022-06-19 06:57:29,413 INFO [train.py:874] (2/4) Epoch 29, batch 900, aishell_loss[loss=0.1322, simple_loss=0.2238, pruned_loss=0.02027, over 4937.00 frames.], tot_loss[loss=0.1397, simple_loss=0.222, pruned_loss=0.02873, over 974127.73 frames.], batch size: 49, aishell_tot_loss[loss=0.1429, simple_loss=0.2293, pruned_loss=0.02821, over 890272.11 frames.], datatang_tot_loss[loss=0.1364, simple_loss=0.2147, pruned_loss=0.02906, over 872597.51 frames.], batch size: 49, lr: 2.84e-04
+2022-06-19 06:57:59,984 INFO [train.py:874] (2/4) Epoch 29, batch 950, aishell_loss[loss=0.1846, simple_loss=0.271, pruned_loss=0.0491, over 4937.00 frames.], tot_loss[loss=0.14, simple_loss=0.2224, pruned_loss=0.02875, over 976740.03 frames.], batch size: 68, aishell_tot_loss[loss=0.1434, simple_loss=0.2302, pruned_loss=0.02831, over 901455.02 frames.], datatang_tot_loss[loss=0.1361, simple_loss=0.2143, pruned_loss=0.02899, over 886027.70 frames.], batch size: 68, lr: 2.84e-04
+2022-06-19 06:58:28,343 INFO [train.py:874] (2/4) Epoch 29, batch 1000, aishell_loss[loss=0.1401, simple_loss=0.23, pruned_loss=0.02509, over 4938.00 frames.], tot_loss[loss=0.1394, simple_loss=0.2218, pruned_loss=0.02846, over 978604.74 frames.], batch size: 45, aishell_tot_loss[loss=0.1432, simple_loss=0.2299, pruned_loss=0.02825, over 910346.54 frames.], datatang_tot_loss[loss=0.1358, simple_loss=0.2141, pruned_loss=0.02875, over 898847.55 frames.], batch size: 45, lr: 2.84e-04
+2022-06-19 06:58:28,344 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 06:58:44,049 INFO [train.py:914] (2/4) Epoch 29, validation: loss=0.165, simple_loss=0.2483, pruned_loss=0.04086, over 1622729.00 frames.
+2022-06-19 06:59:14,568 INFO [train.py:874] (2/4) Epoch 29, batch 1050, aishell_loss[loss=0.166, simple_loss=0.2456, pruned_loss=0.04319, over 4918.00 frames.], tot_loss[loss=0.1401, simple_loss=0.2225, pruned_loss=0.02883, over 980115.45 frames.], batch size: 49, aishell_tot_loss[loss=0.1436, simple_loss=0.2303, pruned_loss=0.02847, over 918674.53 frames.], datatang_tot_loss[loss=0.1362, simple_loss=0.2146, pruned_loss=0.02893, over 909606.26 frames.], batch size: 49, lr: 2.84e-04
+2022-06-19 06:59:43,105 INFO [train.py:874] (2/4) Epoch 29, batch 1100, aishell_loss[loss=0.138, simple_loss=0.219, pruned_loss=0.02848, over 4926.00 frames.], tot_loss[loss=0.1405, simple_loss=0.2233, pruned_loss=0.02884, over 981323.08 frames.], batch size: 33, aishell_tot_loss[loss=0.1445, simple_loss=0.2313, pruned_loss=0.02883, over 926119.58 frames.], datatang_tot_loss[loss=0.1359, simple_loss=0.2146, pruned_loss=0.02863, over 919030.33 frames.], batch size: 33, lr: 2.84e-04
+2022-06-19 07:00:12,937 INFO [train.py:874] (2/4) Epoch 29, batch 1150, aishell_loss[loss=0.1286, simple_loss=0.2231, pruned_loss=0.01706, over 4943.00 frames.], tot_loss[loss=0.1404, simple_loss=0.2234, pruned_loss=0.02865, over 982065.55 frames.], batch size: 45, aishell_tot_loss[loss=0.144, simple_loss=0.2311, pruned_loss=0.02849, over 932362.35 frames.], datatang_tot_loss[loss=0.1364, simple_loss=0.2152, pruned_loss=0.0288, over 927476.35 frames.], batch size: 45, lr: 2.84e-04
+2022-06-19 07:00:42,820 INFO [train.py:874] (2/4) Epoch 29, batch 1200, datatang_loss[loss=0.1566, simple_loss=0.2309, pruned_loss=0.04121, over 4858.00 frames.], tot_loss[loss=0.1406, simple_loss=0.2234, pruned_loss=0.02896, over 982604.52 frames.], batch size: 39, aishell_tot_loss[loss=0.1438, simple_loss=0.2309, pruned_loss=0.02838, over 936541.78 frames.], datatang_tot_loss[loss=0.1372, simple_loss=0.2159, pruned_loss=0.02923, over 936247.84 frames.], batch size: 39, lr: 2.84e-04
+2022-06-19 07:01:11,893 INFO [train.py:874] (2/4) Epoch 29, batch 1250, aishell_loss[loss=0.1349, simple_loss=0.2228, pruned_loss=0.02356, over 4954.00 frames.], tot_loss[loss=0.14, simple_loss=0.2226, pruned_loss=0.02872, over 983028.55 frames.], batch size: 56, aishell_tot_loss[loss=0.1437, simple_loss=0.2306, pruned_loss=0.02837, over 941269.19 frames.], datatang_tot_loss[loss=0.1368, simple_loss=0.2156, pruned_loss=0.02899, over 942841.69 frames.], batch size: 56, lr: 2.84e-04
+2022-06-19 07:01:42,639 INFO [train.py:874] (2/4) Epoch 29, batch 1300, datatang_loss[loss=0.1406, simple_loss=0.2306, pruned_loss=0.02529, over 4922.00 frames.], tot_loss[loss=0.1403, simple_loss=0.2229, pruned_loss=0.02888, over 983554.13 frames.], batch size: 83, aishell_tot_loss[loss=0.1431, simple_loss=0.2301, pruned_loss=0.02805, over 946388.64 frames.], datatang_tot_loss[loss=0.1377, simple_loss=0.2163, pruned_loss=0.02948, over 947921.10 frames.], batch size: 83, lr: 2.84e-04
+2022-06-19 07:02:12,780 INFO [train.py:874] (2/4) Epoch 29, batch 1350, aishell_loss[loss=0.1317, simple_loss=0.2153, pruned_loss=0.02404, over 4926.00 frames.], tot_loss[loss=0.1405, simple_loss=0.223, pruned_loss=0.02901, over 984088.18 frames.], batch size: 32, aishell_tot_loss[loss=0.1425, simple_loss=0.2295, pruned_loss=0.02776, over 950591.68 frames.], datatang_tot_loss[loss=0.1385, simple_loss=0.2171, pruned_loss=0.02993, over 952853.89 frames.], batch size: 32, lr: 2.83e-04
+2022-06-19 07:02:41,609 INFO [train.py:874] (2/4) Epoch 29, batch 1400, datatang_loss[loss=0.1547, simple_loss=0.2378, pruned_loss=0.03583, over 4990.00 frames.], tot_loss[loss=0.1408, simple_loss=0.2235, pruned_loss=0.02912, over 984363.93 frames.], batch size: 25, aishell_tot_loss[loss=0.1427, simple_loss=0.2297, pruned_loss=0.02785, over 954993.70 frames.], datatang_tot_loss[loss=0.1386, simple_loss=0.2172, pruned_loss=0.03002, over 956369.96 frames.], batch size: 25, lr: 2.83e-04
+2022-06-19 07:03:11,147 INFO [train.py:874] (2/4) Epoch 29, batch 1450, aishell_loss[loss=0.1731, simple_loss=0.2505, pruned_loss=0.04789, over 4899.00 frames.], tot_loss[loss=0.1398, simple_loss=0.2223, pruned_loss=0.02868, over 984611.72 frames.], batch size: 60, aishell_tot_loss[loss=0.1425, simple_loss=0.2293, pruned_loss=0.02787, over 958776.22 frames.], datatang_tot_loss[loss=0.1377, simple_loss=0.2163, pruned_loss=0.02957, over 959615.95 frames.], batch size: 60, lr: 2.83e-04
+2022-06-19 07:03:40,957 INFO [train.py:874] (2/4) Epoch 29, batch 1500, datatang_loss[loss=0.1275, simple_loss=0.2048, pruned_loss=0.02514, over 4921.00 frames.], tot_loss[loss=0.1398, simple_loss=0.2226, pruned_loss=0.02856, over 984797.04 frames.], batch size: 81, aishell_tot_loss[loss=0.1429, simple_loss=0.2299, pruned_loss=0.02793, over 961526.04 frames.], datatang_tot_loss[loss=0.1373, simple_loss=0.216, pruned_loss=0.02935, over 963036.90 frames.], batch size: 81, lr: 2.83e-04
+2022-06-19 07:04:10,511 INFO [train.py:874] (2/4) Epoch 29, batch 1550, datatang_loss[loss=0.1386, simple_loss=0.2233, pruned_loss=0.02694, over 4956.00 frames.], tot_loss[loss=0.1406, simple_loss=0.2232, pruned_loss=0.02897, over 985024.80 frames.], batch size: 45, aishell_tot_loss[loss=0.1434, simple_loss=0.2305, pruned_loss=0.02817, over 964308.42 frames.], datatang_tot_loss[loss=0.1375, simple_loss=0.216, pruned_loss=0.02952, over 965804.81 frames.], batch size: 45, lr: 2.83e-04
+2022-06-19 07:04:40,809 INFO [train.py:874] (2/4) Epoch 29, batch 1600, aishell_loss[loss=0.1464, simple_loss=0.235, pruned_loss=0.02889, over 4917.00 frames.], tot_loss[loss=0.1403, simple_loss=0.2229, pruned_loss=0.02886, over 985414.03 frames.], batch size: 79, aishell_tot_loss[loss=0.1431, simple_loss=0.2303, pruned_loss=0.02792, over 966411.47 frames.], datatang_tot_loss[loss=0.1378, simple_loss=0.2161, pruned_loss=0.02968, over 968774.47 frames.], batch size: 79, lr: 2.83e-04
+2022-06-19 07:05:09,472 INFO [train.py:874] (2/4) Epoch 29, batch 1650, datatang_loss[loss=0.17, simple_loss=0.2488, pruned_loss=0.04561, over 4934.00 frames.], tot_loss[loss=0.1396, simple_loss=0.222, pruned_loss=0.02853, over 985030.13 frames.], batch size: 109, aishell_tot_loss[loss=0.1426, simple_loss=0.2299, pruned_loss=0.0277, over 968272.48 frames.], datatang_tot_loss[loss=0.1374, simple_loss=0.2156, pruned_loss=0.02955, over 970687.70 frames.], batch size: 109, lr: 2.83e-04
+2022-06-19 07:05:37,786 INFO [train.py:874] (2/4) Epoch 29, batch 1700, aishell_loss[loss=0.1199, simple_loss=0.2062, pruned_loss=0.01675, over 4822.00 frames.], tot_loss[loss=0.1393, simple_loss=0.2218, pruned_loss=0.02844, over 984841.00 frames.], batch size: 29, aishell_tot_loss[loss=0.1426, simple_loss=0.2299, pruned_loss=0.02763, over 970211.98 frames.], datatang_tot_loss[loss=0.137, simple_loss=0.2151, pruned_loss=0.0295, over 972221.82 frames.], batch size: 29, lr: 2.83e-04
+2022-06-19 07:06:08,731 INFO [train.py:874] (2/4) Epoch 29, batch 1750, datatang_loss[loss=0.1338, simple_loss=0.2165, pruned_loss=0.02551, over 4919.00 frames.], tot_loss[loss=0.1398, simple_loss=0.2221, pruned_loss=0.02875, over 985047.76 frames.], batch size: 83, aishell_tot_loss[loss=0.1424, simple_loss=0.2297, pruned_loss=0.02756, over 971965.91 frames.], datatang_tot_loss[loss=0.1375, simple_loss=0.2154, pruned_loss=0.02986, over 973909.44 frames.], batch size: 83, lr: 2.83e-04
+2022-06-19 07:06:37,232 INFO [train.py:874] (2/4) Epoch 29, batch 1800, datatang_loss[loss=0.1711, simple_loss=0.26, pruned_loss=0.04113, over 4923.00 frames.], tot_loss[loss=0.1391, simple_loss=0.2214, pruned_loss=0.02845, over 984914.66 frames.], batch size: 57, aishell_tot_loss[loss=0.1419, simple_loss=0.2292, pruned_loss=0.02735, over 973169.64 frames.], datatang_tot_loss[loss=0.1373, simple_loss=0.2151, pruned_loss=0.02972, over 975402.11 frames.], batch size: 57, lr: 2.83e-04
+2022-06-19 07:07:05,211 INFO [train.py:874] (2/4) Epoch 29, batch 1850, datatang_loss[loss=0.1372, simple_loss=0.2181, pruned_loss=0.02814, over 4957.00 frames.], tot_loss[loss=0.1386, simple_loss=0.2213, pruned_loss=0.02794, over 985433.89 frames.], batch size: 91, aishell_tot_loss[loss=0.1413, simple_loss=0.2287, pruned_loss=0.02697, over 975239.76 frames.], datatang_tot_loss[loss=0.137, simple_loss=0.2148, pruned_loss=0.02963, over 976437.32 frames.], batch size: 91, lr: 2.83e-04
+2022-06-19 07:07:34,100 INFO [train.py:874] (2/4) Epoch 29, batch 1900, aishell_loss[loss=0.1148, simple_loss=0.1949, pruned_loss=0.01739, over 4816.00 frames.], tot_loss[loss=0.1397, simple_loss=0.2226, pruned_loss=0.02843, over 985265.80 frames.], batch size: 24, aishell_tot_loss[loss=0.1421, simple_loss=0.2294, pruned_loss=0.02737, over 976385.65 frames.], datatang_tot_loss[loss=0.1372, simple_loss=0.215, pruned_loss=0.02972, over 977410.10 frames.], batch size: 24, lr: 2.83e-04
+2022-06-19 07:08:03,014 INFO [train.py:874] (2/4) Epoch 29, batch 1950, aishell_loss[loss=0.1383, simple_loss=0.2294, pruned_loss=0.02365, over 4859.00 frames.], tot_loss[loss=0.1399, simple_loss=0.2227, pruned_loss=0.02856, over 985401.95 frames.], batch size: 38, aishell_tot_loss[loss=0.1426, simple_loss=0.2297, pruned_loss=0.02772, over 977575.81 frames.], datatang_tot_loss[loss=0.1369, simple_loss=0.2147, pruned_loss=0.02952, over 978354.52 frames.], batch size: 38, lr: 2.83e-04
+2022-06-19 07:08:32,530 INFO [train.py:874] (2/4) Epoch 29, batch 2000, aishell_loss[loss=0.1483, simple_loss=0.2286, pruned_loss=0.03405, over 4958.00 frames.], tot_loss[loss=0.1402, simple_loss=0.2232, pruned_loss=0.02861, over 985379.96 frames.], batch size: 40, aishell_tot_loss[loss=0.1426, simple_loss=0.2295, pruned_loss=0.0278, over 978635.37 frames.], datatang_tot_loss[loss=0.137, simple_loss=0.2148, pruned_loss=0.02957, over 979050.72 frames.], batch size: 40, lr: 2.83e-04
+2022-06-19 07:08:32,531 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 07:08:49,525 INFO [train.py:914] (2/4) Epoch 29, validation: loss=0.1647, simple_loss=0.2482, pruned_loss=0.04062, over 1622729.00 frames.
+2022-06-19 07:09:18,419 INFO [train.py:874] (2/4) Epoch 29, batch 2050, datatang_loss[loss=0.1524, simple_loss=0.2286, pruned_loss=0.03812, over 4942.00 frames.], tot_loss[loss=0.1398, simple_loss=0.2224, pruned_loss=0.02862, over 985172.39 frames.], batch size: 88, aishell_tot_loss[loss=0.1421, simple_loss=0.229, pruned_loss=0.02757, over 979220.21 frames.], datatang_tot_loss[loss=0.1372, simple_loss=0.2148, pruned_loss=0.02977, over 979770.88 frames.], batch size: 88, lr: 2.83e-04
+2022-06-19 07:09:48,712 INFO [train.py:874] (2/4) Epoch 29, batch 2100, datatang_loss[loss=0.1097, simple_loss=0.1921, pruned_loss=0.01364, over 4918.00 frames.], tot_loss[loss=0.1399, simple_loss=0.222, pruned_loss=0.02885, over 985167.84 frames.], batch size: 71, aishell_tot_loss[loss=0.142, simple_loss=0.2288, pruned_loss=0.0276, over 979864.53 frames.], datatang_tot_loss[loss=0.1374, simple_loss=0.215, pruned_loss=0.02992, over 980453.14 frames.], batch size: 71, lr: 2.83e-04
+2022-06-19 07:10:18,285 INFO [train.py:874] (2/4) Epoch 29, batch 2150, datatang_loss[loss=0.1358, simple_loss=0.2264, pruned_loss=0.02258, over 4930.00 frames.], tot_loss[loss=0.1401, simple_loss=0.2225, pruned_loss=0.0288, over 985648.93 frames.], batch size: 57, aishell_tot_loss[loss=0.1419, simple_loss=0.2288, pruned_loss=0.02751, over 980619.84 frames.], datatang_tot_loss[loss=0.1377, simple_loss=0.2154, pruned_loss=0.03, over 981392.04 frames.], batch size: 57, lr: 2.82e-04
+2022-06-19 07:10:47,507 INFO [train.py:874] (2/4) Epoch 29, batch 2200, datatang_loss[loss=0.1307, simple_loss=0.2062, pruned_loss=0.02763, over 4986.00 frames.], tot_loss[loss=0.1411, simple_loss=0.2234, pruned_loss=0.02942, over 985721.37 frames.], batch size: 37, aishell_tot_loss[loss=0.1431, simple_loss=0.2299, pruned_loss=0.02819, over 981268.70 frames.], datatang_tot_loss[loss=0.1377, simple_loss=0.2154, pruned_loss=0.02998, over 981906.73 frames.], batch size: 37, lr: 2.82e-04
+2022-06-19 07:11:17,484 INFO [train.py:874] (2/4) Epoch 29, batch 2250, datatang_loss[loss=0.1438, simple_loss=0.2188, pruned_loss=0.03444, over 4976.00 frames.], tot_loss[loss=0.1405, simple_loss=0.2228, pruned_loss=0.02915, over 985486.32 frames.], batch size: 45, aishell_tot_loss[loss=0.1431, simple_loss=0.2296, pruned_loss=0.02831, over 981698.88 frames.], datatang_tot_loss[loss=0.1372, simple_loss=0.2152, pruned_loss=0.02965, over 982199.48 frames.], batch size: 45, lr: 2.82e-04
+2022-06-19 07:11:46,209 INFO [train.py:874] (2/4) Epoch 29, batch 2300, datatang_loss[loss=0.1241, simple_loss=0.2076, pruned_loss=0.02028, over 4916.00 frames.], tot_loss[loss=0.1402, simple_loss=0.2225, pruned_loss=0.02897, over 985355.94 frames.], batch size: 75, aishell_tot_loss[loss=0.1434, simple_loss=0.23, pruned_loss=0.02837, over 981903.03 frames.], datatang_tot_loss[loss=0.137, simple_loss=0.2152, pruned_loss=0.02941, over 982668.87 frames.], batch size: 75, lr: 2.82e-04
+2022-06-19 07:12:16,282 INFO [train.py:874] (2/4) Epoch 29, batch 2350, datatang_loss[loss=0.1418, simple_loss=0.2248, pruned_loss=0.0294, over 4953.00 frames.], tot_loss[loss=0.1405, simple_loss=0.2228, pruned_loss=0.02913, over 984821.86 frames.], batch size: 91, aishell_tot_loss[loss=0.1433, simple_loss=0.2301, pruned_loss=0.02829, over 981920.11 frames.], datatang_tot_loss[loss=0.1374, simple_loss=0.2155, pruned_loss=0.02967, over 982800.70 frames.], batch size: 91, lr: 2.82e-04
+2022-06-19 07:12:46,285 INFO [train.py:874] (2/4) Epoch 29, batch 2400, aishell_loss[loss=0.1607, simple_loss=0.2506, pruned_loss=0.03546, over 4910.00 frames.], tot_loss[loss=0.1402, simple_loss=0.2227, pruned_loss=0.02885, over 985028.70 frames.], batch size: 41, aishell_tot_loss[loss=0.1431, simple_loss=0.23, pruned_loss=0.02807, over 982435.34 frames.], datatang_tot_loss[loss=0.1375, simple_loss=0.2159, pruned_loss=0.02958, over 983057.95 frames.], batch size: 41, lr: 2.82e-04
+2022-06-19 07:13:14,701 INFO [train.py:874] (2/4) Epoch 29, batch 2450, aishell_loss[loss=0.1238, simple_loss=0.2073, pruned_loss=0.02015, over 4976.00 frames.], tot_loss[loss=0.14, simple_loss=0.2224, pruned_loss=0.0288, over 985224.63 frames.], batch size: 31, aishell_tot_loss[loss=0.1432, simple_loss=0.2299, pruned_loss=0.02824, over 982513.19 frames.], datatang_tot_loss[loss=0.1372, simple_loss=0.2156, pruned_loss=0.02936, over 983728.25 frames.], batch size: 31, lr: 2.82e-04
+2022-06-19 07:13:44,298 INFO [train.py:874] (2/4) Epoch 29, batch 2500, aishell_loss[loss=0.1317, simple_loss=0.219, pruned_loss=0.02224, over 4941.00 frames.], tot_loss[loss=0.1402, simple_loss=0.2229, pruned_loss=0.02876, over 985424.53 frames.], batch size: 54, aishell_tot_loss[loss=0.1434, simple_loss=0.2302, pruned_loss=0.02828, over 983113.38 frames.], datatang_tot_loss[loss=0.1371, simple_loss=0.2156, pruned_loss=0.02931, over 983845.32 frames.], batch size: 54, lr: 2.82e-04
+2022-06-19 07:14:14,372 INFO [train.py:874] (2/4) Epoch 29, batch 2550, aishell_loss[loss=0.1622, simple_loss=0.2483, pruned_loss=0.038, over 4905.00 frames.], tot_loss[loss=0.14, simple_loss=0.2228, pruned_loss=0.02859, over 985497.24 frames.], batch size: 34, aishell_tot_loss[loss=0.1434, simple_loss=0.2302, pruned_loss=0.02827, over 983346.83 frames.], datatang_tot_loss[loss=0.1367, simple_loss=0.215, pruned_loss=0.02914, over 984162.64 frames.], batch size: 34, lr: 2.82e-04
+2022-06-19 07:14:43,098 INFO [train.py:874] (2/4) Epoch 29, batch 2600, aishell_loss[loss=0.1504, simple_loss=0.2417, pruned_loss=0.02949, over 4919.00 frames.], tot_loss[loss=0.1404, simple_loss=0.2231, pruned_loss=0.02886, over 985312.84 frames.], batch size: 68, aishell_tot_loss[loss=0.1439, simple_loss=0.2312, pruned_loss=0.02829, over 983504.41 frames.], datatang_tot_loss[loss=0.1367, simple_loss=0.2147, pruned_loss=0.02937, over 984197.53 frames.], batch size: 68, lr: 2.82e-04
+2022-06-19 07:15:12,750 INFO [train.py:874] (2/4) Epoch 29, batch 2650, datatang_loss[loss=0.1439, simple_loss=0.2311, pruned_loss=0.02828, over 4918.00 frames.], tot_loss[loss=0.1404, simple_loss=0.2237, pruned_loss=0.02853, over 985359.80 frames.], batch size: 77, aishell_tot_loss[loss=0.1441, simple_loss=0.2316, pruned_loss=0.02832, over 983769.39 frames.], datatang_tot_loss[loss=0.1363, simple_loss=0.2146, pruned_loss=0.02901, over 984331.57 frames.], batch size: 77, lr: 2.82e-04
+2022-06-19 07:15:41,968 INFO [train.py:874] (2/4) Epoch 29, batch 2700, datatang_loss[loss=0.158, simple_loss=0.2152, pruned_loss=0.05035, over 4891.00 frames.], tot_loss[loss=0.1411, simple_loss=0.2246, pruned_loss=0.02879, over 985348.18 frames.], batch size: 39, aishell_tot_loss[loss=0.1444, simple_loss=0.2321, pruned_loss=0.02834, over 983807.71 frames.], datatang_tot_loss[loss=0.1365, simple_loss=0.2146, pruned_loss=0.02923, over 984616.89 frames.], batch size: 39, lr: 2.82e-04
+2022-06-19 07:16:09,395 INFO [train.py:874] (2/4) Epoch 29, batch 2750, aishell_loss[loss=0.1325, simple_loss=0.2255, pruned_loss=0.01978, over 4941.00 frames.], tot_loss[loss=0.1409, simple_loss=0.2244, pruned_loss=0.02866, over 985596.25 frames.], batch size: 56, aishell_tot_loss[loss=0.1443, simple_loss=0.2322, pruned_loss=0.02824, over 984192.45 frames.], datatang_tot_loss[loss=0.1365, simple_loss=0.2146, pruned_loss=0.0292, over 984771.88 frames.], batch size: 56, lr: 2.82e-04
+2022-06-19 07:16:40,493 INFO [train.py:874] (2/4) Epoch 29, batch 2800, datatang_loss[loss=0.1154, simple_loss=0.1996, pruned_loss=0.01559, over 4971.00 frames.], tot_loss[loss=0.14, simple_loss=0.2233, pruned_loss=0.02839, over 985527.70 frames.], batch size: 67, aishell_tot_loss[loss=0.1437, simple_loss=0.2317, pruned_loss=0.0279, over 984223.91 frames.], datatang_tot_loss[loss=0.1365, simple_loss=0.2145, pruned_loss=0.0292, over 984918.51 frames.], batch size: 67, lr: 2.82e-04
+2022-06-19 07:17:10,569 INFO [train.py:874] (2/4) Epoch 29, batch 2850, aishell_loss[loss=0.1494, simple_loss=0.2494, pruned_loss=0.02473, over 4963.00 frames.], tot_loss[loss=0.1399, simple_loss=0.2231, pruned_loss=0.02832, over 985584.60 frames.], batch size: 69, aishell_tot_loss[loss=0.1438, simple_loss=0.2315, pruned_loss=0.02802, over 984399.77 frames.], datatang_tot_loss[loss=0.1362, simple_loss=0.2144, pruned_loss=0.02898, over 985042.59 frames.], batch size: 69, lr: 2.82e-04
+2022-06-19 07:17:38,887 INFO [train.py:874] (2/4) Epoch 29, batch 2900, aishell_loss[loss=0.145, simple_loss=0.2287, pruned_loss=0.03065, over 4869.00 frames.], tot_loss[loss=0.1394, simple_loss=0.2225, pruned_loss=0.02811, over 985241.16 frames.], batch size: 35, aishell_tot_loss[loss=0.1439, simple_loss=0.2315, pruned_loss=0.02809, over 984087.63 frames.], datatang_tot_loss[loss=0.1355, simple_loss=0.2136, pruned_loss=0.02867, over 985200.32 frames.], batch size: 35, lr: 2.82e-04
+2022-06-19 07:18:09,828 INFO [train.py:874] (2/4) Epoch 29, batch 2950, datatang_loss[loss=0.1313, simple_loss=0.2098, pruned_loss=0.02642, over 4922.00 frames.], tot_loss[loss=0.1387, simple_loss=0.2219, pruned_loss=0.02773, over 985099.19 frames.], batch size: 73, aishell_tot_loss[loss=0.1437, simple_loss=0.2314, pruned_loss=0.02799, over 984013.84 frames.], datatang_tot_loss[loss=0.1349, simple_loss=0.2132, pruned_loss=0.02831, over 985266.59 frames.], batch size: 73, lr: 2.82e-04
+2022-06-19 07:18:38,429 INFO [train.py:874] (2/4) Epoch 29, batch 3000, datatang_loss[loss=0.1473, simple_loss=0.2255, pruned_loss=0.03455, over 4915.00 frames.], tot_loss[loss=0.139, simple_loss=0.2223, pruned_loss=0.02783, over 985074.26 frames.], batch size: 42, aishell_tot_loss[loss=0.1435, simple_loss=0.2313, pruned_loss=0.02791, over 984187.27 frames.], datatang_tot_loss[loss=0.1351, simple_loss=0.2134, pruned_loss=0.0284, over 985190.28 frames.], batch size: 42, lr: 2.81e-04
+2022-06-19 07:18:38,430 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 07:18:54,222 INFO [train.py:914] (2/4) Epoch 29, validation: loss=0.165, simple_loss=0.2488, pruned_loss=0.0406, over 1622729.00 frames.
+2022-06-19 07:19:24,186 INFO [train.py:874] (2/4) Epoch 29, batch 3050, datatang_loss[loss=0.1494, simple_loss=0.2085, pruned_loss=0.04515, over 4920.00 frames.], tot_loss[loss=0.1396, simple_loss=0.222, pruned_loss=0.02854, over 985086.92 frames.], batch size: 34, aishell_tot_loss[loss=0.144, simple_loss=0.2315, pruned_loss=0.02821, over 984353.81 frames.], datatang_tot_loss[loss=0.1355, simple_loss=0.2135, pruned_loss=0.02879, over 985103.18 frames.], batch size: 34, lr: 2.81e-04
+2022-06-19 07:19:51,651 INFO [train.py:874] (2/4) Epoch 29, batch 3100, aishell_loss[loss=0.1334, simple_loss=0.2233, pruned_loss=0.02175, over 4932.00 frames.], tot_loss[loss=0.14, simple_loss=0.2228, pruned_loss=0.02865, over 985003.67 frames.], batch size: 32, aishell_tot_loss[loss=0.1439, simple_loss=0.2316, pruned_loss=0.02808, over 984488.76 frames.], datatang_tot_loss[loss=0.1361, simple_loss=0.214, pruned_loss=0.02905, over 984973.26 frames.], batch size: 32, lr: 2.81e-04
+2022-06-19 07:20:22,493 INFO [train.py:874] (2/4) Epoch 29, batch 3150, datatang_loss[loss=0.1533, simple_loss=0.2315, pruned_loss=0.03757, over 4921.00 frames.], tot_loss[loss=0.1393, simple_loss=0.2224, pruned_loss=0.02814, over 985020.24 frames.], batch size: 83, aishell_tot_loss[loss=0.1434, simple_loss=0.2312, pruned_loss=0.02783, over 984313.72 frames.], datatang_tot_loss[loss=0.1358, simple_loss=0.2139, pruned_loss=0.02879, over 985222.55 frames.], batch size: 83, lr: 2.81e-04
+2022-06-19 07:20:52,232 INFO [train.py:874] (2/4) Epoch 29, batch 3200, aishell_loss[loss=0.1609, simple_loss=0.2514, pruned_loss=0.03518, over 4936.00 frames.], tot_loss[loss=0.1395, simple_loss=0.2227, pruned_loss=0.02814, over 985215.31 frames.], batch size: 79, aishell_tot_loss[loss=0.1439, simple_loss=0.2317, pruned_loss=0.02801, over 984471.99 frames.], datatang_tot_loss[loss=0.1354, simple_loss=0.2137, pruned_loss=0.02857, over 985345.40 frames.], batch size: 79, lr: 2.81e-04
+2022-06-19 07:21:20,882 INFO [train.py:874] (2/4) Epoch 29, batch 3250, aishell_loss[loss=0.143, simple_loss=0.2306, pruned_loss=0.02766, over 4931.00 frames.], tot_loss[loss=0.14, simple_loss=0.2233, pruned_loss=0.02836, over 985385.82 frames.], batch size: 56, aishell_tot_loss[loss=0.1438, simple_loss=0.2317, pruned_loss=0.02793, over 984791.05 frames.], datatang_tot_loss[loss=0.136, simple_loss=0.2142, pruned_loss=0.02885, over 985280.25 frames.], batch size: 56, lr: 2.81e-04
+2022-06-19 07:21:51,293 INFO [train.py:874] (2/4) Epoch 29, batch 3300, aishell_loss[loss=0.1229, simple_loss=0.2169, pruned_loss=0.01444, over 4918.00 frames.], tot_loss[loss=0.1403, simple_loss=0.2234, pruned_loss=0.02857, over 985405.85 frames.], batch size: 52, aishell_tot_loss[loss=0.1434, simple_loss=0.2312, pruned_loss=0.02783, over 984848.64 frames.], datatang_tot_loss[loss=0.1367, simple_loss=0.215, pruned_loss=0.02917, over 985326.97 frames.], batch size: 52, lr: 2.81e-04
+2022-06-19 07:22:20,992 INFO [train.py:874] (2/4) Epoch 29, batch 3350, datatang_loss[loss=0.1456, simple_loss=0.2223, pruned_loss=0.03446, over 4955.00 frames.], tot_loss[loss=0.1394, simple_loss=0.2225, pruned_loss=0.02813, over 985573.32 frames.], batch size: 45, aishell_tot_loss[loss=0.1429, simple_loss=0.2307, pruned_loss=0.02754, over 985116.58 frames.], datatang_tot_loss[loss=0.1362, simple_loss=0.2145, pruned_loss=0.02899, over 985315.71 frames.], batch size: 45, lr: 2.81e-04
+2022-06-19 07:22:48,911 INFO [train.py:874] (2/4) Epoch 29, batch 3400, datatang_loss[loss=0.1258, simple_loss=0.2114, pruned_loss=0.02006, over 4921.00 frames.], tot_loss[loss=0.1396, simple_loss=0.2222, pruned_loss=0.02844, over 985601.31 frames.], batch size: 81, aishell_tot_loss[loss=0.1433, simple_loss=0.231, pruned_loss=0.02777, over 985005.31 frames.], datatang_tot_loss[loss=0.1362, simple_loss=0.2143, pruned_loss=0.02903, over 985532.64 frames.], batch size: 81, lr: 2.81e-04
+2022-06-19 07:23:20,398 INFO [train.py:874] (2/4) Epoch 29, batch 3450, datatang_loss[loss=0.1356, simple_loss=0.21, pruned_loss=0.03064, over 4941.00 frames.], tot_loss[loss=0.1394, simple_loss=0.2222, pruned_loss=0.02827, over 985330.08 frames.], batch size: 69, aishell_tot_loss[loss=0.1428, simple_loss=0.2304, pruned_loss=0.02755, over 984850.29 frames.], datatang_tot_loss[loss=0.1363, simple_loss=0.2145, pruned_loss=0.02909, over 985482.70 frames.], batch size: 69, lr: 2.81e-04
+2022-06-19 07:23:50,334 INFO [train.py:874] (2/4) Epoch 29, batch 3500, aishell_loss[loss=0.1397, simple_loss=0.2308, pruned_loss=0.0243, over 4964.00 frames.], tot_loss[loss=0.1395, simple_loss=0.2223, pruned_loss=0.02832, over 985632.01 frames.], batch size: 40, aishell_tot_loss[loss=0.1421, simple_loss=0.2297, pruned_loss=0.02725, over 985052.53 frames.], datatang_tot_loss[loss=0.137, simple_loss=0.2151, pruned_loss=0.02945, over 985639.96 frames.], batch size: 40, lr: 2.81e-04
+2022-06-19 07:24:19,340 INFO [train.py:874] (2/4) Epoch 29, batch 3550, aishell_loss[loss=0.1269, simple_loss=0.2204, pruned_loss=0.01667, over 4942.00 frames.], tot_loss[loss=0.1392, simple_loss=0.2221, pruned_loss=0.02814, over 985776.35 frames.], batch size: 49, aishell_tot_loss[loss=0.1423, simple_loss=0.2298, pruned_loss=0.02743, over 985134.26 frames.], datatang_tot_loss[loss=0.1364, simple_loss=0.2147, pruned_loss=0.02909, over 985787.90 frames.], batch size: 49, lr: 2.81e-04
+2022-06-19 07:24:50,025 INFO [train.py:874] (2/4) Epoch 29, batch 3600, datatang_loss[loss=0.1262, simple_loss=0.2066, pruned_loss=0.02291, over 4922.00 frames.], tot_loss[loss=0.1386, simple_loss=0.2212, pruned_loss=0.02798, over 985805.35 frames.], batch size: 81, aishell_tot_loss[loss=0.1418, simple_loss=0.2292, pruned_loss=0.02723, over 985015.03 frames.], datatang_tot_loss[loss=0.1363, simple_loss=0.2144, pruned_loss=0.02908, over 986003.00 frames.], batch size: 81, lr: 2.81e-04
+2022-06-19 07:25:18,914 INFO [train.py:874] (2/4) Epoch 29, batch 3650, datatang_loss[loss=0.1624, simple_loss=0.2425, pruned_loss=0.04116, over 4957.00 frames.], tot_loss[loss=0.1388, simple_loss=0.2212, pruned_loss=0.02823, over 985716.14 frames.], batch size: 99, aishell_tot_loss[loss=0.1419, simple_loss=0.2291, pruned_loss=0.02739, over 984916.56 frames.], datatang_tot_loss[loss=0.1364, simple_loss=0.2146, pruned_loss=0.02912, over 986056.86 frames.], batch size: 99, lr: 2.81e-04
+2022-06-19 07:25:46,583 INFO [train.py:874] (2/4) Epoch 29, batch 3700, aishell_loss[loss=0.1328, simple_loss=0.2173, pruned_loss=0.02419, over 4868.00 frames.], tot_loss[loss=0.1383, simple_loss=0.2209, pruned_loss=0.02783, over 985545.40 frames.], batch size: 28, aishell_tot_loss[loss=0.1421, simple_loss=0.2293, pruned_loss=0.02749, over 984806.91 frames.], datatang_tot_loss[loss=0.1356, simple_loss=0.2141, pruned_loss=0.02856, over 986026.61 frames.], batch size: 28, lr: 2.81e-04
+2022-06-19 07:26:16,107 INFO [train.py:874] (2/4) Epoch 29, batch 3750, aishell_loss[loss=0.1457, simple_loss=0.2468, pruned_loss=0.02227, over 4911.00 frames.], tot_loss[loss=0.1394, simple_loss=0.2228, pruned_loss=0.028, over 985367.66 frames.], batch size: 68, aishell_tot_loss[loss=0.1423, simple_loss=0.2298, pruned_loss=0.02739, over 984769.38 frames.], datatang_tot_loss[loss=0.1362, simple_loss=0.2147, pruned_loss=0.02885, over 985962.04 frames.], batch size: 68, lr: 2.81e-04
+2022-06-19 07:26:43,585 INFO [train.py:874] (2/4) Epoch 29, batch 3800, datatang_loss[loss=0.1259, simple_loss=0.2104, pruned_loss=0.02072, over 4951.00 frames.], tot_loss[loss=0.1392, simple_loss=0.2223, pruned_loss=0.02799, over 984883.50 frames.], batch size: 62, aishell_tot_loss[loss=0.1426, simple_loss=0.23, pruned_loss=0.02758, over 984334.98 frames.], datatang_tot_loss[loss=0.1358, simple_loss=0.2143, pruned_loss=0.02863, over 985873.29 frames.], batch size: 62, lr: 2.81e-04
+2022-06-19 07:27:13,067 INFO [train.py:874] (2/4) Epoch 29, batch 3850, aishell_loss[loss=0.1047, simple_loss=0.1907, pruned_loss=0.00933, over 4865.00 frames.], tot_loss[loss=0.1386, simple_loss=0.222, pruned_loss=0.0276, over 984954.68 frames.], batch size: 28, aishell_tot_loss[loss=0.1422, simple_loss=0.2297, pruned_loss=0.02738, over 984372.10 frames.], datatang_tot_loss[loss=0.1354, simple_loss=0.214, pruned_loss=0.02838, over 985877.11 frames.], batch size: 28, lr: 2.80e-04
+2022-06-19 07:27:40,473 INFO [train.py:874] (2/4) Epoch 29, batch 3900, datatang_loss[loss=0.1366, simple_loss=0.2245, pruned_loss=0.02437, over 4957.00 frames.], tot_loss[loss=0.1384, simple_loss=0.2213, pruned_loss=0.0278, over 984924.89 frames.], batch size: 91, aishell_tot_loss[loss=0.1421, simple_loss=0.2294, pruned_loss=0.0274, over 984289.95 frames.], datatang_tot_loss[loss=0.1356, simple_loss=0.2143, pruned_loss=0.02844, over 985822.83 frames.], batch size: 91, lr: 2.80e-04
+2022-06-19 07:28:09,834 INFO [train.py:874] (2/4) Epoch 29, batch 3950, aishell_loss[loss=0.1484, simple_loss=0.2449, pruned_loss=0.02592, over 4973.00 frames.], tot_loss[loss=0.1386, simple_loss=0.2215, pruned_loss=0.02782, over 984927.21 frames.], batch size: 51, aishell_tot_loss[loss=0.1425, simple_loss=0.23, pruned_loss=0.0275, over 984317.64 frames.], datatang_tot_loss[loss=0.1353, simple_loss=0.2139, pruned_loss=0.02833, over 985758.52 frames.], batch size: 51, lr: 2.80e-04
+2022-06-19 07:28:37,199 INFO [train.py:874] (2/4) Epoch 29, batch 4000, aishell_loss[loss=0.1754, simple_loss=0.2614, pruned_loss=0.0447, over 4973.00 frames.], tot_loss[loss=0.1389, simple_loss=0.2218, pruned_loss=0.02798, over 985041.06 frames.], batch size: 48, aishell_tot_loss[loss=0.1428, simple_loss=0.2304, pruned_loss=0.0276, over 984279.64 frames.], datatang_tot_loss[loss=0.1353, simple_loss=0.214, pruned_loss=0.02836, over 985876.22 frames.], batch size: 48, lr: 2.80e-04
+2022-06-19 07:28:37,200 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 07:28:54,126 INFO [train.py:914] (2/4) Epoch 29, validation: loss=0.165, simple_loss=0.2485, pruned_loss=0.04075, over 1622729.00 frames.
+2022-06-19 07:29:22,708 INFO [train.py:874] (2/4) Epoch 29, batch 4050, aishell_loss[loss=0.1499, simple_loss=0.2339, pruned_loss=0.03289, over 4958.00 frames.], tot_loss[loss=0.1388, simple_loss=0.2219, pruned_loss=0.02789, over 985090.62 frames.], batch size: 61, aishell_tot_loss[loss=0.1425, simple_loss=0.2299, pruned_loss=0.02757, over 984562.77 frames.], datatang_tot_loss[loss=0.1353, simple_loss=0.2141, pruned_loss=0.0283, over 985664.80 frames.], batch size: 61, lr: 2.80e-04
+2022-06-19 07:29:48,239 INFO [train.py:874] (2/4) Epoch 29, batch 4100, aishell_loss[loss=0.1296, simple_loss=0.214, pruned_loss=0.0226, over 4917.00 frames.], tot_loss[loss=0.1391, simple_loss=0.2221, pruned_loss=0.02799, over 984914.93 frames.], batch size: 41, aishell_tot_loss[loss=0.1432, simple_loss=0.2305, pruned_loss=0.02794, over 984516.11 frames.], datatang_tot_loss[loss=0.1349, simple_loss=0.2137, pruned_loss=0.02804, over 985538.05 frames.], batch size: 41, lr: 2.80e-04
+2022-06-19 07:30:56,064 INFO [train.py:874] (2/4) Epoch 30, batch 50, datatang_loss[loss=0.1218, simple_loss=0.1903, pruned_loss=0.02662, over 4939.00 frames.], tot_loss[loss=0.1345, simple_loss=0.2167, pruned_loss=0.02611, over 218620.66 frames.], batch size: 50, aishell_tot_loss[loss=0.1422, simple_loss=0.2294, pruned_loss=0.02747, over 120370.80 frames.], datatang_tot_loss[loss=0.1263, simple_loss=0.2034, pruned_loss=0.02467, over 111908.11 frames.], batch size: 50, lr: 2.75e-04
+2022-06-19 07:31:24,113 INFO [train.py:874] (2/4) Epoch 30, batch 100, datatang_loss[loss=0.1202, simple_loss=0.204, pruned_loss=0.01821, over 4912.00 frames.], tot_loss[loss=0.1361, simple_loss=0.2186, pruned_loss=0.02683, over 389023.29 frames.], batch size: 64, aishell_tot_loss[loss=0.1418, simple_loss=0.2297, pruned_loss=0.02693, over 218767.88 frames.], datatang_tot_loss[loss=0.1303, simple_loss=0.2072, pruned_loss=0.02666, over 218736.86 frames.], batch size: 64, lr: 2.75e-04
+2022-06-19 07:31:54,717 INFO [train.py:874] (2/4) Epoch 30, batch 150, aishell_loss[loss=0.1364, simple_loss=0.2212, pruned_loss=0.02577, over 4963.00 frames.], tot_loss[loss=0.1368, simple_loss=0.2195, pruned_loss=0.02699, over 521258.49 frames.], batch size: 31, aishell_tot_loss[loss=0.1416, simple_loss=0.2284, pruned_loss=0.0274, over 329088.65 frames.], datatang_tot_loss[loss=0.1307, simple_loss=0.2087, pruned_loss=0.02641, over 288421.91 frames.], batch size: 31, lr: 2.75e-04
+2022-06-19 07:32:24,879 INFO [train.py:874] (2/4) Epoch 30, batch 200, aishell_loss[loss=0.1569, simple_loss=0.231, pruned_loss=0.04138, over 4929.00 frames.], tot_loss[loss=0.1376, simple_loss=0.2205, pruned_loss=0.02738, over 624289.95 frames.], batch size: 32, aishell_tot_loss[loss=0.1431, simple_loss=0.2301, pruned_loss=0.02805, over 397664.94 frames.], datatang_tot_loss[loss=0.1314, simple_loss=0.2098, pruned_loss=0.02652, over 379779.51 frames.], batch size: 32, lr: 2.75e-04
+2022-06-19 07:32:53,104 INFO [train.py:874] (2/4) Epoch 30, batch 250, datatang_loss[loss=0.1399, simple_loss=0.2226, pruned_loss=0.02859, over 4937.00 frames.], tot_loss[loss=0.1386, simple_loss=0.2213, pruned_loss=0.0279, over 704103.74 frames.], batch size: 88, aishell_tot_loss[loss=0.1439, simple_loss=0.2309, pruned_loss=0.02845, over 461279.60 frames.], datatang_tot_loss[loss=0.1326, simple_loss=0.2111, pruned_loss=0.02703, over 456561.88 frames.], batch size: 88, lr: 2.75e-04
+2022-06-19 07:33:24,729 INFO [train.py:874] (2/4) Epoch 30, batch 300, aishell_loss[loss=0.1491, simple_loss=0.2389, pruned_loss=0.02968, over 4973.00 frames.], tot_loss[loss=0.1388, simple_loss=0.222, pruned_loss=0.02781, over 766347.10 frames.], batch size: 48, aishell_tot_loss[loss=0.144, simple_loss=0.2314, pruned_loss=0.02828, over 522848.00 frames.], datatang_tot_loss[loss=0.133, simple_loss=0.2117, pruned_loss=0.02714, over 518869.46 frames.], batch size: 48, lr: 2.75e-04
+2022-06-19 07:33:54,678 INFO [train.py:874] (2/4) Epoch 30, batch 350, aishell_loss[loss=0.15, simple_loss=0.2368, pruned_loss=0.03155, over 4951.00 frames.], tot_loss[loss=0.1395, simple_loss=0.2232, pruned_loss=0.02783, over 815117.84 frames.], batch size: 32, aishell_tot_loss[loss=0.1439, simple_loss=0.2317, pruned_loss=0.02805, over 593374.44 frames.], datatang_tot_loss[loss=0.1334, simple_loss=0.2121, pruned_loss=0.02738, over 557207.62 frames.], batch size: 32, lr: 2.75e-04
+2022-06-19 07:34:23,749 INFO [train.py:874] (2/4) Epoch 30, batch 400, datatang_loss[loss=0.1268, simple_loss=0.2016, pruned_loss=0.026, over 4981.00 frames.], tot_loss[loss=0.1389, simple_loss=0.2223, pruned_loss=0.02772, over 852471.99 frames.], batch size: 31, aishell_tot_loss[loss=0.1431, simple_loss=0.2304, pruned_loss=0.02785, over 642646.19 frames.], datatang_tot_loss[loss=0.1336, simple_loss=0.2123, pruned_loss=0.02747, over 603770.86 frames.], batch size: 31, lr: 2.75e-04
+2022-06-19 07:34:53,344 INFO [train.py:874] (2/4) Epoch 30, batch 450, datatang_loss[loss=0.1331, simple_loss=0.2144, pruned_loss=0.02588, over 4945.00 frames.], tot_loss[loss=0.1387, simple_loss=0.2215, pruned_loss=0.028, over 881608.78 frames.], batch size: 62, aishell_tot_loss[loss=0.1428, simple_loss=0.2299, pruned_loss=0.02785, over 681121.23 frames.], datatang_tot_loss[loss=0.134, simple_loss=0.2122, pruned_loss=0.02791, over 650419.04 frames.], batch size: 62, lr: 2.75e-04
+2022-06-19 07:35:22,817 INFO [train.py:874] (2/4) Epoch 30, batch 500, datatang_loss[loss=0.1201, simple_loss=0.2029, pruned_loss=0.01867, over 4909.00 frames.], tot_loss[loss=0.1391, simple_loss=0.2218, pruned_loss=0.02817, over 904690.29 frames.], batch size: 64, aishell_tot_loss[loss=0.1424, simple_loss=0.2295, pruned_loss=0.02767, over 722326.79 frames.], datatang_tot_loss[loss=0.1348, simple_loss=0.2127, pruned_loss=0.0284, over 683922.61 frames.], batch size: 64, lr: 2.75e-04
+2022-06-19 07:35:50,079 INFO [train.py:874] (2/4) Epoch 30, batch 550, aishell_loss[loss=0.1598, simple_loss=0.2563, pruned_loss=0.03163, over 4963.00 frames.], tot_loss[loss=0.1403, simple_loss=0.2234, pruned_loss=0.02865, over 922521.50 frames.], batch size: 61, aishell_tot_loss[loss=0.1439, simple_loss=0.231, pruned_loss=0.02843, over 762350.11 frames.], datatang_tot_loss[loss=0.1346, simple_loss=0.2127, pruned_loss=0.02832, over 708563.64 frames.], batch size: 61, lr: 2.75e-04
+2022-06-19 07:36:21,122 INFO [train.py:874] (2/4) Epoch 30, batch 600, datatang_loss[loss=0.1288, simple_loss=0.2099, pruned_loss=0.0239, over 4916.00 frames.], tot_loss[loss=0.1402, simple_loss=0.2232, pruned_loss=0.02861, over 936534.37 frames.], batch size: 75, aishell_tot_loss[loss=0.1447, simple_loss=0.2317, pruned_loss=0.02884, over 789689.70 frames.], datatang_tot_loss[loss=0.1339, simple_loss=0.2119, pruned_loss=0.02792, over 739922.12 frames.], batch size: 75, lr: 2.75e-04
+2022-06-19 07:36:51,307 INFO [train.py:874] (2/4) Epoch 30, batch 650, datatang_loss[loss=0.1808, simple_loss=0.2624, pruned_loss=0.0496, over 4926.00 frames.], tot_loss[loss=0.1397, simple_loss=0.2228, pruned_loss=0.02827, over 947266.83 frames.], batch size: 109, aishell_tot_loss[loss=0.144, simple_loss=0.2312, pruned_loss=0.02841, over 812484.42 frames.], datatang_tot_loss[loss=0.1342, simple_loss=0.2124, pruned_loss=0.028, over 768984.94 frames.], batch size: 109, lr: 2.75e-04
+2022-06-19 07:37:20,502 INFO [train.py:874] (2/4) Epoch 30, batch 700, datatang_loss[loss=0.1404, simple_loss=0.2161, pruned_loss=0.03232, over 4957.00 frames.], tot_loss[loss=0.1392, simple_loss=0.2225, pruned_loss=0.028, over 955829.63 frames.], batch size: 55, aishell_tot_loss[loss=0.1435, simple_loss=0.2308, pruned_loss=0.02806, over 830553.20 frames.], datatang_tot_loss[loss=0.1345, simple_loss=0.213, pruned_loss=0.02804, over 797411.97 frames.], batch size: 55, lr: 2.75e-04
+2022-06-19 07:37:50,961 INFO [train.py:874] (2/4) Epoch 30, batch 750, aishell_loss[loss=0.1711, simple_loss=0.2592, pruned_loss=0.04153, over 4959.00 frames.], tot_loss[loss=0.1388, simple_loss=0.2222, pruned_loss=0.02771, over 962280.56 frames.], batch size: 40, aishell_tot_loss[loss=0.1433, simple_loss=0.2306, pruned_loss=0.02796, over 849993.71 frames.], datatang_tot_loss[loss=0.1341, simple_loss=0.2127, pruned_loss=0.02773, over 817913.63 frames.], batch size: 40, lr: 2.75e-04
+2022-06-19 07:38:21,560 INFO [train.py:874] (2/4) Epoch 30, batch 800, datatang_loss[loss=0.1624, simple_loss=0.229, pruned_loss=0.04797, over 4948.00 frames.], tot_loss[loss=0.1388, simple_loss=0.2223, pruned_loss=0.02765, over 967771.54 frames.], batch size: 55, aishell_tot_loss[loss=0.1428, simple_loss=0.2303, pruned_loss=0.02762, over 869184.19 frames.], datatang_tot_loss[loss=0.1344, simple_loss=0.2128, pruned_loss=0.02798, over 833980.42 frames.], batch size: 55, lr: 2.75e-04
+2022-06-19 07:38:51,806 INFO [train.py:874] (2/4) Epoch 30, batch 850, datatang_loss[loss=0.131, simple_loss=0.2163, pruned_loss=0.02286, over 4937.00 frames.], tot_loss[loss=0.1381, simple_loss=0.2211, pruned_loss=0.02758, over 971463.53 frames.], batch size: 88, aishell_tot_loss[loss=0.1425, simple_loss=0.2301, pruned_loss=0.02743, over 879292.42 frames.], datatang_tot_loss[loss=0.1342, simple_loss=0.2124, pruned_loss=0.02804, over 855966.62 frames.], batch size: 88, lr: 2.75e-04
+2022-06-19 07:39:22,124 INFO [train.py:874] (2/4) Epoch 30, batch 900, datatang_loss[loss=0.1455, simple_loss=0.2197, pruned_loss=0.03564, over 4978.00 frames.], tot_loss[loss=0.1388, simple_loss=0.2216, pruned_loss=0.02799, over 974520.37 frames.], batch size: 26, aishell_tot_loss[loss=0.1429, simple_loss=0.2305, pruned_loss=0.02767, over 889338.39 frames.], datatang_tot_loss[loss=0.1347, simple_loss=0.2129, pruned_loss=0.02822, over 874087.20 frames.], batch size: 26, lr: 2.74e-04
+2022-06-19 07:39:50,712 INFO [train.py:874] (2/4) Epoch 30, batch 950, aishell_loss[loss=0.1414, simple_loss=0.2352, pruned_loss=0.0238, over 4860.00 frames.], tot_loss[loss=0.1392, simple_loss=0.2222, pruned_loss=0.02806, over 977267.66 frames.], batch size: 35, aishell_tot_loss[loss=0.143, simple_loss=0.2308, pruned_loss=0.02763, over 901525.48 frames.], datatang_tot_loss[loss=0.135, simple_loss=0.2132, pruned_loss=0.02836, over 886543.67 frames.], batch size: 35, lr: 2.74e-04
+2022-06-19 07:40:20,532 INFO [train.py:874] (2/4) Epoch 30, batch 1000, datatang_loss[loss=0.1442, simple_loss=0.2267, pruned_loss=0.0308, over 4958.00 frames.], tot_loss[loss=0.1383, simple_loss=0.2217, pruned_loss=0.02745, over 979193.79 frames.], batch size: 45, aishell_tot_loss[loss=0.1426, simple_loss=0.2304, pruned_loss=0.02736, over 912329.62 frames.], datatang_tot_loss[loss=0.1343, simple_loss=0.2128, pruned_loss=0.02795, over 897239.38 frames.], batch size: 45, lr: 2.74e-04
+2022-06-19 07:40:20,532 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 07:40:38,846 INFO [train.py:914] (2/4) Epoch 30, validation: loss=0.1647, simple_loss=0.2491, pruned_loss=0.04015, over 1622729.00 frames.
+2022-06-19 07:41:06,100 INFO [train.py:874] (2/4) Epoch 30, batch 1050, datatang_loss[loss=0.1322, simple_loss=0.2152, pruned_loss=0.02457, over 4942.00 frames.], tot_loss[loss=0.1381, simple_loss=0.2217, pruned_loss=0.02725, over 980966.44 frames.], batch size: 69, aishell_tot_loss[loss=0.1421, simple_loss=0.2301, pruned_loss=0.0271, over 922010.28 frames.], datatang_tot_loss[loss=0.1345, simple_loss=0.2131, pruned_loss=0.02794, over 906830.69 frames.], batch size: 69, lr: 2.74e-04
+2022-06-19 07:41:34,503 INFO [train.py:874] (2/4) Epoch 30, batch 1100, datatang_loss[loss=0.1204, simple_loss=0.2012, pruned_loss=0.01977, over 4917.00 frames.], tot_loss[loss=0.1391, simple_loss=0.2227, pruned_loss=0.02778, over 982095.66 frames.], batch size: 75, aishell_tot_loss[loss=0.1427, simple_loss=0.2305, pruned_loss=0.0274, over 930664.99 frames.], datatang_tot_loss[loss=0.1349, simple_loss=0.2134, pruned_loss=0.0282, over 914752.19 frames.], batch size: 75, lr: 2.74e-04
+2022-06-19 07:42:04,640 INFO [train.py:874] (2/4) Epoch 30, batch 1150, datatang_loss[loss=0.1409, simple_loss=0.2072, pruned_loss=0.03728, over 4949.00 frames.], tot_loss[loss=0.1391, simple_loss=0.2219, pruned_loss=0.02813, over 982679.04 frames.], batch size: 55, aishell_tot_loss[loss=0.1425, simple_loss=0.23, pruned_loss=0.0275, over 934899.67 frames.], datatang_tot_loss[loss=0.1354, simple_loss=0.2139, pruned_loss=0.02847, over 925584.40 frames.], batch size: 55, lr: 2.74e-04
+2022-06-19 07:42:34,011 INFO [train.py:874] (2/4) Epoch 30, batch 1200, datatang_loss[loss=0.1332, simple_loss=0.215, pruned_loss=0.02573, over 4927.00 frames.], tot_loss[loss=0.1382, simple_loss=0.221, pruned_loss=0.02775, over 983441.57 frames.], batch size: 42, aishell_tot_loss[loss=0.142, simple_loss=0.2294, pruned_loss=0.0273, over 939630.50 frames.], datatang_tot_loss[loss=0.1352, simple_loss=0.2138, pruned_loss=0.02825, over 934187.33 frames.], batch size: 42, lr: 2.74e-04
+2022-06-19 07:43:03,423 INFO [train.py:874] (2/4) Epoch 30, batch 1250, aishell_loss[loss=0.1353, simple_loss=0.2198, pruned_loss=0.0254, over 4939.00 frames.], tot_loss[loss=0.138, simple_loss=0.2209, pruned_loss=0.02759, over 984294.64 frames.], batch size: 32, aishell_tot_loss[loss=0.1416, simple_loss=0.2288, pruned_loss=0.02718, over 946712.12 frames.], datatang_tot_loss[loss=0.1351, simple_loss=0.2137, pruned_loss=0.02823, over 938744.54 frames.], batch size: 32, lr: 2.74e-04
+2022-06-19 07:43:31,155 INFO [train.py:874] (2/4) Epoch 30, batch 1300, aishell_loss[loss=0.1498, simple_loss=0.2429, pruned_loss=0.02838, over 4910.00 frames.], tot_loss[loss=0.1381, simple_loss=0.2214, pruned_loss=0.02745, over 984256.13 frames.], batch size: 41, aishell_tot_loss[loss=0.141, simple_loss=0.2283, pruned_loss=0.02686, over 952605.23 frames.], datatang_tot_loss[loss=0.1354, simple_loss=0.2139, pruned_loss=0.02843, over 942169.60 frames.], batch size: 41, lr: 2.74e-04
+2022-06-19 07:44:01,932 INFO [train.py:874] (2/4) Epoch 30, batch 1350, datatang_loss[loss=0.1489, simple_loss=0.2337, pruned_loss=0.03205, over 4920.00 frames.], tot_loss[loss=0.1388, simple_loss=0.222, pruned_loss=0.02783, over 984549.58 frames.], batch size: 83, aishell_tot_loss[loss=0.1414, simple_loss=0.2289, pruned_loss=0.02699, over 956103.64 frames.], datatang_tot_loss[loss=0.1357, simple_loss=0.2141, pruned_loss=0.02865, over 947747.56 frames.], batch size: 83, lr: 2.74e-04
+2022-06-19 07:44:33,760 INFO [train.py:874] (2/4) Epoch 30, batch 1400, aishell_loss[loss=0.1427, simple_loss=0.2301, pruned_loss=0.02767, over 4859.00 frames.], tot_loss[loss=0.1385, simple_loss=0.2218, pruned_loss=0.02764, over 984476.77 frames.], batch size: 37, aishell_tot_loss[loss=0.1408, simple_loss=0.228, pruned_loss=0.02682, over 960310.67 frames.], datatang_tot_loss[loss=0.1359, simple_loss=0.2144, pruned_loss=0.02868, over 950822.41 frames.], batch size: 37, lr: 2.74e-04
+2022-06-19 07:45:02,032 INFO [train.py:874] (2/4) Epoch 30, batch 1450, aishell_loss[loss=0.1162, simple_loss=0.209, pruned_loss=0.01175, over 4966.00 frames.], tot_loss[loss=0.1384, simple_loss=0.2217, pruned_loss=0.02753, over 984631.09 frames.], batch size: 30, aishell_tot_loss[loss=0.141, simple_loss=0.2283, pruned_loss=0.02685, over 963313.05 frames.], datatang_tot_loss[loss=0.1356, simple_loss=0.2141, pruned_loss=0.02851, over 954751.48 frames.], batch size: 30, lr: 2.74e-04
+2022-06-19 07:45:32,013 INFO [train.py:874] (2/4) Epoch 30, batch 1500, datatang_loss[loss=0.1377, simple_loss=0.2139, pruned_loss=0.03073, over 4919.00 frames.], tot_loss[loss=0.1387, simple_loss=0.2219, pruned_loss=0.02779, over 985149.72 frames.], batch size: 79, aishell_tot_loss[loss=0.1414, simple_loss=0.2287, pruned_loss=0.02703, over 966193.92 frames.], datatang_tot_loss[loss=0.1356, simple_loss=0.214, pruned_loss=0.0286, over 958434.66 frames.], batch size: 79, lr: 2.74e-04
+2022-06-19 07:46:02,916 INFO [train.py:874] (2/4) Epoch 30, batch 1550, datatang_loss[loss=0.1347, simple_loss=0.2041, pruned_loss=0.03258, over 4892.00 frames.], tot_loss[loss=0.1387, simple_loss=0.2212, pruned_loss=0.02807, over 985137.07 frames.], batch size: 47, aishell_tot_loss[loss=0.1411, simple_loss=0.228, pruned_loss=0.02711, over 968181.58 frames.], datatang_tot_loss[loss=0.1358, simple_loss=0.2141, pruned_loss=0.02879, over 961853.29 frames.], batch size: 47, lr: 2.74e-04
+2022-06-19 07:46:30,635 INFO [train.py:874] (2/4) Epoch 30, batch 1600, datatang_loss[loss=0.1363, simple_loss=0.2117, pruned_loss=0.03044, over 4965.00 frames.], tot_loss[loss=0.1381, simple_loss=0.2205, pruned_loss=0.02783, over 985457.22 frames.], batch size: 45, aishell_tot_loss[loss=0.1406, simple_loss=0.2273, pruned_loss=0.02694, over 970130.42 frames.], datatang_tot_loss[loss=0.1358, simple_loss=0.2141, pruned_loss=0.02872, over 965013.60 frames.], batch size: 45, lr: 2.74e-04
+2022-06-19 07:47:01,036 INFO [train.py:874] (2/4) Epoch 30, batch 1650, aishell_loss[loss=0.1668, simple_loss=0.2512, pruned_loss=0.04123, over 4944.00 frames.], tot_loss[loss=0.1385, simple_loss=0.2205, pruned_loss=0.02821, over 985587.06 frames.], batch size: 64, aishell_tot_loss[loss=0.141, simple_loss=0.2276, pruned_loss=0.02718, over 971514.39 frames.], datatang_tot_loss[loss=0.1359, simple_loss=0.2141, pruned_loss=0.02883, over 968084.23 frames.], batch size: 64, lr: 2.74e-04
+2022-06-19 07:47:31,286 INFO [train.py:874] (2/4) Epoch 30, batch 1700, datatang_loss[loss=0.1362, simple_loss=0.2153, pruned_loss=0.0285, over 4910.00 frames.], tot_loss[loss=0.1381, simple_loss=0.2204, pruned_loss=0.02791, over 985586.61 frames.], batch size: 71, aishell_tot_loss[loss=0.141, simple_loss=0.2277, pruned_loss=0.02714, over 973279.00 frames.], datatang_tot_loss[loss=0.1354, simple_loss=0.2137, pruned_loss=0.02859, over 970033.61 frames.], batch size: 71, lr: 2.74e-04
+2022-06-19 07:47:59,750 INFO [train.py:874] (2/4) Epoch 30, batch 1750, datatang_loss[loss=0.1402, simple_loss=0.2235, pruned_loss=0.02839, over 4916.00 frames.], tot_loss[loss=0.1382, simple_loss=0.2208, pruned_loss=0.0278, over 985457.63 frames.], batch size: 57, aishell_tot_loss[loss=0.1412, simple_loss=0.228, pruned_loss=0.02721, over 974623.78 frames.], datatang_tot_loss[loss=0.1353, simple_loss=0.2138, pruned_loss=0.02842, over 971830.36 frames.], batch size: 57, lr: 2.74e-04
+2022-06-19 07:48:29,644 INFO [train.py:874] (2/4) Epoch 30, batch 1800, datatang_loss[loss=0.1599, simple_loss=0.2304, pruned_loss=0.04466, over 4955.00 frames.], tot_loss[loss=0.1388, simple_loss=0.2212, pruned_loss=0.02822, over 985413.08 frames.], batch size: 86, aishell_tot_loss[loss=0.1418, simple_loss=0.2286, pruned_loss=0.02749, over 975850.93 frames.], datatang_tot_loss[loss=0.1354, simple_loss=0.2136, pruned_loss=0.02857, over 973449.08 frames.], batch size: 86, lr: 2.73e-04
+2022-06-19 07:49:00,478 INFO [train.py:874] (2/4) Epoch 30, batch 1850, datatang_loss[loss=0.1137, simple_loss=0.1988, pruned_loss=0.01433, over 4919.00 frames.], tot_loss[loss=0.1384, simple_loss=0.2207, pruned_loss=0.028, over 985191.86 frames.], batch size: 73, aishell_tot_loss[loss=0.1419, simple_loss=0.2288, pruned_loss=0.02754, over 976460.05 frames.], datatang_tot_loss[loss=0.135, simple_loss=0.2133, pruned_loss=0.02834, over 975164.52 frames.], batch size: 73, lr: 2.73e-04
+2022-06-19 07:49:28,727 INFO [train.py:874] (2/4) Epoch 30, batch 1900, datatang_loss[loss=0.166, simple_loss=0.2413, pruned_loss=0.04532, over 4950.00 frames.], tot_loss[loss=0.1383, simple_loss=0.2207, pruned_loss=0.02797, over 984882.21 frames.], batch size: 88, aishell_tot_loss[loss=0.1417, simple_loss=0.2284, pruned_loss=0.0275, over 977245.95 frames.], datatang_tot_loss[loss=0.135, simple_loss=0.2133, pruned_loss=0.02838, over 976241.42 frames.], batch size: 88, lr: 2.73e-04
+2022-06-19 07:49:58,517 INFO [train.py:874] (2/4) Epoch 30, batch 1950, aishell_loss[loss=0.138, simple_loss=0.2376, pruned_loss=0.01925, over 4922.00 frames.], tot_loss[loss=0.1386, simple_loss=0.2212, pruned_loss=0.02797, over 984876.09 frames.], batch size: 46, aishell_tot_loss[loss=0.1418, simple_loss=0.2287, pruned_loss=0.02748, over 978150.65 frames.], datatang_tot_loss[loss=0.135, simple_loss=0.2133, pruned_loss=0.02841, over 977221.36 frames.], batch size: 46, lr: 2.73e-04
+2022-06-19 07:50:27,890 INFO [train.py:874] (2/4) Epoch 30, batch 2000, datatang_loss[loss=0.1469, simple_loss=0.2293, pruned_loss=0.03223, over 4955.00 frames.], tot_loss[loss=0.1386, simple_loss=0.2214, pruned_loss=0.02789, over 984720.08 frames.], batch size: 91, aishell_tot_loss[loss=0.141, simple_loss=0.228, pruned_loss=0.02702, over 979041.84 frames.], datatang_tot_loss[loss=0.1357, simple_loss=0.2137, pruned_loss=0.02882, over 977817.38 frames.], batch size: 91, lr: 2.73e-04
+2022-06-19 07:50:27,891 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 07:50:43,694 INFO [train.py:914] (2/4) Epoch 30, validation: loss=0.1643, simple_loss=0.2485, pruned_loss=0.04007, over 1622729.00 frames.
+2022-06-19 07:51:13,819 INFO [train.py:874] (2/4) Epoch 30, batch 2050, datatang_loss[loss=0.1598, simple_loss=0.2435, pruned_loss=0.03812, over 4960.00 frames.], tot_loss[loss=0.1393, simple_loss=0.2219, pruned_loss=0.0284, over 984977.75 frames.], batch size: 99, aishell_tot_loss[loss=0.1416, simple_loss=0.2284, pruned_loss=0.02742, over 979535.80 frames.], datatang_tot_loss[loss=0.1359, simple_loss=0.214, pruned_loss=0.02896, over 979087.18 frames.], batch size: 99, lr: 2.73e-04
+2022-06-19 07:51:43,740 INFO [train.py:874] (2/4) Epoch 30, batch 2100, datatang_loss[loss=0.1648, simple_loss=0.2483, pruned_loss=0.04069, over 4913.00 frames.], tot_loss[loss=0.1402, simple_loss=0.2228, pruned_loss=0.02884, over 985383.24 frames.], batch size: 98, aishell_tot_loss[loss=0.1421, simple_loss=0.2289, pruned_loss=0.02769, over 980408.72 frames.], datatang_tot_loss[loss=0.1366, simple_loss=0.2148, pruned_loss=0.02918, over 979970.08 frames.], batch size: 98, lr: 2.73e-04
+2022-06-19 07:52:12,615 INFO [train.py:874] (2/4) Epoch 30, batch 2150, datatang_loss[loss=0.1409, simple_loss=0.214, pruned_loss=0.03392, over 4907.00 frames.], tot_loss[loss=0.1401, simple_loss=0.2227, pruned_loss=0.02875, over 985365.37 frames.], batch size: 59, aishell_tot_loss[loss=0.1422, simple_loss=0.229, pruned_loss=0.02774, over 980883.18 frames.], datatang_tot_loss[loss=0.1365, simple_loss=0.2148, pruned_loss=0.02911, over 980709.07 frames.], batch size: 59, lr: 2.73e-04
+2022-06-19 07:52:42,925 INFO [train.py:874] (2/4) Epoch 30, batch 2200, aishell_loss[loss=0.146, simple_loss=0.2338, pruned_loss=0.02906, over 4943.00 frames.], tot_loss[loss=0.1392, simple_loss=0.222, pruned_loss=0.02822, over 985416.29 frames.], batch size: 54, aishell_tot_loss[loss=0.1422, simple_loss=0.229, pruned_loss=0.0277, over 981126.31 frames.], datatang_tot_loss[loss=0.1359, simple_loss=0.2144, pruned_loss=0.02866, over 981586.17 frames.], batch size: 54, lr: 2.73e-04
+2022-06-19 07:53:12,733 INFO [train.py:874] (2/4) Epoch 30, batch 2250, aishell_loss[loss=0.1566, simple_loss=0.2397, pruned_loss=0.03682, over 4922.00 frames.], tot_loss[loss=0.1393, simple_loss=0.2224, pruned_loss=0.02808, over 985185.86 frames.], batch size: 33, aishell_tot_loss[loss=0.1424, simple_loss=0.2294, pruned_loss=0.02769, over 981606.18 frames.], datatang_tot_loss[loss=0.1356, simple_loss=0.2142, pruned_loss=0.02855, over 981823.72 frames.], batch size: 33, lr: 2.73e-04
+2022-06-19 07:53:43,998 INFO [train.py:874] (2/4) Epoch 30, batch 2300, aishell_loss[loss=0.1565, simple_loss=0.2435, pruned_loss=0.03477, over 4956.00 frames.], tot_loss[loss=0.1388, simple_loss=0.222, pruned_loss=0.02777, over 985446.00 frames.], batch size: 32, aishell_tot_loss[loss=0.1421, simple_loss=0.2291, pruned_loss=0.02748, over 982164.06 frames.], datatang_tot_loss[loss=0.1356, simple_loss=0.2145, pruned_loss=0.02838, over 982349.41 frames.], batch size: 32, lr: 2.73e-04
+2022-06-19 07:54:14,770 INFO [train.py:874] (2/4) Epoch 30, batch 2350, aishell_loss[loss=0.1596, simple_loss=0.243, pruned_loss=0.03812, over 4967.00 frames.], tot_loss[loss=0.139, simple_loss=0.2219, pruned_loss=0.02806, over 985744.08 frames.], batch size: 61, aishell_tot_loss[loss=0.1418, simple_loss=0.2289, pruned_loss=0.02737, over 982573.63 frames.], datatang_tot_loss[loss=0.1362, simple_loss=0.215, pruned_loss=0.02872, over 982997.22 frames.], batch size: 61, lr: 2.73e-04
+2022-06-19 07:54:42,738 INFO [train.py:874] (2/4) Epoch 30, batch 2400, datatang_loss[loss=0.1266, simple_loss=0.2061, pruned_loss=0.02358, over 4968.00 frames.], tot_loss[loss=0.1389, simple_loss=0.2219, pruned_loss=0.02799, over 985753.83 frames.], batch size: 60, aishell_tot_loss[loss=0.1416, simple_loss=0.2288, pruned_loss=0.02722, over 982909.44 frames.], datatang_tot_loss[loss=0.1363, simple_loss=0.2151, pruned_loss=0.0288, over 983356.02 frames.], batch size: 60, lr: 2.73e-04
+2022-06-19 07:55:12,749 INFO [train.py:874] (2/4) Epoch 30, batch 2450, datatang_loss[loss=0.1423, simple_loss=0.2243, pruned_loss=0.03016, over 4941.00 frames.], tot_loss[loss=0.1382, simple_loss=0.2208, pruned_loss=0.02777, over 985747.72 frames.], batch size: 88, aishell_tot_loss[loss=0.1417, simple_loss=0.2287, pruned_loss=0.02736, over 983046.25 frames.], datatang_tot_loss[loss=0.1356, simple_loss=0.2144, pruned_loss=0.02838, over 983831.86 frames.], batch size: 88, lr: 2.73e-04
+2022-06-19 07:55:43,668 INFO [train.py:874] (2/4) Epoch 30, batch 2500, aishell_loss[loss=0.119, simple_loss=0.2091, pruned_loss=0.01449, over 4951.00 frames.], tot_loss[loss=0.1382, simple_loss=0.221, pruned_loss=0.02775, over 985600.58 frames.], batch size: 45, aishell_tot_loss[loss=0.1416, simple_loss=0.2285, pruned_loss=0.02729, over 983168.13 frames.], datatang_tot_loss[loss=0.1356, simple_loss=0.2144, pruned_loss=0.02841, over 984098.39 frames.], batch size: 45, lr: 2.73e-04
+2022-06-19 07:56:12,506 INFO [train.py:874] (2/4) Epoch 30, batch 2550, datatang_loss[loss=0.1247, simple_loss=0.2016, pruned_loss=0.02389, over 4917.00 frames.], tot_loss[loss=0.1387, simple_loss=0.2216, pruned_loss=0.02787, over 985615.33 frames.], batch size: 75, aishell_tot_loss[loss=0.1423, simple_loss=0.2293, pruned_loss=0.0276, over 983492.02 frames.], datatang_tot_loss[loss=0.1352, simple_loss=0.214, pruned_loss=0.02822, over 984262.87 frames.], batch size: 75, lr: 2.73e-04
+2022-06-19 07:56:42,336 INFO [train.py:874] (2/4) Epoch 30, batch 2600, aishell_loss[loss=0.1692, simple_loss=0.261, pruned_loss=0.03874, over 4971.00 frames.], tot_loss[loss=0.1395, simple_loss=0.2219, pruned_loss=0.02849, over 985414.64 frames.], batch size: 51, aishell_tot_loss[loss=0.1425, simple_loss=0.2293, pruned_loss=0.02788, over 983292.87 frames.], datatang_tot_loss[loss=0.1359, simple_loss=0.2146, pruned_loss=0.02858, over 984653.43 frames.], batch size: 51, lr: 2.73e-04
+2022-06-19 07:57:13,793 INFO [train.py:874] (2/4) Epoch 30, batch 2650, aishell_loss[loss=0.1418, simple_loss=0.2275, pruned_loss=0.02807, over 4872.00 frames.], tot_loss[loss=0.1392, simple_loss=0.2215, pruned_loss=0.02844, over 985671.31 frames.], batch size: 35, aishell_tot_loss[loss=0.1428, simple_loss=0.2295, pruned_loss=0.02802, over 983495.61 frames.], datatang_tot_loss[loss=0.1355, simple_loss=0.2141, pruned_loss=0.02846, over 985038.79 frames.], batch size: 35, lr: 2.73e-04
+2022-06-19 07:57:41,433 INFO [train.py:874] (2/4) Epoch 30, batch 2700, datatang_loss[loss=0.1664, simple_loss=0.2477, pruned_loss=0.04254, over 4943.00 frames.], tot_loss[loss=0.1401, simple_loss=0.2223, pruned_loss=0.02892, over 985951.44 frames.], batch size: 109, aishell_tot_loss[loss=0.1427, simple_loss=0.2292, pruned_loss=0.02806, over 983912.26 frames.], datatang_tot_loss[loss=0.1365, simple_loss=0.215, pruned_loss=0.029, over 985262.36 frames.], batch size: 109, lr: 2.72e-04
+2022-06-19 07:58:11,544 INFO [train.py:874] (2/4) Epoch 30, batch 2750, aishell_loss[loss=0.1669, simple_loss=0.2583, pruned_loss=0.0377, over 4938.00 frames.], tot_loss[loss=0.1401, simple_loss=0.2225, pruned_loss=0.02889, over 985511.05 frames.], batch size: 58, aishell_tot_loss[loss=0.1425, simple_loss=0.2291, pruned_loss=0.02798, over 983963.99 frames.], datatang_tot_loss[loss=0.1367, simple_loss=0.215, pruned_loss=0.02916, over 985085.13 frames.], batch size: 58, lr: 2.72e-04
+2022-06-19 07:58:40,960 INFO [train.py:874] (2/4) Epoch 30, batch 2800, aishell_loss[loss=0.1542, simple_loss=0.2381, pruned_loss=0.03514, over 4987.00 frames.], tot_loss[loss=0.1398, simple_loss=0.2222, pruned_loss=0.02873, over 985770.19 frames.], batch size: 38, aishell_tot_loss[loss=0.1423, simple_loss=0.2289, pruned_loss=0.02789, over 984604.59 frames.], datatang_tot_loss[loss=0.1366, simple_loss=0.215, pruned_loss=0.02914, over 984958.92 frames.], batch size: 38, lr: 2.72e-04
+2022-06-19 07:59:11,148 INFO [train.py:874] (2/4) Epoch 30, batch 2850, datatang_loss[loss=0.1079, simple_loss=0.1827, pruned_loss=0.01655, over 4869.00 frames.], tot_loss[loss=0.1394, simple_loss=0.2219, pruned_loss=0.02849, over 985611.92 frames.], batch size: 39, aishell_tot_loss[loss=0.1423, simple_loss=0.2291, pruned_loss=0.02779, over 984801.32 frames.], datatang_tot_loss[loss=0.1364, simple_loss=0.2148, pruned_loss=0.02901, over 984818.52 frames.], batch size: 39, lr: 2.72e-04
+2022-06-19 07:59:41,808 INFO [train.py:874] (2/4) Epoch 30, batch 2900, aishell_loss[loss=0.1194, simple_loss=0.1802, pruned_loss=0.02931, over 4877.00 frames.], tot_loss[loss=0.139, simple_loss=0.2216, pruned_loss=0.02821, over 985483.43 frames.], batch size: 20, aishell_tot_loss[loss=0.1424, simple_loss=0.2291, pruned_loss=0.02784, over 984802.64 frames.], datatang_tot_loss[loss=0.1359, simple_loss=0.2144, pruned_loss=0.02869, over 984876.90 frames.], batch size: 20, lr: 2.72e-04
+2022-06-19 08:00:11,897 INFO [train.py:874] (2/4) Epoch 30, batch 2950, datatang_loss[loss=0.1129, simple_loss=0.1921, pruned_loss=0.01685, over 4924.00 frames.], tot_loss[loss=0.139, simple_loss=0.2216, pruned_loss=0.02822, over 985407.95 frames.], batch size: 77, aishell_tot_loss[loss=0.1426, simple_loss=0.2295, pruned_loss=0.02781, over 984744.79 frames.], datatang_tot_loss[loss=0.1358, simple_loss=0.2142, pruned_loss=0.02869, over 985001.19 frames.], batch size: 77, lr: 2.72e-04
+2022-06-19 08:00:40,793 INFO [train.py:874] (2/4) Epoch 30, batch 3000, datatang_loss[loss=0.1175, simple_loss=0.1968, pruned_loss=0.01913, over 4939.00 frames.], tot_loss[loss=0.1386, simple_loss=0.2214, pruned_loss=0.0279, over 985405.37 frames.], batch size: 69, aishell_tot_loss[loss=0.1423, simple_loss=0.2293, pruned_loss=0.02764, over 984811.67 frames.], datatang_tot_loss[loss=0.1354, simple_loss=0.2138, pruned_loss=0.02854, over 985081.49 frames.], batch size: 69, lr: 2.72e-04
+2022-06-19 08:00:40,794 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 08:00:57,061 INFO [train.py:914] (2/4) Epoch 30, validation: loss=0.165, simple_loss=0.2488, pruned_loss=0.04058, over 1622729.00 frames.
+2022-06-19 08:01:25,990 INFO [train.py:874] (2/4) Epoch 30, batch 3050, datatang_loss[loss=0.1315, simple_loss=0.2106, pruned_loss=0.02622, over 4903.00 frames.], tot_loss[loss=0.1389, simple_loss=0.2218, pruned_loss=0.02803, over 985539.23 frames.], batch size: 64, aishell_tot_loss[loss=0.1427, simple_loss=0.2297, pruned_loss=0.02786, over 985176.54 frames.], datatang_tot_loss[loss=0.1352, simple_loss=0.2135, pruned_loss=0.02843, over 984969.08 frames.], batch size: 64, lr: 2.72e-04
+2022-06-19 08:01:55,867 INFO [train.py:874] (2/4) Epoch 30, batch 3100, datatang_loss[loss=0.1445, simple_loss=0.2285, pruned_loss=0.03029, over 4969.00 frames.], tot_loss[loss=0.1382, simple_loss=0.221, pruned_loss=0.0277, over 985450.76 frames.], batch size: 40, aishell_tot_loss[loss=0.1421, simple_loss=0.2292, pruned_loss=0.02752, over 985204.55 frames.], datatang_tot_loss[loss=0.1352, simple_loss=0.2137, pruned_loss=0.02838, over 984964.79 frames.], batch size: 40, lr: 2.72e-04
+2022-06-19 08:02:25,658 INFO [train.py:874] (2/4) Epoch 30, batch 3150, datatang_loss[loss=0.1367, simple_loss=0.2093, pruned_loss=0.03201, over 4873.00 frames.], tot_loss[loss=0.1386, simple_loss=0.2216, pruned_loss=0.0278, over 985522.79 frames.], batch size: 39, aishell_tot_loss[loss=0.1431, simple_loss=0.2304, pruned_loss=0.02789, over 985374.61 frames.], datatang_tot_loss[loss=0.1348, simple_loss=0.2134, pruned_loss=0.02807, over 984965.99 frames.], batch size: 39, lr: 2.72e-04
+2022-06-19 08:02:54,298 INFO [train.py:874] (2/4) Epoch 30, batch 3200, datatang_loss[loss=0.124, simple_loss=0.2043, pruned_loss=0.02181, over 4924.00 frames.], tot_loss[loss=0.1391, simple_loss=0.2222, pruned_loss=0.02803, over 985817.60 frames.], batch size: 73, aishell_tot_loss[loss=0.1432, simple_loss=0.2305, pruned_loss=0.02795, over 985695.08 frames.], datatang_tot_loss[loss=0.135, simple_loss=0.2136, pruned_loss=0.02823, over 985035.56 frames.], batch size: 73, lr: 2.72e-04
+2022-06-19 08:03:24,480 INFO [train.py:874] (2/4) Epoch 30, batch 3250, datatang_loss[loss=0.1382, simple_loss=0.2217, pruned_loss=0.02738, over 4933.00 frames.], tot_loss[loss=0.1397, simple_loss=0.2226, pruned_loss=0.0284, over 985728.96 frames.], batch size: 94, aishell_tot_loss[loss=0.1433, simple_loss=0.2304, pruned_loss=0.02812, over 985558.62 frames.], datatang_tot_loss[loss=0.1354, simple_loss=0.214, pruned_loss=0.02843, over 985174.66 frames.], batch size: 94, lr: 2.72e-04
+2022-06-19 08:03:54,501 INFO [train.py:874] (2/4) Epoch 30, batch 3300, datatang_loss[loss=0.1166, simple_loss=0.203, pruned_loss=0.01514, over 4939.00 frames.], tot_loss[loss=0.1396, simple_loss=0.2222, pruned_loss=0.02848, over 985415.52 frames.], batch size: 88, aishell_tot_loss[loss=0.1428, simple_loss=0.2298, pruned_loss=0.02787, over 985276.93 frames.], datatang_tot_loss[loss=0.1361, simple_loss=0.2146, pruned_loss=0.02879, over 985221.57 frames.], batch size: 88, lr: 2.72e-04
+2022-06-19 08:04:23,847 INFO [train.py:874] (2/4) Epoch 30, batch 3350, aishell_loss[loss=0.1279, simple_loss=0.2207, pruned_loss=0.01758, over 4906.00 frames.], tot_loss[loss=0.1386, simple_loss=0.2216, pruned_loss=0.0278, over 985635.93 frames.], batch size: 46, aishell_tot_loss[loss=0.1421, simple_loss=0.2291, pruned_loss=0.02751, over 985359.22 frames.], datatang_tot_loss[loss=0.1357, simple_loss=0.2143, pruned_loss=0.0285, over 985413.28 frames.], batch size: 46, lr: 2.72e-04
+2022-06-19 08:04:55,169 INFO [train.py:874] (2/4) Epoch 30, batch 3400, aishell_loss[loss=0.1483, simple_loss=0.2232, pruned_loss=0.03667, over 4948.00 frames.], tot_loss[loss=0.1391, simple_loss=0.2221, pruned_loss=0.02803, over 985498.62 frames.], batch size: 31, aishell_tot_loss[loss=0.1422, simple_loss=0.2293, pruned_loss=0.02759, over 985109.27 frames.], datatang_tot_loss[loss=0.1359, simple_loss=0.2146, pruned_loss=0.02861, over 985587.56 frames.], batch size: 31, lr: 2.72e-04
+2022-06-19 08:05:24,109 INFO [train.py:874] (2/4) Epoch 30, batch 3450, aishell_loss[loss=0.1312, simple_loss=0.2248, pruned_loss=0.01881, over 4896.00 frames.], tot_loss[loss=0.1394, simple_loss=0.2223, pruned_loss=0.02823, over 985138.06 frames.], batch size: 34, aishell_tot_loss[loss=0.1422, simple_loss=0.2293, pruned_loss=0.02753, over 984760.61 frames.], datatang_tot_loss[loss=0.1363, simple_loss=0.2148, pruned_loss=0.02888, over 985603.77 frames.], batch size: 34, lr: 2.72e-04
+2022-06-19 08:05:54,699 INFO [train.py:874] (2/4) Epoch 30, batch 3500, datatang_loss[loss=0.1509, simple_loss=0.2321, pruned_loss=0.03481, over 4944.00 frames.], tot_loss[loss=0.1383, simple_loss=0.221, pruned_loss=0.02778, over 985371.75 frames.], batch size: 88, aishell_tot_loss[loss=0.1416, simple_loss=0.2286, pruned_loss=0.02727, over 984964.23 frames.], datatang_tot_loss[loss=0.1358, simple_loss=0.2143, pruned_loss=0.02864, over 985622.66 frames.], batch size: 88, lr: 2.72e-04
+2022-06-19 08:06:24,510 INFO [train.py:874] (2/4) Epoch 30, batch 3550, datatang_loss[loss=0.09695, simple_loss=0.1677, pruned_loss=0.01309, over 4842.00 frames.], tot_loss[loss=0.1382, simple_loss=0.2203, pruned_loss=0.02807, over 985183.63 frames.], batch size: 30, aishell_tot_loss[loss=0.1416, simple_loss=0.2284, pruned_loss=0.02737, over 984675.29 frames.], datatang_tot_loss[loss=0.1357, simple_loss=0.2139, pruned_loss=0.02876, over 985716.42 frames.], batch size: 30, lr: 2.72e-04
+2022-06-19 08:06:53,795 INFO [train.py:874] (2/4) Epoch 30, batch 3600, datatang_loss[loss=0.1264, simple_loss=0.2087, pruned_loss=0.02205, over 4953.00 frames.], tot_loss[loss=0.1383, simple_loss=0.2205, pruned_loss=0.02805, over 985182.27 frames.], batch size: 55, aishell_tot_loss[loss=0.1415, simple_loss=0.2282, pruned_loss=0.02744, over 984506.83 frames.], datatang_tot_loss[loss=0.1357, simple_loss=0.214, pruned_loss=0.02868, over 985883.33 frames.], batch size: 55, lr: 2.71e-04
+2022-06-19 08:07:23,925 INFO [train.py:874] (2/4) Epoch 30, batch 3650, datatang_loss[loss=0.1438, simple_loss=0.2215, pruned_loss=0.03304, over 4943.00 frames.], tot_loss[loss=0.1386, simple_loss=0.2207, pruned_loss=0.02826, over 985549.50 frames.], batch size: 88, aishell_tot_loss[loss=0.1421, simple_loss=0.2288, pruned_loss=0.02771, over 984795.34 frames.], datatang_tot_loss[loss=0.1355, simple_loss=0.2139, pruned_loss=0.02858, over 985946.24 frames.], batch size: 88, lr: 2.71e-04
+2022-06-19 08:07:54,535 INFO [train.py:874] (2/4) Epoch 30, batch 3700, aishell_loss[loss=0.1478, simple_loss=0.2511, pruned_loss=0.02221, over 4866.00 frames.], tot_loss[loss=0.1372, simple_loss=0.2193, pruned_loss=0.02756, over 985472.11 frames.], batch size: 37, aishell_tot_loss[loss=0.142, simple_loss=0.229, pruned_loss=0.02751, over 984755.43 frames.], datatang_tot_loss[loss=0.1343, simple_loss=0.2126, pruned_loss=0.02802, over 985897.93 frames.], batch size: 37, lr: 2.71e-04
+2022-06-19 08:08:23,938 INFO [train.py:874] (2/4) Epoch 30, batch 3750, aishell_loss[loss=0.1334, simple_loss=0.2292, pruned_loss=0.0188, over 4886.00 frames.], tot_loss[loss=0.137, simple_loss=0.2194, pruned_loss=0.0273, over 985220.96 frames.], batch size: 42, aishell_tot_loss[loss=0.1418, simple_loss=0.229, pruned_loss=0.0273, over 984645.71 frames.], datatang_tot_loss[loss=0.1341, simple_loss=0.2124, pruned_loss=0.02791, over 985768.34 frames.], batch size: 42, lr: 2.71e-04
+2022-06-19 08:08:52,768 INFO [train.py:874] (2/4) Epoch 30, batch 3800, datatang_loss[loss=0.1194, simple_loss=0.1904, pruned_loss=0.02414, over 4904.00 frames.], tot_loss[loss=0.1377, simple_loss=0.2201, pruned_loss=0.02767, over 985342.72 frames.], batch size: 42, aishell_tot_loss[loss=0.142, simple_loss=0.2291, pruned_loss=0.02743, over 984824.17 frames.], datatang_tot_loss[loss=0.1344, simple_loss=0.2126, pruned_loss=0.02811, over 985744.60 frames.], batch size: 42, lr: 2.71e-04
+2022-06-19 08:09:20,699 INFO [train.py:874] (2/4) Epoch 30, batch 3850, datatang_loss[loss=0.13, simple_loss=0.2202, pruned_loss=0.01987, over 4928.00 frames.], tot_loss[loss=0.1383, simple_loss=0.2213, pruned_loss=0.02769, over 985179.89 frames.], batch size: 94, aishell_tot_loss[loss=0.1422, simple_loss=0.2295, pruned_loss=0.02745, over 984720.27 frames.], datatang_tot_loss[loss=0.1346, simple_loss=0.2131, pruned_loss=0.02809, over 985706.70 frames.], batch size: 94, lr: 2.71e-04
+2022-06-19 08:09:50,490 INFO [train.py:874] (2/4) Epoch 30, batch 3900, aishell_loss[loss=0.1715, simple_loss=0.26, pruned_loss=0.04152, over 4931.00 frames.], tot_loss[loss=0.1386, simple_loss=0.222, pruned_loss=0.02761, over 984742.02 frames.], batch size: 58, aishell_tot_loss[loss=0.1423, simple_loss=0.2296, pruned_loss=0.02747, over 984349.40 frames.], datatang_tot_loss[loss=0.1345, simple_loss=0.213, pruned_loss=0.02796, over 985657.57 frames.], batch size: 58, lr: 2.71e-04
+2022-06-19 08:10:17,983 INFO [train.py:874] (2/4) Epoch 30, batch 3950, datatang_loss[loss=0.1081, simple_loss=0.1787, pruned_loss=0.01868, over 4887.00 frames.], tot_loss[loss=0.1382, simple_loss=0.2212, pruned_loss=0.02762, over 985135.66 frames.], batch size: 52, aishell_tot_loss[loss=0.1417, simple_loss=0.2286, pruned_loss=0.02734, over 984566.94 frames.], datatang_tot_loss[loss=0.1347, simple_loss=0.2132, pruned_loss=0.02808, over 985817.41 frames.], batch size: 52, lr: 2.71e-04
+2022-06-19 08:10:51,420 INFO [train.py:874] (2/4) Epoch 30, batch 4000, aishell_loss[loss=0.1469, simple_loss=0.2318, pruned_loss=0.03103, over 4862.00 frames.], tot_loss[loss=0.1383, simple_loss=0.2212, pruned_loss=0.02771, over 985464.57 frames.], batch size: 36, aishell_tot_loss[loss=0.1416, simple_loss=0.2286, pruned_loss=0.02724, over 984747.54 frames.], datatang_tot_loss[loss=0.1348, simple_loss=0.2132, pruned_loss=0.02826, over 985983.74 frames.], batch size: 36, lr: 2.71e-04
+2022-06-19 08:10:51,421 INFO [train.py:905] (2/4) Computing validation loss
+2022-06-19 08:11:08,527 INFO [train.py:914] (2/4) Epoch 30, validation: loss=0.1651, simple_loss=0.249, pruned_loss=0.04054, over 1622729.00 frames.
+2022-06-19 08:11:36,709 INFO [train.py:874] (2/4) Epoch 30, batch 4050, aishell_loss[loss=0.12, simple_loss=0.2052, pruned_loss=0.01741, over 4859.00 frames.], tot_loss[loss=0.1384, simple_loss=0.2214, pruned_loss=0.02772, over 985020.13 frames.], batch size: 28, aishell_tot_loss[loss=0.1416, simple_loss=0.2287, pruned_loss=0.02724, over 984439.31 frames.], datatang_tot_loss[loss=0.135, simple_loss=0.2135, pruned_loss=0.02825, over 985846.14 frames.], batch size: 28, lr: 2.71e-04
+2022-06-19 08:11:58,112 INFO [train.py:1125] (2/4) Done!