diff --git "a/log/log-train-2023-02-08-23-42-53-3" "b/log/log-train-2023-02-08-23-42-53-3" new file mode 100644--- /dev/null +++ "b/log/log-train-2023-02-08-23-42-53-3" @@ -0,0 +1,2623 @@ +2023-02-08 23:42:53,786 INFO [train.py:973] (3/4) Training started +2023-02-08 23:42:53,786 INFO [train.py:983] (3/4) Device: cuda:3 +2023-02-08 23:42:53,850 INFO [train.py:992] (3/4) {'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': 3000, 'feature_dim': 80, 'subsampling_factor': 4, 'warm_step': 2000, 'env_info': {'k2-version': '1.23.3', 'k2-build-type': 'Debug', 'k2-with-cuda': True, 'k2-git-sha1': '3b81ac9686aee539d447bb2085b2cdfc131c7c91', 'k2-git-date': 'Thu Jan 26 20:40:25 2023', 'lhotse-version': '1.9.0.dev+git.97bf4b0.dirty', 'torch-version': '1.10.0+cu102', 'torch-cuda-available': True, 'torch-cuda-version': '10.2', 'python-version': '3.8', 'icefall-git-branch': 'surt', 'icefall-git-sha1': 'b3d0d34-dirty', 'icefall-git-date': 'Sat Feb 4 14:53:48 2023', 'icefall-path': '/exp/draj/mini_scale_2022/icefall', 'k2-path': '/exp/draj/mini_scale_2022/k2/k2/python/k2/__init__.py', 'lhotse-path': '/exp/draj/mini_scale_2022/lhotse/lhotse/__init__.py', 'hostname': 'r8n07', 'IP address': '10.1.8.7'}, 'world_size': 4, 'master_port': 12354, 'tensorboard': True, 'num_epochs': 30, 'start_epoch': 28, 'start_batch': 0, 'exp_dir': PosixPath('pruned_transducer_stateless7_streaming/exp/v1'), 'bpe_model': 'data/lang_bpe_500/bpe.model', 'base_lr': 0.05, 'lr_batches': 5000, 'lr_epochs': 3.5, 'context_size': 2, 'prune_range': 5, 'lm_scale': 0.25, 'am_scale': 0.0, 'simple_loss_scale': 0.5, 'seed': 42, 'print_diagnostics': False, 'inf_check': False, 'save_every_n': 2000, 'keep_last_k': 10, 'average_period': 200, 'use_fp16': True, 'num_encoder_layers': '2,2,2,2,2', 'feedforward_dims': '768,768,768,768,768', 'nhead': '8,8,8,8,8', 'encoder_dims': '256,256,256,256,256', 'attention_dims': '192,192,192,192,192', 'encoder_unmasked_dims': '192,192,192,192,192', 'zipformer_downsampling_factors': '1,2,4,8,2', 'cnn_module_kernels': '31,31,31,31,31', 'decoder_dim': 512, 'joiner_dim': 512, 'short_chunk_size': 50, 'num_left_chunks': 4, 'decode_chunk_len': 32, 'full_libri': True, 'manifest_dir': PosixPath('data/manifests'), 'max_duration': 500, 'bucketing_sampler': True, 'num_buckets': 30, 'concatenate_cuts': False, 'duration_factor': 1.0, 'gap': 1.0, 'on_the_fly_feats': False, 'shuffle': True, 'drop_last': True, 'return_cuts': True, 'num_workers': 2, 'enable_spec_aug': True, 'spec_aug_time_warp_factor': 80, 'enable_musan': True, 'input_strategy': 'PrecomputedFeatures', 'blank_id': 0, 'vocab_size': 500} +2023-02-08 23:42:53,850 INFO [train.py:994] (3/4) About to create model +2023-02-08 23:42:54,169 INFO [zipformer.py:402] (3/4) At encoder stack 4, which has downsampling_factor=2, we will combine the outputs of layers 1 and 3, with downsampling_factors=2 and 8. +2023-02-08 23:42:54,181 INFO [train.py:998] (3/4) Number of model parameters: 20697573 +2023-02-08 23:42:54,181 INFO [checkpoint.py:112] (3/4) Loading checkpoint from pruned_transducer_stateless7_streaming/exp/v1/epoch-27.pt +2023-02-08 23:43:03,645 INFO [train.py:1013] (3/4) Using DDP +2023-02-08 23:43:03,870 INFO [train.py:1030] (3/4) Loading optimizer state dict +2023-02-08 23:43:04,078 INFO [train.py:1038] (3/4) Loading scheduler state dict +2023-02-08 23:43:04,078 INFO [asr_datamodule.py:420] (3/4) About to get the shuffled train-clean-100, train-clean-360 and train-other-500 cuts +2023-02-08 23:43:04,265 INFO [asr_datamodule.py:224] (3/4) Enable MUSAN +2023-02-08 23:43:04,265 INFO [asr_datamodule.py:225] (3/4) About to get Musan cuts +2023-02-08 23:43:05,836 INFO [asr_datamodule.py:249] (3/4) Enable SpecAugment +2023-02-08 23:43:05,836 INFO [asr_datamodule.py:250] (3/4) Time warp factor: 80 +2023-02-08 23:43:05,836 INFO [asr_datamodule.py:260] (3/4) Num frame mask: 10 +2023-02-08 23:43:05,836 INFO [asr_datamodule.py:273] (3/4) About to create train dataset +2023-02-08 23:43:05,837 INFO [asr_datamodule.py:300] (3/4) Using DynamicBucketingSampler. +2023-02-08 23:43:05,857 WARNING [train.py:1067] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-02-08 23:43:08,000 INFO [asr_datamodule.py:316] (3/4) About to create train dataloader +2023-02-08 23:43:08,001 INFO [asr_datamodule.py:430] (3/4) About to get dev-clean cuts +2023-02-08 23:43:08,002 INFO [asr_datamodule.py:437] (3/4) About to get dev-other cuts +2023-02-08 23:43:08,003 INFO [asr_datamodule.py:347] (3/4) About to create dev dataset +2023-02-08 23:43:08,360 INFO [asr_datamodule.py:364] (3/4) About to create dev dataloader +2023-02-08 23:43:08,360 INFO [train.py:1122] (3/4) Loading grad scaler state dict +2023-02-08 23:43:20,252 WARNING [train.py:1067] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-02-08 23:43:25,773 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.82 vs. limit=2.0 +2023-02-08 23:43:26,062 INFO [train.py:901] (3/4) Epoch 28, batch 0, loss[loss=0.2773, simple_loss=0.3445, pruned_loss=0.105, over 8338.00 frames. ], tot_loss[loss=0.2773, simple_loss=0.3445, pruned_loss=0.105, over 8338.00 frames. ], batch size: 26, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:43:26,063 INFO [train.py:926] (3/4) Computing validation loss +2023-02-08 23:43:38,187 INFO [train.py:935] (3/4) Epoch 28, validation: loss=0.1714, simple_loss=0.2712, pruned_loss=0.03579, over 944034.00 frames. +2023-02-08 23:43:38,189 INFO [train.py:936] (3/4) Maximum memory allocated so far is 6172MB +2023-02-08 23:43:48,606 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=218250.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:43:59,124 WARNING [train.py:1067] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-02-08 23:43:59,745 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=218260.0, num_to_drop=1, layers_to_drop={0} +2023-02-08 23:44:26,827 INFO [train.py:901] (3/4) Epoch 28, batch 50, loss[loss=0.1722, simple_loss=0.2535, pruned_loss=0.04542, over 7681.00 frames. ], tot_loss[loss=0.2039, simple_loss=0.2884, pruned_loss=0.05966, over 368830.33 frames. ], batch size: 18, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:44:44,702 WARNING [train.py:1067] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-02-08 23:44:48,220 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.737e+02 2.503e+02 3.099e+02 3.838e+02 3.677e+03, threshold=6.198e+02, percent-clipped=7.0 +2023-02-08 23:45:09,768 INFO [train.py:901] (3/4) Epoch 28, batch 100, loss[loss=0.1583, simple_loss=0.2476, pruned_loss=0.03449, over 7250.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.2869, pruned_loss=0.05862, over 646435.40 frames. ], batch size: 16, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:45:12,261 WARNING [train.py:1067] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-02-08 23:45:21,937 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.67 vs. limit=5.0 +2023-02-08 23:45:42,217 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=218375.0, num_to_drop=1, layers_to_drop={0} +2023-02-08 23:45:52,964 INFO [train.py:901] (3/4) Epoch 28, batch 150, loss[loss=0.2217, simple_loss=0.3054, pruned_loss=0.06899, over 8029.00 frames. ], tot_loss[loss=0.203, simple_loss=0.2872, pruned_loss=0.05942, over 861149.95 frames. ], batch size: 22, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:46:01,150 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=218397.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:46:12,823 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.615e+02 2.274e+02 2.796e+02 3.416e+02 5.816e+02, threshold=5.591e+02, percent-clipped=0.0 +2023-02-08 23:46:19,722 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=218422.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:46:32,312 INFO [train.py:901] (3/4) Epoch 28, batch 200, loss[loss=0.2025, simple_loss=0.2894, pruned_loss=0.05781, over 8294.00 frames. ], tot_loss[loss=0.1997, simple_loss=0.2842, pruned_loss=0.05762, over 1028911.67 frames. ], batch size: 23, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:46:50,636 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=218462.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:47:10,726 INFO [train.py:901] (3/4) Epoch 28, batch 250, loss[loss=0.1621, simple_loss=0.2421, pruned_loss=0.04104, over 7655.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2819, pruned_loss=0.05681, over 1159440.38 frames. ], batch size: 19, lr: 2.71e-03, grad_scale: 16.0 +2023-02-08 23:47:23,075 WARNING [train.py:1067] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-02-08 23:47:31,295 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.545e+02 2.405e+02 2.917e+02 3.543e+02 7.929e+02, threshold=5.833e+02, percent-clipped=6.0 +2023-02-08 23:47:33,429 WARNING [train.py:1067] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-02-08 23:47:41,366 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=218527.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:47:48,884 INFO [train.py:901] (3/4) Epoch 28, batch 300, loss[loss=0.2228, simple_loss=0.3082, pruned_loss=0.06874, over 8480.00 frames. ], tot_loss[loss=0.1992, simple_loss=0.2829, pruned_loss=0.05782, over 1261879.64 frames. ], batch size: 29, lr: 2.71e-03, grad_scale: 16.0 +2023-02-08 23:47:53,404 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=218544.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:48:14,322 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=218572.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:48:18,159 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=218577.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:48:25,696 INFO [train.py:901] (3/4) Epoch 28, batch 350, loss[loss=0.1862, simple_loss=0.2743, pruned_loss=0.04904, over 8590.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.2838, pruned_loss=0.05819, over 1342752.24 frames. ], batch size: 49, lr: 2.71e-03, grad_scale: 16.0 +2023-02-08 23:48:28,674 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=218592.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:48:29,483 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.0123, 2.0399, 3.2701, 2.5324, 2.8620, 2.1814, 1.8781, 1.7359], + device='cuda:3'), covar=tensor([0.7750, 0.6949, 0.2414, 0.4547, 0.3588, 0.4418, 0.2957, 0.6298], + device='cuda:3'), in_proj_covar=tensor([0.0959, 0.1017, 0.0823, 0.0986, 0.1018, 0.0922, 0.0763, 0.0846], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-08 23:48:43,888 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.664e+02 2.330e+02 2.853e+02 3.797e+02 9.826e+02, threshold=5.707e+02, percent-clipped=4.0 +2023-02-08 23:49:00,083 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=218631.0, num_to_drop=1, layers_to_drop={1} +2023-02-08 23:49:04,816 INFO [train.py:901] (3/4) Epoch 28, batch 400, loss[loss=0.1988, simple_loss=0.2864, pruned_loss=0.05563, over 8514.00 frames. ], tot_loss[loss=0.1997, simple_loss=0.2837, pruned_loss=0.05786, over 1409164.74 frames. ], batch size: 28, lr: 2.71e-03, grad_scale: 16.0 +2023-02-08 23:49:16,414 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.5512, 1.8825, 2.8553, 1.4712, 2.0254, 1.9280, 1.7093, 2.2373], + device='cuda:3'), covar=tensor([0.2033, 0.2853, 0.0921, 0.4836, 0.2102, 0.3463, 0.2543, 0.2301], + device='cuda:3'), in_proj_covar=tensor([0.0542, 0.0637, 0.0566, 0.0672, 0.0660, 0.0615, 0.0568, 0.0648], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-08 23:49:17,853 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=218656.0, num_to_drop=1, layers_to_drop={0} +2023-02-08 23:49:19,982 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=218659.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:49:40,567 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=218687.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:49:41,128 INFO [train.py:901] (3/4) Epoch 28, batch 450, loss[loss=0.198, simple_loss=0.2753, pruned_loss=0.06029, over 8295.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2828, pruned_loss=0.05749, over 1453736.92 frames. ], batch size: 23, lr: 2.71e-03, grad_scale: 16.0 +2023-02-08 23:49:59,789 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.572e+02 2.362e+02 2.836e+02 3.643e+02 9.062e+02, threshold=5.672e+02, percent-clipped=2.0 +2023-02-08 23:50:18,546 INFO [train.py:901] (3/4) Epoch 28, batch 500, loss[loss=0.2004, simple_loss=0.2856, pruned_loss=0.05758, over 8547.00 frames. ], tot_loss[loss=0.1999, simple_loss=0.2838, pruned_loss=0.05804, over 1491748.28 frames. ], batch size: 39, lr: 2.71e-03, grad_scale: 16.0 +2023-02-08 23:50:57,133 INFO [train.py:901] (3/4) Epoch 28, batch 550, loss[loss=0.1777, simple_loss=0.2683, pruned_loss=0.04354, over 8325.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.2823, pruned_loss=0.05738, over 1521291.96 frames. ], batch size: 25, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:51:05,273 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.5919, 2.4267, 3.1531, 2.5908, 3.1196, 2.6485, 2.5440, 1.8463], + device='cuda:3'), covar=tensor([0.5409, 0.4994, 0.2192, 0.4103, 0.2785, 0.3351, 0.1891, 0.5938], + device='cuda:3'), in_proj_covar=tensor([0.0963, 0.1021, 0.0829, 0.0991, 0.1024, 0.0928, 0.0768, 0.0849], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-08 23:51:16,041 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.820e+02 2.392e+02 2.925e+02 3.560e+02 1.211e+03, threshold=5.850e+02, percent-clipped=4.0 +2023-02-08 23:51:29,401 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.7925, 2.6443, 1.9672, 2.4746, 2.3655, 1.7605, 2.2783, 2.3314], + device='cuda:3'), covar=tensor([0.1487, 0.0434, 0.1228, 0.0648, 0.0771, 0.1549, 0.0982, 0.0974], + device='cuda:3'), in_proj_covar=tensor([0.0359, 0.0244, 0.0342, 0.0315, 0.0305, 0.0349, 0.0352, 0.0325], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-02-08 23:51:30,193 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=218833.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:51:33,473 INFO [train.py:901] (3/4) Epoch 28, batch 600, loss[loss=0.2595, simple_loss=0.3358, pruned_loss=0.09164, over 8669.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.2838, pruned_loss=0.05819, over 1544857.98 frames. ], batch size: 34, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:51:33,881 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.79 vs. limit=5.0 +2023-02-08 23:51:53,194 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=218858.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:51:56,611 WARNING [train.py:1067] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-02-08 23:52:04,161 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=218871.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:52:06,123 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.77 vs. limit=5.0 +2023-02-08 23:52:10,233 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6356, 1.9088, 1.9845, 1.4125, 2.1495, 1.4600, 0.6641, 1.9078], + device='cuda:3'), covar=tensor([0.0737, 0.0416, 0.0328, 0.0667, 0.0425, 0.0956, 0.0990, 0.0360], + device='cuda:3'), in_proj_covar=tensor([0.0474, 0.0412, 0.0366, 0.0458, 0.0395, 0.0553, 0.0402, 0.0442], + device='cuda:3'), out_proj_covar=tensor([1.2539e-04, 1.0696e-04, 9.5266e-05, 1.1974e-04, 1.0344e-04, 1.5414e-04, + 1.0732e-04, 1.1585e-04], device='cuda:3') +2023-02-08 23:52:18,546 INFO [train.py:901] (3/4) Epoch 28, batch 650, loss[loss=0.2359, simple_loss=0.3177, pruned_loss=0.07701, over 8533.00 frames. ], tot_loss[loss=0.1999, simple_loss=0.2835, pruned_loss=0.05814, over 1558626.51 frames. ], batch size: 31, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:52:40,038 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.733e+02 2.221e+02 2.637e+02 3.403e+02 7.509e+02, threshold=5.274e+02, percent-clipped=1.0 +2023-02-08 23:52:41,079 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=218915.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:52:54,643 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.8751, 3.8076, 3.5321, 1.9273, 3.3667, 3.4842, 3.4282, 3.3284], + device='cuda:3'), covar=tensor([0.0832, 0.0595, 0.0951, 0.4371, 0.0978, 0.1191, 0.1368, 0.0952], + device='cuda:3'), in_proj_covar=tensor([0.0540, 0.0461, 0.0445, 0.0555, 0.0444, 0.0463, 0.0439, 0.0406], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-08 23:52:55,975 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=218936.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:52:57,374 INFO [train.py:901] (3/4) Epoch 28, batch 700, loss[loss=0.1914, simple_loss=0.2904, pruned_loss=0.04625, over 8282.00 frames. ], tot_loss[loss=0.1996, simple_loss=0.2833, pruned_loss=0.05791, over 1571576.20 frames. ], batch size: 23, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:52:59,064 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=218940.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:53:01,202 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=218943.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:53:18,867 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=218968.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:53:31,049 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=218983.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:53:33,197 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=218986.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:53:34,531 INFO [train.py:901] (3/4) Epoch 28, batch 750, loss[loss=0.2104, simple_loss=0.2976, pruned_loss=0.06157, over 8246.00 frames. ], tot_loss[loss=0.2005, simple_loss=0.2845, pruned_loss=0.05822, over 1583600.64 frames. ], batch size: 24, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:53:46,823 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=219002.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:53:55,127 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.438e+02 2.280e+02 2.810e+02 3.388e+02 7.203e+02, threshold=5.620e+02, percent-clipped=6.0 +2023-02-08 23:53:55,168 WARNING [train.py:1067] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-02-08 23:54:04,608 WARNING [train.py:1067] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-02-08 23:54:12,584 INFO [train.py:901] (3/4) Epoch 28, batch 800, loss[loss=0.2244, simple_loss=0.2999, pruned_loss=0.07445, over 8359.00 frames. ], tot_loss[loss=0.2006, simple_loss=0.2844, pruned_loss=0.05837, over 1591836.39 frames. ], batch size: 24, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:54:12,707 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.2426, 3.1519, 2.9383, 1.6002, 2.8151, 2.9595, 2.7911, 2.8058], + device='cuda:3'), covar=tensor([0.1225, 0.0913, 0.1338, 0.4692, 0.1310, 0.1233, 0.1784, 0.1181], + device='cuda:3'), in_proj_covar=tensor([0.0541, 0.0460, 0.0445, 0.0555, 0.0444, 0.0464, 0.0438, 0.0405], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-08 23:54:13,435 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=219039.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:54:22,038 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=219051.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:54:49,140 INFO [train.py:901] (3/4) Epoch 28, batch 850, loss[loss=0.199, simple_loss=0.2879, pruned_loss=0.05504, over 8655.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.284, pruned_loss=0.0581, over 1598090.78 frames. ], batch size: 39, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:54:50,769 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=219090.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:55:10,262 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.529e+02 2.432e+02 3.183e+02 3.929e+02 8.024e+02, threshold=6.365e+02, percent-clipped=6.0 +2023-02-08 23:55:11,247 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.6794, 2.4975, 3.1594, 2.6463, 3.0651, 2.6764, 2.5707, 2.4329], + device='cuda:3'), covar=tensor([0.4526, 0.4376, 0.1780, 0.3074, 0.2193, 0.2558, 0.1567, 0.4160], + device='cuda:3'), in_proj_covar=tensor([0.0965, 0.1023, 0.0830, 0.0991, 0.1022, 0.0928, 0.0769, 0.0848], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-08 23:55:15,817 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-02-08 23:55:27,563 INFO [train.py:901] (3/4) Epoch 28, batch 900, loss[loss=0.1898, simple_loss=0.2844, pruned_loss=0.04757, over 7816.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.2837, pruned_loss=0.05825, over 1601094.93 frames. ], batch size: 20, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:56:03,866 INFO [train.py:901] (3/4) Epoch 28, batch 950, loss[loss=0.1683, simple_loss=0.2524, pruned_loss=0.04215, over 7697.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.2827, pruned_loss=0.05712, over 1605667.69 frames. ], batch size: 18, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:56:15,311 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.2411, 1.4883, 3.3809, 1.0372, 2.9879, 2.7682, 3.0725, 2.9842], + device='cuda:3'), covar=tensor([0.0898, 0.4131, 0.0825, 0.4573, 0.1365, 0.1233, 0.0833, 0.0928], + device='cuda:3'), in_proj_covar=tensor([0.0676, 0.0659, 0.0729, 0.0654, 0.0739, 0.0631, 0.0637, 0.0711], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-08 23:56:22,891 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.715e+02 2.524e+02 3.053e+02 4.249e+02 9.516e+02, threshold=6.106e+02, percent-clipped=7.0 +2023-02-08 23:56:29,705 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=219221.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:56:34,873 WARNING [train.py:1067] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-02-08 23:56:40,766 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.3514, 1.4002, 4.5469, 1.7276, 4.0187, 3.7492, 4.0902, 3.9557], + device='cuda:3'), covar=tensor([0.0657, 0.5109, 0.0537, 0.4381, 0.1153, 0.1049, 0.0633, 0.0736], + device='cuda:3'), in_proj_covar=tensor([0.0676, 0.0659, 0.0727, 0.0653, 0.0739, 0.0631, 0.0636, 0.0711], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-08 23:56:43,595 INFO [train.py:901] (3/4) Epoch 28, batch 1000, loss[loss=0.2055, simple_loss=0.2913, pruned_loss=0.05981, over 8241.00 frames. ], tot_loss[loss=0.1986, simple_loss=0.2829, pruned_loss=0.05718, over 1606567.52 frames. ], batch size: 24, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:56:46,622 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=219242.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:57:04,459 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=219267.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:57:11,566 WARNING [train.py:1067] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-02-08 23:57:19,442 INFO [train.py:901] (3/4) Epoch 28, batch 1050, loss[loss=0.2046, simple_loss=0.287, pruned_loss=0.0611, over 8458.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2824, pruned_loss=0.05682, over 1613101.38 frames. ], batch size: 27, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:57:23,683 WARNING [train.py:1067] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-02-08 23:57:33,336 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=219307.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:57:38,279 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.554e+02 2.456e+02 2.957e+02 3.788e+02 8.190e+02, threshold=5.915e+02, percent-clipped=1.0 +2023-02-08 23:57:47,827 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=219327.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:57:52,100 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=219332.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:57:52,297 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-02-08 23:57:56,811 INFO [train.py:901] (3/4) Epoch 28, batch 1100, loss[loss=0.2463, simple_loss=0.337, pruned_loss=0.07775, over 8466.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2822, pruned_loss=0.05658, over 1616541.25 frames. ], batch size: 25, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:58:03,512 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=219346.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:58:13,605 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=219360.0, num_to_drop=1, layers_to_drop={1} +2023-02-08 23:58:30,142 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=219383.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:58:33,477 INFO [train.py:901] (3/4) Epoch 28, batch 1150, loss[loss=0.2127, simple_loss=0.3006, pruned_loss=0.0624, over 8096.00 frames. ], tot_loss[loss=0.1983, simple_loss=0.2824, pruned_loss=0.05708, over 1613806.73 frames. ], batch size: 23, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:58:37,170 WARNING [train.py:1067] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-02-08 23:58:52,534 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.530e+02 2.386e+02 3.071e+02 3.782e+02 1.293e+03, threshold=6.141e+02, percent-clipped=2.0 +2023-02-08 23:59:07,140 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=219434.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:59:10,014 INFO [train.py:901] (3/4) Epoch 28, batch 1200, loss[loss=0.175, simple_loss=0.253, pruned_loss=0.04852, over 7810.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2814, pruned_loss=0.05638, over 1617818.17 frames. ], batch size: 20, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:59:13,032 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=219442.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:59:27,990 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=219461.0, num_to_drop=0, layers_to_drop=set() +2023-02-08 23:59:47,599 INFO [train.py:901] (3/4) Epoch 28, batch 1250, loss[loss=0.1831, simple_loss=0.2496, pruned_loss=0.05837, over 7709.00 frames. ], tot_loss[loss=0.1961, simple_loss=0.2802, pruned_loss=0.05597, over 1613637.68 frames. ], batch size: 18, lr: 2.71e-03, grad_scale: 8.0 +2023-02-08 23:59:48,520 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.7938, 1.6702, 1.9062, 1.7113, 0.9458, 1.6368, 2.2519, 2.2658], + device='cuda:3'), covar=tensor([0.0452, 0.1278, 0.1605, 0.1416, 0.0649, 0.1510, 0.0654, 0.0526], + device='cuda:3'), in_proj_covar=tensor([0.0099, 0.0153, 0.0188, 0.0161, 0.0101, 0.0163, 0.0112, 0.0146], + device='cuda:3'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:3') +2023-02-08 23:59:55,045 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=219498.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:00:02,378 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=219508.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:00:06,615 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.598e+02 2.357e+02 2.809e+02 3.466e+02 7.121e+02, threshold=5.618e+02, percent-clipped=3.0 +2023-02-09 00:00:23,411 INFO [train.py:901] (3/4) Epoch 28, batch 1300, loss[loss=0.1924, simple_loss=0.2824, pruned_loss=0.05124, over 7808.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2823, pruned_loss=0.05685, over 1614870.73 frames. ], batch size: 20, lr: 2.71e-03, grad_scale: 8.0 +2023-02-09 00:00:25,027 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.7824, 2.9232, 2.5733, 4.1047, 1.8562, 2.3339, 2.7348, 2.9686], + device='cuda:3'), covar=tensor([0.0622, 0.0779, 0.0717, 0.0203, 0.0975, 0.1086, 0.0792, 0.0750], + device='cuda:3'), in_proj_covar=tensor([0.0227, 0.0191, 0.0240, 0.0209, 0.0199, 0.0242, 0.0245, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0006, 0.0005], + device='cuda:3') +2023-02-09 00:00:31,523 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=219549.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:00:42,960 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=219565.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:01:02,340 INFO [train.py:901] (3/4) Epoch 28, batch 1350, loss[loss=0.2085, simple_loss=0.277, pruned_loss=0.07001, over 7694.00 frames. ], tot_loss[loss=0.1983, simple_loss=0.2824, pruned_loss=0.05705, over 1616427.95 frames. ], batch size: 18, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:01:22,006 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.513e+02 2.365e+02 2.856e+02 3.377e+02 7.819e+02, threshold=5.713e+02, percent-clipped=4.0 +2023-02-09 00:01:39,672 INFO [train.py:901] (3/4) Epoch 28, batch 1400, loss[loss=0.1877, simple_loss=0.2677, pruned_loss=0.05387, over 8082.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2818, pruned_loss=0.05716, over 1614995.59 frames. ], batch size: 21, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:02:09,801 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=219680.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:02:15,329 INFO [train.py:901] (3/4) Epoch 28, batch 1450, loss[loss=0.1794, simple_loss=0.2502, pruned_loss=0.0543, over 7667.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2802, pruned_loss=0.0563, over 1613169.78 frames. ], batch size: 19, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:02:24,266 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=219698.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:02:25,400 WARNING [train.py:1067] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-02-09 00:02:28,351 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=219704.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 00:02:33,318 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.1875, 3.0798, 2.8727, 1.6156, 2.7867, 2.8906, 2.7197, 2.7696], + device='cuda:3'), covar=tensor([0.1137, 0.0863, 0.1219, 0.4508, 0.1191, 0.1246, 0.1650, 0.1062], + device='cuda:3'), in_proj_covar=tensor([0.0540, 0.0460, 0.0448, 0.0555, 0.0444, 0.0465, 0.0439, 0.0406], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 00:02:36,733 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.725e+02 2.298e+02 2.874e+02 3.536e+02 7.746e+02, threshold=5.748e+02, percent-clipped=3.0 +2023-02-09 00:02:39,220 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=219717.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:02:43,526 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=219723.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:02:54,207 INFO [train.py:901] (3/4) Epoch 28, batch 1500, loss[loss=0.2276, simple_loss=0.3089, pruned_loss=0.07316, over 8548.00 frames. ], tot_loss[loss=0.1961, simple_loss=0.2799, pruned_loss=0.0561, over 1615795.38 frames. ], batch size: 31, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:02:57,277 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=219742.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:03:05,719 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=219754.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:03:23,844 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=219779.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:03:25,274 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=219781.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:03:29,883 INFO [train.py:901] (3/4) Epoch 28, batch 1550, loss[loss=0.233, simple_loss=0.3129, pruned_loss=0.07653, over 8307.00 frames. ], tot_loss[loss=0.1961, simple_loss=0.2797, pruned_loss=0.05626, over 1607980.28 frames. ], batch size: 25, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:03:42,633 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=219805.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:03:49,436 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.606e+02 2.435e+02 2.945e+02 3.827e+02 6.900e+02, threshold=5.889e+02, percent-clipped=4.0 +2023-02-09 00:03:54,634 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=219819.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 00:03:55,271 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.5664, 1.5010, 1.7966, 1.3755, 0.9357, 1.5573, 1.5583, 1.4053], + device='cuda:3'), covar=tensor([0.0584, 0.1153, 0.1571, 0.1464, 0.0572, 0.1361, 0.0676, 0.0674], + device='cuda:3'), in_proj_covar=tensor([0.0098, 0.0152, 0.0188, 0.0161, 0.0101, 0.0162, 0.0112, 0.0145], + device='cuda:3'), out_proj_covar=tensor([0.0006, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:3') +2023-02-09 00:04:03,145 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=219830.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:04:08,591 INFO [train.py:901] (3/4) Epoch 28, batch 1600, loss[loss=0.1936, simple_loss=0.2714, pruned_loss=0.0579, over 7793.00 frames. ], tot_loss[loss=0.1957, simple_loss=0.2796, pruned_loss=0.05595, over 1605847.78 frames. ], batch size: 20, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:04:18,707 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=219852.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:04:35,336 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=219875.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:04:44,471 INFO [train.py:901] (3/4) Epoch 28, batch 1650, loss[loss=0.2549, simple_loss=0.3282, pruned_loss=0.09081, over 8536.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2801, pruned_loss=0.05673, over 1606000.28 frames. ], batch size: 49, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:05:02,695 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.357e+02 2.482e+02 2.898e+02 3.443e+02 5.647e+02, threshold=5.797e+02, percent-clipped=0.0 +2023-02-09 00:05:20,308 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=219936.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:05:21,479 INFO [train.py:901] (3/4) Epoch 28, batch 1700, loss[loss=0.1707, simple_loss=0.2572, pruned_loss=0.0421, over 7823.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2805, pruned_loss=0.05682, over 1608587.82 frames. ], batch size: 20, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:05:39,123 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=219961.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:05:43,286 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=219967.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:05:57,781 INFO [train.py:901] (3/4) Epoch 28, batch 1750, loss[loss=0.2328, simple_loss=0.3233, pruned_loss=0.07114, over 8622.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2805, pruned_loss=0.0568, over 1609755.99 frames. ], batch size: 31, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:06:17,596 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.415e+02 2.339e+02 2.848e+02 3.606e+02 1.047e+03, threshold=5.695e+02, percent-clipped=4.0 +2023-02-09 00:06:34,452 INFO [train.py:901] (3/4) Epoch 28, batch 1800, loss[loss=0.1711, simple_loss=0.2607, pruned_loss=0.04071, over 8027.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2798, pruned_loss=0.05644, over 1610687.75 frames. ], batch size: 22, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:07:02,982 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=220075.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 00:07:11,953 INFO [train.py:901] (3/4) Epoch 28, batch 1850, loss[loss=0.2119, simple_loss=0.2825, pruned_loss=0.07064, over 6841.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2815, pruned_loss=0.05683, over 1616688.50 frames. ], batch size: 15, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:07:20,521 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=220100.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 00:07:23,908 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=220105.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:07:30,313 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.818e+02 2.328e+02 2.682e+02 3.608e+02 8.535e+02, threshold=5.364e+02, percent-clipped=7.0 +2023-02-09 00:07:31,752 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=220116.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:07:38,072 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=220125.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:07:47,172 INFO [train.py:901] (3/4) Epoch 28, batch 1900, loss[loss=0.1657, simple_loss=0.2466, pruned_loss=0.04235, over 7785.00 frames. ], tot_loss[loss=0.1982, simple_loss=0.2822, pruned_loss=0.05711, over 1618024.59 frames. ], batch size: 19, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:08:19,369 WARNING [train.py:1067] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-02-09 00:08:25,768 INFO [train.py:901] (3/4) Epoch 28, batch 1950, loss[loss=0.174, simple_loss=0.2618, pruned_loss=0.04314, over 8246.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2816, pruned_loss=0.05665, over 1619529.97 frames. ], batch size: 22, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:08:32,998 WARNING [train.py:1067] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-02-09 00:08:44,716 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.518e+02 2.461e+02 2.916e+02 3.869e+02 7.609e+02, threshold=5.833e+02, percent-clipped=8.0 +2023-02-09 00:08:48,262 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=220219.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:08:51,070 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=220223.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:08:53,654 WARNING [train.py:1067] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-02-09 00:09:01,370 INFO [train.py:901] (3/4) Epoch 28, batch 2000, loss[loss=0.1922, simple_loss=0.2829, pruned_loss=0.05073, over 8492.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2804, pruned_loss=0.05558, over 1619103.65 frames. ], batch size: 28, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:09:02,869 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=220240.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:09:07,814 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=220247.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:09:08,597 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=220248.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:09:29,133 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=220276.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:09:37,452 INFO [train.py:901] (3/4) Epoch 28, batch 2050, loss[loss=0.1516, simple_loss=0.2355, pruned_loss=0.0339, over 7254.00 frames. ], tot_loss[loss=0.1957, simple_loss=0.2797, pruned_loss=0.05581, over 1613233.72 frames. ], batch size: 16, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:09:58,201 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.689e+02 2.398e+02 2.757e+02 3.324e+02 6.340e+02, threshold=5.514e+02, percent-clipped=2.0 +2023-02-09 00:10:03,996 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.4675, 2.3891, 3.0722, 2.5694, 2.9423, 2.5385, 2.3516, 1.9293], + device='cuda:3'), covar=tensor([0.5609, 0.5131, 0.2152, 0.3899, 0.2714, 0.3230, 0.1930, 0.5577], + device='cuda:3'), in_proj_covar=tensor([0.0970, 0.1029, 0.0835, 0.0997, 0.1028, 0.0934, 0.0773, 0.0853], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 00:10:12,779 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=220334.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:10:15,437 INFO [train.py:901] (3/4) Epoch 28, batch 2100, loss[loss=0.1886, simple_loss=0.2692, pruned_loss=0.05398, over 7905.00 frames. ], tot_loss[loss=0.1962, simple_loss=0.2803, pruned_loss=0.05607, over 1614485.37 frames. ], batch size: 20, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:10:20,353 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-02-09 00:10:42,393 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.59 vs. limit=2.0 +2023-02-09 00:10:51,258 INFO [train.py:901] (3/4) Epoch 28, batch 2150, loss[loss=0.2704, simple_loss=0.3445, pruned_loss=0.09815, over 8589.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2812, pruned_loss=0.05647, over 1616417.39 frames. ], batch size: 34, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:11:11,485 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.499e+02 2.504e+02 2.973e+02 4.041e+02 1.001e+03, threshold=5.945e+02, percent-clipped=8.0 +2023-02-09 00:11:28,331 INFO [train.py:901] (3/4) Epoch 28, batch 2200, loss[loss=0.2017, simple_loss=0.2853, pruned_loss=0.05904, over 8467.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2815, pruned_loss=0.05671, over 1612601.39 frames. ], batch size: 29, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:11:36,287 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=220449.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:11:44,017 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=220460.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:12:03,396 INFO [train.py:901] (3/4) Epoch 28, batch 2250, loss[loss=0.2743, simple_loss=0.3386, pruned_loss=0.105, over 7206.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2817, pruned_loss=0.05685, over 1614702.28 frames. ], batch size: 72, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:12:09,262 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=220496.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:12:15,027 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.43 vs. limit=5.0 +2023-02-09 00:12:22,281 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.608e+02 2.331e+02 2.835e+02 3.325e+02 7.200e+02, threshold=5.671e+02, percent-clipped=3.0 +2023-02-09 00:12:27,565 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=220521.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:12:41,523 INFO [train.py:901] (3/4) Epoch 28, batch 2300, loss[loss=0.1816, simple_loss=0.2662, pruned_loss=0.04851, over 8344.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2817, pruned_loss=0.057, over 1612323.09 frames. ], batch size: 26, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:12:44,394 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.8360, 1.4281, 3.9883, 1.4481, 3.5645, 3.2466, 3.6172, 3.5074], + device='cuda:3'), covar=tensor([0.0661, 0.4735, 0.0671, 0.4530, 0.1209, 0.1126, 0.0704, 0.0765], + device='cuda:3'), in_proj_covar=tensor([0.0682, 0.0670, 0.0739, 0.0664, 0.0752, 0.0639, 0.0646, 0.0722], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 00:12:51,188 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.3801, 1.1766, 2.3311, 1.2983, 2.1837, 2.4952, 2.7022, 2.0701], + device='cuda:3'), covar=tensor([0.1177, 0.1559, 0.0419, 0.2015, 0.0707, 0.0406, 0.0611, 0.0721], + device='cuda:3'), in_proj_covar=tensor([0.0308, 0.0327, 0.0293, 0.0321, 0.0323, 0.0277, 0.0441, 0.0308], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:3') +2023-02-09 00:12:59,748 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=220564.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:13:06,034 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=220573.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:13:07,482 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=220575.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:13:16,616 INFO [train.py:901] (3/4) Epoch 28, batch 2350, loss[loss=0.1867, simple_loss=0.2878, pruned_loss=0.04283, over 8613.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2818, pruned_loss=0.05713, over 1612678.44 frames. ], batch size: 31, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:13:18,250 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=220590.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:13:18,835 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=220591.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:13:35,683 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.370e+02 2.329e+02 2.956e+02 3.826e+02 8.837e+02, threshold=5.912e+02, percent-clipped=4.0 +2023-02-09 00:13:36,675 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=220615.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:13:40,248 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=220620.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:13:53,607 INFO [train.py:901] (3/4) Epoch 28, batch 2400, loss[loss=0.1599, simple_loss=0.249, pruned_loss=0.03541, over 7560.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2815, pruned_loss=0.05669, over 1617060.95 frames. ], batch size: 18, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:14:16,684 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=220669.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:14:29,716 INFO [train.py:901] (3/4) Epoch 28, batch 2450, loss[loss=0.1781, simple_loss=0.2798, pruned_loss=0.03817, over 8251.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2816, pruned_loss=0.05697, over 1617346.23 frames. ], batch size: 24, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:14:42,732 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=220706.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:14:48,782 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.600e+02 2.507e+02 3.309e+02 3.917e+02 8.053e+02, threshold=6.618e+02, percent-clipped=4.0 +2023-02-09 00:15:03,093 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=220735.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:15:05,098 INFO [train.py:901] (3/4) Epoch 28, batch 2500, loss[loss=0.1995, simple_loss=0.2915, pruned_loss=0.05372, over 8340.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2827, pruned_loss=0.05745, over 1617550.65 frames. ], batch size: 26, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:15:15,123 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.0298, 1.0437, 0.9850, 1.2303, 0.6495, 0.8976, 1.0149, 1.0678], + device='cuda:3'), covar=tensor([0.0626, 0.0612, 0.0703, 0.0490, 0.0837, 0.1007, 0.0520, 0.0486], + device='cuda:3'), in_proj_covar=tensor([0.0229, 0.0193, 0.0242, 0.0211, 0.0202, 0.0245, 0.0249, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0005, 0.0006, 0.0005], + device='cuda:3') +2023-02-09 00:15:42,751 INFO [train.py:901] (3/4) Epoch 28, batch 2550, loss[loss=0.214, simple_loss=0.3034, pruned_loss=0.06231, over 8361.00 frames. ], tot_loss[loss=0.1995, simple_loss=0.283, pruned_loss=0.05803, over 1612792.32 frames. ], batch size: 24, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:16:02,758 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.451e+02 2.505e+02 3.011e+02 3.782e+02 1.017e+03, threshold=6.023e+02, percent-clipped=3.0 +2023-02-09 00:16:06,707 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=220820.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:16:14,508 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=220831.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:16:19,219 INFO [train.py:901] (3/4) Epoch 28, batch 2600, loss[loss=0.212, simple_loss=0.2933, pruned_loss=0.06531, over 7978.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.2833, pruned_loss=0.05818, over 1617593.27 frames. ], batch size: 21, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:16:20,648 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=220840.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:16:24,298 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=220845.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:16:32,228 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=220856.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:16:57,370 INFO [train.py:901] (3/4) Epoch 28, batch 2650, loss[loss=0.1922, simple_loss=0.2711, pruned_loss=0.05669, over 7657.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.2831, pruned_loss=0.05791, over 1619348.06 frames. ], batch size: 19, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:17:12,920 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.5345, 2.3833, 1.6386, 2.1852, 2.0899, 1.5529, 2.1233, 2.1025], + device='cuda:3'), covar=tensor([0.1653, 0.0513, 0.1467, 0.0706, 0.0836, 0.1725, 0.0991, 0.1107], + device='cuda:3'), in_proj_covar=tensor([0.0356, 0.0242, 0.0339, 0.0311, 0.0301, 0.0345, 0.0346, 0.0321], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-02-09 00:17:16,289 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.526e+02 2.381e+02 2.801e+02 3.642e+02 5.464e+02, threshold=5.602e+02, percent-clipped=0.0 +2023-02-09 00:17:17,801 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=220917.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:17:21,291 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=220922.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:17:32,907 INFO [train.py:901] (3/4) Epoch 28, batch 2700, loss[loss=0.1784, simple_loss=0.2728, pruned_loss=0.04195, over 8325.00 frames. ], tot_loss[loss=0.2, simple_loss=0.284, pruned_loss=0.05798, over 1621892.33 frames. ], batch size: 25, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:17:50,477 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=220962.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:18:09,046 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=220987.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:18:09,541 INFO [train.py:901] (3/4) Epoch 28, batch 2750, loss[loss=0.1913, simple_loss=0.2791, pruned_loss=0.05177, over 8473.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.2833, pruned_loss=0.05776, over 1623938.00 frames. ], batch size: 27, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:18:11,779 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=220991.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:18:29,675 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=221013.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:18:31,036 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.597e+02 2.419e+02 2.908e+02 3.517e+02 7.342e+02, threshold=5.816e+02, percent-clipped=5.0 +2023-02-09 00:18:31,972 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=221016.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:18:43,266 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=221032.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:18:47,246 INFO [train.py:901] (3/4) Epoch 28, batch 2800, loss[loss=0.1789, simple_loss=0.2755, pruned_loss=0.04113, over 8364.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2819, pruned_loss=0.05667, over 1622143.50 frames. ], batch size: 24, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:18:57,815 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.5552, 1.4594, 1.8655, 1.2495, 1.2168, 1.8163, 0.2108, 1.2230], + device='cuda:3'), covar=tensor([0.1387, 0.1220, 0.0380, 0.0822, 0.2293, 0.0440, 0.1846, 0.1104], + device='cuda:3'), in_proj_covar=tensor([0.0201, 0.0205, 0.0136, 0.0224, 0.0277, 0.0146, 0.0174, 0.0199], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 00:19:22,678 INFO [train.py:901] (3/4) Epoch 28, batch 2850, loss[loss=0.1811, simple_loss=0.2735, pruned_loss=0.04434, over 7978.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2825, pruned_loss=0.05713, over 1620327.51 frames. ], batch size: 21, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:19:43,235 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.362e+02 2.366e+02 2.856e+02 3.627e+02 6.501e+02, threshold=5.713e+02, percent-clipped=2.0 +2023-02-09 00:19:53,963 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=221128.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:20:00,671 INFO [train.py:901] (3/4) Epoch 28, batch 2900, loss[loss=0.2189, simple_loss=0.2813, pruned_loss=0.07819, over 7417.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2822, pruned_loss=0.05685, over 1620943.56 frames. ], batch size: 17, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:20:04,298 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.4768, 1.9991, 3.1033, 1.3516, 2.4359, 1.9285, 1.6891, 2.4035], + device='cuda:3'), covar=tensor([0.2244, 0.2868, 0.0987, 0.5290, 0.2021, 0.3747, 0.2709, 0.2433], + device='cuda:3'), in_proj_covar=tensor([0.0541, 0.0639, 0.0566, 0.0673, 0.0663, 0.0612, 0.0566, 0.0647], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 00:20:32,266 WARNING [train.py:1067] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-02-09 00:20:33,724 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=221184.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:20:36,494 INFO [train.py:901] (3/4) Epoch 28, batch 2950, loss[loss=0.1859, simple_loss=0.2632, pruned_loss=0.05426, over 7266.00 frames. ], tot_loss[loss=0.1983, simple_loss=0.2824, pruned_loss=0.05706, over 1621034.03 frames. ], batch size: 16, lr: 2.70e-03, grad_scale: 8.0 +2023-02-09 00:20:43,340 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.8636, 2.3075, 3.7593, 1.7019, 3.0641, 2.3021, 1.9494, 2.8389], + device='cuda:3'), covar=tensor([0.1953, 0.2769, 0.0967, 0.4618, 0.1754, 0.3269, 0.2430, 0.2459], + device='cuda:3'), in_proj_covar=tensor([0.0541, 0.0640, 0.0567, 0.0674, 0.0663, 0.0612, 0.0567, 0.0648], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 00:20:55,452 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.514e+02 2.299e+02 2.993e+02 3.879e+02 1.208e+03, threshold=5.985e+02, percent-clipped=10.0 +2023-02-09 00:21:01,648 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.00 vs. limit=5.0 +2023-02-09 00:21:13,543 INFO [train.py:901] (3/4) Epoch 28, batch 3000, loss[loss=0.2014, simple_loss=0.2826, pruned_loss=0.06011, over 8562.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2809, pruned_loss=0.05648, over 1613237.08 frames. ], batch size: 39, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:21:13,543 INFO [train.py:926] (3/4) Computing validation loss +2023-02-09 00:21:31,972 INFO [train.py:935] (3/4) Epoch 28, validation: loss=0.1712, simple_loss=0.2708, pruned_loss=0.03578, over 944034.00 frames. +2023-02-09 00:21:31,974 INFO [train.py:936] (3/4) Maximum memory allocated so far is 6430MB +2023-02-09 00:21:47,610 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6528, 1.6635, 2.1370, 1.3915, 1.3306, 2.1017, 0.2785, 1.3729], + device='cuda:3'), covar=tensor([0.1583, 0.1142, 0.0394, 0.1002, 0.2206, 0.0452, 0.1815, 0.1187], + device='cuda:3'), in_proj_covar=tensor([0.0201, 0.0205, 0.0137, 0.0225, 0.0278, 0.0147, 0.0174, 0.0199], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 00:21:54,464 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=221266.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:21:55,527 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.60 vs. limit=2.0 +2023-02-09 00:22:03,533 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.12 vs. limit=5.0 +2023-02-09 00:22:10,152 INFO [train.py:901] (3/4) Epoch 28, batch 3050, loss[loss=0.1824, simple_loss=0.2557, pruned_loss=0.0545, over 7432.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2813, pruned_loss=0.05687, over 1613906.86 frames. ], batch size: 17, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:22:10,388 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=221288.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:22:18,084 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=221299.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:22:28,220 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=221313.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:22:29,365 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.568e+02 2.361e+02 2.830e+02 3.600e+02 1.199e+03, threshold=5.660e+02, percent-clipped=4.0 +2023-02-09 00:22:37,155 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.5432, 2.0794, 3.3102, 1.6186, 1.7285, 3.2335, 0.8256, 2.0389], + device='cuda:3'), covar=tensor([0.1368, 0.1155, 0.0246, 0.1544, 0.2263, 0.0349, 0.2026, 0.1235], + device='cuda:3'), in_proj_covar=tensor([0.0201, 0.0205, 0.0137, 0.0224, 0.0277, 0.0146, 0.0174, 0.0199], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 00:22:45,352 INFO [train.py:901] (3/4) Epoch 28, batch 3100, loss[loss=0.2037, simple_loss=0.2997, pruned_loss=0.05385, over 8368.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2813, pruned_loss=0.05689, over 1615725.00 frames. ], batch size: 24, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:23:18,458 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=221381.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:23:20,619 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=221384.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:23:23,254 INFO [train.py:901] (3/4) Epoch 28, batch 3150, loss[loss=0.2251, simple_loss=0.3056, pruned_loss=0.07227, over 8408.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2813, pruned_loss=0.05695, over 1613946.52 frames. ], batch size: 49, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:23:31,492 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.9683, 1.5706, 3.1421, 1.5728, 2.2448, 3.3285, 3.4833, 2.8628], + device='cuda:3'), covar=tensor([0.1174, 0.1785, 0.0357, 0.2044, 0.1111, 0.0279, 0.0563, 0.0567], + device='cuda:3'), in_proj_covar=tensor([0.0308, 0.0329, 0.0294, 0.0323, 0.0326, 0.0278, 0.0443, 0.0309], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:3') +2023-02-09 00:23:34,616 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6453, 2.4648, 1.7415, 2.2605, 2.0845, 1.6182, 2.0474, 2.2555], + device='cuda:3'), covar=tensor([0.1683, 0.0523, 0.1529, 0.0752, 0.0983, 0.1789, 0.1185, 0.1233], + device='cuda:3'), in_proj_covar=tensor([0.0362, 0.0246, 0.0344, 0.0315, 0.0305, 0.0350, 0.0351, 0.0325], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-02-09 00:23:38,998 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=221409.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:23:43,026 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.782e+02 2.344e+02 3.031e+02 3.872e+02 9.124e+02, threshold=6.062e+02, percent-clipped=5.0 +2023-02-09 00:24:00,285 INFO [train.py:901] (3/4) Epoch 28, batch 3200, loss[loss=0.2359, simple_loss=0.3065, pruned_loss=0.08267, over 8635.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2813, pruned_loss=0.05715, over 1609566.04 frames. ], batch size: 39, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:24:21,664 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.8966, 1.4772, 1.6092, 1.4492, 1.1713, 1.5182, 1.8197, 1.7242], + device='cuda:3'), covar=tensor([0.0750, 0.1346, 0.1802, 0.1534, 0.0739, 0.1554, 0.0855, 0.0587], + device='cuda:3'), in_proj_covar=tensor([0.0099, 0.0154, 0.0191, 0.0163, 0.0102, 0.0164, 0.0114, 0.0147], + device='cuda:3'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:3') +2023-02-09 00:24:35,508 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6876, 1.4911, 2.8586, 1.3867, 2.3019, 3.0488, 3.2317, 2.6284], + device='cuda:3'), covar=tensor([0.1284, 0.1728, 0.0382, 0.2255, 0.0937, 0.0313, 0.0595, 0.0585], + device='cuda:3'), in_proj_covar=tensor([0.0307, 0.0328, 0.0294, 0.0323, 0.0325, 0.0278, 0.0441, 0.0309], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:3') +2023-02-09 00:24:36,746 INFO [train.py:901] (3/4) Epoch 28, batch 3250, loss[loss=0.1924, simple_loss=0.2733, pruned_loss=0.05578, over 7656.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2808, pruned_loss=0.05665, over 1609824.14 frames. ], batch size: 19, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:24:56,686 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.554e+02 2.359e+02 2.800e+02 3.771e+02 8.910e+02, threshold=5.600e+02, percent-clipped=3.0 +2023-02-09 00:25:04,064 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=221525.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:25:12,969 INFO [train.py:901] (3/4) Epoch 28, batch 3300, loss[loss=0.1719, simple_loss=0.2631, pruned_loss=0.04039, over 7925.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2809, pruned_loss=0.05687, over 1606973.43 frames. ], batch size: 20, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:25:25,041 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=221555.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:25:42,974 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=221580.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:25:48,463 INFO [train.py:901] (3/4) Epoch 28, batch 3350, loss[loss=0.2098, simple_loss=0.2945, pruned_loss=0.06257, over 8670.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2814, pruned_loss=0.0572, over 1610122.86 frames. ], batch size: 34, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:26:09,960 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.399e+02 2.531e+02 3.062e+02 3.663e+02 8.444e+02, threshold=6.124e+02, percent-clipped=3.0 +2023-02-09 00:26:26,152 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=221637.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:26:26,641 INFO [train.py:901] (3/4) Epoch 28, batch 3400, loss[loss=0.1791, simple_loss=0.268, pruned_loss=0.04504, over 8134.00 frames. ], tot_loss[loss=0.1983, simple_loss=0.2817, pruned_loss=0.05746, over 1607791.35 frames. ], batch size: 22, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:26:43,903 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=221662.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:27:02,351 INFO [train.py:901] (3/4) Epoch 28, batch 3450, loss[loss=0.2402, simple_loss=0.3228, pruned_loss=0.07879, over 8369.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2822, pruned_loss=0.05769, over 1610139.82 frames. ], batch size: 24, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:27:21,425 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.552e+02 2.306e+02 2.763e+02 3.583e+02 8.756e+02, threshold=5.526e+02, percent-clipped=3.0 +2023-02-09 00:27:39,502 INFO [train.py:901] (3/4) Epoch 28, batch 3500, loss[loss=0.241, simple_loss=0.2996, pruned_loss=0.0912, over 7206.00 frames. ], tot_loss[loss=0.2007, simple_loss=0.2838, pruned_loss=0.05874, over 1610103.49 frames. ], batch size: 16, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:27:56,077 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-02-09 00:28:03,549 WARNING [train.py:1067] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-02-09 00:28:07,264 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=221776.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:28:15,813 INFO [train.py:901] (3/4) Epoch 28, batch 3550, loss[loss=0.2063, simple_loss=0.2917, pruned_loss=0.06038, over 8125.00 frames. ], tot_loss[loss=0.1991, simple_loss=0.2825, pruned_loss=0.05785, over 1606290.22 frames. ], batch size: 22, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:28:35,057 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.37 vs. limit=2.0 +2023-02-09 00:28:35,267 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.790e+02 2.405e+02 2.949e+02 3.672e+02 8.337e+02, threshold=5.897e+02, percent-clipped=3.0 +2023-02-09 00:28:52,590 INFO [train.py:901] (3/4) Epoch 28, batch 3600, loss[loss=0.1613, simple_loss=0.2419, pruned_loss=0.04042, over 7225.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2814, pruned_loss=0.05689, over 1611463.75 frames. ], batch size: 16, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:29:08,872 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-02-09 00:29:09,925 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.2560, 3.1582, 2.9212, 1.7723, 2.8836, 2.8975, 2.8365, 2.8051], + device='cuda:3'), covar=tensor([0.1065, 0.0750, 0.1154, 0.4123, 0.1043, 0.1302, 0.1515, 0.1020], + device='cuda:3'), in_proj_covar=tensor([0.0542, 0.0457, 0.0448, 0.0560, 0.0444, 0.0465, 0.0441, 0.0406], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 00:29:15,451 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=221869.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:29:28,471 INFO [train.py:901] (3/4) Epoch 28, batch 3650, loss[loss=0.2425, simple_loss=0.3132, pruned_loss=0.08592, over 7027.00 frames. ], tot_loss[loss=0.1995, simple_loss=0.2831, pruned_loss=0.05796, over 1610307.24 frames. ], batch size: 71, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:29:42,346 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=221908.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:29:47,070 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.635e+02 2.399e+02 3.022e+02 3.885e+02 8.966e+02, threshold=6.044e+02, percent-clipped=2.0 +2023-02-09 00:30:02,955 INFO [train.py:901] (3/4) Epoch 28, batch 3700, loss[loss=0.2572, simple_loss=0.3169, pruned_loss=0.09879, over 6493.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.2832, pruned_loss=0.05854, over 1603722.03 frames. ], batch size: 72, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:30:05,061 WARNING [train.py:1067] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-02-09 00:30:38,683 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=221984.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:30:41,399 INFO [train.py:901] (3/4) Epoch 28, batch 3750, loss[loss=0.1991, simple_loss=0.2917, pruned_loss=0.05325, over 8465.00 frames. ], tot_loss[loss=0.2004, simple_loss=0.2842, pruned_loss=0.05827, over 1609798.14 frames. ], batch size: 27, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:30:48,356 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.4893, 1.4409, 1.7513, 1.3736, 0.9377, 1.4997, 1.5285, 1.4119], + device='cuda:3'), covar=tensor([0.0605, 0.1267, 0.1617, 0.1482, 0.0575, 0.1458, 0.0697, 0.0670], + device='cuda:3'), in_proj_covar=tensor([0.0099, 0.0153, 0.0190, 0.0162, 0.0101, 0.0163, 0.0112, 0.0146], + device='cuda:3'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:3') +2023-02-09 00:31:01,413 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.685e+02 2.662e+02 3.142e+02 4.083e+02 1.270e+03, threshold=6.284e+02, percent-clipped=8.0 +2023-02-09 00:31:10,199 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=222027.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:31:17,963 INFO [train.py:901] (3/4) Epoch 28, batch 3800, loss[loss=0.2195, simple_loss=0.2978, pruned_loss=0.07057, over 8240.00 frames. ], tot_loss[loss=0.1993, simple_loss=0.2829, pruned_loss=0.05787, over 1605959.17 frames. ], batch size: 22, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:31:55,279 INFO [train.py:901] (3/4) Epoch 28, batch 3850, loss[loss=0.2261, simple_loss=0.3147, pruned_loss=0.06875, over 8516.00 frames. ], tot_loss[loss=0.2004, simple_loss=0.2836, pruned_loss=0.0586, over 1610917.02 frames. ], batch size: 26, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:32:11,766 WARNING [train.py:1067] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-02-09 00:32:13,843 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.497e+02 2.208e+02 2.768e+02 3.453e+02 7.901e+02, threshold=5.537e+02, percent-clipped=1.0 +2023-02-09 00:32:15,442 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([5.8801, 6.0035, 5.2115, 2.5589, 5.3085, 5.7276, 5.4786, 5.5338], + device='cuda:3'), covar=tensor([0.0509, 0.0352, 0.0999, 0.4770, 0.0711, 0.0669, 0.1106, 0.0516], + device='cuda:3'), in_proj_covar=tensor([0.0545, 0.0459, 0.0448, 0.0560, 0.0444, 0.0467, 0.0442, 0.0408], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 00:32:17,532 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=222120.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:32:30,077 INFO [train.py:901] (3/4) Epoch 28, batch 3900, loss[loss=0.2019, simple_loss=0.2833, pruned_loss=0.06019, over 6833.00 frames. ], tot_loss[loss=0.2, simple_loss=0.2833, pruned_loss=0.05833, over 1609433.68 frames. ], batch size: 71, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:33:00,577 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.0257, 3.6082, 1.9397, 2.8774, 2.5259, 1.8428, 2.6051, 3.1431], + device='cuda:3'), covar=tensor([0.1837, 0.0464, 0.1566, 0.0866, 0.1068, 0.1936, 0.1284, 0.0842], + device='cuda:3'), in_proj_covar=tensor([0.0359, 0.0245, 0.0343, 0.0315, 0.0305, 0.0348, 0.0350, 0.0323], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-02-09 00:33:06,448 INFO [train.py:901] (3/4) Epoch 28, batch 3950, loss[loss=0.2294, simple_loss=0.2977, pruned_loss=0.08058, over 8106.00 frames. ], tot_loss[loss=0.2002, simple_loss=0.2832, pruned_loss=0.05854, over 1611012.96 frames. ], batch size: 23, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:33:17,904 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=222203.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:33:20,797 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.2003, 3.6695, 2.3613, 3.0314, 2.7147, 2.1228, 2.8334, 3.0917], + device='cuda:3'), covar=tensor([0.1744, 0.0373, 0.1267, 0.0736, 0.0852, 0.1598, 0.1230, 0.1244], + device='cuda:3'), in_proj_covar=tensor([0.0358, 0.0245, 0.0342, 0.0314, 0.0304, 0.0347, 0.0350, 0.0323], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-02-09 00:33:26,082 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.641e+02 2.338e+02 2.821e+02 3.606e+02 1.107e+03, threshold=5.643e+02, percent-clipped=4.0 +2023-02-09 00:33:31,801 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=222223.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:33:35,385 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.5576, 1.5436, 2.0202, 1.3495, 1.2793, 2.0543, 0.3189, 1.3009], + device='cuda:3'), covar=tensor([0.1749, 0.1572, 0.0502, 0.1228, 0.2542, 0.0429, 0.2122, 0.1552], + device='cuda:3'), in_proj_covar=tensor([0.0202, 0.0207, 0.0137, 0.0225, 0.0279, 0.0147, 0.0176, 0.0200], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 00:33:37,687 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.22 vs. limit=2.0 +2023-02-09 00:33:40,243 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=222235.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:33:42,144 INFO [train.py:901] (3/4) Epoch 28, batch 4000, loss[loss=0.1676, simple_loss=0.2505, pruned_loss=0.04234, over 7698.00 frames. ], tot_loss[loss=0.1996, simple_loss=0.2827, pruned_loss=0.0582, over 1612950.18 frames. ], batch size: 18, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:33:43,728 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=222240.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:33:51,748 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=222252.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:34:01,382 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=222265.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:34:17,513 INFO [train.py:901] (3/4) Epoch 28, batch 4050, loss[loss=0.2365, simple_loss=0.3085, pruned_loss=0.08224, over 7541.00 frames. ], tot_loss[loss=0.1996, simple_loss=0.2827, pruned_loss=0.0583, over 1612015.30 frames. ], batch size: 18, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:34:38,325 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.744e+02 2.414e+02 3.092e+02 4.009e+02 1.246e+03, threshold=6.184e+02, percent-clipped=7.0 +2023-02-09 00:34:54,252 INFO [train.py:901] (3/4) Epoch 28, batch 4100, loss[loss=0.1629, simple_loss=0.245, pruned_loss=0.04042, over 7721.00 frames. ], tot_loss[loss=0.2, simple_loss=0.283, pruned_loss=0.05855, over 1608353.77 frames. ], batch size: 18, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:35:14,880 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=222367.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:35:17,648 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=222371.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:35:29,384 INFO [train.py:901] (3/4) Epoch 28, batch 4150, loss[loss=0.1695, simple_loss=0.2495, pruned_loss=0.04477, over 7792.00 frames. ], tot_loss[loss=0.1987, simple_loss=0.2821, pruned_loss=0.05765, over 1611463.78 frames. ], batch size: 19, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:35:49,100 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.589e+02 2.330e+02 2.692e+02 3.176e+02 6.436e+02, threshold=5.384e+02, percent-clipped=1.0 +2023-02-09 00:36:07,144 INFO [train.py:901] (3/4) Epoch 28, batch 4200, loss[loss=0.1973, simple_loss=0.2922, pruned_loss=0.05122, over 8339.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2823, pruned_loss=0.05761, over 1612244.94 frames. ], batch size: 26, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:36:14,736 WARNING [train.py:1067] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-02-09 00:36:37,817 WARNING [train.py:1067] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-02-09 00:36:41,348 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=222486.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:36:42,589 INFO [train.py:901] (3/4) Epoch 28, batch 4250, loss[loss=0.1855, simple_loss=0.2695, pruned_loss=0.05078, over 7930.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.282, pruned_loss=0.05678, over 1616023.45 frames. ], batch size: 20, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:36:44,938 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=222491.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:36:53,800 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-02-09 00:37:00,857 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.734e+02 2.539e+02 3.193e+02 4.198e+02 8.289e+02, threshold=6.386e+02, percent-clipped=5.0 +2023-02-09 00:37:01,771 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=222516.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:37:18,018 INFO [train.py:901] (3/4) Epoch 28, batch 4300, loss[loss=0.2251, simple_loss=0.2998, pruned_loss=0.07516, over 8732.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2828, pruned_loss=0.05698, over 1621161.52 frames. ], batch size: 30, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:37:20,195 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=222541.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:37:25,033 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=222547.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:37:39,471 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=222567.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:37:54,078 INFO [train.py:901] (3/4) Epoch 28, batch 4350, loss[loss=0.2102, simple_loss=0.2873, pruned_loss=0.06656, over 8516.00 frames. ], tot_loss[loss=0.199, simple_loss=0.2832, pruned_loss=0.05742, over 1621529.51 frames. ], batch size: 26, lr: 2.69e-03, grad_scale: 8.0 +2023-02-09 00:37:54,218 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=222588.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:38:11,687 WARNING [train.py:1067] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-02-09 00:38:13,105 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.579e+02 2.501e+02 2.979e+02 3.614e+02 7.360e+02, threshold=5.959e+02, percent-clipped=2.0 +2023-02-09 00:38:18,806 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=222623.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:38:29,001 INFO [train.py:901] (3/4) Epoch 28, batch 4400, loss[loss=0.1705, simple_loss=0.2591, pruned_loss=0.04098, over 7802.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2829, pruned_loss=0.05732, over 1622270.96 frames. ], batch size: 20, lr: 2.69e-03, grad_scale: 16.0 +2023-02-09 00:38:37,330 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=222648.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:38:40,089 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.8849, 1.9115, 3.1352, 2.3275, 2.6487, 1.9373, 1.6926, 1.6980], + device='cuda:3'), covar=tensor([0.8009, 0.6950, 0.2388, 0.4939, 0.4111, 0.5156, 0.3275, 0.6569], + device='cuda:3'), in_proj_covar=tensor([0.0966, 0.1028, 0.0832, 0.0995, 0.1022, 0.0934, 0.0771, 0.0850], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 00:38:47,080 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=222662.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:38:54,356 WARNING [train.py:1067] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-02-09 00:39:01,973 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=222682.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:39:04,786 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.9794, 1.8711, 2.1186, 2.0606, 1.2605, 1.9358, 2.4216, 2.2847], + device='cuda:3'), covar=tensor([0.0445, 0.1076, 0.1495, 0.1235, 0.0516, 0.1286, 0.0558, 0.0553], + device='cuda:3'), in_proj_covar=tensor([0.0099, 0.0153, 0.0190, 0.0162, 0.0101, 0.0163, 0.0113, 0.0146], + device='cuda:3'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:3') +2023-02-09 00:39:05,968 INFO [train.py:901] (3/4) Epoch 28, batch 4450, loss[loss=0.1919, simple_loss=0.2674, pruned_loss=0.05825, over 7702.00 frames. ], tot_loss[loss=0.1993, simple_loss=0.2832, pruned_loss=0.05769, over 1621071.65 frames. ], batch size: 18, lr: 2.69e-03, grad_scale: 16.0 +2023-02-09 00:39:24,968 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.632e+02 2.353e+02 2.798e+02 3.446e+02 6.111e+02, threshold=5.597e+02, percent-clipped=1.0 +2023-02-09 00:39:38,908 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.99 vs. limit=5.0 +2023-02-09 00:39:41,219 INFO [train.py:901] (3/4) Epoch 28, batch 4500, loss[loss=0.201, simple_loss=0.2865, pruned_loss=0.05773, over 8436.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2825, pruned_loss=0.0577, over 1618438.03 frames. ], batch size: 27, lr: 2.69e-03, grad_scale: 16.0 +2023-02-09 00:39:44,227 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=222742.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:39:45,360 WARNING [train.py:1067] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-02-09 00:39:58,786 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.62 vs. limit=2.0 +2023-02-09 00:40:02,733 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=222767.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:40:18,346 INFO [train.py:901] (3/4) Epoch 28, batch 4550, loss[loss=0.1741, simple_loss=0.2685, pruned_loss=0.03986, over 8023.00 frames. ], tot_loss[loss=0.1999, simple_loss=0.2832, pruned_loss=0.05827, over 1617468.33 frames. ], batch size: 22, lr: 2.69e-03, grad_scale: 16.0 +2023-02-09 00:40:26,135 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6728, 1.7611, 1.5852, 2.2026, 1.0480, 1.5260, 1.7022, 1.7673], + device='cuda:3'), covar=tensor([0.0849, 0.0808, 0.0960, 0.0490, 0.1166, 0.1264, 0.0754, 0.0747], + device='cuda:3'), in_proj_covar=tensor([0.0232, 0.0195, 0.0245, 0.0214, 0.0205, 0.0248, 0.0251, 0.0205], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:3') +2023-02-09 00:40:37,186 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.362e+02 2.324e+02 2.721e+02 3.677e+02 6.861e+02, threshold=5.442e+02, percent-clipped=4.0 +2023-02-09 00:40:53,694 INFO [train.py:901] (3/4) Epoch 28, batch 4600, loss[loss=0.2174, simple_loss=0.2996, pruned_loss=0.06758, over 8496.00 frames. ], tot_loss[loss=0.1987, simple_loss=0.2823, pruned_loss=0.05752, over 1620232.45 frames. ], batch size: 26, lr: 2.69e-03, grad_scale: 16.0 +2023-02-09 00:41:11,248 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6123, 2.0556, 3.2506, 1.5177, 2.4475, 2.1312, 1.7679, 2.5833], + device='cuda:3'), covar=tensor([0.2057, 0.2850, 0.0941, 0.4776, 0.1973, 0.3165, 0.2577, 0.2353], + device='cuda:3'), in_proj_covar=tensor([0.0540, 0.0640, 0.0567, 0.0674, 0.0664, 0.0612, 0.0566, 0.0647], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 00:41:27,927 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=222885.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:41:30,003 INFO [train.py:901] (3/4) Epoch 28, batch 4650, loss[loss=0.1864, simple_loss=0.2619, pruned_loss=0.05545, over 7657.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2821, pruned_loss=0.05737, over 1614596.34 frames. ], batch size: 19, lr: 2.68e-03, grad_scale: 16.0 +2023-02-09 00:41:50,684 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.622e+02 2.423e+02 3.099e+02 3.500e+02 7.849e+02, threshold=6.198e+02, percent-clipped=6.0 +2023-02-09 00:41:53,068 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=222918.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:42:02,698 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=222932.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:42:06,747 INFO [train.py:901] (3/4) Epoch 28, batch 4700, loss[loss=0.217, simple_loss=0.3069, pruned_loss=0.06356, over 8461.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2825, pruned_loss=0.05756, over 1617570.34 frames. ], batch size: 25, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:42:06,969 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=222938.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:42:10,454 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=222943.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:42:20,652 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=222958.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:42:24,266 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=222963.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:42:40,841 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=222987.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:42:41,326 INFO [train.py:901] (3/4) Epoch 28, batch 4750, loss[loss=0.2025, simple_loss=0.2876, pruned_loss=0.05871, over 8314.00 frames. ], tot_loss[loss=0.199, simple_loss=0.2821, pruned_loss=0.05792, over 1615654.18 frames. ], batch size: 25, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:42:50,003 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=223000.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:42:52,052 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=223002.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:42:54,706 WARNING [train.py:1067] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-02-09 00:42:58,160 WARNING [train.py:1067] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-02-09 00:43:02,872 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.833e+02 2.412e+02 2.807e+02 3.833e+02 7.869e+02, threshold=5.613e+02, percent-clipped=5.0 +2023-02-09 00:43:18,663 INFO [train.py:901] (3/4) Epoch 28, batch 4800, loss[loss=0.1797, simple_loss=0.2638, pruned_loss=0.04782, over 7276.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.2818, pruned_loss=0.0576, over 1611484.39 frames. ], batch size: 16, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:43:25,201 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=223047.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:43:36,315 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6473, 1.5013, 1.7866, 1.4252, 0.9287, 1.5695, 1.5624, 1.5397], + device='cuda:3'), covar=tensor([0.0566, 0.1173, 0.1560, 0.1411, 0.0549, 0.1366, 0.0674, 0.0623], + device='cuda:3'), in_proj_covar=tensor([0.0099, 0.0153, 0.0190, 0.0161, 0.0101, 0.0163, 0.0113, 0.0146], + device='cuda:3'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:3') +2023-02-09 00:43:48,808 WARNING [train.py:1067] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-02-09 00:43:49,675 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=223082.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:43:50,676 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.91 vs. limit=5.0 +2023-02-09 00:43:53,650 INFO [train.py:901] (3/4) Epoch 28, batch 4850, loss[loss=0.1983, simple_loss=0.2893, pruned_loss=0.05371, over 8352.00 frames. ], tot_loss[loss=0.1995, simple_loss=0.2827, pruned_loss=0.05816, over 1612046.21 frames. ], batch size: 24, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:44:13,745 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.627e+02 2.508e+02 3.332e+02 4.408e+02 9.671e+02, threshold=6.663e+02, percent-clipped=7.0 +2023-02-09 00:44:31,161 INFO [train.py:901] (3/4) Epoch 28, batch 4900, loss[loss=0.1877, simple_loss=0.2614, pruned_loss=0.05705, over 8076.00 frames. ], tot_loss[loss=0.1987, simple_loss=0.2821, pruned_loss=0.05768, over 1612339.43 frames. ], batch size: 21, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:44:56,993 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.38 vs. limit=2.0 +2023-02-09 00:45:07,007 INFO [train.py:901] (3/4) Epoch 28, batch 4950, loss[loss=0.1827, simple_loss=0.2766, pruned_loss=0.04438, over 8253.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2825, pruned_loss=0.05761, over 1616423.90 frames. ], batch size: 22, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:45:09,276 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=223191.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 00:45:26,419 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.723e+02 2.334e+02 2.712e+02 3.560e+02 9.309e+02, threshold=5.424e+02, percent-clipped=3.0 +2023-02-09 00:45:42,295 INFO [train.py:901] (3/4) Epoch 28, batch 5000, loss[loss=0.1864, simple_loss=0.2785, pruned_loss=0.0471, over 8134.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.2833, pruned_loss=0.05814, over 1615630.29 frames. ], batch size: 22, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:45:56,166 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=223256.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:46:14,344 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=223281.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:46:18,934 INFO [train.py:901] (3/4) Epoch 28, batch 5050, loss[loss=0.1686, simple_loss=0.2647, pruned_loss=0.03626, over 8193.00 frames. ], tot_loss[loss=0.1998, simple_loss=0.2831, pruned_loss=0.05822, over 1610545.26 frames. ], batch size: 23, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:46:28,803 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=223302.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:46:29,639 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=223303.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:46:32,883 WARNING [train.py:1067] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-02-09 00:46:38,387 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.592e+02 2.274e+02 2.931e+02 3.573e+02 6.090e+02, threshold=5.862e+02, percent-clipped=1.0 +2023-02-09 00:46:46,952 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=223328.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:46:48,963 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=223331.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:46:53,555 INFO [train.py:901] (3/4) Epoch 28, batch 5100, loss[loss=0.1911, simple_loss=0.2751, pruned_loss=0.05357, over 7810.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2822, pruned_loss=0.05773, over 1606437.29 frames. ], batch size: 19, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:46:59,470 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=223346.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:47:31,192 INFO [train.py:901] (3/4) Epoch 28, batch 5150, loss[loss=0.1588, simple_loss=0.2522, pruned_loss=0.03275, over 8042.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.2814, pruned_loss=0.05738, over 1604413.90 frames. ], batch size: 22, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:47:50,495 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.755e+02 2.349e+02 2.964e+02 3.516e+02 1.122e+03, threshold=5.928e+02, percent-clipped=3.0 +2023-02-09 00:47:51,407 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=223417.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:47:57,709 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=223426.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:48:05,968 INFO [train.py:901] (3/4) Epoch 28, batch 5200, loss[loss=0.185, simple_loss=0.273, pruned_loss=0.04854, over 8105.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2815, pruned_loss=0.0576, over 1607095.93 frames. ], batch size: 23, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:48:11,625 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=223446.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:48:22,800 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=223461.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:48:31,223 WARNING [train.py:1067] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-02-09 00:48:44,075 INFO [train.py:901] (3/4) Epoch 28, batch 5250, loss[loss=0.2355, simple_loss=0.3085, pruned_loss=0.08122, over 7649.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2822, pruned_loss=0.05779, over 1605204.47 frames. ], batch size: 19, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:49:03,772 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.671e+02 2.237e+02 2.837e+02 3.561e+02 7.405e+02, threshold=5.674e+02, percent-clipped=6.0 +2023-02-09 00:49:17,137 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=223535.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 00:49:19,096 INFO [train.py:901] (3/4) Epoch 28, batch 5300, loss[loss=0.2079, simple_loss=0.2944, pruned_loss=0.06075, over 8808.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2807, pruned_loss=0.05705, over 1607978.40 frames. ], batch size: 40, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:49:21,387 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=223541.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:49:22,624 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.9290, 1.7014, 2.5662, 1.5817, 2.2078, 2.8653, 2.8779, 2.4548], + device='cuda:3'), covar=tensor([0.0983, 0.1504, 0.0634, 0.1820, 0.1618, 0.0335, 0.0807, 0.0611], + device='cuda:3'), in_proj_covar=tensor([0.0306, 0.0326, 0.0293, 0.0320, 0.0324, 0.0276, 0.0440, 0.0309], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:3') +2023-02-09 00:49:55,535 INFO [train.py:901] (3/4) Epoch 28, batch 5350, loss[loss=0.2133, simple_loss=0.2962, pruned_loss=0.0652, over 8241.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2808, pruned_loss=0.05701, over 1606787.11 frames. ], batch size: 22, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:50:01,804 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=223596.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:50:15,668 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.554e+02 2.359e+02 2.840e+02 3.657e+02 7.209e+02, threshold=5.681e+02, percent-clipped=3.0 +2023-02-09 00:50:30,884 INFO [train.py:901] (3/4) Epoch 28, batch 5400, loss[loss=0.1839, simple_loss=0.2702, pruned_loss=0.04882, over 8191.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2807, pruned_loss=0.05653, over 1609178.84 frames. ], batch size: 23, lr: 2.68e-03, grad_scale: 4.0 +2023-02-09 00:50:39,708 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=223650.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 00:50:53,017 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.1973, 1.3999, 4.3606, 1.7218, 3.9476, 3.6172, 3.9664, 3.8667], + device='cuda:3'), covar=tensor([0.0614, 0.4649, 0.0519, 0.4039, 0.0953, 0.0941, 0.0587, 0.0612], + device='cuda:3'), in_proj_covar=tensor([0.0683, 0.0665, 0.0738, 0.0659, 0.0741, 0.0634, 0.0644, 0.0715], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 00:50:55,915 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=223673.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:51:06,364 INFO [train.py:901] (3/4) Epoch 28, batch 5450, loss[loss=0.1552, simple_loss=0.2349, pruned_loss=0.03772, over 7702.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2803, pruned_loss=0.0565, over 1610974.74 frames. ], batch size: 18, lr: 2.68e-03, grad_scale: 4.0 +2023-02-09 00:51:13,799 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=223698.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:51:16,760 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=223702.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:51:20,202 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.7718, 1.4877, 1.6794, 1.3926, 1.0362, 1.4596, 1.5990, 1.3313], + device='cuda:3'), covar=tensor([0.0608, 0.1305, 0.1698, 0.1558, 0.0586, 0.1515, 0.0738, 0.0742], + device='cuda:3'), in_proj_covar=tensor([0.0100, 0.0154, 0.0191, 0.0162, 0.0102, 0.0164, 0.0113, 0.0147], + device='cuda:3'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:3') +2023-02-09 00:51:28,421 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.718e+02 2.405e+02 2.886e+02 3.694e+02 6.837e+02, threshold=5.773e+02, percent-clipped=3.0 +2023-02-09 00:51:28,657 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=223717.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:51:31,408 WARNING [train.py:1067] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-02-09 00:51:36,518 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=223727.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:51:44,559 INFO [train.py:901] (3/4) Epoch 28, batch 5500, loss[loss=0.1985, simple_loss=0.2887, pruned_loss=0.05418, over 8602.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.28, pruned_loss=0.05632, over 1611712.74 frames. ], batch size: 34, lr: 2.68e-03, grad_scale: 4.0 +2023-02-09 00:51:47,336 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=223742.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:52:01,377 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.46 vs. limit=2.0 +2023-02-09 00:52:18,977 INFO [train.py:901] (3/4) Epoch 28, batch 5550, loss[loss=0.172, simple_loss=0.2691, pruned_loss=0.03743, over 7972.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2796, pruned_loss=0.05622, over 1611421.55 frames. ], batch size: 21, lr: 2.68e-03, grad_scale: 4.0 +2023-02-09 00:52:25,550 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=223797.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:52:39,666 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.770e+02 2.462e+02 3.010e+02 3.574e+02 1.274e+03, threshold=6.020e+02, percent-clipped=3.0 +2023-02-09 00:52:43,457 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=223822.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:52:55,583 INFO [train.py:901] (3/4) Epoch 28, batch 5600, loss[loss=0.1898, simple_loss=0.2792, pruned_loss=0.05019, over 8243.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2792, pruned_loss=0.05619, over 1611747.92 frames. ], batch size: 22, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:53:31,690 INFO [train.py:901] (3/4) Epoch 28, batch 5650, loss[loss=0.2322, simple_loss=0.3083, pruned_loss=0.078, over 7201.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.281, pruned_loss=0.05724, over 1613078.32 frames. ], batch size: 71, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:53:31,873 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.0288, 1.5090, 3.4412, 1.5972, 2.4482, 3.7847, 3.9026, 3.2698], + device='cuda:3'), covar=tensor([0.1122, 0.1812, 0.0314, 0.1980, 0.0990, 0.0208, 0.0510, 0.0469], + device='cuda:3'), in_proj_covar=tensor([0.0307, 0.0327, 0.0294, 0.0322, 0.0325, 0.0277, 0.0441, 0.0309], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:3') +2023-02-09 00:53:41,207 WARNING [train.py:1067] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-02-09 00:53:44,200 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=223906.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 00:53:51,298 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.465e+02 2.313e+02 2.789e+02 3.752e+02 1.102e+03, threshold=5.578e+02, percent-clipped=3.0 +2023-02-09 00:54:01,593 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=223931.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 00:54:06,144 INFO [train.py:901] (3/4) Epoch 28, batch 5700, loss[loss=0.2235, simple_loss=0.3043, pruned_loss=0.07132, over 8662.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2815, pruned_loss=0.05725, over 1617246.84 frames. ], batch size: 34, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:54:07,521 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=223940.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:54:43,225 INFO [train.py:901] (3/4) Epoch 28, batch 5750, loss[loss=0.2308, simple_loss=0.3089, pruned_loss=0.07635, over 8450.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2802, pruned_loss=0.05676, over 1612390.89 frames. ], batch size: 27, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:54:48,696 WARNING [train.py:1067] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-02-09 00:55:04,141 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.665e+02 2.263e+02 2.713e+02 3.241e+02 8.661e+02, threshold=5.425e+02, percent-clipped=3.0 +2023-02-09 00:55:18,763 INFO [train.py:901] (3/4) Epoch 28, batch 5800, loss[loss=0.1758, simple_loss=0.2621, pruned_loss=0.04472, over 8130.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2811, pruned_loss=0.05731, over 1613336.13 frames. ], batch size: 22, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:55:30,597 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=224055.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:55:55,787 INFO [train.py:901] (3/4) Epoch 28, batch 5850, loss[loss=0.1885, simple_loss=0.2678, pruned_loss=0.05454, over 5111.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2821, pruned_loss=0.05769, over 1608033.69 frames. ], batch size: 11, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:56:04,943 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.7099, 1.9548, 2.1050, 1.3985, 2.2289, 1.5524, 0.6866, 1.9316], + device='cuda:3'), covar=tensor([0.0693, 0.0436, 0.0361, 0.0689, 0.0445, 0.0930, 0.0962, 0.0415], + device='cuda:3'), in_proj_covar=tensor([0.0472, 0.0411, 0.0364, 0.0460, 0.0395, 0.0552, 0.0402, 0.0442], + device='cuda:3'), out_proj_covar=tensor([1.2492e-04, 1.0636e-04, 9.4696e-05, 1.2019e-04, 1.0338e-04, 1.5402e-04, + 1.0734e-04, 1.1575e-04], device='cuda:3') +2023-02-09 00:56:14,224 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.38 vs. limit=2.0 +2023-02-09 00:56:15,668 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.578e+02 2.538e+02 3.148e+02 4.118e+02 7.183e+02, threshold=6.296e+02, percent-clipped=12.0 +2023-02-09 00:56:30,206 INFO [train.py:901] (3/4) Epoch 28, batch 5900, loss[loss=0.2277, simple_loss=0.3009, pruned_loss=0.07728, over 8495.00 frames. ], tot_loss[loss=0.1992, simple_loss=0.2827, pruned_loss=0.0579, over 1610932.11 frames. ], batch size: 26, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:56:41,648 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.81 vs. limit=5.0 +2023-02-09 00:57:06,187 INFO [train.py:901] (3/4) Epoch 28, batch 5950, loss[loss=0.1911, simple_loss=0.2848, pruned_loss=0.04869, over 8473.00 frames. ], tot_loss[loss=0.1992, simple_loss=0.2827, pruned_loss=0.05779, over 1613492.54 frames. ], batch size: 25, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:57:28,301 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.818e+02 2.486e+02 3.110e+02 3.888e+02 7.674e+02, threshold=6.220e+02, percent-clipped=4.0 +2023-02-09 00:57:37,732 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=224230.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:57:43,111 INFO [train.py:901] (3/4) Epoch 28, batch 6000, loss[loss=0.2105, simple_loss=0.2875, pruned_loss=0.06674, over 8463.00 frames. ], tot_loss[loss=0.199, simple_loss=0.2825, pruned_loss=0.05772, over 1616149.27 frames. ], batch size: 49, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:57:43,111 INFO [train.py:926] (3/4) Computing validation loss +2023-02-09 00:57:56,798 INFO [train.py:935] (3/4) Epoch 28, validation: loss=0.1714, simple_loss=0.2708, pruned_loss=0.03603, over 944034.00 frames. +2023-02-09 00:57:56,800 INFO [train.py:936] (3/4) Maximum memory allocated so far is 6473MB +2023-02-09 00:58:33,308 INFO [train.py:901] (3/4) Epoch 28, batch 6050, loss[loss=0.1746, simple_loss=0.2525, pruned_loss=0.04831, over 7436.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.282, pruned_loss=0.05735, over 1612830.82 frames. ], batch size: 17, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:58:46,566 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.58 vs. limit=2.0 +2023-02-09 00:58:49,938 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=224311.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:58:54,040 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.623e+02 2.462e+02 3.109e+02 3.867e+02 1.260e+03, threshold=6.217e+02, percent-clipped=5.0 +2023-02-09 00:59:06,133 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.4678, 2.3665, 3.1498, 2.4474, 2.8798, 2.4862, 2.3388, 1.8115], + device='cuda:3'), covar=tensor([0.5818, 0.5434, 0.2091, 0.4334, 0.3036, 0.3251, 0.1987, 0.5826], + device='cuda:3'), in_proj_covar=tensor([0.0974, 0.1037, 0.0842, 0.1008, 0.1031, 0.0941, 0.0779, 0.0860], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 00:59:08,854 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=224336.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 00:59:10,065 INFO [train.py:901] (3/4) Epoch 28, batch 6100, loss[loss=0.2058, simple_loss=0.3046, pruned_loss=0.05353, over 8024.00 frames. ], tot_loss[loss=0.1992, simple_loss=0.2826, pruned_loss=0.05793, over 1613709.60 frames. ], batch size: 22, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 00:59:14,108 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-02-09 00:59:26,298 WARNING [train.py:1067] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-02-09 00:59:46,736 INFO [train.py:901] (3/4) Epoch 28, batch 6150, loss[loss=0.2095, simple_loss=0.295, pruned_loss=0.06197, over 8442.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2808, pruned_loss=0.05695, over 1614410.16 frames. ], batch size: 49, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 01:00:06,959 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.535e+02 2.362e+02 2.823e+02 3.455e+02 8.158e+02, threshold=5.645e+02, percent-clipped=2.0 +2023-02-09 01:00:21,346 INFO [train.py:901] (3/4) Epoch 28, batch 6200, loss[loss=0.2179, simple_loss=0.3007, pruned_loss=0.06753, over 8494.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2812, pruned_loss=0.05716, over 1618999.99 frames. ], batch size: 28, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 01:00:58,236 INFO [train.py:901] (3/4) Epoch 28, batch 6250, loss[loss=0.153, simple_loss=0.2339, pruned_loss=0.0361, over 7812.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2806, pruned_loss=0.05675, over 1616271.38 frames. ], batch size: 19, lr: 2.68e-03, grad_scale: 8.0 +2023-02-09 01:01:18,588 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.516e+02 2.593e+02 3.043e+02 4.250e+02 9.084e+02, threshold=6.087e+02, percent-clipped=11.0 +2023-02-09 01:01:33,265 INFO [train.py:901] (3/4) Epoch 28, batch 6300, loss[loss=0.1737, simple_loss=0.2709, pruned_loss=0.03824, over 8469.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2803, pruned_loss=0.05693, over 1608620.49 frames. ], batch size: 25, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:02:00,127 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=224574.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:02:04,367 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=224580.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:02:10,459 INFO [train.py:901] (3/4) Epoch 28, batch 6350, loss[loss=0.2139, simple_loss=0.301, pruned_loss=0.06337, over 8552.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2808, pruned_loss=0.05729, over 1607505.01 frames. ], batch size: 31, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:02:17,589 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.2724, 3.1776, 3.0200, 1.7064, 2.9494, 2.9517, 2.8724, 2.8410], + device='cuda:3'), covar=tensor([0.1191, 0.0784, 0.1210, 0.4276, 0.1079, 0.1277, 0.1575, 0.1129], + device='cuda:3'), in_proj_covar=tensor([0.0543, 0.0456, 0.0449, 0.0557, 0.0443, 0.0465, 0.0441, 0.0408], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 01:02:30,928 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.559e+02 2.315e+02 2.720e+02 3.259e+02 6.733e+02, threshold=5.440e+02, percent-clipped=2.0 +2023-02-09 01:02:38,891 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6451, 2.5342, 1.9209, 2.2346, 2.1874, 1.5552, 2.0746, 2.1573], + device='cuda:3'), covar=tensor([0.1565, 0.0446, 0.1267, 0.0725, 0.0781, 0.1672, 0.1060, 0.0957], + device='cuda:3'), in_proj_covar=tensor([0.0360, 0.0245, 0.0342, 0.0316, 0.0303, 0.0348, 0.0352, 0.0325], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-02-09 01:02:45,885 INFO [train.py:901] (3/4) Epoch 28, batch 6400, loss[loss=0.2624, simple_loss=0.3483, pruned_loss=0.08828, over 8333.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2805, pruned_loss=0.0571, over 1610525.66 frames. ], batch size: 26, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:03:06,888 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.5304, 1.6777, 2.1162, 1.4015, 1.5305, 1.7827, 1.5549, 1.5514], + device='cuda:3'), covar=tensor([0.2175, 0.2884, 0.1059, 0.5120, 0.2177, 0.3704, 0.2949, 0.2283], + device='cuda:3'), in_proj_covar=tensor([0.0542, 0.0642, 0.0569, 0.0677, 0.0668, 0.0617, 0.0569, 0.0652], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 01:03:21,510 INFO [train.py:901] (3/4) Epoch 28, batch 6450, loss[loss=0.1866, simple_loss=0.281, pruned_loss=0.04611, over 8465.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2809, pruned_loss=0.0574, over 1611785.06 frames. ], batch size: 25, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:03:22,398 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=224689.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:03:43,002 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.578e+02 2.303e+02 2.784e+02 3.485e+02 7.082e+02, threshold=5.567e+02, percent-clipped=7.0 +2023-02-09 01:03:57,597 INFO [train.py:901] (3/4) Epoch 28, batch 6500, loss[loss=0.1733, simple_loss=0.2717, pruned_loss=0.03746, over 8469.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2803, pruned_loss=0.05694, over 1609083.58 frames. ], batch size: 25, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:04:02,105 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=224744.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:04:11,984 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=224758.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:04:32,261 INFO [train.py:901] (3/4) Epoch 28, batch 6550, loss[loss=0.2205, simple_loss=0.3065, pruned_loss=0.06726, over 8501.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2804, pruned_loss=0.05639, over 1612832.73 frames. ], batch size: 26, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:04:43,338 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6815, 1.9183, 1.9342, 1.3405, 1.9841, 1.5177, 0.4405, 1.9809], + device='cuda:3'), covar=tensor([0.0490, 0.0358, 0.0331, 0.0509, 0.0403, 0.0902, 0.0943, 0.0236], + device='cuda:3'), in_proj_covar=tensor([0.0475, 0.0412, 0.0366, 0.0462, 0.0397, 0.0555, 0.0407, 0.0444], + device='cuda:3'), out_proj_covar=tensor([1.2584e-04, 1.0681e-04, 9.5229e-05, 1.2065e-04, 1.0362e-04, 1.5488e-04, + 1.0854e-04, 1.1618e-04], device='cuda:3') +2023-02-09 01:04:47,800 WARNING [train.py:1067] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-02-09 01:04:53,996 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.746e+02 2.489e+02 3.184e+02 3.768e+02 7.222e+02, threshold=6.368e+02, percent-clipped=1.0 +2023-02-09 01:05:08,042 WARNING [train.py:1067] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-02-09 01:05:09,343 INFO [train.py:901] (3/4) Epoch 28, batch 6600, loss[loss=0.2234, simple_loss=0.3067, pruned_loss=0.07005, over 8197.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.2816, pruned_loss=0.05729, over 1614711.91 frames. ], batch size: 23, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:05:14,990 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.8392, 1.6833, 2.1469, 1.7198, 1.1495, 1.8777, 2.5948, 2.2407], + device='cuda:3'), covar=tensor([0.0454, 0.1254, 0.1490, 0.1404, 0.0559, 0.1399, 0.0560, 0.0583], + device='cuda:3'), in_proj_covar=tensor([0.0099, 0.0153, 0.0189, 0.0161, 0.0101, 0.0163, 0.0113, 0.0147], + device='cuda:3'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:3') +2023-02-09 01:05:19,997 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-02-09 01:05:44,396 INFO [train.py:901] (3/4) Epoch 28, batch 6650, loss[loss=0.2344, simple_loss=0.3168, pruned_loss=0.07594, over 8358.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2811, pruned_loss=0.05719, over 1613302.69 frames. ], batch size: 24, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:06:04,774 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.705e+02 2.463e+02 2.971e+02 3.895e+02 9.422e+02, threshold=5.941e+02, percent-clipped=4.0 +2023-02-09 01:06:10,966 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=224924.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:06:21,166 INFO [train.py:901] (3/4) Epoch 28, batch 6700, loss[loss=0.1785, simple_loss=0.2635, pruned_loss=0.04669, over 8455.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2813, pruned_loss=0.05731, over 1615126.87 frames. ], batch size: 25, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:06:22,724 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.5158, 1.3860, 1.8294, 1.1731, 1.1329, 1.8134, 0.2832, 1.1430], + device='cuda:3'), covar=tensor([0.1532, 0.1188, 0.0361, 0.0850, 0.2136, 0.0406, 0.1747, 0.1144], + device='cuda:3'), in_proj_covar=tensor([0.0203, 0.0208, 0.0138, 0.0225, 0.0279, 0.0149, 0.0176, 0.0201], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 01:06:26,241 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=224945.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:06:40,863 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.8344, 1.8549, 2.5479, 1.5049, 1.2384, 2.5644, 0.5586, 1.4957], + device='cuda:3'), covar=tensor([0.1672, 0.1008, 0.0360, 0.1223, 0.2413, 0.0313, 0.1919, 0.1240], + device='cuda:3'), in_proj_covar=tensor([0.0203, 0.0208, 0.0138, 0.0225, 0.0279, 0.0148, 0.0176, 0.0201], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 01:06:43,079 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.1209, 1.9641, 2.5086, 2.1213, 2.3735, 2.2475, 2.1066, 1.3552], + device='cuda:3'), covar=tensor([0.5935, 0.5038, 0.2067, 0.4039, 0.2818, 0.3454, 0.2016, 0.5481], + device='cuda:3'), in_proj_covar=tensor([0.0970, 0.1035, 0.0839, 0.1001, 0.1028, 0.0939, 0.0776, 0.0855], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 01:06:44,443 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=224970.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:06:57,021 INFO [train.py:901] (3/4) Epoch 28, batch 6750, loss[loss=0.2433, simple_loss=0.3292, pruned_loss=0.0787, over 8448.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.282, pruned_loss=0.05751, over 1613819.79 frames. ], batch size: 27, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:07:17,011 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.474e+02 2.343e+02 2.979e+02 3.883e+02 6.136e+02, threshold=5.958e+02, percent-clipped=2.0 +2023-02-09 01:07:28,015 WARNING [train.py:1067] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-02-09 01:07:32,092 INFO [train.py:901] (3/4) Epoch 28, batch 6800, loss[loss=0.2596, simple_loss=0.3311, pruned_loss=0.09402, over 6986.00 frames. ], tot_loss[loss=0.1983, simple_loss=0.2821, pruned_loss=0.05724, over 1610629.21 frames. ], batch size: 71, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:07:32,972 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=225039.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:08:08,381 INFO [train.py:901] (3/4) Epoch 28, batch 6850, loss[loss=0.1969, simple_loss=0.285, pruned_loss=0.05445, over 8535.00 frames. ], tot_loss[loss=0.1987, simple_loss=0.2823, pruned_loss=0.0575, over 1613129.74 frames. ], batch size: 34, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:08:08,460 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=225088.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:08:18,037 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=225102.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:08:18,710 WARNING [train.py:1067] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-02-09 01:08:28,498 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.612e+02 2.283e+02 2.996e+02 3.907e+02 8.918e+02, threshold=5.992e+02, percent-clipped=3.0 +2023-02-09 01:08:31,367 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=225121.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:08:32,732 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.6284, 1.4062, 3.9215, 1.4958, 3.2373, 3.0999, 3.6073, 3.5555], + device='cuda:3'), covar=tensor([0.1248, 0.6181, 0.1136, 0.5337, 0.2117, 0.1861, 0.0968, 0.1036], + device='cuda:3'), in_proj_covar=tensor([0.0686, 0.0667, 0.0738, 0.0662, 0.0746, 0.0635, 0.0645, 0.0722], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 01:08:42,880 INFO [train.py:901] (3/4) Epoch 28, batch 6900, loss[loss=0.2545, simple_loss=0.3227, pruned_loss=0.09315, over 6876.00 frames. ], tot_loss[loss=0.1995, simple_loss=0.2829, pruned_loss=0.05804, over 1608315.06 frames. ], batch size: 71, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:08:58,671 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=225160.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:09:18,981 INFO [train.py:901] (3/4) Epoch 28, batch 6950, loss[loss=0.1765, simple_loss=0.2581, pruned_loss=0.04744, over 8085.00 frames. ], tot_loss[loss=0.1999, simple_loss=0.2833, pruned_loss=0.05823, over 1610132.81 frames. ], batch size: 21, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:09:30,339 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=225203.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:09:30,868 WARNING [train.py:1067] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-02-09 01:09:40,021 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.583e+02 2.456e+02 2.946e+02 3.977e+02 8.721e+02, threshold=5.892e+02, percent-clipped=6.0 +2023-02-09 01:09:40,234 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=225217.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:09:54,451 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.54 vs. limit=2.0 +2023-02-09 01:09:54,776 INFO [train.py:901] (3/4) Epoch 28, batch 7000, loss[loss=0.2166, simple_loss=0.2845, pruned_loss=0.07431, over 7780.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2822, pruned_loss=0.05733, over 1611616.22 frames. ], batch size: 19, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:10:31,226 INFO [train.py:901] (3/4) Epoch 28, batch 7050, loss[loss=0.2068, simple_loss=0.2713, pruned_loss=0.07114, over 7819.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2812, pruned_loss=0.05688, over 1610760.81 frames. ], batch size: 20, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:10:36,349 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=225295.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:10:52,626 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.739e+02 2.402e+02 2.844e+02 3.449e+02 6.425e+02, threshold=5.688e+02, percent-clipped=2.0 +2023-02-09 01:10:55,666 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=225320.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:11:08,193 INFO [train.py:901] (3/4) Epoch 28, batch 7100, loss[loss=0.1782, simple_loss=0.2462, pruned_loss=0.05512, over 7252.00 frames. ], tot_loss[loss=0.1961, simple_loss=0.28, pruned_loss=0.05612, over 1608729.12 frames. ], batch size: 16, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:11:43,041 INFO [train.py:901] (3/4) Epoch 28, batch 7150, loss[loss=0.1828, simple_loss=0.2739, pruned_loss=0.04581, over 8472.00 frames. ], tot_loss[loss=0.1956, simple_loss=0.2795, pruned_loss=0.05585, over 1609379.06 frames. ], batch size: 25, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:11:54,513 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([5.5798, 1.5734, 5.7089, 2.2861, 5.1844, 4.7798, 5.2434, 5.1505], + device='cuda:3'), covar=tensor([0.0502, 0.4955, 0.0437, 0.4003, 0.0946, 0.0847, 0.0530, 0.0528], + device='cuda:3'), in_proj_covar=tensor([0.0684, 0.0667, 0.0736, 0.0662, 0.0747, 0.0636, 0.0646, 0.0721], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 01:12:05,408 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.501e+02 2.377e+02 2.906e+02 3.542e+02 6.036e+02, threshold=5.811e+02, percent-clipped=2.0 +2023-02-09 01:12:21,602 INFO [train.py:901] (3/4) Epoch 28, batch 7200, loss[loss=0.1825, simple_loss=0.2693, pruned_loss=0.04785, over 8518.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2805, pruned_loss=0.05684, over 1611028.58 frames. ], batch size: 26, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:12:35,183 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=225457.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:12:36,589 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=225459.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:12:40,499 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=225465.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:12:45,517 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.6849, 4.6901, 4.1991, 2.0615, 4.0667, 4.2950, 4.1654, 4.1283], + device='cuda:3'), covar=tensor([0.0598, 0.0445, 0.0883, 0.4423, 0.0886, 0.0802, 0.1126, 0.0815], + device='cuda:3'), in_proj_covar=tensor([0.0544, 0.0459, 0.0452, 0.0557, 0.0444, 0.0466, 0.0444, 0.0411], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 01:12:46,209 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=225473.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:12:51,110 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=225480.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:12:53,936 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=225484.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:12:56,554 INFO [train.py:901] (3/4) Epoch 28, batch 7250, loss[loss=0.172, simple_loss=0.2657, pruned_loss=0.0391, over 8228.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2795, pruned_loss=0.0563, over 1608501.82 frames. ], batch size: 24, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:13:03,606 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=225498.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:13:07,607 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=225504.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:13:13,959 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.4189, 2.3085, 3.0288, 2.4803, 2.9599, 2.5470, 2.3416, 1.9032], + device='cuda:3'), covar=tensor([0.5964, 0.5413, 0.2292, 0.4463, 0.3014, 0.3259, 0.1918, 0.5979], + device='cuda:3'), in_proj_covar=tensor([0.0970, 0.1034, 0.0842, 0.1001, 0.1029, 0.0938, 0.0776, 0.0856], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 01:13:16,362 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.440e+02 2.486e+02 3.022e+02 3.617e+02 8.325e+02, threshold=6.044e+02, percent-clipped=6.0 +2023-02-09 01:13:21,986 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=225523.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:13:29,593 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.79 vs. limit=5.0 +2023-02-09 01:13:32,776 INFO [train.py:901] (3/4) Epoch 28, batch 7300, loss[loss=0.2047, simple_loss=0.2926, pruned_loss=0.05838, over 8526.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2807, pruned_loss=0.05693, over 1612851.94 frames. ], batch size: 28, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:13:34,179 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=225540.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:14:00,402 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=225577.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 01:14:02,411 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=225580.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:14:07,681 INFO [train.py:901] (3/4) Epoch 28, batch 7350, loss[loss=0.1815, simple_loss=0.2779, pruned_loss=0.04253, over 8294.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2804, pruned_loss=0.05695, over 1612369.29 frames. ], batch size: 23, lr: 2.67e-03, grad_scale: 8.0 +2023-02-09 01:14:24,556 WARNING [train.py:1067] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-02-09 01:14:27,897 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.502e+02 2.380e+02 2.753e+02 3.463e+02 7.224e+02, threshold=5.506e+02, percent-clipped=3.0 +2023-02-09 01:14:29,530 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=225619.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:14:42,948 INFO [train.py:901] (3/4) Epoch 28, batch 7400, loss[loss=0.1796, simple_loss=0.2644, pruned_loss=0.0474, over 7922.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2808, pruned_loss=0.05712, over 1613569.94 frames. ], batch size: 20, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:14:42,964 WARNING [train.py:1067] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-02-09 01:14:57,030 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6498, 1.8696, 2.6223, 1.5318, 2.1561, 1.8919, 1.7198, 2.0704], + device='cuda:3'), covar=tensor([0.1507, 0.2213, 0.0701, 0.3651, 0.1453, 0.2675, 0.1941, 0.2104], + device='cuda:3'), in_proj_covar=tensor([0.0539, 0.0638, 0.0562, 0.0673, 0.0664, 0.0614, 0.0564, 0.0648], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 01:15:19,326 INFO [train.py:901] (3/4) Epoch 28, batch 7450, loss[loss=0.208, simple_loss=0.2898, pruned_loss=0.06312, over 8355.00 frames. ], tot_loss[loss=0.1982, simple_loss=0.2814, pruned_loss=0.05746, over 1608405.78 frames. ], batch size: 24, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:15:25,008 WARNING [train.py:1067] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-02-09 01:15:40,012 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.685e+02 2.396e+02 3.007e+02 3.866e+02 7.466e+02, threshold=6.014e+02, percent-clipped=6.0 +2023-02-09 01:15:54,482 INFO [train.py:901] (3/4) Epoch 28, batch 7500, loss[loss=0.1804, simple_loss=0.2733, pruned_loss=0.04381, over 8320.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2819, pruned_loss=0.05789, over 1604934.14 frames. ], batch size: 25, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:16:32,496 INFO [train.py:901] (3/4) Epoch 28, batch 7550, loss[loss=0.1853, simple_loss=0.2548, pruned_loss=0.05794, over 7711.00 frames. ], tot_loss[loss=0.1982, simple_loss=0.2811, pruned_loss=0.05766, over 1603911.87 frames. ], batch size: 18, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:16:41,873 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=225801.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:16:52,716 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.729e+02 2.388e+02 3.127e+02 4.485e+02 1.321e+03, threshold=6.254e+02, percent-clipped=11.0 +2023-02-09 01:16:57,642 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=225824.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:17:05,959 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=225836.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:17:07,134 INFO [train.py:901] (3/4) Epoch 28, batch 7600, loss[loss=0.1703, simple_loss=0.2586, pruned_loss=0.04101, over 8249.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2806, pruned_loss=0.05711, over 1602397.61 frames. ], batch size: 22, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:17:17,014 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.1068, 3.4378, 2.2309, 2.9044, 2.8719, 2.0388, 2.8748, 3.2032], + device='cuda:3'), covar=tensor([0.1855, 0.0425, 0.1279, 0.0827, 0.0760, 0.1628, 0.1062, 0.1065], + device='cuda:3'), in_proj_covar=tensor([0.0358, 0.0246, 0.0342, 0.0314, 0.0302, 0.0347, 0.0349, 0.0322], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-02-09 01:17:23,290 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=225861.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:17:24,625 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.0250, 1.6514, 1.5064, 1.5346, 1.3386, 1.3954, 1.3136, 1.2879], + device='cuda:3'), covar=tensor([0.1200, 0.0501, 0.1311, 0.0614, 0.0847, 0.1528, 0.0930, 0.0826], + device='cuda:3'), in_proj_covar=tensor([0.0358, 0.0246, 0.0342, 0.0314, 0.0302, 0.0347, 0.0349, 0.0322], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-02-09 01:17:27,276 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=225867.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:17:34,160 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=225875.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:17:35,175 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.46 vs. limit=2.0 +2023-02-09 01:17:37,638 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=225880.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:17:40,905 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=225884.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:17:43,473 INFO [train.py:901] (3/4) Epoch 28, batch 7650, loss[loss=0.2061, simple_loss=0.3134, pruned_loss=0.04943, over 8520.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2806, pruned_loss=0.05709, over 1603257.22 frames. ], batch size: 28, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:17:51,607 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=225900.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:18:01,182 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=225913.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:18:03,359 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=225916.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:18:03,830 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.518e+02 2.353e+02 2.789e+02 3.444e+02 7.654e+02, threshold=5.579e+02, percent-clipped=1.0 +2023-02-09 01:18:06,735 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=225921.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 01:18:18,652 INFO [train.py:901] (3/4) Epoch 28, batch 7700, loss[loss=0.1972, simple_loss=0.2811, pruned_loss=0.05664, over 7702.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2807, pruned_loss=0.05746, over 1597323.86 frames. ], batch size: 18, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:18:19,540 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=225939.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:18:44,242 WARNING [train.py:1067] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-02-09 01:18:49,252 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=225982.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:18:53,155 INFO [train.py:901] (3/4) Epoch 28, batch 7750, loss[loss=0.1816, simple_loss=0.263, pruned_loss=0.05008, over 7518.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2802, pruned_loss=0.05672, over 1601286.02 frames. ], batch size: 18, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:19:01,896 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=225999.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:19:15,720 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.697e+02 2.439e+02 2.815e+02 3.514e+02 7.333e+02, threshold=5.630e+02, percent-clipped=1.0 +2023-02-09 01:19:29,680 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=226036.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 01:19:30,737 INFO [train.py:901] (3/4) Epoch 28, batch 7800, loss[loss=0.2559, simple_loss=0.3301, pruned_loss=0.09089, over 8666.00 frames. ], tot_loss[loss=0.1986, simple_loss=0.2817, pruned_loss=0.05776, over 1603203.27 frames. ], batch size: 34, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:19:45,520 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=226059.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:20:01,006 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.2476, 2.0663, 2.6397, 2.2530, 2.6588, 2.3387, 2.1954, 1.5463], + device='cuda:3'), covar=tensor([0.5915, 0.5314, 0.2249, 0.4081, 0.2506, 0.3289, 0.1953, 0.5534], + device='cuda:3'), in_proj_covar=tensor([0.0967, 0.1033, 0.0839, 0.1001, 0.1028, 0.0938, 0.0775, 0.0856], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 01:20:05,562 INFO [train.py:901] (3/4) Epoch 28, batch 7850, loss[loss=0.2099, simple_loss=0.2968, pruned_loss=0.06144, over 8587.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2814, pruned_loss=0.05714, over 1605844.33 frames. ], batch size: 31, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:20:25,277 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.578e+02 2.405e+02 3.030e+02 3.828e+02 1.208e+03, threshold=6.060e+02, percent-clipped=4.0 +2023-02-09 01:20:39,671 INFO [train.py:901] (3/4) Epoch 28, batch 7900, loss[loss=0.2396, simple_loss=0.3155, pruned_loss=0.08185, over 8602.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2822, pruned_loss=0.05734, over 1608888.16 frames. ], batch size: 31, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:21:02,884 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=226172.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:21:13,381 INFO [train.py:901] (3/4) Epoch 28, batch 7950, loss[loss=0.2279, simple_loss=0.303, pruned_loss=0.07644, over 8292.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2813, pruned_loss=0.0569, over 1608347.19 frames. ], batch size: 23, lr: 2.67e-03, grad_scale: 16.0 +2023-02-09 01:21:18,310 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=226195.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:21:19,708 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=226197.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:21:33,035 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.821e+02 2.474e+02 2.920e+02 3.612e+02 7.690e+02, threshold=5.839e+02, percent-clipped=4.0 +2023-02-09 01:21:35,322 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=226220.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:21:37,938 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=226224.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:21:47,691 INFO [train.py:901] (3/4) Epoch 28, batch 8000, loss[loss=0.1662, simple_loss=0.2437, pruned_loss=0.04433, over 7232.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2816, pruned_loss=0.05681, over 1612402.02 frames. ], batch size: 16, lr: 2.66e-03, grad_scale: 16.0 +2023-02-09 01:21:47,896 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=226238.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:21:57,011 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=226251.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:21:59,709 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=226255.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:22:00,950 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=226257.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:22:05,095 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=226263.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:22:17,272 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=226280.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:22:22,556 INFO [train.py:901] (3/4) Epoch 28, batch 8050, loss[loss=0.1461, simple_loss=0.2276, pruned_loss=0.03232, over 7298.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2793, pruned_loss=0.05632, over 1598361.91 frames. ], batch size: 16, lr: 2.66e-03, grad_scale: 16.0 +2023-02-09 01:22:25,399 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=226292.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 01:22:42,674 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.612e+02 2.434e+02 3.095e+02 3.696e+02 6.520e+02, threshold=6.190e+02, percent-clipped=3.0 +2023-02-09 01:22:42,879 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=226317.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 01:22:57,839 WARNING [train.py:1067] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-02-09 01:23:01,698 INFO [train.py:901] (3/4) Epoch 29, batch 0, loss[loss=0.1951, simple_loss=0.2822, pruned_loss=0.05399, over 8024.00 frames. ], tot_loss[loss=0.1951, simple_loss=0.2822, pruned_loss=0.05399, over 8024.00 frames. ], batch size: 22, lr: 2.62e-03, grad_scale: 16.0 +2023-02-09 01:23:01,698 INFO [train.py:926] (3/4) Computing validation loss +2023-02-09 01:23:13,264 INFO [train.py:935] (3/4) Epoch 29, validation: loss=0.1705, simple_loss=0.2705, pruned_loss=0.03528, over 944034.00 frames. +2023-02-09 01:23:13,265 INFO [train.py:936] (3/4) Maximum memory allocated so far is 6473MB +2023-02-09 01:23:26,224 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=226339.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:23:29,645 WARNING [train.py:1067] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-02-09 01:23:39,592 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.90 vs. limit=5.0 +2023-02-09 01:23:49,917 INFO [train.py:901] (3/4) Epoch 29, batch 50, loss[loss=0.2171, simple_loss=0.295, pruned_loss=0.0696, over 8445.00 frames. ], tot_loss[loss=0.2, simple_loss=0.2841, pruned_loss=0.05795, over 363541.11 frames. ], batch size: 27, lr: 2.62e-03, grad_scale: 16.0 +2023-02-09 01:23:50,828 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=226372.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:23:53,661 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.0888, 1.2399, 1.1828, 0.8399, 1.2110, 1.0172, 0.1414, 1.2262], + device='cuda:3'), covar=tensor([0.0474, 0.0431, 0.0411, 0.0576, 0.0486, 0.1056, 0.0953, 0.0351], + device='cuda:3'), in_proj_covar=tensor([0.0478, 0.0414, 0.0368, 0.0462, 0.0398, 0.0555, 0.0406, 0.0443], + device='cuda:3'), out_proj_covar=tensor([1.2657e-04, 1.0724e-04, 9.5925e-05, 1.2073e-04, 1.0402e-04, 1.5462e-04, + 1.0826e-04, 1.1606e-04], device='cuda:3') +2023-02-09 01:24:06,018 WARNING [train.py:1067] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-02-09 01:24:12,516 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=226403.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:24:22,929 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.399e+02 2.293e+02 2.936e+02 3.721e+02 6.222e+02, threshold=5.872e+02, percent-clipped=1.0 +2023-02-09 01:24:25,776 INFO [train.py:901] (3/4) Epoch 29, batch 100, loss[loss=0.1612, simple_loss=0.2382, pruned_loss=0.04215, over 8036.00 frames. ], tot_loss[loss=0.201, simple_loss=0.2852, pruned_loss=0.05841, over 646663.15 frames. ], batch size: 22, lr: 2.62e-03, grad_scale: 16.0 +2023-02-09 01:24:30,616 WARNING [train.py:1067] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-02-09 01:24:33,810 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.3404, 2.0482, 2.5304, 2.1533, 2.4782, 2.3456, 2.2185, 1.4171], + device='cuda:3'), covar=tensor([0.5481, 0.4832, 0.2208, 0.4135, 0.2525, 0.3131, 0.1898, 0.5395], + device='cuda:3'), in_proj_covar=tensor([0.0969, 0.1034, 0.0840, 0.1003, 0.1030, 0.0940, 0.0776, 0.0858], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 01:25:02,686 INFO [train.py:901] (3/4) Epoch 29, batch 150, loss[loss=0.1757, simple_loss=0.2646, pruned_loss=0.04338, over 8256.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.283, pruned_loss=0.05701, over 861909.54 frames. ], batch size: 24, lr: 2.62e-03, grad_scale: 16.0 +2023-02-09 01:25:34,578 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.592e+02 2.438e+02 2.916e+02 4.111e+02 7.524e+02, threshold=5.832e+02, percent-clipped=2.0 +2023-02-09 01:25:35,534 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=226518.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:25:37,497 INFO [train.py:901] (3/4) Epoch 29, batch 200, loss[loss=0.2152, simple_loss=0.2947, pruned_loss=0.06783, over 8444.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.2835, pruned_loss=0.05765, over 1033869.09 frames. ], batch size: 29, lr: 2.62e-03, grad_scale: 16.0 +2023-02-09 01:26:12,528 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.5851, 1.6624, 2.0940, 1.6425, 0.9810, 1.6717, 2.1241, 2.2030], + device='cuda:3'), covar=tensor([0.0487, 0.1176, 0.1516, 0.1380, 0.0602, 0.1396, 0.0650, 0.0581], + device='cuda:3'), in_proj_covar=tensor([0.0099, 0.0153, 0.0189, 0.0161, 0.0102, 0.0163, 0.0113, 0.0147], + device='cuda:3'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:3') +2023-02-09 01:26:12,614 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.4069, 2.3460, 3.0723, 2.4468, 2.9403, 2.5693, 2.4642, 2.0684], + device='cuda:3'), covar=tensor([0.5876, 0.5170, 0.2132, 0.4836, 0.3128, 0.3160, 0.1856, 0.5815], + device='cuda:3'), in_proj_covar=tensor([0.0965, 0.1030, 0.0837, 0.0999, 0.1025, 0.0935, 0.0773, 0.0852], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 01:26:13,673 INFO [train.py:901] (3/4) Epoch 29, batch 250, loss[loss=0.2481, simple_loss=0.3333, pruned_loss=0.08142, over 8506.00 frames. ], tot_loss[loss=0.2005, simple_loss=0.2847, pruned_loss=0.05819, over 1169184.76 frames. ], batch size: 26, lr: 2.62e-03, grad_scale: 16.0 +2023-02-09 01:26:26,060 WARNING [train.py:1067] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-02-09 01:26:31,160 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=226595.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:26:31,310 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=226595.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:26:33,827 WARNING [train.py:1067] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-02-09 01:26:46,401 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.753e+02 2.514e+02 3.003e+02 3.645e+02 8.891e+02, threshold=6.006e+02, percent-clipped=9.0 +2023-02-09 01:26:48,784 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=226620.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:26:49,255 INFO [train.py:901] (3/4) Epoch 29, batch 300, loss[loss=0.2028, simple_loss=0.2794, pruned_loss=0.06311, over 7926.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.284, pruned_loss=0.05806, over 1269378.88 frames. ], batch size: 20, lr: 2.62e-03, grad_scale: 16.0 +2023-02-09 01:26:54,305 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=226628.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:27:12,772 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=226653.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:27:25,799 INFO [train.py:901] (3/4) Epoch 29, batch 350, loss[loss=0.19, simple_loss=0.286, pruned_loss=0.04699, over 8467.00 frames. ], tot_loss[loss=0.2006, simple_loss=0.2846, pruned_loss=0.05829, over 1347240.17 frames. ], batch size: 25, lr: 2.62e-03, grad_scale: 16.0 +2023-02-09 01:27:45,005 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.2254, 2.6196, 2.7870, 1.5549, 3.2960, 1.8669, 1.5887, 2.2208], + device='cuda:3'), covar=tensor([0.0921, 0.0477, 0.0385, 0.1027, 0.0514, 0.1071, 0.1055, 0.0738], + device='cuda:3'), in_proj_covar=tensor([0.0483, 0.0418, 0.0372, 0.0466, 0.0401, 0.0559, 0.0408, 0.0447], + device='cuda:3'), out_proj_covar=tensor([1.2785e-04, 1.0831e-04, 9.7018e-05, 1.2160e-04, 1.0475e-04, 1.5584e-04, + 1.0894e-04, 1.1705e-04], device='cuda:3') +2023-02-09 01:27:46,744 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.4178, 1.6234, 2.0476, 1.3625, 1.4244, 1.6989, 1.4793, 1.4819], + device='cuda:3'), covar=tensor([0.2012, 0.2677, 0.1151, 0.4678, 0.2211, 0.3539, 0.2624, 0.2318], + device='cuda:3'), in_proj_covar=tensor([0.0540, 0.0637, 0.0562, 0.0671, 0.0664, 0.0614, 0.0564, 0.0646], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 01:27:54,192 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=226710.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:27:58,935 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.666e+02 2.400e+02 2.862e+02 3.557e+02 6.632e+02, threshold=5.725e+02, percent-clipped=2.0 +2023-02-09 01:28:01,716 INFO [train.py:901] (3/4) Epoch 29, batch 400, loss[loss=0.1489, simple_loss=0.2251, pruned_loss=0.03636, over 7203.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2815, pruned_loss=0.05659, over 1405216.33 frames. ], batch size: 16, lr: 2.61e-03, grad_scale: 16.0 +2023-02-09 01:28:37,562 INFO [train.py:901] (3/4) Epoch 29, batch 450, loss[loss=0.2331, simple_loss=0.3137, pruned_loss=0.0763, over 8499.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2818, pruned_loss=0.05688, over 1453364.50 frames. ], batch size: 26, lr: 2.61e-03, grad_scale: 16.0 +2023-02-09 01:28:38,565 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.19 vs. limit=5.0 +2023-02-09 01:28:39,871 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=226774.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:28:53,563 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.6752, 2.1627, 3.5492, 1.8182, 1.7358, 3.5356, 0.5984, 2.2267], + device='cuda:3'), covar=tensor([0.1203, 0.1227, 0.0204, 0.1380, 0.2313, 0.0353, 0.2008, 0.1134], + device='cuda:3'), in_proj_covar=tensor([0.0202, 0.0208, 0.0138, 0.0224, 0.0280, 0.0148, 0.0174, 0.0200], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 01:28:56,591 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.35 vs. limit=2.0 +2023-02-09 01:28:58,207 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=226799.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:29:11,123 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.602e+02 2.449e+02 2.964e+02 3.856e+02 9.700e+02, threshold=5.929e+02, percent-clipped=9.0 +2023-02-09 01:29:13,785 INFO [train.py:901] (3/4) Epoch 29, batch 500, loss[loss=0.2069, simple_loss=0.3035, pruned_loss=0.05512, over 8096.00 frames. ], tot_loss[loss=0.2003, simple_loss=0.2839, pruned_loss=0.05829, over 1493689.32 frames. ], batch size: 23, lr: 2.61e-03, grad_scale: 16.0 +2023-02-09 01:29:48,207 INFO [train.py:901] (3/4) Epoch 29, batch 550, loss[loss=0.1894, simple_loss=0.276, pruned_loss=0.05142, over 8461.00 frames. ], tot_loss[loss=0.2012, simple_loss=0.2845, pruned_loss=0.05901, over 1521808.05 frames. ], batch size: 27, lr: 2.61e-03, grad_scale: 16.0 +2023-02-09 01:29:51,494 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-02-09 01:30:21,915 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.374e+02 2.486e+02 3.156e+02 4.092e+02 1.034e+03, threshold=6.313e+02, percent-clipped=6.0 +2023-02-09 01:30:24,629 INFO [train.py:901] (3/4) Epoch 29, batch 600, loss[loss=0.1813, simple_loss=0.2622, pruned_loss=0.05027, over 7202.00 frames. ], tot_loss[loss=0.2009, simple_loss=0.284, pruned_loss=0.05887, over 1538431.20 frames. ], batch size: 16, lr: 2.61e-03, grad_scale: 16.0 +2023-02-09 01:30:28,544 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.16 vs. limit=5.0 +2023-02-09 01:30:35,115 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.2652, 1.5441, 3.4314, 1.5177, 2.4533, 3.7300, 3.8785, 3.2660], + device='cuda:3'), covar=tensor([0.1025, 0.1875, 0.0365, 0.2214, 0.1173, 0.0242, 0.0534, 0.0488], + device='cuda:3'), in_proj_covar=tensor([0.0310, 0.0327, 0.0295, 0.0325, 0.0327, 0.0278, 0.0444, 0.0309], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:3') +2023-02-09 01:30:38,689 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.3377, 2.2167, 2.7488, 2.4044, 2.7458, 2.3821, 2.2299, 1.7625], + device='cuda:3'), covar=tensor([0.5493, 0.5182, 0.2149, 0.3754, 0.2536, 0.3249, 0.1917, 0.5201], + device='cuda:3'), in_proj_covar=tensor([0.0965, 0.1033, 0.0838, 0.0999, 0.1026, 0.0936, 0.0772, 0.0852], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 01:30:43,053 WARNING [train.py:1067] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-02-09 01:30:56,508 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=226966.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:30:59,759 INFO [train.py:901] (3/4) Epoch 29, batch 650, loss[loss=0.1866, simple_loss=0.2604, pruned_loss=0.05641, over 7667.00 frames. ], tot_loss[loss=0.2, simple_loss=0.283, pruned_loss=0.05853, over 1557331.46 frames. ], batch size: 19, lr: 2.61e-03, grad_scale: 16.0 +2023-02-09 01:31:13,841 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=226991.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:31:25,429 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=227007.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:31:32,762 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.693e+02 2.425e+02 2.942e+02 3.859e+02 6.314e+02, threshold=5.885e+02, percent-clipped=1.0 +2023-02-09 01:31:36,256 INFO [train.py:901] (3/4) Epoch 29, batch 700, loss[loss=0.1828, simple_loss=0.269, pruned_loss=0.04833, over 8132.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2818, pruned_loss=0.05755, over 1566709.54 frames. ], batch size: 22, lr: 2.61e-03, grad_scale: 16.0 +2023-02-09 01:31:56,846 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.9297, 2.1983, 1.8084, 3.0759, 2.0406, 1.9062, 2.2135, 2.3798], + device='cuda:3'), covar=tensor([0.1250, 0.1099, 0.1527, 0.0431, 0.1025, 0.1542, 0.0859, 0.0888], + device='cuda:3'), in_proj_covar=tensor([0.0231, 0.0196, 0.0245, 0.0214, 0.0204, 0.0249, 0.0251, 0.0206], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:3') +2023-02-09 01:32:12,699 INFO [train.py:901] (3/4) Epoch 29, batch 750, loss[loss=0.2001, simple_loss=0.293, pruned_loss=0.05363, over 8311.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2809, pruned_loss=0.05737, over 1575170.98 frames. ], batch size: 25, lr: 2.61e-03, grad_scale: 16.0 +2023-02-09 01:32:31,397 WARNING [train.py:1067] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-02-09 01:32:40,262 WARNING [train.py:1067] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-02-09 01:32:44,348 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.694e+02 2.547e+02 3.055e+02 4.023e+02 1.198e+03, threshold=6.109e+02, percent-clipped=3.0 +2023-02-09 01:32:47,755 INFO [train.py:901] (3/4) Epoch 29, batch 800, loss[loss=0.169, simple_loss=0.2503, pruned_loss=0.04387, over 7920.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2819, pruned_loss=0.05795, over 1582414.66 frames. ], batch size: 20, lr: 2.61e-03, grad_scale: 16.0 +2023-02-09 01:33:24,592 INFO [train.py:901] (3/4) Epoch 29, batch 850, loss[loss=0.1708, simple_loss=0.2419, pruned_loss=0.04991, over 7552.00 frames. ], tot_loss[loss=0.1982, simple_loss=0.2813, pruned_loss=0.0576, over 1587660.51 frames. ], batch size: 18, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:33:57,024 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.572e+02 2.463e+02 2.886e+02 3.949e+02 8.845e+02, threshold=5.773e+02, percent-clipped=3.0 +2023-02-09 01:33:59,146 INFO [train.py:901] (3/4) Epoch 29, batch 900, loss[loss=0.2075, simple_loss=0.2856, pruned_loss=0.06474, over 8470.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2803, pruned_loss=0.05721, over 1591229.17 frames. ], batch size: 25, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:34:35,609 INFO [train.py:901] (3/4) Epoch 29, batch 950, loss[loss=0.2124, simple_loss=0.2916, pruned_loss=0.06658, over 8625.00 frames. ], tot_loss[loss=0.1992, simple_loss=0.282, pruned_loss=0.0582, over 1602537.10 frames. ], batch size: 31, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:35:04,851 WARNING [train.py:1067] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-02-09 01:35:08,996 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.642e+02 2.581e+02 3.033e+02 3.905e+02 1.035e+03, threshold=6.066e+02, percent-clipped=4.0 +2023-02-09 01:35:11,166 INFO [train.py:901] (3/4) Epoch 29, batch 1000, loss[loss=0.2059, simple_loss=0.2843, pruned_loss=0.06378, over 7969.00 frames. ], tot_loss[loss=0.1992, simple_loss=0.2823, pruned_loss=0.05803, over 1606033.10 frames. ], batch size: 21, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:35:32,312 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=227351.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:35:39,814 WARNING [train.py:1067] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-02-09 01:35:47,242 INFO [train.py:901] (3/4) Epoch 29, batch 1050, loss[loss=0.2824, simple_loss=0.3549, pruned_loss=0.105, over 8598.00 frames. ], tot_loss[loss=0.1992, simple_loss=0.2824, pruned_loss=0.05796, over 1608084.84 frames. ], batch size: 34, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:35:52,703 WARNING [train.py:1067] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-02-09 01:36:22,043 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.565e+02 2.373e+02 3.020e+02 3.651e+02 1.051e+03, threshold=6.040e+02, percent-clipped=1.0 +2023-02-09 01:36:23,013 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.4810, 1.9876, 3.0609, 1.7259, 1.6753, 3.0068, 0.9627, 2.1739], + device='cuda:3'), covar=tensor([0.1298, 0.1212, 0.0245, 0.1177, 0.2153, 0.0314, 0.1762, 0.1132], + device='cuda:3'), in_proj_covar=tensor([0.0202, 0.0207, 0.0137, 0.0224, 0.0278, 0.0147, 0.0173, 0.0199], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 01:36:24,309 INFO [train.py:901] (3/4) Epoch 29, batch 1100, loss[loss=0.183, simple_loss=0.2707, pruned_loss=0.04765, over 8462.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2813, pruned_loss=0.05725, over 1606817.24 frames. ], batch size: 25, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:36:30,382 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=227429.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:36:56,362 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=227466.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:36:59,078 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.7459, 1.4991, 1.7733, 1.3948, 0.9592, 1.5256, 1.6187, 1.5775], + device='cuda:3'), covar=tensor([0.0554, 0.1206, 0.1637, 0.1457, 0.0582, 0.1450, 0.0718, 0.0667], + device='cuda:3'), in_proj_covar=tensor([0.0100, 0.0154, 0.0191, 0.0163, 0.0102, 0.0164, 0.0114, 0.0147], + device='cuda:3'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:3') +2023-02-09 01:36:59,552 INFO [train.py:901] (3/4) Epoch 29, batch 1150, loss[loss=0.1973, simple_loss=0.2664, pruned_loss=0.06406, over 7434.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2795, pruned_loss=0.05664, over 1605842.14 frames. ], batch size: 17, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:37:06,449 WARNING [train.py:1067] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-02-09 01:37:34,273 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.484e+02 2.320e+02 2.698e+02 3.625e+02 7.425e+02, threshold=5.396e+02, percent-clipped=3.0 +2023-02-09 01:37:36,426 INFO [train.py:901] (3/4) Epoch 29, batch 1200, loss[loss=0.1361, simple_loss=0.2188, pruned_loss=0.02668, over 7421.00 frames. ], tot_loss[loss=0.1948, simple_loss=0.2783, pruned_loss=0.05572, over 1608458.91 frames. ], batch size: 17, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:37:53,821 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.9843, 2.0883, 1.7566, 2.7312, 1.2514, 1.6200, 1.9197, 2.1688], + device='cuda:3'), covar=tensor([0.0719, 0.0786, 0.0928, 0.0353, 0.1139, 0.1337, 0.0852, 0.0795], + device='cuda:3'), in_proj_covar=tensor([0.0230, 0.0195, 0.0244, 0.0213, 0.0203, 0.0247, 0.0250, 0.0205], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:3') +2023-02-09 01:38:09,059 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.0715, 1.4765, 1.7312, 1.3919, 1.0189, 1.5103, 1.8689, 1.7231], + device='cuda:3'), covar=tensor([0.0541, 0.1285, 0.1680, 0.1511, 0.0620, 0.1508, 0.0702, 0.0634], + device='cuda:3'), in_proj_covar=tensor([0.0100, 0.0154, 0.0190, 0.0162, 0.0102, 0.0163, 0.0113, 0.0147], + device='cuda:3'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:3') +2023-02-09 01:38:11,539 INFO [train.py:901] (3/4) Epoch 29, batch 1250, loss[loss=0.1985, simple_loss=0.2856, pruned_loss=0.0557, over 8103.00 frames. ], tot_loss[loss=0.1953, simple_loss=0.2788, pruned_loss=0.05587, over 1613091.10 frames. ], batch size: 23, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:38:13,783 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.4977, 2.3458, 1.8142, 2.1604, 1.9762, 1.5795, 1.9447, 1.9519], + device='cuda:3'), covar=tensor([0.1425, 0.0493, 0.1171, 0.0617, 0.0782, 0.1547, 0.1000, 0.0980], + device='cuda:3'), in_proj_covar=tensor([0.0356, 0.0246, 0.0344, 0.0314, 0.0303, 0.0349, 0.0351, 0.0323], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-02-09 01:38:19,537 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6009, 2.0620, 3.1442, 1.5097, 2.3864, 2.0395, 1.7558, 2.3702], + device='cuda:3'), covar=tensor([0.1985, 0.2675, 0.0886, 0.4868, 0.2074, 0.3551, 0.2590, 0.2408], + device='cuda:3'), in_proj_covar=tensor([0.0539, 0.0639, 0.0564, 0.0672, 0.0667, 0.0617, 0.0565, 0.0648], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 01:38:46,454 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.451e+02 2.398e+02 2.828e+02 3.393e+02 7.704e+02, threshold=5.657e+02, percent-clipped=4.0 +2023-02-09 01:38:48,670 INFO [train.py:901] (3/4) Epoch 29, batch 1300, loss[loss=0.2033, simple_loss=0.2821, pruned_loss=0.06222, over 8238.00 frames. ], tot_loss[loss=0.1951, simple_loss=0.2789, pruned_loss=0.05565, over 1608240.56 frames. ], batch size: 22, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:39:24,555 INFO [train.py:901] (3/4) Epoch 29, batch 1350, loss[loss=0.1712, simple_loss=0.249, pruned_loss=0.0467, over 7435.00 frames. ], tot_loss[loss=0.1942, simple_loss=0.2781, pruned_loss=0.05519, over 1605185.38 frames. ], batch size: 17, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:39:52,810 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=227711.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:39:58,202 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.624e+02 2.335e+02 2.785e+02 3.650e+02 1.055e+03, threshold=5.570e+02, percent-clipped=4.0 +2023-02-09 01:40:00,379 INFO [train.py:901] (3/4) Epoch 29, batch 1400, loss[loss=0.2634, simple_loss=0.3418, pruned_loss=0.09253, over 8360.00 frames. ], tot_loss[loss=0.1954, simple_loss=0.2792, pruned_loss=0.05581, over 1610138.09 frames. ], batch size: 24, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:40:01,311 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=227722.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:40:20,607 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=227747.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:40:38,114 INFO [train.py:901] (3/4) Epoch 29, batch 1450, loss[loss=0.2238, simple_loss=0.3128, pruned_loss=0.06742, over 8470.00 frames. ], tot_loss[loss=0.1951, simple_loss=0.2795, pruned_loss=0.0554, over 1613677.53 frames. ], batch size: 25, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:40:39,494 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=227773.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:40:47,996 WARNING [train.py:1067] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-02-09 01:41:11,911 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.751e+02 2.369e+02 2.795e+02 3.434e+02 1.018e+03, threshold=5.589e+02, percent-clipped=3.0 +2023-02-09 01:41:14,106 INFO [train.py:901] (3/4) Epoch 29, batch 1500, loss[loss=0.1793, simple_loss=0.2824, pruned_loss=0.03809, over 8324.00 frames. ], tot_loss[loss=0.1953, simple_loss=0.2797, pruned_loss=0.05546, over 1616061.13 frames. ], batch size: 25, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:41:32,536 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=227845.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:41:40,143 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([5.8341, 6.0642, 5.0794, 2.5640, 5.2300, 5.7161, 5.5229, 5.5039], + device='cuda:3'), covar=tensor([0.0491, 0.0335, 0.0931, 0.4437, 0.0780, 0.0622, 0.0956, 0.0641], + device='cuda:3'), in_proj_covar=tensor([0.0545, 0.0458, 0.0456, 0.0563, 0.0446, 0.0470, 0.0442, 0.0414], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 01:41:51,482 INFO [train.py:901] (3/4) Epoch 29, batch 1550, loss[loss=0.1892, simple_loss=0.2767, pruned_loss=0.05088, over 8354.00 frames. ], tot_loss[loss=0.1952, simple_loss=0.2794, pruned_loss=0.05555, over 1617643.84 frames. ], batch size: 24, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:42:04,416 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=227888.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:42:25,539 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.467e+02 2.452e+02 3.275e+02 4.864e+02 1.208e+03, threshold=6.551e+02, percent-clipped=17.0 +2023-02-09 01:42:27,688 INFO [train.py:901] (3/4) Epoch 29, batch 1600, loss[loss=0.1751, simple_loss=0.2562, pruned_loss=0.04699, over 7770.00 frames. ], tot_loss[loss=0.1953, simple_loss=0.2797, pruned_loss=0.05551, over 1615827.29 frames. ], batch size: 19, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:42:27,878 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.4076, 1.3941, 1.3762, 1.8373, 0.6174, 1.2617, 1.3238, 1.4971], + device='cuda:3'), covar=tensor([0.1000, 0.0835, 0.1205, 0.0531, 0.1215, 0.1479, 0.0733, 0.0789], + device='cuda:3'), in_proj_covar=tensor([0.0231, 0.0195, 0.0245, 0.0213, 0.0204, 0.0247, 0.0251, 0.0206], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:3') +2023-02-09 01:43:04,965 INFO [train.py:901] (3/4) Epoch 29, batch 1650, loss[loss=0.2033, simple_loss=0.291, pruned_loss=0.05776, over 8038.00 frames. ], tot_loss[loss=0.1951, simple_loss=0.2791, pruned_loss=0.05553, over 1610620.91 frames. ], batch size: 22, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:43:39,886 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.526e+02 2.401e+02 2.687e+02 3.396e+02 6.045e+02, threshold=5.374e+02, percent-clipped=0.0 +2023-02-09 01:43:42,091 INFO [train.py:901] (3/4) Epoch 29, batch 1700, loss[loss=0.1778, simple_loss=0.2579, pruned_loss=0.04885, over 7232.00 frames. ], tot_loss[loss=0.1938, simple_loss=0.278, pruned_loss=0.05484, over 1606840.55 frames. ], batch size: 16, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:43:42,959 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.4175, 4.3993, 3.9918, 2.0595, 3.9061, 4.0311, 3.9189, 3.8397], + device='cuda:3'), covar=tensor([0.0697, 0.0503, 0.0947, 0.4812, 0.0870, 0.0940, 0.1238, 0.0817], + device='cuda:3'), in_proj_covar=tensor([0.0548, 0.0461, 0.0457, 0.0567, 0.0447, 0.0472, 0.0445, 0.0415], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 01:43:44,503 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.9568, 1.4315, 4.1786, 1.8199, 2.5600, 4.6852, 4.8966, 4.1334], + device='cuda:3'), covar=tensor([0.1406, 0.2127, 0.0306, 0.2173, 0.1172, 0.0221, 0.0564, 0.0546], + device='cuda:3'), in_proj_covar=tensor([0.0312, 0.0329, 0.0298, 0.0327, 0.0329, 0.0280, 0.0450, 0.0312], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:3') +2023-02-09 01:43:45,212 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=228025.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:44:06,284 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=228055.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:44:18,101 INFO [train.py:901] (3/4) Epoch 29, batch 1750, loss[loss=0.19, simple_loss=0.2754, pruned_loss=0.05229, over 8079.00 frames. ], tot_loss[loss=0.1955, simple_loss=0.2796, pruned_loss=0.0557, over 1610423.76 frames. ], batch size: 21, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:44:28,146 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.0430, 1.8765, 2.2583, 1.9739, 2.2971, 2.1625, 2.0349, 1.1759], + device='cuda:3'), covar=tensor([0.6227, 0.5114, 0.2262, 0.3921, 0.2645, 0.3478, 0.2095, 0.5633], + device='cuda:3'), in_proj_covar=tensor([0.0974, 0.1038, 0.0842, 0.1005, 0.1030, 0.0942, 0.0776, 0.0857], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 01:44:38,679 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=228098.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:44:52,744 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.427e+02 2.425e+02 2.925e+02 3.463e+02 6.679e+02, threshold=5.849e+02, percent-clipped=2.0 +2023-02-09 01:44:55,485 INFO [train.py:901] (3/4) Epoch 29, batch 1800, loss[loss=0.2006, simple_loss=0.2813, pruned_loss=0.05989, over 8301.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2796, pruned_loss=0.05596, over 1606725.84 frames. ], batch size: 23, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:45:11,255 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=228144.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:45:12,124 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.75 vs. limit=2.0 +2023-02-09 01:45:28,618 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=228169.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:45:29,267 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=228170.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:45:29,728 INFO [train.py:901] (3/4) Epoch 29, batch 1850, loss[loss=0.2119, simple_loss=0.2964, pruned_loss=0.06366, over 8507.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.2806, pruned_loss=0.0564, over 1611486.22 frames. ], batch size: 39, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:45:42,841 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=228189.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:46:03,753 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.759e+02 2.509e+02 2.932e+02 3.424e+02 5.958e+02, threshold=5.864e+02, percent-clipped=1.0 +2023-02-09 01:46:05,869 INFO [train.py:901] (3/4) Epoch 29, batch 1900, loss[loss=0.1814, simple_loss=0.2672, pruned_loss=0.0478, over 8670.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.2809, pruned_loss=0.05628, over 1613564.50 frames. ], batch size: 34, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:46:29,137 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.92 vs. limit=5.0 +2023-02-09 01:46:32,893 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.37 vs. limit=2.0 +2023-02-09 01:46:38,538 WARNING [train.py:1067] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-02-09 01:46:41,414 INFO [train.py:901] (3/4) Epoch 29, batch 1950, loss[loss=0.2092, simple_loss=0.2971, pruned_loss=0.06063, over 8358.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2804, pruned_loss=0.0564, over 1611577.03 frames. ], batch size: 24, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:46:50,972 WARNING [train.py:1067] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-02-09 01:47:05,322 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=228304.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:47:10,130 WARNING [train.py:1067] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-02-09 01:47:16,244 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.886e+02 2.422e+02 3.056e+02 3.762e+02 8.552e+02, threshold=6.111e+02, percent-clipped=4.0 +2023-02-09 01:47:18,246 INFO [train.py:901] (3/4) Epoch 29, batch 2000, loss[loss=0.1861, simple_loss=0.278, pruned_loss=0.04711, over 7820.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2799, pruned_loss=0.05609, over 1612381.39 frames. ], batch size: 20, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:47:36,284 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=228347.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:47:39,108 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=228351.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:47:44,567 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.8999, 2.0782, 2.1519, 1.5069, 2.3263, 1.5658, 0.7695, 2.0364], + device='cuda:3'), covar=tensor([0.0792, 0.0413, 0.0399, 0.0724, 0.0570, 0.1089, 0.1133, 0.0403], + device='cuda:3'), in_proj_covar=tensor([0.0476, 0.0414, 0.0369, 0.0460, 0.0397, 0.0552, 0.0403, 0.0443], + device='cuda:3'), out_proj_covar=tensor([1.2611e-04, 1.0730e-04, 9.6024e-05, 1.2017e-04, 1.0384e-04, 1.5379e-04, + 1.0767e-04, 1.1589e-04], device='cuda:3') +2023-02-09 01:47:52,203 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=228369.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:47:53,547 INFO [train.py:901] (3/4) Epoch 29, batch 2050, loss[loss=0.1833, simple_loss=0.2698, pruned_loss=0.0484, over 8088.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2799, pruned_loss=0.05586, over 1616032.15 frames. ], batch size: 21, lr: 2.61e-03, grad_scale: 8.0 +2023-02-09 01:48:25,963 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.725e+02 2.333e+02 2.866e+02 3.600e+02 5.490e+02, threshold=5.733e+02, percent-clipped=0.0 +2023-02-09 01:48:28,128 INFO [train.py:901] (3/4) Epoch 29, batch 2100, loss[loss=0.174, simple_loss=0.2552, pruned_loss=0.04645, over 7250.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2803, pruned_loss=0.05618, over 1616587.79 frames. ], batch size: 16, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:48:32,344 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=228426.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:48:44,591 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=228442.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:48:48,010 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=228447.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 01:48:50,740 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=228451.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:48:59,117 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.8044, 2.2010, 3.4005, 1.6958, 2.6640, 2.2863, 1.8939, 2.7264], + device='cuda:3'), covar=tensor([0.1913, 0.2627, 0.0838, 0.4629, 0.1760, 0.3197, 0.2453, 0.2092], + device='cuda:3'), in_proj_covar=tensor([0.0543, 0.0642, 0.0568, 0.0675, 0.0667, 0.0617, 0.0566, 0.0650], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 01:49:04,308 INFO [train.py:901] (3/4) Epoch 29, batch 2150, loss[loss=0.2264, simple_loss=0.3039, pruned_loss=0.0744, over 8539.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2802, pruned_loss=0.05629, over 1614733.87 frames. ], batch size: 28, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:49:05,158 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.7699, 1.5545, 3.9477, 1.5444, 3.4760, 3.3249, 3.5637, 3.4711], + device='cuda:3'), covar=tensor([0.0764, 0.4517, 0.0771, 0.4459, 0.1280, 0.1074, 0.0744, 0.0860], + device='cuda:3'), in_proj_covar=tensor([0.0686, 0.0669, 0.0744, 0.0664, 0.0750, 0.0639, 0.0647, 0.0721], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 01:49:14,153 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=228484.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:49:37,700 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.627e+02 2.659e+02 3.270e+02 3.951e+02 1.171e+03, threshold=6.540e+02, percent-clipped=10.0 +2023-02-09 01:49:39,906 INFO [train.py:901] (3/4) Epoch 29, batch 2200, loss[loss=0.196, simple_loss=0.2892, pruned_loss=0.05139, over 8597.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2808, pruned_loss=0.05637, over 1617269.82 frames. ], batch size: 34, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:49:47,014 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.2532, 2.0045, 4.4438, 2.0736, 3.9882, 3.7435, 4.0374, 3.9465], + device='cuda:3'), covar=tensor([0.0698, 0.3991, 0.0627, 0.3967, 0.0966, 0.0929, 0.0607, 0.0691], + device='cuda:3'), in_proj_covar=tensor([0.0684, 0.0666, 0.0742, 0.0662, 0.0747, 0.0637, 0.0644, 0.0719], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 01:50:06,275 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=228557.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:50:08,439 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=228560.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:50:16,510 INFO [train.py:901] (3/4) Epoch 29, batch 2250, loss[loss=0.2047, simple_loss=0.2925, pruned_loss=0.05843, over 8244.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2806, pruned_loss=0.05614, over 1618913.40 frames. ], batch size: 24, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:50:26,598 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=228585.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:50:44,246 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.8354, 1.7157, 2.7769, 1.3249, 2.4069, 3.0984, 3.2999, 2.3687], + device='cuda:3'), covar=tensor([0.1465, 0.1979, 0.0714, 0.2646, 0.1580, 0.0432, 0.0745, 0.0969], + device='cuda:3'), in_proj_covar=tensor([0.0311, 0.0327, 0.0297, 0.0325, 0.0329, 0.0279, 0.0447, 0.0312], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:3') +2023-02-09 01:50:50,353 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.529e+02 2.350e+02 2.877e+02 3.583e+02 6.549e+02, threshold=5.755e+02, percent-clipped=1.0 +2023-02-09 01:50:52,580 INFO [train.py:901] (3/4) Epoch 29, batch 2300, loss[loss=0.2347, simple_loss=0.3115, pruned_loss=0.07897, over 8685.00 frames. ], tot_loss[loss=0.195, simple_loss=0.2796, pruned_loss=0.05524, over 1613873.57 frames. ], batch size: 39, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:50:53,467 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.8489, 1.7560, 2.4881, 1.6876, 1.3702, 2.4255, 0.5901, 1.5323], + device='cuda:3'), covar=tensor([0.1365, 0.1280, 0.0296, 0.1030, 0.2354, 0.0384, 0.1856, 0.1158], + device='cuda:3'), in_proj_covar=tensor([0.0202, 0.0206, 0.0138, 0.0223, 0.0278, 0.0149, 0.0173, 0.0199], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 01:51:06,297 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.68 vs. limit=2.0 +2023-02-09 01:51:22,808 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.8563, 1.5627, 4.0109, 1.4350, 3.5961, 3.3373, 3.6141, 3.4980], + device='cuda:3'), covar=tensor([0.0644, 0.4471, 0.0639, 0.4247, 0.1158, 0.1009, 0.0647, 0.0786], + device='cuda:3'), in_proj_covar=tensor([0.0686, 0.0669, 0.0746, 0.0666, 0.0751, 0.0640, 0.0647, 0.0722], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 01:51:28,782 INFO [train.py:901] (3/4) Epoch 29, batch 2350, loss[loss=0.2202, simple_loss=0.2986, pruned_loss=0.07094, over 8583.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2803, pruned_loss=0.05559, over 1616296.28 frames. ], batch size: 49, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:51:43,130 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=228691.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:51:45,886 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=228695.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:52:02,331 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.786e+02 2.434e+02 3.194e+02 3.873e+02 7.294e+02, threshold=6.388e+02, percent-clipped=4.0 +2023-02-09 01:52:04,282 INFO [train.py:901] (3/4) Epoch 29, batch 2400, loss[loss=0.2204, simple_loss=0.3117, pruned_loss=0.06453, over 8199.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2808, pruned_loss=0.05622, over 1615623.44 frames. ], batch size: 23, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:52:17,433 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=228740.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:52:34,957 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=228765.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:52:39,615 INFO [train.py:901] (3/4) Epoch 29, batch 2450, loss[loss=0.1739, simple_loss=0.2546, pruned_loss=0.04658, over 7961.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2811, pruned_loss=0.05655, over 1615666.64 frames. ], batch size: 21, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:52:55,045 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=228791.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 01:53:05,517 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=228806.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:53:08,252 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=228810.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:53:10,296 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=228813.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:53:13,508 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.470e+02 2.278e+02 2.630e+02 3.527e+02 5.802e+02, threshold=5.259e+02, percent-clipped=0.0 +2023-02-09 01:53:16,199 INFO [train.py:901] (3/4) Epoch 29, batch 2500, loss[loss=0.2011, simple_loss=0.2768, pruned_loss=0.06271, over 6807.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2815, pruned_loss=0.05648, over 1615032.10 frames. ], batch size: 15, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:53:28,181 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=228838.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:53:31,516 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.5363, 2.4528, 1.8300, 2.2911, 2.1480, 1.5360, 2.0604, 2.1358], + device='cuda:3'), covar=tensor([0.1513, 0.0492, 0.1269, 0.0598, 0.0706, 0.1634, 0.0965, 0.1020], + device='cuda:3'), in_proj_covar=tensor([0.0356, 0.0244, 0.0343, 0.0315, 0.0302, 0.0349, 0.0350, 0.0320], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-02-09 01:53:51,014 INFO [train.py:901] (3/4) Epoch 29, batch 2550, loss[loss=0.2023, simple_loss=0.2876, pruned_loss=0.05848, over 8459.00 frames. ], tot_loss[loss=0.198, simple_loss=0.282, pruned_loss=0.05696, over 1614060.80 frames. ], batch size: 25, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:54:05,689 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.07 vs. limit=5.0 +2023-02-09 01:54:17,018 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=228906.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 01:54:25,705 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.542e+02 2.320e+02 2.761e+02 3.420e+02 6.403e+02, threshold=5.523e+02, percent-clipped=2.0 +2023-02-09 01:54:27,841 INFO [train.py:901] (3/4) Epoch 29, batch 2600, loss[loss=0.2336, simple_loss=0.3136, pruned_loss=0.07679, over 8687.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2811, pruned_loss=0.05657, over 1615783.28 frames. ], batch size: 34, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:54:51,311 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=228953.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:55:03,876 INFO [train.py:901] (3/4) Epoch 29, batch 2650, loss[loss=0.1795, simple_loss=0.2524, pruned_loss=0.05332, over 7782.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.281, pruned_loss=0.05679, over 1615105.72 frames. ], batch size: 19, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:55:12,101 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.75 vs. limit=2.0 +2023-02-09 01:55:37,355 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.671e+02 2.477e+02 2.922e+02 3.574e+02 8.790e+02, threshold=5.845e+02, percent-clipped=2.0 +2023-02-09 01:55:39,297 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.3266, 1.2804, 3.4071, 0.9785, 3.0514, 2.8142, 3.1116, 3.0241], + device='cuda:3'), covar=tensor([0.0693, 0.4460, 0.0792, 0.4461, 0.1221, 0.1147, 0.0690, 0.0779], + device='cuda:3'), in_proj_covar=tensor([0.0682, 0.0663, 0.0742, 0.0662, 0.0744, 0.0636, 0.0641, 0.0716], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 01:55:39,870 INFO [train.py:901] (3/4) Epoch 29, batch 2700, loss[loss=0.1928, simple_loss=0.2798, pruned_loss=0.05287, over 7814.00 frames. ], tot_loss[loss=0.1962, simple_loss=0.2798, pruned_loss=0.05634, over 1612190.97 frames. ], batch size: 20, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:56:09,693 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([5.2323, 2.0329, 5.4105, 2.5623, 4.9177, 4.5744, 4.9485, 4.8750], + device='cuda:3'), covar=tensor([0.0522, 0.4397, 0.0448, 0.3801, 0.0953, 0.0878, 0.0512, 0.0553], + device='cuda:3'), in_proj_covar=tensor([0.0680, 0.0662, 0.0740, 0.0661, 0.0743, 0.0635, 0.0640, 0.0715], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 01:56:11,148 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=229062.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:56:14,634 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=229066.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:56:17,775 INFO [train.py:901] (3/4) Epoch 29, batch 2750, loss[loss=0.1941, simple_loss=0.2722, pruned_loss=0.058, over 8089.00 frames. ], tot_loss[loss=0.1954, simple_loss=0.2791, pruned_loss=0.05587, over 1613000.86 frames. ], batch size: 21, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:56:29,445 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=229087.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:56:32,358 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=229091.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:56:51,299 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.676e+02 2.406e+02 2.857e+02 3.824e+02 7.570e+02, threshold=5.715e+02, percent-clipped=1.0 +2023-02-09 01:56:53,448 INFO [train.py:901] (3/4) Epoch 29, batch 2800, loss[loss=0.1607, simple_loss=0.244, pruned_loss=0.03871, over 8095.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2796, pruned_loss=0.05617, over 1605576.43 frames. ], batch size: 21, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 01:56:57,317 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.63 vs. limit=2.0 +2023-02-09 01:57:24,974 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=229162.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 01:57:31,032 INFO [train.py:901] (3/4) Epoch 29, batch 2850, loss[loss=0.1825, simple_loss=0.2685, pruned_loss=0.04825, over 8093.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2803, pruned_loss=0.05633, over 1608699.98 frames. ], batch size: 23, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 01:57:43,240 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=229187.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 01:58:05,224 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.341e+02 2.443e+02 3.095e+02 3.828e+02 9.615e+02, threshold=6.189e+02, percent-clipped=4.0 +2023-02-09 01:58:07,410 INFO [train.py:901] (3/4) Epoch 29, batch 2900, loss[loss=0.2254, simple_loss=0.2996, pruned_loss=0.07562, over 8440.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.28, pruned_loss=0.05625, over 1607562.12 frames. ], batch size: 27, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 01:58:15,665 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.69 vs. limit=2.0 +2023-02-09 01:58:44,220 INFO [train.py:901] (3/4) Epoch 29, batch 2950, loss[loss=0.1642, simple_loss=0.2384, pruned_loss=0.04503, over 7208.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2806, pruned_loss=0.05655, over 1609947.25 frames. ], batch size: 16, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 01:58:47,090 WARNING [train.py:1067] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-02-09 01:59:02,255 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=229297.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:59:17,278 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.610e+02 2.481e+02 3.005e+02 3.741e+02 9.617e+02, threshold=6.010e+02, percent-clipped=3.0 +2023-02-09 01:59:17,409 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=229318.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 01:59:19,282 INFO [train.py:901] (3/4) Epoch 29, batch 3000, loss[loss=0.1983, simple_loss=0.2958, pruned_loss=0.05035, over 8198.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.2807, pruned_loss=0.05633, over 1610017.50 frames. ], batch size: 23, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 01:59:19,282 INFO [train.py:926] (3/4) Computing validation loss +2023-02-09 01:59:34,610 INFO [train.py:935] (3/4) Epoch 29, validation: loss=0.17, simple_loss=0.2699, pruned_loss=0.03504, over 944034.00 frames. +2023-02-09 01:59:34,611 INFO [train.py:936] (3/4) Maximum memory allocated so far is 6473MB +2023-02-09 02:00:09,406 INFO [train.py:901] (3/4) Epoch 29, batch 3050, loss[loss=0.2106, simple_loss=0.2902, pruned_loss=0.06555, over 8347.00 frames. ], tot_loss[loss=0.1961, simple_loss=0.2801, pruned_loss=0.05603, over 1610431.45 frames. ], batch size: 26, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 02:00:24,138 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.3325, 3.5833, 2.5138, 3.1379, 3.0727, 2.0210, 2.9668, 3.2414], + device='cuda:3'), covar=tensor([0.1685, 0.0436, 0.1206, 0.0746, 0.0732, 0.1682, 0.1022, 0.1076], + device='cuda:3'), in_proj_covar=tensor([0.0354, 0.0243, 0.0342, 0.0313, 0.0300, 0.0348, 0.0349, 0.0318], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-02-09 02:00:40,613 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=229412.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:00:44,471 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.402e+02 2.505e+02 2.815e+02 3.570e+02 7.212e+02, threshold=5.630e+02, percent-clipped=4.0 +2023-02-09 02:00:46,509 INFO [train.py:901] (3/4) Epoch 29, batch 3100, loss[loss=0.2184, simple_loss=0.3056, pruned_loss=0.06558, over 8506.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2805, pruned_loss=0.05651, over 1609162.67 frames. ], batch size: 28, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 02:01:15,193 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=229461.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:01:21,908 INFO [train.py:901] (3/4) Epoch 29, batch 3150, loss[loss=0.1918, simple_loss=0.2852, pruned_loss=0.0492, over 8474.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2812, pruned_loss=0.05695, over 1608961.18 frames. ], batch size: 25, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 02:01:24,683 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.79 vs. limit=2.0 +2023-02-09 02:01:39,335 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=229495.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:01:56,480 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.645e+02 2.412e+02 3.144e+02 3.923e+02 1.015e+03, threshold=6.289e+02, percent-clipped=11.0 +2023-02-09 02:01:58,593 INFO [train.py:901] (3/4) Epoch 29, batch 3200, loss[loss=0.1792, simple_loss=0.2899, pruned_loss=0.03424, over 8234.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2802, pruned_loss=0.05623, over 1611447.37 frames. ], batch size: 24, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 02:02:25,102 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-02-09 02:02:35,456 INFO [train.py:901] (3/4) Epoch 29, batch 3250, loss[loss=0.1564, simple_loss=0.2363, pruned_loss=0.03824, over 7424.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2814, pruned_loss=0.05718, over 1610389.19 frames. ], batch size: 17, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 02:02:52,634 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=229595.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:03:09,578 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.589e+02 2.257e+02 2.746e+02 3.374e+02 9.131e+02, threshold=5.492e+02, percent-clipped=1.0 +2023-02-09 02:03:11,691 INFO [train.py:901] (3/4) Epoch 29, batch 3300, loss[loss=0.2143, simple_loss=0.3106, pruned_loss=0.05899, over 8487.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2807, pruned_loss=0.05645, over 1610663.06 frames. ], batch size: 26, lr: 2.60e-03, grad_scale: 16.0 +2023-02-09 02:03:41,141 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=229662.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:03:45,541 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=229668.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:03:47,460 INFO [train.py:901] (3/4) Epoch 29, batch 3350, loss[loss=0.1882, simple_loss=0.2797, pruned_loss=0.0484, over 8544.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2809, pruned_loss=0.05673, over 1611806.59 frames. ], batch size: 49, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 02:03:52,169 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=229677.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 02:04:03,324 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=229693.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:04:20,882 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.769e+02 2.561e+02 2.994e+02 3.779e+02 7.703e+02, threshold=5.989e+02, percent-clipped=7.0 +2023-02-09 02:04:22,314 INFO [train.py:901] (3/4) Epoch 29, batch 3400, loss[loss=0.2196, simple_loss=0.2897, pruned_loss=0.07473, over 8443.00 frames. ], tot_loss[loss=0.1986, simple_loss=0.2819, pruned_loss=0.05766, over 1612854.54 frames. ], batch size: 27, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 02:04:59,076 INFO [train.py:901] (3/4) Epoch 29, batch 3450, loss[loss=0.1807, simple_loss=0.2682, pruned_loss=0.04665, over 8031.00 frames. ], tot_loss[loss=0.1982, simple_loss=0.2811, pruned_loss=0.05762, over 1604894.72 frames. ], batch size: 22, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 02:05:03,519 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=229777.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:05:23,717 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=229805.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:05:25,167 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.0712, 2.0917, 1.7993, 2.7638, 1.3180, 1.6991, 1.9189, 2.1374], + device='cuda:3'), covar=tensor([0.0635, 0.0749, 0.0836, 0.0334, 0.1071, 0.1196, 0.0848, 0.0716], + device='cuda:3'), in_proj_covar=tensor([0.0230, 0.0195, 0.0244, 0.0212, 0.0202, 0.0245, 0.0250, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0005, 0.0006, 0.0005], + device='cuda:3') +2023-02-09 02:05:33,259 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.613e+02 2.314e+02 2.735e+02 3.470e+02 1.051e+03, threshold=5.470e+02, percent-clipped=3.0 +2023-02-09 02:05:34,613 INFO [train.py:901] (3/4) Epoch 29, batch 3500, loss[loss=0.1907, simple_loss=0.27, pruned_loss=0.05571, over 8095.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.2812, pruned_loss=0.05753, over 1605372.52 frames. ], batch size: 21, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 02:05:34,872 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.1266, 1.9681, 2.4617, 2.1138, 2.4687, 2.2356, 2.0980, 1.3966], + device='cuda:3'), covar=tensor([0.6161, 0.5278, 0.2175, 0.4054, 0.2743, 0.3332, 0.1953, 0.5451], + device='cuda:3'), in_proj_covar=tensor([0.0970, 0.1034, 0.0841, 0.1005, 0.1029, 0.0942, 0.0777, 0.0858], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 02:05:46,951 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=229839.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:05:58,496 WARNING [train.py:1067] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-02-09 02:06:12,024 INFO [train.py:901] (3/4) Epoch 29, batch 3550, loss[loss=0.1865, simple_loss=0.2651, pruned_loss=0.05391, over 7434.00 frames. ], tot_loss[loss=0.1962, simple_loss=0.2798, pruned_loss=0.05633, over 1608511.37 frames. ], batch size: 17, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 02:06:41,728 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=229912.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:06:46,512 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.708e+02 2.466e+02 3.015e+02 3.666e+02 8.686e+02, threshold=6.030e+02, percent-clipped=2.0 +2023-02-09 02:06:47,438 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=229920.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:06:47,948 INFO [train.py:901] (3/4) Epoch 29, batch 3600, loss[loss=0.1874, simple_loss=0.2706, pruned_loss=0.05211, over 8284.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2809, pruned_loss=0.05713, over 1609442.75 frames. ], batch size: 23, lr: 2.60e-03, grad_scale: 8.0 +2023-02-09 02:07:00,989 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=229939.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:07:11,758 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=229954.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:07:24,207 INFO [train.py:901] (3/4) Epoch 29, batch 3650, loss[loss=0.2456, simple_loss=0.3152, pruned_loss=0.08798, over 8610.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2791, pruned_loss=0.05624, over 1607757.44 frames. ], batch size: 34, lr: 2.60e-03, grad_scale: 4.0 +2023-02-09 02:07:50,385 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=230005.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:08:00,958 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.417e+02 2.398e+02 2.862e+02 3.372e+02 7.881e+02, threshold=5.724e+02, percent-clipped=3.0 +2023-02-09 02:08:01,709 INFO [train.py:901] (3/4) Epoch 29, batch 3700, loss[loss=0.169, simple_loss=0.2581, pruned_loss=0.03997, over 8244.00 frames. ], tot_loss[loss=0.1941, simple_loss=0.2777, pruned_loss=0.05524, over 1606943.44 frames. ], batch size: 22, lr: 2.60e-03, grad_scale: 4.0 +2023-02-09 02:08:02,485 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=230021.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 02:08:06,536 WARNING [train.py:1067] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-02-09 02:08:10,859 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=230033.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:08:25,153 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=230054.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:08:28,458 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=230058.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:08:37,283 INFO [train.py:901] (3/4) Epoch 29, batch 3750, loss[loss=0.1863, simple_loss=0.2617, pruned_loss=0.05548, over 7657.00 frames. ], tot_loss[loss=0.1945, simple_loss=0.2776, pruned_loss=0.05564, over 1601777.38 frames. ], batch size: 19, lr: 2.60e-03, grad_scale: 4.0 +2023-02-09 02:09:13,034 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.639e+02 2.469e+02 3.110e+02 3.966e+02 1.066e+03, threshold=6.219e+02, percent-clipped=4.0 +2023-02-09 02:09:13,774 INFO [train.py:901] (3/4) Epoch 29, batch 3800, loss[loss=0.1976, simple_loss=0.281, pruned_loss=0.05706, over 8246.00 frames. ], tot_loss[loss=0.1951, simple_loss=0.2783, pruned_loss=0.05593, over 1605963.22 frames. ], batch size: 22, lr: 2.60e-03, grad_scale: 4.0 +2023-02-09 02:09:24,875 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=230136.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 02:09:29,249 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.5869, 2.0660, 2.9040, 1.4102, 2.1468, 1.8591, 1.8128, 2.1704], + device='cuda:3'), covar=tensor([0.2248, 0.2943, 0.1288, 0.5485, 0.2493, 0.4067, 0.2813, 0.2937], + device='cuda:3'), in_proj_covar=tensor([0.0542, 0.0641, 0.0567, 0.0675, 0.0668, 0.0616, 0.0567, 0.0648], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 02:09:40,095 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-02-09 02:09:47,386 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=230168.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:09:49,383 INFO [train.py:901] (3/4) Epoch 29, batch 3850, loss[loss=0.1713, simple_loss=0.2616, pruned_loss=0.04052, over 8016.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2793, pruned_loss=0.0563, over 1608035.83 frames. ], batch size: 22, lr: 2.60e-03, grad_scale: 4.0 +2023-02-09 02:09:53,096 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=230176.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:10:10,990 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=230201.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:10:12,892 WARNING [train.py:1067] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-02-09 02:10:17,565 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=230210.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:10:24,394 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.322e+02 2.511e+02 3.297e+02 3.973e+02 1.066e+03, threshold=6.594e+02, percent-clipped=6.0 +2023-02-09 02:10:25,128 INFO [train.py:901] (3/4) Epoch 29, batch 3900, loss[loss=0.1921, simple_loss=0.2635, pruned_loss=0.06037, over 7791.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.28, pruned_loss=0.05631, over 1611232.48 frames. ], batch size: 19, lr: 2.59e-03, grad_scale: 4.0 +2023-02-09 02:10:36,234 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=230235.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:10:51,873 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=230256.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:10:54,165 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=230259.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:11:02,680 INFO [train.py:901] (3/4) Epoch 29, batch 3950, loss[loss=0.1587, simple_loss=0.2436, pruned_loss=0.03692, over 8248.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2802, pruned_loss=0.05622, over 1611644.92 frames. ], batch size: 22, lr: 2.59e-03, grad_scale: 4.0 +2023-02-09 02:11:15,909 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.7005, 2.5371, 1.8461, 2.2974, 2.2375, 1.5740, 2.1600, 2.2911], + device='cuda:3'), covar=tensor([0.1530, 0.0509, 0.1382, 0.0687, 0.0755, 0.1689, 0.1034, 0.0947], + device='cuda:3'), in_proj_covar=tensor([0.0359, 0.0246, 0.0347, 0.0317, 0.0305, 0.0351, 0.0354, 0.0323], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-02-09 02:11:30,500 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=230310.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:11:37,967 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.749e+02 2.564e+02 3.192e+02 4.107e+02 9.729e+02, threshold=6.384e+02, percent-clipped=2.0 +2023-02-09 02:11:38,727 INFO [train.py:901] (3/4) Epoch 29, batch 4000, loss[loss=0.1711, simple_loss=0.2503, pruned_loss=0.046, over 7967.00 frames. ], tot_loss[loss=0.1957, simple_loss=0.2792, pruned_loss=0.05603, over 1606652.72 frames. ], batch size: 21, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:11:43,478 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.5234, 2.2756, 2.7777, 2.3831, 2.7847, 2.4213, 2.3908, 2.0531], + device='cuda:3'), covar=tensor([0.4125, 0.4218, 0.1847, 0.3354, 0.2129, 0.2797, 0.1633, 0.4146], + device='cuda:3'), in_proj_covar=tensor([0.0963, 0.1032, 0.0837, 0.1002, 0.1027, 0.0937, 0.0774, 0.0854], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 02:11:50,444 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=230335.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:11:59,739 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=230349.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:12:16,076 INFO [train.py:901] (3/4) Epoch 29, batch 4050, loss[loss=0.1939, simple_loss=0.275, pruned_loss=0.05643, over 7968.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2796, pruned_loss=0.05605, over 1610772.24 frames. ], batch size: 21, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:12:16,260 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=230371.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:12:31,107 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=230392.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 02:12:48,557 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=230417.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 02:12:50,317 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.613e+02 2.354e+02 2.823e+02 3.644e+02 9.834e+02, threshold=5.645e+02, percent-clipped=2.0 +2023-02-09 02:12:51,008 INFO [train.py:901] (3/4) Epoch 29, batch 4100, loss[loss=0.189, simple_loss=0.2749, pruned_loss=0.05159, over 8138.00 frames. ], tot_loss[loss=0.1962, simple_loss=0.2798, pruned_loss=0.05632, over 1614167.90 frames. ], batch size: 22, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:13:22,439 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=230464.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:13:26,983 INFO [train.py:901] (3/4) Epoch 29, batch 4150, loss[loss=0.2095, simple_loss=0.2971, pruned_loss=0.06097, over 8480.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2815, pruned_loss=0.0572, over 1614574.93 frames. ], batch size: 29, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:13:38,474 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=230486.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:13:56,275 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=230512.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:14:01,809 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.657e+02 2.471e+02 3.120e+02 4.398e+02 1.014e+03, threshold=6.241e+02, percent-clipped=11.0 +2023-02-09 02:14:02,509 INFO [train.py:901] (3/4) Epoch 29, batch 4200, loss[loss=0.1872, simple_loss=0.276, pruned_loss=0.04923, over 8194.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2807, pruned_loss=0.05663, over 1612680.03 frames. ], batch size: 23, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:14:16,934 WARNING [train.py:1067] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-02-09 02:14:17,778 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=230543.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:14:24,144 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=230551.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:14:28,484 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.8239, 1.4330, 4.0129, 1.4423, 3.6030, 3.3801, 3.6704, 3.5493], + device='cuda:3'), covar=tensor([0.0648, 0.4590, 0.0648, 0.4300, 0.1100, 0.0966, 0.0609, 0.0731], + device='cuda:3'), in_proj_covar=tensor([0.0682, 0.0664, 0.0744, 0.0659, 0.0746, 0.0638, 0.0642, 0.0718], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 02:14:38,390 INFO [train.py:901] (3/4) Epoch 29, batch 4250, loss[loss=0.1939, simple_loss=0.2745, pruned_loss=0.05666, over 8701.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.2819, pruned_loss=0.05754, over 1616130.18 frames. ], batch size: 39, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:14:41,222 WARNING [train.py:1067] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-02-09 02:15:01,588 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=230603.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:15:09,498 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.38 vs. limit=5.0 +2023-02-09 02:15:13,975 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.550e+02 2.532e+02 3.108e+02 3.832e+02 7.900e+02, threshold=6.217e+02, percent-clipped=4.0 +2023-02-09 02:15:14,728 INFO [train.py:901] (3/4) Epoch 29, batch 4300, loss[loss=0.1962, simple_loss=0.2784, pruned_loss=0.05703, over 8125.00 frames. ], tot_loss[loss=0.1986, simple_loss=0.282, pruned_loss=0.0576, over 1615099.64 frames. ], batch size: 22, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:15:19,490 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=230627.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:15:19,532 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=230627.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:15:37,544 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=230652.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:15:51,724 INFO [train.py:901] (3/4) Epoch 29, batch 4350, loss[loss=0.1941, simple_loss=0.2755, pruned_loss=0.05635, over 7814.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2816, pruned_loss=0.05721, over 1617173.55 frames. ], batch size: 20, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:15:51,991 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.2067, 2.0376, 2.6077, 2.1819, 2.6904, 2.2996, 2.1256, 1.5410], + device='cuda:3'), covar=tensor([0.5976, 0.5372, 0.2294, 0.4078, 0.2615, 0.3422, 0.1993, 0.5641], + device='cuda:3'), in_proj_covar=tensor([0.0966, 0.1036, 0.0839, 0.1005, 0.1028, 0.0939, 0.0777, 0.0856], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 02:16:17,533 WARNING [train.py:1067] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-02-09 02:16:22,179 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.0663, 1.2270, 1.2317, 0.8642, 1.2243, 1.0939, 0.0685, 1.1927], + device='cuda:3'), covar=tensor([0.0525, 0.0482, 0.0409, 0.0632, 0.0505, 0.1063, 0.1030, 0.0385], + device='cuda:3'), in_proj_covar=tensor([0.0480, 0.0417, 0.0372, 0.0466, 0.0401, 0.0557, 0.0408, 0.0447], + device='cuda:3'), out_proj_covar=tensor([1.2716e-04, 1.0794e-04, 9.6898e-05, 1.2189e-04, 1.0498e-04, 1.5509e-04, + 1.0882e-04, 1.1685e-04], device='cuda:3') +2023-02-09 02:16:27,810 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=230718.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:16:29,089 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.602e+02 2.462e+02 3.042e+02 3.743e+02 1.027e+03, threshold=6.085e+02, percent-clipped=1.0 +2023-02-09 02:16:29,340 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=230720.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:16:29,848 INFO [train.py:901] (3/4) Epoch 29, batch 4400, loss[loss=0.1742, simple_loss=0.2627, pruned_loss=0.04283, over 8133.00 frames. ], tot_loss[loss=0.1987, simple_loss=0.2819, pruned_loss=0.05776, over 1611815.71 frames. ], batch size: 22, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:16:31,537 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.7118, 2.4337, 1.8222, 2.3516, 2.2576, 1.6503, 2.2386, 2.2081], + device='cuda:3'), covar=tensor([0.1374, 0.0427, 0.1224, 0.0594, 0.0656, 0.1460, 0.0843, 0.0859], + device='cuda:3'), in_proj_covar=tensor([0.0355, 0.0244, 0.0344, 0.0314, 0.0301, 0.0347, 0.0351, 0.0320], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-02-09 02:16:47,627 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=230745.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:16:57,328 WARNING [train.py:1067] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-02-09 02:17:06,600 INFO [train.py:901] (3/4) Epoch 29, batch 4450, loss[loss=0.1607, simple_loss=0.254, pruned_loss=0.03372, over 7971.00 frames. ], tot_loss[loss=0.199, simple_loss=0.2823, pruned_loss=0.05785, over 1612376.64 frames. ], batch size: 21, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:17:36,254 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=230810.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:17:43,245 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.628e+02 2.373e+02 2.913e+02 3.542e+02 8.795e+02, threshold=5.826e+02, percent-clipped=1.0 +2023-02-09 02:17:43,976 INFO [train.py:901] (3/4) Epoch 29, batch 4500, loss[loss=0.2299, simple_loss=0.3075, pruned_loss=0.07615, over 8466.00 frames. ], tot_loss[loss=0.2008, simple_loss=0.2837, pruned_loss=0.05894, over 1614455.11 frames. ], batch size: 25, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:17:51,397 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=230830.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:17:54,893 WARNING [train.py:1067] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-02-09 02:18:19,708 INFO [train.py:901] (3/4) Epoch 29, batch 4550, loss[loss=0.205, simple_loss=0.2876, pruned_loss=0.06123, over 8184.00 frames. ], tot_loss[loss=0.2002, simple_loss=0.2835, pruned_loss=0.05843, over 1616579.98 frames. ], batch size: 23, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:18:28,417 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=230883.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:18:31,626 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=230887.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:18:37,214 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=230895.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:18:46,428 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=230908.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:18:46,487 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=230908.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:18:55,058 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.629e+02 2.340e+02 2.935e+02 3.730e+02 9.176e+02, threshold=5.869e+02, percent-clipped=5.0 +2023-02-09 02:18:55,763 INFO [train.py:901] (3/4) Epoch 29, batch 4600, loss[loss=0.2027, simple_loss=0.2799, pruned_loss=0.06279, over 6853.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.283, pruned_loss=0.05793, over 1612488.83 frames. ], batch size: 71, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:19:04,456 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([5.2474, 1.8430, 5.4242, 2.5654, 4.9338, 4.6043, 5.0118, 4.8925], + device='cuda:3'), covar=tensor([0.0523, 0.4541, 0.0405, 0.3555, 0.0900, 0.0844, 0.0505, 0.0551], + device='cuda:3'), in_proj_covar=tensor([0.0680, 0.0660, 0.0742, 0.0654, 0.0742, 0.0634, 0.0639, 0.0717], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 02:19:13,588 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=230945.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:19:31,776 INFO [train.py:901] (3/4) Epoch 29, batch 4650, loss[loss=0.2496, simple_loss=0.3167, pruned_loss=0.0913, over 7518.00 frames. ], tot_loss[loss=0.1993, simple_loss=0.2826, pruned_loss=0.05802, over 1608355.50 frames. ], batch size: 71, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:19:33,927 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=230974.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:19:51,179 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=230999.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:19:53,938 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=231002.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:19:59,423 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=231010.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:20:06,387 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.812e+02 2.522e+02 3.157e+02 3.845e+02 7.559e+02, threshold=6.314e+02, percent-clipped=7.0 +2023-02-09 02:20:07,122 INFO [train.py:901] (3/4) Epoch 29, batch 4700, loss[loss=0.164, simple_loss=0.2414, pruned_loss=0.04329, over 7803.00 frames. ], tot_loss[loss=0.1987, simple_loss=0.2817, pruned_loss=0.05787, over 1608864.32 frames. ], batch size: 19, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:20:43,871 INFO [train.py:901] (3/4) Epoch 29, batch 4750, loss[loss=0.1931, simple_loss=0.2787, pruned_loss=0.05373, over 8104.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2804, pruned_loss=0.05685, over 1608452.41 frames. ], batch size: 23, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:20:56,678 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.5360, 1.6363, 4.7599, 1.8081, 4.2183, 3.9965, 4.3155, 4.1760], + device='cuda:3'), covar=tensor([0.0586, 0.4522, 0.0506, 0.4234, 0.1081, 0.0912, 0.0537, 0.0675], + device='cuda:3'), in_proj_covar=tensor([0.0678, 0.0656, 0.0739, 0.0652, 0.0738, 0.0632, 0.0636, 0.0715], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 02:21:00,626 WARNING [train.py:1067] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-02-09 02:21:02,802 WARNING [train.py:1067] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-02-09 02:21:11,158 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.7952, 1.6682, 2.5220, 1.5319, 1.3950, 2.4511, 0.5183, 1.4908], + device='cuda:3'), covar=tensor([0.1669, 0.1262, 0.0341, 0.1188, 0.2452, 0.0362, 0.1902, 0.1335], + device='cuda:3'), in_proj_covar=tensor([0.0204, 0.0207, 0.0139, 0.0225, 0.0281, 0.0150, 0.0174, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 02:21:18,721 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.727e+02 2.440e+02 2.924e+02 3.615e+02 6.392e+02, threshold=5.847e+02, percent-clipped=1.0 +2023-02-09 02:21:19,433 INFO [train.py:901] (3/4) Epoch 29, batch 4800, loss[loss=0.1852, simple_loss=0.2714, pruned_loss=0.04948, over 8043.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.2801, pruned_loss=0.05667, over 1611096.63 frames. ], batch size: 22, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:21:43,702 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=231154.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:21:55,944 INFO [train.py:901] (3/4) Epoch 29, batch 4850, loss[loss=0.1625, simple_loss=0.2459, pruned_loss=0.03951, over 7432.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2805, pruned_loss=0.05692, over 1606210.37 frames. ], batch size: 17, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:21:55,953 WARNING [train.py:1067] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-02-09 02:22:17,919 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=231201.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:22:31,026 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.570e+02 2.490e+02 3.133e+02 4.186e+02 8.287e+02, threshold=6.266e+02, percent-clipped=6.0 +2023-02-09 02:22:31,772 INFO [train.py:901] (3/4) Epoch 29, batch 4900, loss[loss=0.1479, simple_loss=0.2369, pruned_loss=0.02941, over 7546.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2809, pruned_loss=0.05736, over 1607586.65 frames. ], batch size: 18, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:22:35,614 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=231226.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:22:40,495 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.6794, 2.4514, 3.2269, 2.6831, 3.3150, 2.7209, 2.6000, 2.0839], + device='cuda:3'), covar=tensor([0.5624, 0.5471, 0.2138, 0.4188, 0.2568, 0.3186, 0.1918, 0.5903], + device='cuda:3'), in_proj_covar=tensor([0.0964, 0.1030, 0.0838, 0.1002, 0.1024, 0.0935, 0.0775, 0.0854], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 02:22:54,383 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=231252.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:22:59,245 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=231258.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:23:04,603 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=231266.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:23:06,615 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=231269.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:23:06,654 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=231269.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:23:07,919 INFO [train.py:901] (3/4) Epoch 29, batch 4950, loss[loss=0.1972, simple_loss=0.2909, pruned_loss=0.05179, over 8682.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2813, pruned_loss=0.05777, over 1609380.44 frames. ], batch size: 34, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:23:17,259 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=231283.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:23:19,709 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.29 vs. limit=2.0 +2023-02-09 02:23:22,763 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=231291.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:23:43,160 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.636e+02 2.476e+02 2.910e+02 3.581e+02 7.956e+02, threshold=5.820e+02, percent-clipped=2.0 +2023-02-09 02:23:43,854 INFO [train.py:901] (3/4) Epoch 29, batch 5000, loss[loss=0.2374, simple_loss=0.3186, pruned_loss=0.07814, over 8520.00 frames. ], tot_loss[loss=0.1991, simple_loss=0.2818, pruned_loss=0.05816, over 1609368.64 frames. ], batch size: 28, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:23:51,804 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.8169, 1.6311, 2.3788, 1.4658, 1.4257, 2.3269, 0.7320, 1.5556], + device='cuda:3'), covar=tensor([0.1536, 0.1201, 0.0359, 0.1158, 0.2270, 0.0439, 0.1832, 0.1317], + device='cuda:3'), in_proj_covar=tensor([0.0203, 0.0206, 0.0138, 0.0224, 0.0281, 0.0149, 0.0174, 0.0201], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 02:24:17,060 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=231367.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:24:19,688 INFO [train.py:901] (3/4) Epoch 29, batch 5050, loss[loss=0.2246, simple_loss=0.3022, pruned_loss=0.07356, over 7255.00 frames. ], tot_loss[loss=0.1996, simple_loss=0.283, pruned_loss=0.05814, over 1611175.00 frames. ], batch size: 71, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:24:30,653 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.81 vs. limit=5.0 +2023-02-09 02:24:37,786 WARNING [train.py:1067] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-02-09 02:24:55,716 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.552e+02 2.315e+02 2.860e+02 3.491e+02 8.708e+02, threshold=5.721e+02, percent-clipped=8.0 +2023-02-09 02:24:56,376 INFO [train.py:901] (3/4) Epoch 29, batch 5100, loss[loss=0.229, simple_loss=0.3145, pruned_loss=0.07178, over 8353.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.2825, pruned_loss=0.05811, over 1608761.46 frames. ], batch size: 24, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:25:32,301 INFO [train.py:901] (3/4) Epoch 29, batch 5150, loss[loss=0.2454, simple_loss=0.3257, pruned_loss=0.08257, over 6715.00 frames. ], tot_loss[loss=0.1992, simple_loss=0.2822, pruned_loss=0.05811, over 1603223.06 frames. ], batch size: 72, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:26:08,162 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.585e+02 2.313e+02 2.816e+02 3.592e+02 7.666e+02, threshold=5.632e+02, percent-clipped=6.0 +2023-02-09 02:26:08,942 INFO [train.py:901] (3/4) Epoch 29, batch 5200, loss[loss=0.2207, simple_loss=0.3055, pruned_loss=0.06795, over 8281.00 frames. ], tot_loss[loss=0.1987, simple_loss=0.2819, pruned_loss=0.05773, over 1603236.88 frames. ], batch size: 49, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:26:12,091 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=231525.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:26:18,580 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.72 vs. limit=2.0 +2023-02-09 02:26:30,212 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=231550.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:26:39,123 WARNING [train.py:1067] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-02-09 02:26:44,753 INFO [train.py:901] (3/4) Epoch 29, batch 5250, loss[loss=0.2123, simple_loss=0.3028, pruned_loss=0.06096, over 8355.00 frames. ], tot_loss[loss=0.199, simple_loss=0.2825, pruned_loss=0.05778, over 1608187.53 frames. ], batch size: 24, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:27:00,050 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.44 vs. limit=2.0 +2023-02-09 02:27:15,800 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=231613.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:27:20,505 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.520e+02 2.485e+02 2.947e+02 3.559e+02 7.815e+02, threshold=5.893e+02, percent-clipped=2.0 +2023-02-09 02:27:21,231 INFO [train.py:901] (3/4) Epoch 29, batch 5300, loss[loss=0.2046, simple_loss=0.2824, pruned_loss=0.06344, over 7781.00 frames. ], tot_loss[loss=0.1986, simple_loss=0.2825, pruned_loss=0.05738, over 1605616.03 frames. ], batch size: 19, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:27:22,864 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=231623.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:27:41,036 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=231648.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:27:53,040 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=231665.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:27:57,010 INFO [train.py:901] (3/4) Epoch 29, batch 5350, loss[loss=0.2417, simple_loss=0.3295, pruned_loss=0.07691, over 8563.00 frames. ], tot_loss[loss=0.1993, simple_loss=0.2828, pruned_loss=0.05789, over 1604638.78 frames. ], batch size: 34, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:28:32,684 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.493e+02 2.535e+02 3.138e+02 3.956e+02 6.651e+02, threshold=6.276e+02, percent-clipped=5.0 +2023-02-09 02:28:33,441 INFO [train.py:901] (3/4) Epoch 29, batch 5400, loss[loss=0.1961, simple_loss=0.2788, pruned_loss=0.05663, over 8631.00 frames. ], tot_loss[loss=0.1999, simple_loss=0.2831, pruned_loss=0.05833, over 1606596.50 frames. ], batch size: 39, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:28:38,398 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=231728.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:28:52,251 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.7432, 4.7099, 4.3029, 2.4628, 4.2663, 4.3415, 4.2791, 4.1785], + device='cuda:3'), covar=tensor([0.0597, 0.0428, 0.0888, 0.4034, 0.0790, 0.1009, 0.1196, 0.0739], + device='cuda:3'), in_proj_covar=tensor([0.0543, 0.0458, 0.0449, 0.0561, 0.0440, 0.0468, 0.0440, 0.0411], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 02:29:09,665 INFO [train.py:901] (3/4) Epoch 29, batch 5450, loss[loss=0.1965, simple_loss=0.2791, pruned_loss=0.05691, over 8081.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.2821, pruned_loss=0.05742, over 1607365.20 frames. ], batch size: 21, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:29:29,511 WARNING [train.py:1067] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-02-09 02:29:44,870 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.390e+02 2.416e+02 2.826e+02 3.521e+02 6.915e+02, threshold=5.653e+02, percent-clipped=1.0 +2023-02-09 02:29:45,620 INFO [train.py:901] (3/4) Epoch 29, batch 5500, loss[loss=0.1739, simple_loss=0.2518, pruned_loss=0.04799, over 7563.00 frames. ], tot_loss[loss=0.198, simple_loss=0.282, pruned_loss=0.05704, over 1611838.16 frames. ], batch size: 18, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:30:18,916 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6014, 1.9430, 3.1443, 1.4399, 2.5588, 2.0610, 1.6195, 2.6114], + device='cuda:3'), covar=tensor([0.2094, 0.3007, 0.0943, 0.5169, 0.1964, 0.3558, 0.2668, 0.2355], + device='cuda:3'), in_proj_covar=tensor([0.0542, 0.0641, 0.0566, 0.0674, 0.0664, 0.0611, 0.0568, 0.0646], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 02:30:21,385 INFO [train.py:901] (3/4) Epoch 29, batch 5550, loss[loss=0.2097, simple_loss=0.2993, pruned_loss=0.06008, over 8328.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2817, pruned_loss=0.05699, over 1612461.44 frames. ], batch size: 25, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:30:56,568 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.629e+02 2.336e+02 2.917e+02 3.506e+02 1.057e+03, threshold=5.834e+02, percent-clipped=5.0 +2023-02-09 02:30:57,329 INFO [train.py:901] (3/4) Epoch 29, batch 5600, loss[loss=0.229, simple_loss=0.2963, pruned_loss=0.08081, over 7218.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2808, pruned_loss=0.05657, over 1607179.83 frames. ], batch size: 71, lr: 2.59e-03, grad_scale: 8.0 +2023-02-09 02:31:20,492 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.5674, 1.6322, 2.3717, 1.2776, 1.0546, 2.3397, 0.3603, 1.3991], + device='cuda:3'), covar=tensor([0.1834, 0.1204, 0.0383, 0.1476, 0.2745, 0.0404, 0.2057, 0.1285], + device='cuda:3'), in_proj_covar=tensor([0.0202, 0.0205, 0.0137, 0.0225, 0.0279, 0.0149, 0.0173, 0.0200], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 02:31:34,510 INFO [train.py:901] (3/4) Epoch 29, batch 5650, loss[loss=0.2013, simple_loss=0.2872, pruned_loss=0.05767, over 8334.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2815, pruned_loss=0.05645, over 1613155.03 frames. ], batch size: 25, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:31:38,815 WARNING [train.py:1067] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-02-09 02:31:43,823 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=231984.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:31:58,720 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.5566, 1.4559, 1.6596, 1.3952, 0.8242, 1.4352, 1.4754, 1.3584], + device='cuda:3'), covar=tensor([0.0610, 0.1226, 0.1686, 0.1502, 0.0618, 0.1440, 0.0737, 0.0678], + device='cuda:3'), in_proj_covar=tensor([0.0100, 0.0154, 0.0189, 0.0162, 0.0102, 0.0163, 0.0114, 0.0147], + device='cuda:3'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0009, 0.0007, 0.0008], + device='cuda:3') +2023-02-09 02:32:02,833 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=232009.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:32:02,964 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=232009.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:32:10,319 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.554e+02 2.351e+02 2.719e+02 3.536e+02 6.635e+02, threshold=5.438e+02, percent-clipped=1.0 +2023-02-09 02:32:11,030 INFO [train.py:901] (3/4) Epoch 29, batch 5700, loss[loss=0.1865, simple_loss=0.2641, pruned_loss=0.05447, over 8091.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2825, pruned_loss=0.05676, over 1617636.24 frames. ], batch size: 21, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:32:31,552 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.8246, 3.7874, 3.4238, 1.9515, 3.3603, 3.5017, 3.3485, 3.3800], + device='cuda:3'), covar=tensor([0.0855, 0.0609, 0.1100, 0.4498, 0.1005, 0.1202, 0.1419, 0.0972], + device='cuda:3'), in_proj_covar=tensor([0.0544, 0.0457, 0.0447, 0.0560, 0.0441, 0.0466, 0.0440, 0.0410], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 02:32:44,169 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.82 vs. limit=2.0 +2023-02-09 02:32:45,198 WARNING [train.py:1067] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-02-09 02:32:47,263 INFO [train.py:901] (3/4) Epoch 29, batch 5750, loss[loss=0.2615, simple_loss=0.3369, pruned_loss=0.09309, over 8256.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2812, pruned_loss=0.05644, over 1616308.97 frames. ], batch size: 24, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:32:54,477 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.7015, 1.5318, 3.1938, 1.4850, 2.4485, 3.5058, 3.7394, 2.6410], + device='cuda:3'), covar=tensor([0.1530, 0.1998, 0.0402, 0.2370, 0.1000, 0.0352, 0.0522, 0.0878], + device='cuda:3'), in_proj_covar=tensor([0.0310, 0.0328, 0.0296, 0.0325, 0.0326, 0.0278, 0.0446, 0.0310], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:3') +2023-02-09 02:33:02,540 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.87 vs. limit=2.0 +2023-02-09 02:33:23,816 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.487e+02 2.417e+02 3.013e+02 3.730e+02 1.097e+03, threshold=6.026e+02, percent-clipped=6.0 +2023-02-09 02:33:24,556 INFO [train.py:901] (3/4) Epoch 29, batch 5800, loss[loss=0.2071, simple_loss=0.2899, pruned_loss=0.06213, over 8557.00 frames. ], tot_loss[loss=0.1962, simple_loss=0.2802, pruned_loss=0.05612, over 1610751.17 frames. ], batch size: 48, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:33:26,860 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=232124.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:33:34,978 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([5.5989, 5.7265, 5.0393, 2.5870, 4.9748, 5.3380, 5.2578, 5.1361], + device='cuda:3'), covar=tensor([0.0502, 0.0333, 0.0786, 0.3953, 0.0749, 0.0757, 0.0934, 0.0616], + device='cuda:3'), in_proj_covar=tensor([0.0544, 0.0457, 0.0448, 0.0560, 0.0441, 0.0467, 0.0440, 0.0411], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 02:33:59,561 INFO [train.py:901] (3/4) Epoch 29, batch 5850, loss[loss=0.1768, simple_loss=0.269, pruned_loss=0.04223, over 8367.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2803, pruned_loss=0.05579, over 1612919.91 frames. ], batch size: 24, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:34:34,659 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.378e+02 2.421e+02 2.869e+02 3.503e+02 7.290e+02, threshold=5.737e+02, percent-clipped=3.0 +2023-02-09 02:34:35,358 INFO [train.py:901] (3/4) Epoch 29, batch 5900, loss[loss=0.1782, simple_loss=0.2668, pruned_loss=0.04483, over 8352.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.281, pruned_loss=0.05604, over 1616797.51 frames. ], batch size: 24, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:35:06,618 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=232265.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:35:10,675 INFO [train.py:901] (3/4) Epoch 29, batch 5950, loss[loss=0.2766, simple_loss=0.3489, pruned_loss=0.1022, over 6978.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2814, pruned_loss=0.05628, over 1619803.28 frames. ], batch size: 71, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:35:46,608 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.609e+02 2.503e+02 2.837e+02 3.700e+02 9.228e+02, threshold=5.675e+02, percent-clipped=4.0 +2023-02-09 02:35:47,358 INFO [train.py:901] (3/4) Epoch 29, batch 6000, loss[loss=0.2634, simple_loss=0.3421, pruned_loss=0.09232, over 8318.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2804, pruned_loss=0.05582, over 1614279.02 frames. ], batch size: 25, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:35:47,358 INFO [train.py:926] (3/4) Computing validation loss +2023-02-09 02:36:01,201 INFO [train.py:935] (3/4) Epoch 29, validation: loss=0.1708, simple_loss=0.2701, pruned_loss=0.03577, over 944034.00 frames. +2023-02-09 02:36:01,202 INFO [train.py:936] (3/4) Maximum memory allocated so far is 6473MB +2023-02-09 02:36:37,558 INFO [train.py:901] (3/4) Epoch 29, batch 6050, loss[loss=0.1931, simple_loss=0.2918, pruned_loss=0.04721, over 8510.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2802, pruned_loss=0.05577, over 1614637.50 frames. ], batch size: 26, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:36:44,009 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=232380.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:36:57,841 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=232399.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:37:02,393 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=232405.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:37:12,725 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.752e+02 2.608e+02 3.073e+02 4.031e+02 7.869e+02, threshold=6.145e+02, percent-clipped=3.0 +2023-02-09 02:37:13,453 INFO [train.py:901] (3/4) Epoch 29, batch 6100, loss[loss=0.1767, simple_loss=0.282, pruned_loss=0.03565, over 8696.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2808, pruned_loss=0.05611, over 1615342.90 frames. ], batch size: 39, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:37:27,129 WARNING [train.py:1067] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-02-09 02:37:49,995 INFO [train.py:901] (3/4) Epoch 29, batch 6150, loss[loss=0.1847, simple_loss=0.2675, pruned_loss=0.05096, over 8241.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2803, pruned_loss=0.05613, over 1614584.55 frames. ], batch size: 22, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:38:06,464 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=232494.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:38:25,294 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.683e+02 2.548e+02 2.901e+02 3.636e+02 6.365e+02, threshold=5.801e+02, percent-clipped=1.0 +2023-02-09 02:38:25,882 INFO [train.py:901] (3/4) Epoch 29, batch 6200, loss[loss=0.2143, simple_loss=0.3027, pruned_loss=0.06296, over 8102.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2812, pruned_loss=0.05693, over 1611772.82 frames. ], batch size: 23, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:38:36,019 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.8524, 1.4296, 2.9099, 1.4910, 2.2338, 3.1211, 3.2632, 2.6680], + device='cuda:3'), covar=tensor([0.1121, 0.1717, 0.0344, 0.2045, 0.0872, 0.0278, 0.0595, 0.0555], + device='cuda:3'), in_proj_covar=tensor([0.0309, 0.0326, 0.0295, 0.0325, 0.0324, 0.0278, 0.0444, 0.0309], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:3') +2023-02-09 02:38:36,692 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=232536.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:39:02,790 INFO [train.py:901] (3/4) Epoch 29, batch 6250, loss[loss=0.2022, simple_loss=0.2923, pruned_loss=0.05601, over 8495.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.281, pruned_loss=0.05695, over 1612859.85 frames. ], batch size: 39, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:39:06,499 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.5958, 1.9931, 3.1058, 1.4705, 2.4554, 2.0883, 1.6997, 2.5706], + device='cuda:3'), covar=tensor([0.2068, 0.2857, 0.0834, 0.4966, 0.1931, 0.3461, 0.2613, 0.2130], + device='cuda:3'), in_proj_covar=tensor([0.0544, 0.0642, 0.0565, 0.0676, 0.0667, 0.0617, 0.0570, 0.0648], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 02:39:30,166 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=232609.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:39:36,610 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.7400, 1.9345, 2.0221, 1.4440, 2.2109, 1.4469, 0.7763, 1.9724], + device='cuda:3'), covar=tensor([0.0764, 0.0474, 0.0397, 0.0718, 0.0500, 0.1224, 0.1071, 0.0387], + device='cuda:3'), in_proj_covar=tensor([0.0477, 0.0419, 0.0373, 0.0465, 0.0400, 0.0560, 0.0409, 0.0444], + device='cuda:3'), out_proj_covar=tensor([1.2637e-04, 1.0845e-04, 9.7002e-05, 1.2166e-04, 1.0481e-04, 1.5611e-04, + 1.0894e-04, 1.1617e-04], device='cuda:3') +2023-02-09 02:39:37,821 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.584e+02 2.332e+02 2.762e+02 3.491e+02 8.673e+02, threshold=5.523e+02, percent-clipped=4.0 +2023-02-09 02:39:39,185 INFO [train.py:901] (3/4) Epoch 29, batch 6300, loss[loss=0.2046, simple_loss=0.2743, pruned_loss=0.06744, over 7566.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2803, pruned_loss=0.05678, over 1609004.60 frames. ], batch size: 18, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:39:43,048 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.78 vs. limit=2.0 +2023-02-09 02:40:14,943 INFO [train.py:901] (3/4) Epoch 29, batch 6350, loss[loss=0.1947, simple_loss=0.2707, pruned_loss=0.05929, over 8090.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2806, pruned_loss=0.05664, over 1610245.13 frames. ], batch size: 21, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:40:50,669 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.474e+02 2.465e+02 3.018e+02 3.778e+02 1.284e+03, threshold=6.036e+02, percent-clipped=3.0 +2023-02-09 02:40:51,316 INFO [train.py:901] (3/4) Epoch 29, batch 6400, loss[loss=0.1946, simple_loss=0.2799, pruned_loss=0.05463, over 7909.00 frames. ], tot_loss[loss=0.1961, simple_loss=0.28, pruned_loss=0.05612, over 1610204.62 frames. ], batch size: 20, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:40:53,593 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=232724.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:41:08,031 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=232743.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:41:28,283 INFO [train.py:901] (3/4) Epoch 29, batch 6450, loss[loss=0.1984, simple_loss=0.2934, pruned_loss=0.05166, over 8497.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.2804, pruned_loss=0.05654, over 1614395.85 frames. ], batch size: 28, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:42:05,009 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.603e+02 2.358e+02 3.084e+02 4.266e+02 9.590e+02, threshold=6.168e+02, percent-clipped=5.0 +2023-02-09 02:42:05,756 INFO [train.py:901] (3/4) Epoch 29, batch 6500, loss[loss=0.2039, simple_loss=0.2951, pruned_loss=0.05632, over 8247.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2805, pruned_loss=0.05694, over 1613095.60 frames. ], batch size: 24, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:42:07,055 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.37 vs. limit=2.0 +2023-02-09 02:42:17,442 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=232838.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:42:23,947 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-02-09 02:42:31,263 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=232858.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:42:40,782 INFO [train.py:901] (3/4) Epoch 29, batch 6550, loss[loss=0.2341, simple_loss=0.3272, pruned_loss=0.07049, over 8355.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2811, pruned_loss=0.05727, over 1610584.67 frames. ], batch size: 24, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:42:47,140 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=232880.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:42:47,794 WARNING [train.py:1067] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-02-09 02:42:48,128 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.20 vs. limit=2.0 +2023-02-09 02:43:08,163 WARNING [train.py:1067] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-02-09 02:43:16,617 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.606e+02 2.398e+02 2.846e+02 3.696e+02 7.042e+02, threshold=5.692e+02, percent-clipped=2.0 +2023-02-09 02:43:17,342 INFO [train.py:901] (3/4) Epoch 29, batch 6600, loss[loss=0.1948, simple_loss=0.2849, pruned_loss=0.0524, over 8247.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2798, pruned_loss=0.05635, over 1609339.20 frames. ], batch size: 24, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:43:32,636 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-02-09 02:43:40,631 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=232953.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:43:52,849 INFO [train.py:901] (3/4) Epoch 29, batch 6650, loss[loss=0.2316, simple_loss=0.3208, pruned_loss=0.07118, over 8255.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2806, pruned_loss=0.05648, over 1611993.81 frames. ], batch size: 24, lr: 2.58e-03, grad_scale: 16.0 +2023-02-09 02:43:59,979 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=232980.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:44:10,262 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=232995.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:44:17,494 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=233005.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:44:19,564 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=233008.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:44:28,482 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.750e+02 2.388e+02 2.820e+02 3.481e+02 6.998e+02, threshold=5.640e+02, percent-clipped=2.0 +2023-02-09 02:44:28,502 INFO [train.py:901] (3/4) Epoch 29, batch 6700, loss[loss=0.1948, simple_loss=0.2881, pruned_loss=0.05074, over 8239.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2798, pruned_loss=0.05592, over 1613859.74 frames. ], batch size: 22, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:44:40,621 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=233036.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:44:42,172 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6393, 2.3393, 3.9603, 1.5748, 2.7196, 2.1887, 1.8507, 2.8999], + device='cuda:3'), covar=tensor([0.2012, 0.2701, 0.0973, 0.4895, 0.2160, 0.3555, 0.2583, 0.2505], + device='cuda:3'), in_proj_covar=tensor([0.0543, 0.0640, 0.0564, 0.0674, 0.0664, 0.0613, 0.0567, 0.0646], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 02:45:06,278 INFO [train.py:901] (3/4) Epoch 29, batch 6750, loss[loss=0.2162, simple_loss=0.2925, pruned_loss=0.06997, over 8286.00 frames. ], tot_loss[loss=0.1951, simple_loss=0.2789, pruned_loss=0.05569, over 1608599.71 frames. ], batch size: 23, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:45:30,754 WARNING [train.py:1067] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-02-09 02:45:38,426 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=233114.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:45:41,383 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6438, 2.1650, 3.2041, 1.5075, 2.5736, 1.9950, 1.8273, 2.5908], + device='cuda:3'), covar=tensor([0.2068, 0.2833, 0.0965, 0.5096, 0.1964, 0.3749, 0.2597, 0.2393], + device='cuda:3'), in_proj_covar=tensor([0.0543, 0.0640, 0.0565, 0.0675, 0.0663, 0.0615, 0.0567, 0.0646], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 02:45:42,082 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.0161, 2.4170, 3.7002, 1.8991, 2.0800, 3.6282, 0.6331, 2.2286], + device='cuda:3'), covar=tensor([0.1165, 0.1126, 0.0209, 0.1630, 0.2120, 0.0284, 0.2140, 0.1282], + device='cuda:3'), in_proj_covar=tensor([0.0204, 0.0206, 0.0138, 0.0226, 0.0280, 0.0149, 0.0173, 0.0201], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 02:45:43,343 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.753e+02 2.536e+02 3.042e+02 3.988e+02 8.675e+02, threshold=6.084e+02, percent-clipped=8.0 +2023-02-09 02:45:43,363 INFO [train.py:901] (3/4) Epoch 29, batch 6800, loss[loss=0.199, simple_loss=0.2811, pruned_loss=0.0584, over 8289.00 frames. ], tot_loss[loss=0.1947, simple_loss=0.2789, pruned_loss=0.05527, over 1613366.02 frames. ], batch size: 23, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:45:56,023 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=233139.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:46:19,947 INFO [train.py:901] (3/4) Epoch 29, batch 6850, loss[loss=0.1587, simple_loss=0.2486, pruned_loss=0.03442, over 7540.00 frames. ], tot_loss[loss=0.1952, simple_loss=0.2797, pruned_loss=0.05533, over 1613452.80 frames. ], batch size: 18, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:46:22,005 WARNING [train.py:1067] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-02-09 02:46:46,505 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=233209.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:46:55,109 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.759e+02 2.470e+02 3.068e+02 3.817e+02 7.038e+02, threshold=6.136e+02, percent-clipped=5.0 +2023-02-09 02:46:55,130 INFO [train.py:901] (3/4) Epoch 29, batch 6900, loss[loss=0.1771, simple_loss=0.2652, pruned_loss=0.04453, over 7653.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.281, pruned_loss=0.05624, over 1614088.57 frames. ], batch size: 19, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:47:04,474 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=233234.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:47:16,262 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=233251.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:47:31,598 INFO [train.py:901] (3/4) Epoch 29, batch 6950, loss[loss=0.1902, simple_loss=0.271, pruned_loss=0.05472, over 7810.00 frames. ], tot_loss[loss=0.1956, simple_loss=0.2798, pruned_loss=0.05572, over 1614873.46 frames. ], batch size: 20, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:47:33,687 WARNING [train.py:1067] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-02-09 02:47:35,275 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=233276.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:47:35,910 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=233277.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:48:03,958 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.4529, 2.8480, 2.2696, 3.8562, 1.8496, 2.2501, 2.5788, 2.9013], + device='cuda:3'), covar=tensor([0.0685, 0.0782, 0.0789, 0.0238, 0.1100, 0.1226, 0.0873, 0.0752], + device='cuda:3'), in_proj_covar=tensor([0.0231, 0.0195, 0.0245, 0.0214, 0.0202, 0.0246, 0.0249, 0.0205], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:3') +2023-02-09 02:48:07,353 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.648e+02 2.388e+02 2.860e+02 3.725e+02 6.106e+02, threshold=5.720e+02, percent-clipped=0.0 +2023-02-09 02:48:07,373 INFO [train.py:901] (3/4) Epoch 29, batch 7000, loss[loss=0.1695, simple_loss=0.2658, pruned_loss=0.0366, over 8134.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.28, pruned_loss=0.05586, over 1615161.22 frames. ], batch size: 22, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:48:30,364 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=233352.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:48:44,309 INFO [train.py:901] (3/4) Epoch 29, batch 7050, loss[loss=0.1681, simple_loss=0.2514, pruned_loss=0.04246, over 7972.00 frames. ], tot_loss[loss=0.1951, simple_loss=0.279, pruned_loss=0.05565, over 1612217.77 frames. ], batch size: 21, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:48:51,050 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=233380.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:49:22,023 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.696e+02 2.400e+02 3.088e+02 3.796e+02 6.683e+02, threshold=6.176e+02, percent-clipped=2.0 +2023-02-09 02:49:22,043 INFO [train.py:901] (3/4) Epoch 29, batch 7100, loss[loss=0.2031, simple_loss=0.2814, pruned_loss=0.06239, over 7908.00 frames. ], tot_loss[loss=0.1953, simple_loss=0.2793, pruned_loss=0.05561, over 1613836.09 frames. ], batch size: 20, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:49:30,716 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.5252, 1.4172, 1.8282, 1.1942, 1.1774, 1.7985, 0.2826, 1.2131], + device='cuda:3'), covar=tensor([0.1451, 0.1113, 0.0395, 0.0919, 0.2255, 0.0463, 0.1766, 0.1121], + device='cuda:3'), in_proj_covar=tensor([0.0202, 0.0205, 0.0137, 0.0225, 0.0279, 0.0148, 0.0172, 0.0200], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 02:49:48,018 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.29 vs. limit=2.0 +2023-02-09 02:49:49,428 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.80 vs. limit=2.0 +2023-02-09 02:49:54,759 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=233467.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:49:57,421 INFO [train.py:901] (3/4) Epoch 29, batch 7150, loss[loss=0.1666, simple_loss=0.2437, pruned_loss=0.04473, over 7715.00 frames. ], tot_loss[loss=0.1955, simple_loss=0.2798, pruned_loss=0.05562, over 1615878.44 frames. ], batch size: 18, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:50:14,748 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=233495.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:50:17,637 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.0816, 2.2334, 1.8553, 2.8200, 1.3744, 1.6729, 2.1004, 2.2091], + device='cuda:3'), covar=tensor([0.0670, 0.0749, 0.0853, 0.0340, 0.0957, 0.1173, 0.0662, 0.0760], + device='cuda:3'), in_proj_covar=tensor([0.0231, 0.0195, 0.0244, 0.0214, 0.0202, 0.0246, 0.0249, 0.0206], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:3') +2023-02-09 02:50:34,192 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.538e+02 2.385e+02 2.900e+02 3.377e+02 5.605e+02, threshold=5.800e+02, percent-clipped=0.0 +2023-02-09 02:50:34,212 INFO [train.py:901] (3/4) Epoch 29, batch 7200, loss[loss=0.222, simple_loss=0.3009, pruned_loss=0.07156, over 8651.00 frames. ], tot_loss[loss=0.1954, simple_loss=0.2797, pruned_loss=0.05555, over 1615001.86 frames. ], batch size: 34, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:51:10,563 INFO [train.py:901] (3/4) Epoch 29, batch 7250, loss[loss=0.1833, simple_loss=0.2679, pruned_loss=0.04938, over 8025.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2803, pruned_loss=0.05586, over 1615776.06 frames. ], batch size: 22, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:51:15,245 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.7435, 1.6301, 2.4443, 1.5053, 1.3668, 2.3241, 0.5819, 1.4428], + device='cuda:3'), covar=tensor([0.1565, 0.1194, 0.0292, 0.1169, 0.2241, 0.0413, 0.1790, 0.1276], + device='cuda:3'), in_proj_covar=tensor([0.0204, 0.0207, 0.0139, 0.0228, 0.0281, 0.0150, 0.0174, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 02:51:46,282 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.552e+02 2.384e+02 2.959e+02 3.398e+02 1.041e+03, threshold=5.918e+02, percent-clipped=4.0 +2023-02-09 02:51:46,302 INFO [train.py:901] (3/4) Epoch 29, batch 7300, loss[loss=0.1867, simple_loss=0.2792, pruned_loss=0.04706, over 8349.00 frames. ], tot_loss[loss=0.1945, simple_loss=0.2785, pruned_loss=0.0552, over 1611492.12 frames. ], batch size: 26, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:51:46,383 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=233621.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:52:13,023 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=233657.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:52:22,373 INFO [train.py:901] (3/4) Epoch 29, batch 7350, loss[loss=0.2009, simple_loss=0.299, pruned_loss=0.0514, over 8458.00 frames. ], tot_loss[loss=0.1953, simple_loss=0.2794, pruned_loss=0.05566, over 1612410.21 frames. ], batch size: 27, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:52:28,050 WARNING [train.py:1067] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-02-09 02:52:43,007 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.5811, 1.9922, 1.9818, 1.2976, 2.1158, 1.3900, 0.7574, 1.7902], + device='cuda:3'), covar=tensor([0.1124, 0.0476, 0.0503, 0.0908, 0.0630, 0.1345, 0.1271, 0.0535], + device='cuda:3'), in_proj_covar=tensor([0.0478, 0.0417, 0.0374, 0.0464, 0.0400, 0.0558, 0.0409, 0.0447], + device='cuda:3'), out_proj_covar=tensor([1.2646e-04, 1.0781e-04, 9.7219e-05, 1.2119e-04, 1.0461e-04, 1.5514e-04, + 1.0893e-04, 1.1695e-04], device='cuda:3') +2023-02-09 02:52:47,223 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.79 vs. limit=2.0 +2023-02-09 02:52:48,258 WARNING [train.py:1067] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-02-09 02:52:58,116 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.683e+02 2.663e+02 3.074e+02 4.100e+02 9.512e+02, threshold=6.147e+02, percent-clipped=7.0 +2023-02-09 02:52:58,137 INFO [train.py:901] (3/4) Epoch 29, batch 7400, loss[loss=0.1787, simple_loss=0.2613, pruned_loss=0.04799, over 8032.00 frames. ], tot_loss[loss=0.1961, simple_loss=0.2804, pruned_loss=0.05588, over 1614905.22 frames. ], batch size: 22, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:52:59,679 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=233723.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:53:03,228 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=233728.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:53:09,649 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=233736.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:53:18,945 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=233748.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:53:21,098 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=233751.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:53:26,159 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.3416, 2.7108, 2.1259, 3.5732, 1.9185, 2.0312, 2.4629, 2.6295], + device='cuda:3'), covar=tensor([0.0737, 0.0737, 0.0830, 0.0368, 0.0937, 0.1172, 0.0875, 0.0771], + device='cuda:3'), in_proj_covar=tensor([0.0231, 0.0195, 0.0245, 0.0214, 0.0202, 0.0246, 0.0250, 0.0206], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:3') +2023-02-09 02:53:32,427 WARNING [train.py:1067] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-02-09 02:53:36,088 INFO [train.py:901] (3/4) Epoch 29, batch 7450, loss[loss=0.1666, simple_loss=0.2451, pruned_loss=0.04405, over 7705.00 frames. ], tot_loss[loss=0.195, simple_loss=0.2795, pruned_loss=0.0553, over 1613620.00 frames. ], batch size: 18, lr: 2.58e-03, grad_scale: 8.0 +2023-02-09 02:53:39,821 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=233776.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:53:45,832 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.20 vs. limit=2.0 +2023-02-09 02:54:07,541 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.8398, 1.4916, 1.6973, 1.3612, 0.9661, 1.4785, 1.6417, 1.4794], + device='cuda:3'), covar=tensor([0.0622, 0.1284, 0.1759, 0.1565, 0.0641, 0.1517, 0.0753, 0.0697], + device='cuda:3'), in_proj_covar=tensor([0.0101, 0.0155, 0.0191, 0.0163, 0.0102, 0.0165, 0.0115, 0.0148], + device='cuda:3'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0008, 0.0008], + device='cuda:3') +2023-02-09 02:54:12,670 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.619e+02 2.228e+02 2.637e+02 3.402e+02 7.399e+02, threshold=5.273e+02, percent-clipped=2.0 +2023-02-09 02:54:12,692 INFO [train.py:901] (3/4) Epoch 29, batch 7500, loss[loss=0.1907, simple_loss=0.2836, pruned_loss=0.04893, over 8479.00 frames. ], tot_loss[loss=0.1953, simple_loss=0.2796, pruned_loss=0.05557, over 1614163.36 frames. ], batch size: 25, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:54:12,803 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.4021, 4.3997, 3.9516, 1.8834, 3.8596, 4.0445, 3.9426, 3.8268], + device='cuda:3'), covar=tensor([0.0705, 0.0512, 0.0958, 0.4913, 0.0968, 0.0942, 0.1168, 0.0801], + device='cuda:3'), in_proj_covar=tensor([0.0544, 0.0458, 0.0449, 0.0560, 0.0444, 0.0465, 0.0443, 0.0411], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 02:54:27,537 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.7351, 4.7482, 4.2065, 2.0971, 4.1456, 4.3899, 4.2908, 4.2072], + device='cuda:3'), covar=tensor([0.0596, 0.0474, 0.0933, 0.4467, 0.0872, 0.0876, 0.1128, 0.0823], + device='cuda:3'), in_proj_covar=tensor([0.0544, 0.0458, 0.0450, 0.0560, 0.0445, 0.0465, 0.0443, 0.0411], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 02:54:48,749 INFO [train.py:901] (3/4) Epoch 29, batch 7550, loss[loss=0.1925, simple_loss=0.2828, pruned_loss=0.0511, over 8507.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2803, pruned_loss=0.05618, over 1615407.49 frames. ], batch size: 26, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:55:24,355 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.668e+02 2.424e+02 2.935e+02 3.534e+02 7.288e+02, threshold=5.870e+02, percent-clipped=3.0 +2023-02-09 02:55:24,376 INFO [train.py:901] (3/4) Epoch 29, batch 7600, loss[loss=0.2227, simple_loss=0.3072, pruned_loss=0.06906, over 8247.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2808, pruned_loss=0.05639, over 1613992.24 frames. ], batch size: 24, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:55:47,835 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.64 vs. limit=5.0 +2023-02-09 02:56:01,026 INFO [train.py:901] (3/4) Epoch 29, batch 7650, loss[loss=0.1948, simple_loss=0.2933, pruned_loss=0.04816, over 8339.00 frames. ], tot_loss[loss=0.1962, simple_loss=0.2806, pruned_loss=0.05584, over 1611723.02 frames. ], batch size: 26, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:56:16,595 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=233992.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:56:23,685 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=234001.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:56:35,529 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=234017.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:56:38,045 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.742e+02 2.208e+02 2.731e+02 3.331e+02 6.993e+02, threshold=5.462e+02, percent-clipped=2.0 +2023-02-09 02:56:38,066 INFO [train.py:901] (3/4) Epoch 29, batch 7700, loss[loss=0.1704, simple_loss=0.2508, pruned_loss=0.04502, over 7811.00 frames. ], tot_loss[loss=0.1956, simple_loss=0.2803, pruned_loss=0.05549, over 1612733.83 frames. ], batch size: 19, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:56:49,503 WARNING [train.py:1067] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-02-09 02:56:53,160 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=234043.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:57:12,874 INFO [train.py:901] (3/4) Epoch 29, batch 7750, loss[loss=0.1827, simple_loss=0.2666, pruned_loss=0.04943, over 7810.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2811, pruned_loss=0.05579, over 1618503.09 frames. ], batch size: 20, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:57:13,648 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=234072.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:57:45,238 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=234116.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:57:48,461 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.737e+02 2.355e+02 2.809e+02 3.505e+02 7.382e+02, threshold=5.617e+02, percent-clipped=2.0 +2023-02-09 02:57:48,481 INFO [train.py:901] (3/4) Epoch 29, batch 7800, loss[loss=0.1839, simple_loss=0.2808, pruned_loss=0.04351, over 8464.00 frames. ], tot_loss[loss=0.1955, simple_loss=0.2802, pruned_loss=0.05544, over 1620578.92 frames. ], batch size: 25, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:58:21,909 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.60 vs. limit=2.0 +2023-02-09 02:58:24,429 INFO [train.py:901] (3/4) Epoch 29, batch 7850, loss[loss=0.1916, simple_loss=0.2717, pruned_loss=0.05574, over 7928.00 frames. ], tot_loss[loss=0.1945, simple_loss=0.2792, pruned_loss=0.05484, over 1619207.99 frames. ], batch size: 20, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:58:36,022 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=234187.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:58:58,948 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=234220.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:58:59,503 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.576e+02 2.264e+02 2.745e+02 3.500e+02 8.048e+02, threshold=5.490e+02, percent-clipped=4.0 +2023-02-09 02:58:59,523 INFO [train.py:901] (3/4) Epoch 29, batch 7900, loss[loss=0.2169, simple_loss=0.3037, pruned_loss=0.06511, over 8294.00 frames. ], tot_loss[loss=0.1942, simple_loss=0.2786, pruned_loss=0.05485, over 1616104.42 frames. ], batch size: 23, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:59:19,163 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=234249.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 02:59:34,070 INFO [train.py:901] (3/4) Epoch 29, batch 7950, loss[loss=0.1715, simple_loss=0.26, pruned_loss=0.04153, over 8474.00 frames. ], tot_loss[loss=0.1949, simple_loss=0.2789, pruned_loss=0.05543, over 1613963.24 frames. ], batch size: 27, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 02:59:38,565 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.7767, 1.5279, 2.6395, 1.4083, 2.2269, 2.8627, 2.9570, 2.5348], + device='cuda:3'), covar=tensor([0.1203, 0.1709, 0.0452, 0.2163, 0.1254, 0.0311, 0.0769, 0.0509], + device='cuda:3'), in_proj_covar=tensor([0.0312, 0.0331, 0.0298, 0.0328, 0.0330, 0.0282, 0.0452, 0.0311], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:3') +2023-02-09 02:59:40,745 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.4526, 4.4436, 4.1083, 2.1945, 3.9976, 4.1443, 4.0671, 3.8626], + device='cuda:3'), covar=tensor([0.0772, 0.0496, 0.0939, 0.4172, 0.0908, 0.0802, 0.1239, 0.0685], + device='cuda:3'), in_proj_covar=tensor([0.0548, 0.0459, 0.0450, 0.0559, 0.0443, 0.0466, 0.0446, 0.0413], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 02:59:51,754 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.7790, 1.9058, 1.6346, 2.3431, 0.9813, 1.4560, 1.6645, 1.8655], + device='cuda:3'), covar=tensor([0.0742, 0.0720, 0.0943, 0.0370, 0.1107, 0.1321, 0.0790, 0.0767], + device='cuda:3'), in_proj_covar=tensor([0.0231, 0.0196, 0.0246, 0.0214, 0.0203, 0.0248, 0.0251, 0.0206], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:3') +2023-02-09 03:00:10,405 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.560e+02 2.366e+02 2.676e+02 3.633e+02 8.832e+02, threshold=5.352e+02, percent-clipped=6.0 +2023-02-09 03:00:10,425 INFO [train.py:901] (3/4) Epoch 29, batch 8000, loss[loss=0.1954, simple_loss=0.2806, pruned_loss=0.05506, over 8294.00 frames. ], tot_loss[loss=0.1951, simple_loss=0.2791, pruned_loss=0.05553, over 1617753.55 frames. ], batch size: 23, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 03:00:34,732 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.0011, 1.5571, 3.4802, 1.5158, 2.4294, 3.7644, 3.9333, 3.2355], + device='cuda:3'), covar=tensor([0.1166, 0.1840, 0.0285, 0.2018, 0.0973, 0.0240, 0.0543, 0.0510], + device='cuda:3'), in_proj_covar=tensor([0.0312, 0.0331, 0.0298, 0.0329, 0.0330, 0.0282, 0.0452, 0.0311], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:3') +2023-02-09 03:00:41,455 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.8160, 1.5046, 1.7630, 1.4425, 1.1144, 1.5152, 1.8005, 1.4448], + device='cuda:3'), covar=tensor([0.0585, 0.1272, 0.1673, 0.1513, 0.0590, 0.1518, 0.0678, 0.0732], + device='cuda:3'), in_proj_covar=tensor([0.0101, 0.0154, 0.0191, 0.0162, 0.0102, 0.0165, 0.0114, 0.0148], + device='cuda:3'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0007, 0.0008], + device='cuda:3') +2023-02-09 03:00:44,626 INFO [train.py:901] (3/4) Epoch 29, batch 8050, loss[loss=0.1822, simple_loss=0.2557, pruned_loss=0.05438, over 7817.00 frames. ], tot_loss[loss=0.1938, simple_loss=0.2767, pruned_loss=0.05542, over 1594933.64 frames. ], batch size: 20, lr: 2.57e-03, grad_scale: 8.0 +2023-02-09 03:00:45,524 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=234372.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:00:55,754 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=234387.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:01:02,566 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=234397.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:01:19,887 WARNING [train.py:1067] (3/4) Exclude cut with ID 3488-85273-0017-111273_sp0.9 from training. Duration: 27.47775 +2023-02-09 03:01:23,757 INFO [train.py:901] (3/4) Epoch 30, batch 0, loss[loss=0.177, simple_loss=0.2614, pruned_loss=0.04627, over 8237.00 frames. ], tot_loss[loss=0.177, simple_loss=0.2614, pruned_loss=0.04627, over 8237.00 frames. ], batch size: 22, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:01:23,757 INFO [train.py:926] (3/4) Computing validation loss +2023-02-09 03:01:35,938 INFO [train.py:935] (3/4) Epoch 30, validation: loss=0.1704, simple_loss=0.27, pruned_loss=0.03537, over 944034.00 frames. +2023-02-09 03:01:35,940 INFO [train.py:936] (3/4) Maximum memory allocated so far is 6543MB +2023-02-09 03:01:47,832 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.533e+02 2.332e+02 2.743e+02 3.464e+02 7.498e+02, threshold=5.486e+02, percent-clipped=3.0 +2023-02-09 03:01:52,036 WARNING [train.py:1067] (3/4) Exclude cut with ID 8291-282929-0007-9590_sp0.9 from training. Duration: 28.72225 +2023-02-09 03:02:04,481 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=234443.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:02:12,492 INFO [train.py:901] (3/4) Epoch 30, batch 50, loss[loss=0.2163, simple_loss=0.3061, pruned_loss=0.06324, over 8612.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2815, pruned_loss=0.05704, over 366781.02 frames. ], batch size: 49, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:02:23,086 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=234468.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:02:28,104 WARNING [train.py:1067] (3/4) Exclude cut with ID 6709-74022-0004-57021_sp1.1 from training. Duration: 0.9409375 +2023-02-09 03:02:49,596 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=234502.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:02:50,885 INFO [train.py:901] (3/4) Epoch 30, batch 100, loss[loss=0.1751, simple_loss=0.2532, pruned_loss=0.04853, over 7970.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2827, pruned_loss=0.05616, over 648997.20 frames. ], batch size: 21, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:02:54,576 WARNING [train.py:1067] (3/4) Exclude cut with ID 497-129325-0061-9566_sp1.1 from training. Duration: 0.97725 +2023-02-09 03:03:03,338 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.539e+02 2.294e+02 2.794e+02 3.449e+02 7.855e+02, threshold=5.588e+02, percent-clipped=7.0 +2023-02-09 03:03:28,113 INFO [train.py:901] (3/4) Epoch 30, batch 150, loss[loss=0.2075, simple_loss=0.2748, pruned_loss=0.07016, over 7814.00 frames. ], tot_loss[loss=0.1982, simple_loss=0.2825, pruned_loss=0.05697, over 861691.10 frames. ], batch size: 20, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:03:35,359 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=234564.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:03:56,932 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=234593.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:04:04,598 INFO [train.py:901] (3/4) Epoch 30, batch 200, loss[loss=0.2565, simple_loss=0.3212, pruned_loss=0.09589, over 8548.00 frames. ], tot_loss[loss=0.2001, simple_loss=0.2838, pruned_loss=0.05824, over 1027488.03 frames. ], batch size: 31, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:04:13,774 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.1999, 2.1306, 2.6419, 2.2975, 2.7978, 2.3176, 2.1441, 1.7058], + device='cuda:3'), covar=tensor([0.6176, 0.5334, 0.2308, 0.4212, 0.2646, 0.3584, 0.2113, 0.5527], + device='cuda:3'), in_proj_covar=tensor([0.0973, 0.1039, 0.0847, 0.1012, 0.1034, 0.0945, 0.0781, 0.0861], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 03:04:16,943 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.375e+02 2.329e+02 2.797e+02 3.759e+02 1.341e+03, threshold=5.593e+02, percent-clipped=8.0 +2023-02-09 03:04:40,210 INFO [train.py:901] (3/4) Epoch 30, batch 250, loss[loss=0.2108, simple_loss=0.2931, pruned_loss=0.06421, over 7805.00 frames. ], tot_loss[loss=0.2015, simple_loss=0.2847, pruned_loss=0.05912, over 1160082.22 frames. ], batch size: 19, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:04:48,451 WARNING [train.py:1067] (3/4) Exclude cut with ID 6951-79737-0043-83149_sp0.9 from training. Duration: 28.0944375 +2023-02-09 03:04:57,629 WARNING [train.py:1067] (3/4) Exclude cut with ID 4278-13270-0009-62705_sp0.9 from training. Duration: 25.45 +2023-02-09 03:04:58,550 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=234679.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:05:04,658 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=234687.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:05:16,612 INFO [train.py:901] (3/4) Epoch 30, batch 300, loss[loss=0.1893, simple_loss=0.278, pruned_loss=0.05031, over 8656.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2819, pruned_loss=0.05699, over 1262947.99 frames. ], batch size: 39, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:05:19,745 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=234708.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:05:29,385 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.648e+02 2.381e+02 2.888e+02 3.640e+02 7.253e+02, threshold=5.776e+02, percent-clipped=4.0 +2023-02-09 03:05:45,727 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=234744.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:05:46,513 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6946, 1.9407, 1.9410, 1.3636, 2.1002, 1.5418, 0.5552, 1.9201], + device='cuda:3'), covar=tensor([0.0652, 0.0445, 0.0403, 0.0672, 0.0485, 0.0994, 0.1019, 0.0310], + device='cuda:3'), in_proj_covar=tensor([0.0477, 0.0416, 0.0372, 0.0465, 0.0400, 0.0555, 0.0405, 0.0443], + device='cuda:3'), out_proj_covar=tensor([1.2618e-04, 1.0768e-04, 9.6869e-05, 1.2151e-04, 1.0482e-04, 1.5420e-04, + 1.0793e-04, 1.1575e-04], device='cuda:3') +2023-02-09 03:05:52,651 INFO [train.py:901] (3/4) Epoch 30, batch 350, loss[loss=0.206, simple_loss=0.296, pruned_loss=0.05797, over 8677.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2809, pruned_loss=0.05667, over 1337525.99 frames. ], batch size: 34, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:05:55,593 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=234758.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:06:14,216 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=234783.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:06:27,361 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=234800.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:06:30,080 INFO [train.py:901] (3/4) Epoch 30, batch 400, loss[loss=0.1764, simple_loss=0.2548, pruned_loss=0.049, over 7795.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2813, pruned_loss=0.05659, over 1401330.12 frames. ], batch size: 19, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:06:42,234 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.545e+02 2.513e+02 2.926e+02 3.697e+02 1.204e+03, threshold=5.852e+02, percent-clipped=7.0 +2023-02-09 03:07:06,517 INFO [train.py:901] (3/4) Epoch 30, batch 450, loss[loss=0.2128, simple_loss=0.2826, pruned_loss=0.07149, over 8043.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2807, pruned_loss=0.05617, over 1449740.27 frames. ], batch size: 22, lr: 2.53e-03, grad_scale: 8.0 +2023-02-09 03:07:38,962 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=234900.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:07:42,330 INFO [train.py:901] (3/4) Epoch 30, batch 500, loss[loss=0.2031, simple_loss=0.2755, pruned_loss=0.06537, over 7677.00 frames. ], tot_loss[loss=0.1956, simple_loss=0.2802, pruned_loss=0.05545, over 1484993.77 frames. ], batch size: 19, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:07:54,768 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.531e+02 2.393e+02 2.948e+02 3.833e+02 6.284e+02, threshold=5.896e+02, percent-clipped=1.0 +2023-02-09 03:08:04,864 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=234935.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:08:18,303 INFO [train.py:901] (3/4) Epoch 30, batch 550, loss[loss=0.1937, simple_loss=0.2735, pruned_loss=0.05697, over 7808.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2816, pruned_loss=0.05627, over 1519630.17 frames. ], batch size: 20, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:08:22,869 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=234960.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:08:26,398 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=234964.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:08:34,799 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=234976.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:08:36,257 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=234978.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:08:44,127 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=234989.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:08:54,548 INFO [train.py:901] (3/4) Epoch 30, batch 600, loss[loss=0.225, simple_loss=0.3102, pruned_loss=0.06994, over 8513.00 frames. ], tot_loss[loss=0.1956, simple_loss=0.2804, pruned_loss=0.05545, over 1544180.71 frames. ], batch size: 28, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:09:05,997 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.582e+02 2.485e+02 2.961e+02 3.544e+02 6.861e+02, threshold=5.922e+02, percent-clipped=1.0 +2023-02-09 03:09:10,829 WARNING [train.py:1067] (3/4) Exclude cut with ID 7255-291500-0003-9465_sp0.9 from training. Duration: 29.816625 +2023-02-09 03:09:13,652 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=235031.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:09:29,975 INFO [train.py:901] (3/4) Epoch 30, batch 650, loss[loss=0.2099, simple_loss=0.3, pruned_loss=0.05993, over 8500.00 frames. ], tot_loss[loss=0.1942, simple_loss=0.2789, pruned_loss=0.05476, over 1557865.69 frames. ], batch size: 28, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:09:54,569 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=235088.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:10:05,671 INFO [train.py:901] (3/4) Epoch 30, batch 700, loss[loss=0.1994, simple_loss=0.2875, pruned_loss=0.05565, over 8509.00 frames. ], tot_loss[loss=0.1945, simple_loss=0.2789, pruned_loss=0.05509, over 1565758.31 frames. ], batch size: 26, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:10:12,449 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.51 vs. limit=2.0 +2023-02-09 03:10:17,690 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.543e+02 2.477e+02 3.055e+02 3.959e+02 7.285e+02, threshold=6.109e+02, percent-clipped=6.0 +2023-02-09 03:10:33,426 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=235144.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:10:34,913 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=235146.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:10:41,067 INFO [train.py:901] (3/4) Epoch 30, batch 750, loss[loss=0.2065, simple_loss=0.2992, pruned_loss=0.05688, over 8584.00 frames. ], tot_loss[loss=0.195, simple_loss=0.2794, pruned_loss=0.05532, over 1580201.40 frames. ], batch size: 34, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:10:59,028 WARNING [train.py:1067] (3/4) Exclude cut with ID 5239-32139-0047-92994_sp0.9 from training. Duration: 30.1555625 +2023-02-09 03:10:59,128 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.2573, 3.1797, 2.9842, 1.5663, 2.8835, 3.0004, 2.8311, 2.9046], + device='cuda:3'), covar=tensor([0.1224, 0.0772, 0.1234, 0.4820, 0.1199, 0.1163, 0.1591, 0.1023], + device='cuda:3'), in_proj_covar=tensor([0.0551, 0.0461, 0.0451, 0.0562, 0.0445, 0.0470, 0.0446, 0.0414], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 03:11:08,043 WARNING [train.py:1067] (3/4) Exclude cut with ID 8631-249866-0030-64025_sp0.9 from training. Duration: 26.32775 +2023-02-09 03:11:17,084 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=235203.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:11:17,654 INFO [train.py:901] (3/4) Epoch 30, batch 800, loss[loss=0.2071, simple_loss=0.2929, pruned_loss=0.06061, over 8336.00 frames. ], tot_loss[loss=0.1946, simple_loss=0.2786, pruned_loss=0.05532, over 1584321.56 frames. ], batch size: 25, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:11:30,474 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.721e+02 2.417e+02 2.836e+02 3.328e+02 8.160e+02, threshold=5.671e+02, percent-clipped=2.0 +2023-02-09 03:11:46,749 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=235244.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:11:53,489 INFO [train.py:901] (3/4) Epoch 30, batch 850, loss[loss=0.2042, simple_loss=0.3056, pruned_loss=0.05143, over 8260.00 frames. ], tot_loss[loss=0.1941, simple_loss=0.2784, pruned_loss=0.05488, over 1593098.21 frames. ], batch size: 24, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:11:57,057 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=235259.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:12:07,384 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=235273.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:12:20,619 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.6217, 4.6951, 4.1889, 2.1792, 4.0927, 4.3678, 4.1782, 4.1542], + device='cuda:3'), covar=tensor([0.0722, 0.0496, 0.1018, 0.4535, 0.0928, 0.0941, 0.1125, 0.0772], + device='cuda:3'), in_proj_covar=tensor([0.0547, 0.0458, 0.0448, 0.0559, 0.0443, 0.0467, 0.0443, 0.0411], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 03:12:25,087 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.50 vs. limit=2.0 +2023-02-09 03:12:29,572 INFO [train.py:901] (3/4) Epoch 30, batch 900, loss[loss=0.2095, simple_loss=0.2934, pruned_loss=0.06284, over 8605.00 frames. ], tot_loss[loss=0.1944, simple_loss=0.279, pruned_loss=0.05495, over 1599810.32 frames. ], batch size: 34, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:12:41,044 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=235320.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:12:41,602 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.572e+02 2.438e+02 3.006e+02 3.865e+02 6.238e+02, threshold=6.012e+02, percent-clipped=6.0 +2023-02-09 03:12:43,011 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=235322.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:13:05,257 INFO [train.py:901] (3/4) Epoch 30, batch 950, loss[loss=0.2024, simple_loss=0.2921, pruned_loss=0.05636, over 8729.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2802, pruned_loss=0.0557, over 1607640.80 frames. ], batch size: 40, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:13:08,761 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=235359.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:13:21,827 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=235378.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:13:32,472 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.3185, 3.5899, 2.5820, 3.1642, 2.9561, 2.1585, 3.0597, 3.1603], + device='cuda:3'), covar=tensor([0.1554, 0.0371, 0.1036, 0.0593, 0.0732, 0.1469, 0.0934, 0.1047], + device='cuda:3'), in_proj_covar=tensor([0.0357, 0.0243, 0.0344, 0.0314, 0.0302, 0.0349, 0.0350, 0.0321], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-02-09 03:13:32,983 WARNING [train.py:1067] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp1.1 from training. Duration: 25.3818125 +2023-02-09 03:13:39,417 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=235402.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:13:40,160 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.3908, 2.1979, 2.6647, 2.3917, 2.6719, 2.3892, 2.3184, 1.9747], + device='cuda:3'), covar=tensor([0.4786, 0.4597, 0.1988, 0.3380, 0.2225, 0.3017, 0.1792, 0.4576], + device='cuda:3'), in_proj_covar=tensor([0.0978, 0.1045, 0.0853, 0.1015, 0.1038, 0.0949, 0.0782, 0.0865], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 03:13:40,616 INFO [train.py:901] (3/4) Epoch 30, batch 1000, loss[loss=0.2096, simple_loss=0.2893, pruned_loss=0.06499, over 7489.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2811, pruned_loss=0.05594, over 1615674.18 frames. ], batch size: 73, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:13:52,268 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.648e+02 2.479e+02 3.055e+02 4.205e+02 7.814e+02, threshold=6.110e+02, percent-clipped=3.0 +2023-02-09 03:13:56,464 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=235427.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:14:02,533 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=235435.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:14:03,948 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=235437.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:14:08,461 WARNING [train.py:1067] (3/4) Exclude cut with ID 6951-79737-0043-83149 from training. Duration: 25.285 +2023-02-09 03:14:15,938 INFO [train.py:901] (3/4) Epoch 30, batch 1050, loss[loss=0.2468, simple_loss=0.3322, pruned_loss=0.08067, over 8107.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2808, pruned_loss=0.05606, over 1613846.04 frames. ], batch size: 23, lr: 2.52e-03, grad_scale: 16.0 +2023-02-09 03:14:19,402 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=235459.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:14:21,156 WARNING [train.py:1067] (3/4) Exclude cut with ID 2929-85685-0079-61403 from training. Duration: 29.735 +2023-02-09 03:14:36,667 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=235484.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:14:41,817 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.61 vs. limit=2.0 +2023-02-09 03:14:50,595 INFO [train.py:901] (3/4) Epoch 30, batch 1100, loss[loss=0.1945, simple_loss=0.2897, pruned_loss=0.04962, over 8498.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2805, pruned_loss=0.05632, over 1610272.25 frames. ], batch size: 28, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:14:59,100 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=235515.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:15:03,868 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.639e+02 2.444e+02 3.135e+02 3.900e+02 6.752e+02, threshold=6.270e+02, percent-clipped=2.0 +2023-02-09 03:15:04,825 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.9035, 2.1833, 2.2803, 1.5292, 2.3401, 1.7221, 0.8092, 2.1728], + device='cuda:3'), covar=tensor([0.0719, 0.0387, 0.0322, 0.0685, 0.0519, 0.1001, 0.1023, 0.0350], + device='cuda:3'), in_proj_covar=tensor([0.0483, 0.0421, 0.0375, 0.0470, 0.0404, 0.0561, 0.0410, 0.0448], + device='cuda:3'), out_proj_covar=tensor([1.2799e-04, 1.0897e-04, 9.7569e-05, 1.2288e-04, 1.0573e-04, 1.5611e-04, + 1.0930e-04, 1.1726e-04], device='cuda:3') +2023-02-09 03:15:09,971 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.72 vs. limit=5.0 +2023-02-09 03:15:12,924 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.86 vs. limit=5.0 +2023-02-09 03:15:17,338 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=235540.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:15:26,829 INFO [train.py:901] (3/4) Epoch 30, batch 1150, loss[loss=0.1968, simple_loss=0.2758, pruned_loss=0.05892, over 7936.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2811, pruned_loss=0.05674, over 1614730.09 frames. ], batch size: 20, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:15:35,845 WARNING [train.py:1067] (3/4) Exclude cut with ID 7255-291500-0005-9467_sp0.9 from training. Duration: 27.8166875 +2023-02-09 03:15:39,057 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.52 vs. limit=2.0 +2023-02-09 03:16:03,363 INFO [train.py:901] (3/4) Epoch 30, batch 1200, loss[loss=0.1753, simple_loss=0.2504, pruned_loss=0.05011, over 7235.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.28, pruned_loss=0.05591, over 1617291.54 frames. ], batch size: 16, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:16:11,493 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=235615.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:16:12,764 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=235617.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:16:16,061 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.630e+02 2.378e+02 2.945e+02 3.640e+02 8.540e+02, threshold=5.890e+02, percent-clipped=4.0 +2023-02-09 03:16:30,042 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=235640.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:16:32,796 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.4432, 2.4030, 3.1310, 2.6229, 3.0669, 2.5080, 2.4292, 2.0790], + device='cuda:3'), covar=tensor([0.6027, 0.5297, 0.2105, 0.3925, 0.2608, 0.3255, 0.1938, 0.5490], + device='cuda:3'), in_proj_covar=tensor([0.0970, 0.1037, 0.0846, 0.1008, 0.1030, 0.0943, 0.0777, 0.0860], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 03:16:39,500 INFO [train.py:901] (3/4) Epoch 30, batch 1250, loss[loss=0.2007, simple_loss=0.2978, pruned_loss=0.05182, over 8575.00 frames. ], tot_loss[loss=0.195, simple_loss=0.2791, pruned_loss=0.05545, over 1612625.47 frames. ], batch size: 31, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:17:06,190 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=235691.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:17:07,623 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=235693.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:17:10,371 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=235697.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:17:15,249 INFO [train.py:901] (3/4) Epoch 30, batch 1300, loss[loss=0.2289, simple_loss=0.3065, pruned_loss=0.07562, over 8127.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2804, pruned_loss=0.0563, over 1616985.43 frames. ], batch size: 22, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:17:23,824 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=235716.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:17:25,221 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=235718.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:17:27,375 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.3746, 2.3095, 2.9742, 2.4613, 2.8567, 2.4450, 2.3448, 1.9016], + device='cuda:3'), covar=tensor([0.6076, 0.5472, 0.2283, 0.4455, 0.3097, 0.3326, 0.1869, 0.5824], + device='cuda:3'), in_proj_covar=tensor([0.0964, 0.1030, 0.0841, 0.1002, 0.1025, 0.0939, 0.0773, 0.0855], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 03:17:27,741 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.629e+02 2.444e+02 2.774e+02 3.314e+02 6.214e+02, threshold=5.548e+02, percent-clipped=2.0 +2023-02-09 03:17:27,830 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=235722.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:17:34,608 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=235732.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:17:35,434 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.76 vs. limit=5.0 +2023-02-09 03:17:50,125 INFO [train.py:901] (3/4) Epoch 30, batch 1350, loss[loss=0.2339, simple_loss=0.31, pruned_loss=0.07885, over 8449.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2809, pruned_loss=0.05671, over 1614901.54 frames. ], batch size: 29, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:18:26,587 INFO [train.py:901] (3/4) Epoch 30, batch 1400, loss[loss=0.1932, simple_loss=0.2731, pruned_loss=0.05666, over 8556.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2802, pruned_loss=0.05646, over 1615352.04 frames. ], batch size: 49, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:18:30,243 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=235809.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:18:39,099 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.499e+02 2.338e+02 2.741e+02 3.583e+02 7.907e+02, threshold=5.482e+02, percent-clipped=6.0 +2023-02-09 03:18:49,552 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=235837.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:19:00,849 INFO [train.py:901] (3/4) Epoch 30, batch 1450, loss[loss=0.1689, simple_loss=0.251, pruned_loss=0.0434, over 7431.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2807, pruned_loss=0.0565, over 1618089.77 frames. ], batch size: 17, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:19:08,256 WARNING [train.py:1067] (3/4) Exclude cut with ID 8291-282929-0007-9590 from training. Duration: 25.85 +2023-02-09 03:19:22,538 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.57 vs. limit=2.0 +2023-02-09 03:19:38,201 INFO [train.py:901] (3/4) Epoch 30, batch 1500, loss[loss=0.181, simple_loss=0.2758, pruned_loss=0.04313, over 8203.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2809, pruned_loss=0.05591, over 1619559.66 frames. ], batch size: 23, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:19:42,081 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-02-09 03:19:51,219 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.723e+02 2.306e+02 2.900e+02 3.560e+02 8.272e+02, threshold=5.801e+02, percent-clipped=7.0 +2023-02-09 03:19:56,730 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.84 vs. limit=2.0 +2023-02-09 03:20:14,187 INFO [train.py:901] (3/4) Epoch 30, batch 1550, loss[loss=0.1752, simple_loss=0.2647, pruned_loss=0.04287, over 7819.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2803, pruned_loss=0.05565, over 1617597.77 frames. ], batch size: 20, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:20:17,113 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.8606, 1.2785, 3.4979, 1.5438, 2.5920, 3.8906, 4.0503, 3.3836], + device='cuda:3'), covar=tensor([0.1213, 0.2113, 0.0290, 0.2080, 0.0912, 0.0213, 0.0535, 0.0542], + device='cuda:3'), in_proj_covar=tensor([0.0309, 0.0329, 0.0296, 0.0329, 0.0327, 0.0282, 0.0447, 0.0310], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:3') +2023-02-09 03:20:22,056 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-02-09 03:20:27,557 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.3063, 1.2842, 1.5443, 1.0294, 1.0326, 1.5487, 0.2603, 1.0790], + device='cuda:3'), covar=tensor([0.1112, 0.0887, 0.0323, 0.0659, 0.1864, 0.0392, 0.1467, 0.1058], + device='cuda:3'), in_proj_covar=tensor([0.0203, 0.0208, 0.0139, 0.0226, 0.0280, 0.0149, 0.0175, 0.0202], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 03:20:38,890 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=235988.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:20:51,283 INFO [train.py:901] (3/4) Epoch 30, batch 1600, loss[loss=0.2051, simple_loss=0.2936, pruned_loss=0.05828, over 8521.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2801, pruned_loss=0.05581, over 1613555.40 frames. ], batch size: 28, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:20:57,831 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=236013.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:21:00,085 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.2902, 2.1319, 1.7227, 1.9954, 1.8181, 1.5505, 1.7248, 1.7408], + device='cuda:3'), covar=tensor([0.1283, 0.0418, 0.1141, 0.0504, 0.0702, 0.1457, 0.0920, 0.0836], + device='cuda:3'), in_proj_covar=tensor([0.0360, 0.0246, 0.0349, 0.0317, 0.0304, 0.0352, 0.0354, 0.0325], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-02-09 03:21:04,931 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.669e+02 2.693e+02 3.134e+02 4.092e+02 8.333e+02, threshold=6.267e+02, percent-clipped=7.0 +2023-02-09 03:21:15,458 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=236036.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:21:19,101 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=236041.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:21:28,385 INFO [train.py:901] (3/4) Epoch 30, batch 1650, loss[loss=0.21, simple_loss=0.2935, pruned_loss=0.0633, over 8106.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.281, pruned_loss=0.05598, over 1618438.42 frames. ], batch size: 23, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:21:56,790 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=236093.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:22:04,816 INFO [train.py:901] (3/4) Epoch 30, batch 1700, loss[loss=0.1843, simple_loss=0.2748, pruned_loss=0.04686, over 8109.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2814, pruned_loss=0.05665, over 1615550.08 frames. ], batch size: 23, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:22:15,382 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=236118.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:22:17,731 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.648e+02 2.438e+02 2.804e+02 3.459e+02 5.840e+02, threshold=5.608e+02, percent-clipped=0.0 +2023-02-09 03:22:34,514 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.31 vs. limit=5.0 +2023-02-09 03:22:40,514 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=236153.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:22:41,155 INFO [train.py:901] (3/4) Epoch 30, batch 1750, loss[loss=0.18, simple_loss=0.2755, pruned_loss=0.04224, over 8243.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2807, pruned_loss=0.05646, over 1615241.55 frames. ], batch size: 24, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:22:42,663 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=236156.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:23:07,927 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.0424, 1.8915, 2.2858, 1.9947, 2.2603, 2.0638, 2.0045, 1.6026], + device='cuda:3'), covar=tensor([0.4324, 0.4071, 0.1875, 0.3251, 0.2104, 0.2791, 0.1587, 0.4168], + device='cuda:3'), in_proj_covar=tensor([0.0972, 0.1037, 0.0848, 0.1008, 0.1032, 0.0946, 0.0778, 0.0860], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 03:23:16,215 INFO [train.py:901] (3/4) Epoch 30, batch 1800, loss[loss=0.1682, simple_loss=0.2568, pruned_loss=0.03982, over 7709.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.28, pruned_loss=0.05592, over 1613154.61 frames. ], batch size: 18, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:23:29,389 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.538e+02 2.503e+02 3.165e+02 3.852e+02 7.294e+02, threshold=6.329e+02, percent-clipped=5.0 +2023-02-09 03:23:52,514 INFO [train.py:901] (3/4) Epoch 30, batch 1850, loss[loss=0.3074, simple_loss=0.385, pruned_loss=0.1149, over 8572.00 frames. ], tot_loss[loss=0.1962, simple_loss=0.2804, pruned_loss=0.05605, over 1609491.91 frames. ], batch size: 39, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:24:03,965 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=236268.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:24:28,804 INFO [train.py:901] (3/4) Epoch 30, batch 1900, loss[loss=0.2564, simple_loss=0.3195, pruned_loss=0.0967, over 7065.00 frames. ], tot_loss[loss=0.1954, simple_loss=0.2795, pruned_loss=0.05569, over 1607829.58 frames. ], batch size: 72, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:24:41,167 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.65 vs. limit=2.0 +2023-02-09 03:24:41,427 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.517e+02 2.325e+02 3.004e+02 3.832e+02 8.674e+02, threshold=6.008e+02, percent-clipped=3.0 +2023-02-09 03:25:01,728 WARNING [train.py:1067] (3/4) Exclude cut with ID 7357-94126-0021-44397_sp0.9 from training. Duration: 27.511125 +2023-02-09 03:25:05,198 INFO [train.py:901] (3/4) Epoch 30, batch 1950, loss[loss=0.1908, simple_loss=0.2824, pruned_loss=0.04957, over 8350.00 frames. ], tot_loss[loss=0.1945, simple_loss=0.2787, pruned_loss=0.05515, over 1610727.10 frames. ], batch size: 24, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:25:13,610 WARNING [train.py:1067] (3/4) Exclude cut with ID 7357-94126-0014-44390 from training. Duration: 27.92 +2023-02-09 03:25:24,201 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=236380.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:25:25,043 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.4670, 2.7327, 3.0191, 1.8797, 3.1745, 1.9923, 1.7120, 2.3853], + device='cuda:3'), covar=tensor([0.0773, 0.0416, 0.0329, 0.0810, 0.0542, 0.0835, 0.1046, 0.0579], + device='cuda:3'), in_proj_covar=tensor([0.0477, 0.0418, 0.0373, 0.0465, 0.0400, 0.0555, 0.0405, 0.0442], + device='cuda:3'), out_proj_covar=tensor([1.2627e-04, 1.0814e-04, 9.7134e-05, 1.2130e-04, 1.0478e-04, 1.5435e-04, + 1.0794e-04, 1.1561e-04], device='cuda:3') +2023-02-09 03:25:25,065 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.8308, 2.3670, 4.0771, 1.6923, 3.0264, 2.4743, 1.8610, 3.0156], + device='cuda:3'), covar=tensor([0.1881, 0.2781, 0.0749, 0.4826, 0.1895, 0.3229, 0.2633, 0.2459], + device='cuda:3'), in_proj_covar=tensor([0.0546, 0.0646, 0.0569, 0.0681, 0.0670, 0.0622, 0.0573, 0.0651], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 03:25:33,317 WARNING [train.py:1067] (3/4) Exclude cut with ID 4964-30587-0040-138716_sp0.9 from training. Duration: 25.0944375 +2023-02-09 03:25:41,056 INFO [train.py:901] (3/4) Epoch 30, batch 2000, loss[loss=0.2087, simple_loss=0.3011, pruned_loss=0.05816, over 8451.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.2801, pruned_loss=0.05581, over 1616316.87 frames. ], batch size: 39, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:25:43,381 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.9382, 1.7217, 2.0409, 1.8608, 2.0481, 2.0105, 1.8869, 0.9202], + device='cuda:3'), covar=tensor([0.6179, 0.4873, 0.2324, 0.3639, 0.2515, 0.3272, 0.2014, 0.5044], + device='cuda:3'), in_proj_covar=tensor([0.0970, 0.1036, 0.0846, 0.1007, 0.1032, 0.0947, 0.0778, 0.0857], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 03:25:46,882 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=236412.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:25:53,568 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.624e+02 2.406e+02 2.956e+02 3.756e+02 9.982e+02, threshold=5.913e+02, percent-clipped=8.0 +2023-02-09 03:26:04,549 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=236437.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:26:16,287 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6675, 1.7882, 1.5317, 2.2689, 1.0002, 1.3946, 1.6685, 1.8466], + device='cuda:3'), covar=tensor([0.0765, 0.0696, 0.0985, 0.0388, 0.0975, 0.1304, 0.0701, 0.0723], + device='cuda:3'), in_proj_covar=tensor([0.0231, 0.0194, 0.0244, 0.0215, 0.0201, 0.0247, 0.0251, 0.0206], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:3') +2023-02-09 03:26:16,759 INFO [train.py:901] (3/4) Epoch 30, batch 2050, loss[loss=0.2484, simple_loss=0.3204, pruned_loss=0.08822, over 8521.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2805, pruned_loss=0.05603, over 1618286.81 frames. ], batch size: 26, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:26:28,377 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.3273, 2.1348, 1.7249, 1.9965, 1.8477, 1.5312, 1.7736, 1.6834], + device='cuda:3'), covar=tensor([0.1300, 0.0481, 0.1293, 0.0586, 0.0788, 0.1637, 0.0934, 0.0930], + device='cuda:3'), in_proj_covar=tensor([0.0359, 0.0245, 0.0348, 0.0316, 0.0304, 0.0351, 0.0354, 0.0323], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-02-09 03:26:47,588 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=236495.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:26:53,532 INFO [train.py:901] (3/4) Epoch 30, batch 2100, loss[loss=0.2061, simple_loss=0.2905, pruned_loss=0.06087, over 8248.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2805, pruned_loss=0.05676, over 1614542.96 frames. ], batch size: 24, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:27:06,244 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.478e+02 2.477e+02 3.089e+02 3.892e+02 8.089e+02, threshold=6.178e+02, percent-clipped=3.0 +2023-02-09 03:27:07,830 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=236524.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:27:16,745 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.4163, 1.2575, 2.3765, 1.4654, 2.2089, 2.5338, 2.7344, 2.1546], + device='cuda:3'), covar=tensor([0.1186, 0.1576, 0.0426, 0.2045, 0.0800, 0.0395, 0.0652, 0.0659], + device='cuda:3'), in_proj_covar=tensor([0.0310, 0.0330, 0.0298, 0.0329, 0.0329, 0.0282, 0.0449, 0.0312], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:3') +2023-02-09 03:27:25,243 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=236549.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:27:28,493 INFO [train.py:901] (3/4) Epoch 30, batch 2150, loss[loss=0.1848, simple_loss=0.2696, pruned_loss=0.04996, over 7816.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2802, pruned_loss=0.05631, over 1617684.80 frames. ], batch size: 20, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:27:59,355 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.75 vs. limit=5.0 +2023-02-09 03:28:04,582 INFO [train.py:901] (3/4) Epoch 30, batch 2200, loss[loss=0.2296, simple_loss=0.3067, pruned_loss=0.07629, over 8247.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2805, pruned_loss=0.05616, over 1619329.29 frames. ], batch size: 24, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:28:06,919 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.3894, 4.3695, 4.0041, 1.9793, 3.8845, 4.0432, 3.8827, 3.8512], + device='cuda:3'), covar=tensor([0.0811, 0.0522, 0.0965, 0.4690, 0.0927, 0.0884, 0.1243, 0.0840], + device='cuda:3'), in_proj_covar=tensor([0.0552, 0.0462, 0.0450, 0.0565, 0.0447, 0.0474, 0.0450, 0.0415], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 03:28:18,768 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.663e+02 2.435e+02 2.816e+02 3.564e+02 9.413e+02, threshold=5.632e+02, percent-clipped=3.0 +2023-02-09 03:28:40,981 INFO [train.py:901] (3/4) Epoch 30, batch 2250, loss[loss=0.1909, simple_loss=0.2745, pruned_loss=0.05369, over 7921.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.281, pruned_loss=0.05705, over 1616615.39 frames. ], batch size: 20, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:29:04,191 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.58 vs. limit=2.0 +2023-02-09 03:29:16,922 INFO [train.py:901] (3/4) Epoch 30, batch 2300, loss[loss=0.1957, simple_loss=0.2776, pruned_loss=0.05689, over 7623.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2807, pruned_loss=0.05678, over 1614636.32 frames. ], batch size: 17, lr: 2.52e-03, grad_scale: 8.0 +2023-02-09 03:29:29,252 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.819e+02 2.489e+02 3.036e+02 4.215e+02 7.962e+02, threshold=6.071e+02, percent-clipped=6.0 +2023-02-09 03:29:50,611 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=236751.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:29:52,586 INFO [train.py:901] (3/4) Epoch 30, batch 2350, loss[loss=0.1824, simple_loss=0.2686, pruned_loss=0.04808, over 8332.00 frames. ], tot_loss[loss=0.1961, simple_loss=0.2795, pruned_loss=0.05635, over 1609530.31 frames. ], batch size: 26, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:30:08,613 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=236776.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:30:29,061 INFO [train.py:901] (3/4) Epoch 30, batch 2400, loss[loss=0.1678, simple_loss=0.2674, pruned_loss=0.03413, over 8127.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.281, pruned_loss=0.05674, over 1609803.39 frames. ], batch size: 22, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:30:42,226 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.757e+02 2.281e+02 2.685e+02 3.729e+02 8.099e+02, threshold=5.371e+02, percent-clipped=9.0 +2023-02-09 03:31:05,196 INFO [train.py:901] (3/4) Epoch 30, batch 2450, loss[loss=0.1914, simple_loss=0.2831, pruned_loss=0.04981, over 8241.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2808, pruned_loss=0.05681, over 1610683.25 frames. ], batch size: 24, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:31:39,706 INFO [train.py:901] (3/4) Epoch 30, batch 2500, loss[loss=0.1695, simple_loss=0.2491, pruned_loss=0.04497, over 7242.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2814, pruned_loss=0.05713, over 1613911.94 frames. ], batch size: 16, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:31:43,939 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=236910.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:31:52,883 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.501e+02 2.373e+02 3.086e+02 3.767e+02 7.222e+02, threshold=6.171e+02, percent-clipped=6.0 +2023-02-09 03:32:13,799 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.4078, 3.8271, 2.6439, 3.2047, 2.9957, 2.2074, 3.1309, 3.2861], + device='cuda:3'), covar=tensor([0.1648, 0.0355, 0.1126, 0.0676, 0.0838, 0.1505, 0.1103, 0.1064], + device='cuda:3'), in_proj_covar=tensor([0.0356, 0.0242, 0.0345, 0.0314, 0.0302, 0.0350, 0.0351, 0.0320], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-02-09 03:32:16,575 INFO [train.py:901] (3/4) Epoch 30, batch 2550, loss[loss=0.1791, simple_loss=0.2594, pruned_loss=0.04943, over 7669.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2806, pruned_loss=0.05683, over 1612238.18 frames. ], batch size: 19, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:32:54,178 INFO [train.py:901] (3/4) Epoch 30, batch 2600, loss[loss=0.1887, simple_loss=0.2812, pruned_loss=0.0481, over 8203.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2806, pruned_loss=0.05672, over 1611598.60 frames. ], batch size: 23, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:33:06,912 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.551e+02 2.458e+02 3.021e+02 3.974e+02 8.394e+02, threshold=6.042e+02, percent-clipped=5.0 +2023-02-09 03:33:12,328 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.61 vs. limit=5.0 +2023-02-09 03:33:30,300 INFO [train.py:901] (3/4) Epoch 30, batch 2650, loss[loss=0.1987, simple_loss=0.292, pruned_loss=0.05275, over 8318.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2817, pruned_loss=0.057, over 1613770.72 frames. ], batch size: 26, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:34:06,434 INFO [train.py:901] (3/4) Epoch 30, batch 2700, loss[loss=0.2586, simple_loss=0.3407, pruned_loss=0.08829, over 8468.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.281, pruned_loss=0.05639, over 1609543.06 frames. ], batch size: 27, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:34:07,585 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-02-09 03:34:18,973 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.680e+02 2.464e+02 3.015e+02 4.068e+02 7.247e+02, threshold=6.030e+02, percent-clipped=1.0 +2023-02-09 03:34:41,474 INFO [train.py:901] (3/4) Epoch 30, batch 2750, loss[loss=0.1855, simple_loss=0.2693, pruned_loss=0.05091, over 7794.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2804, pruned_loss=0.05612, over 1606025.25 frames. ], batch size: 19, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:35:18,239 INFO [train.py:901] (3/4) Epoch 30, batch 2800, loss[loss=0.1779, simple_loss=0.2673, pruned_loss=0.04424, over 7928.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2806, pruned_loss=0.05668, over 1604167.53 frames. ], batch size: 20, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:35:20,497 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.9612, 1.4497, 1.6878, 1.3582, 1.0075, 1.3980, 1.7048, 1.5585], + device='cuda:3'), covar=tensor([0.0563, 0.1289, 0.1720, 0.1540, 0.0633, 0.1602, 0.0724, 0.0684], + device='cuda:3'), in_proj_covar=tensor([0.0101, 0.0155, 0.0192, 0.0163, 0.0102, 0.0165, 0.0114, 0.0149], + device='cuda:3'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0008, 0.0008], + device='cuda:3') +2023-02-09 03:35:31,345 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.593e+02 2.300e+02 2.824e+02 3.573e+02 8.919e+02, threshold=5.648e+02, percent-clipped=3.0 +2023-02-09 03:35:45,300 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.8405, 1.3972, 4.0049, 1.5864, 3.5485, 3.3807, 3.6386, 3.5690], + device='cuda:3'), covar=tensor([0.0713, 0.4573, 0.0614, 0.4190, 0.1113, 0.0965, 0.0723, 0.0697], + device='cuda:3'), in_proj_covar=tensor([0.0694, 0.0671, 0.0752, 0.0670, 0.0756, 0.0645, 0.0655, 0.0729], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 03:35:48,279 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-02-09 03:35:53,018 INFO [train.py:901] (3/4) Epoch 30, batch 2850, loss[loss=0.2225, simple_loss=0.2994, pruned_loss=0.07277, over 8360.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2808, pruned_loss=0.05689, over 1606538.97 frames. ], batch size: 24, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:35:53,088 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=237254.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:36:18,914 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.84 vs. limit=2.0 +2023-02-09 03:36:27,596 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.60 vs. limit=2.0 +2023-02-09 03:36:29,227 INFO [train.py:901] (3/4) Epoch 30, batch 2900, loss[loss=0.2433, simple_loss=0.3173, pruned_loss=0.08464, over 7444.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2813, pruned_loss=0.05722, over 1611423.89 frames. ], batch size: 71, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:36:42,593 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.704e+02 2.592e+02 3.021e+02 4.387e+02 8.419e+02, threshold=6.042e+02, percent-clipped=5.0 +2023-02-09 03:37:04,018 WARNING [train.py:1067] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp1.1 from training. Duration: 0.7545625 +2023-02-09 03:37:05,362 INFO [train.py:901] (3/4) Epoch 30, batch 2950, loss[loss=0.2158, simple_loss=0.3094, pruned_loss=0.06105, over 8353.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2809, pruned_loss=0.05658, over 1613641.59 frames. ], batch size: 24, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:37:15,789 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=237369.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:37:36,850 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.2815, 3.1483, 2.9730, 1.4347, 2.8913, 2.9292, 2.8328, 2.8276], + device='cuda:3'), covar=tensor([0.1148, 0.0754, 0.1252, 0.4599, 0.1111, 0.1186, 0.1481, 0.0980], + device='cuda:3'), in_proj_covar=tensor([0.0550, 0.0461, 0.0450, 0.0562, 0.0445, 0.0473, 0.0447, 0.0415], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 03:37:40,341 INFO [train.py:901] (3/4) Epoch 30, batch 3000, loss[loss=0.1815, simple_loss=0.2577, pruned_loss=0.05271, over 7767.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.2814, pruned_loss=0.05737, over 1610430.00 frames. ], batch size: 19, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:37:40,341 INFO [train.py:926] (3/4) Computing validation loss +2023-02-09 03:37:54,058 INFO [train.py:935] (3/4) Epoch 30, validation: loss=0.1704, simple_loss=0.2697, pruned_loss=0.0356, over 944034.00 frames. +2023-02-09 03:37:54,059 INFO [train.py:936] (3/4) Maximum memory allocated so far is 6543MB +2023-02-09 03:38:07,361 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.697e+02 2.369e+02 2.918e+02 3.560e+02 6.316e+02, threshold=5.836e+02, percent-clipped=1.0 +2023-02-09 03:38:31,181 INFO [train.py:901] (3/4) Epoch 30, batch 3050, loss[loss=0.2387, simple_loss=0.3141, pruned_loss=0.08162, over 8324.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2808, pruned_loss=0.05728, over 1606288.78 frames. ], batch size: 25, lr: 2.51e-03, grad_scale: 8.0 +2023-02-09 03:39:01,958 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.8957, 1.9773, 1.9677, 1.5736, 2.0758, 1.7234, 1.2600, 1.9573], + device='cuda:3'), covar=tensor([0.0549, 0.0386, 0.0280, 0.0564, 0.0370, 0.0704, 0.0786, 0.0362], + device='cuda:3'), in_proj_covar=tensor([0.0483, 0.0422, 0.0376, 0.0467, 0.0403, 0.0562, 0.0409, 0.0446], + device='cuda:3'), out_proj_covar=tensor([1.2778e-04, 1.0912e-04, 9.7950e-05, 1.2182e-04, 1.0531e-04, 1.5642e-04, + 1.0891e-04, 1.1676e-04], device='cuda:3') +2023-02-09 03:39:07,127 INFO [train.py:901] (3/4) Epoch 30, batch 3100, loss[loss=0.1873, simple_loss=0.2709, pruned_loss=0.05187, over 7445.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.28, pruned_loss=0.0572, over 1606049.79 frames. ], batch size: 17, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:39:10,844 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=237509.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:39:14,042 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.34 vs. limit=2.0 +2023-02-09 03:39:19,648 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.408e+02 2.429e+02 3.016e+02 3.485e+02 6.483e+02, threshold=6.032e+02, percent-clipped=4.0 +2023-02-09 03:39:20,784 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-02-09 03:39:43,980 INFO [train.py:901] (3/4) Epoch 30, batch 3150, loss[loss=0.1971, simple_loss=0.2935, pruned_loss=0.05032, over 8483.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2796, pruned_loss=0.05677, over 1611375.67 frames. ], batch size: 28, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:40:03,668 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=237581.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:40:21,025 INFO [train.py:901] (3/4) Epoch 30, batch 3200, loss[loss=0.1431, simple_loss=0.2207, pruned_loss=0.03279, over 7936.00 frames. ], tot_loss[loss=0.1962, simple_loss=0.2791, pruned_loss=0.05661, over 1613666.68 frames. ], batch size: 20, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:40:33,294 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.611e+02 2.320e+02 2.861e+02 3.592e+02 8.186e+02, threshold=5.722e+02, percent-clipped=5.0 +2023-02-09 03:40:35,499 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=237625.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:40:53,394 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=237650.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:40:56,034 INFO [train.py:901] (3/4) Epoch 30, batch 3250, loss[loss=0.2033, simple_loss=0.2845, pruned_loss=0.06108, over 6781.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2811, pruned_loss=0.05712, over 1615734.53 frames. ], batch size: 15, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:41:32,173 INFO [train.py:901] (3/4) Epoch 30, batch 3300, loss[loss=0.1475, simple_loss=0.2326, pruned_loss=0.0312, over 7428.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2816, pruned_loss=0.05723, over 1616689.00 frames. ], batch size: 17, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:41:44,069 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.38 vs. limit=5.0 +2023-02-09 03:41:45,789 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.739e+02 2.392e+02 2.907e+02 3.818e+02 6.093e+02, threshold=5.813e+02, percent-clipped=2.0 +2023-02-09 03:42:07,977 INFO [train.py:901] (3/4) Epoch 30, batch 3350, loss[loss=0.2024, simple_loss=0.2888, pruned_loss=0.05795, over 8572.00 frames. ], tot_loss[loss=0.1981, simple_loss=0.2814, pruned_loss=0.0574, over 1616838.98 frames. ], batch size: 31, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:42:44,179 INFO [train.py:901] (3/4) Epoch 30, batch 3400, loss[loss=0.1671, simple_loss=0.2539, pruned_loss=0.04015, over 8087.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2807, pruned_loss=0.05692, over 1618574.38 frames. ], batch size: 21, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:42:47,135 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=237808.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:42:57,411 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.758e+02 2.484e+02 3.245e+02 4.483e+02 9.283e+02, threshold=6.490e+02, percent-clipped=12.0 +2023-02-09 03:43:00,493 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.4091, 1.5899, 1.5848, 1.1771, 1.6510, 1.2304, 0.3458, 1.6357], + device='cuda:3'), covar=tensor([0.0615, 0.0460, 0.0388, 0.0600, 0.0549, 0.1177, 0.1103, 0.0321], + device='cuda:3'), in_proj_covar=tensor([0.0486, 0.0425, 0.0380, 0.0470, 0.0406, 0.0566, 0.0410, 0.0449], + device='cuda:3'), out_proj_covar=tensor([1.2850e-04, 1.0992e-04, 9.8809e-05, 1.2263e-04, 1.0614e-04, 1.5748e-04, + 1.0928e-04, 1.1749e-04], device='cuda:3') +2023-02-09 03:43:19,447 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=237853.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:43:20,121 INFO [train.py:901] (3/4) Epoch 30, batch 3450, loss[loss=0.155, simple_loss=0.2469, pruned_loss=0.03161, over 8238.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2805, pruned_loss=0.05699, over 1612923.52 frames. ], batch size: 22, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:43:56,083 INFO [train.py:901] (3/4) Epoch 30, batch 3500, loss[loss=0.1973, simple_loss=0.2764, pruned_loss=0.05907, over 7547.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2813, pruned_loss=0.05729, over 1614486.15 frames. ], batch size: 18, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:44:08,738 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.712e+02 2.505e+02 3.010e+02 3.725e+02 8.965e+02, threshold=6.019e+02, percent-clipped=4.0 +2023-02-09 03:44:11,005 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=237925.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:44:11,721 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.9200, 1.4879, 3.3178, 1.7276, 2.4183, 3.6290, 3.7385, 3.0864], + device='cuda:3'), covar=tensor([0.1180, 0.1851, 0.0286, 0.1820, 0.1014, 0.0234, 0.0522, 0.0471], + device='cuda:3'), in_proj_covar=tensor([0.0308, 0.0327, 0.0296, 0.0326, 0.0328, 0.0281, 0.0447, 0.0308], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:3') +2023-02-09 03:44:14,960 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=237930.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:44:22,371 WARNING [train.py:1067] (3/4) Exclude cut with ID 453-131332-0000-131866_sp0.9 from training. Duration: 25.3333125 +2023-02-09 03:44:32,809 INFO [train.py:901] (3/4) Epoch 30, batch 3550, loss[loss=0.2041, simple_loss=0.2754, pruned_loss=0.0664, over 8084.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2807, pruned_loss=0.05651, over 1615663.23 frames. ], batch size: 21, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:44:43,166 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=237968.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:44:49,855 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-02-09 03:45:09,425 INFO [train.py:901] (3/4) Epoch 30, batch 3600, loss[loss=0.2144, simple_loss=0.3072, pruned_loss=0.06083, over 8241.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2816, pruned_loss=0.05666, over 1615663.78 frames. ], batch size: 48, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:45:22,458 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.684e+02 2.286e+02 2.832e+02 3.360e+02 7.556e+02, threshold=5.664e+02, percent-clipped=2.0 +2023-02-09 03:45:35,411 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=238040.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:45:40,890 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.5664, 4.5644, 4.1420, 2.0995, 4.0357, 4.2320, 4.0847, 4.1033], + device='cuda:3'), covar=tensor([0.0692, 0.0497, 0.0882, 0.4622, 0.0836, 0.1036, 0.1327, 0.0912], + device='cuda:3'), in_proj_covar=tensor([0.0551, 0.0463, 0.0452, 0.0563, 0.0445, 0.0475, 0.0451, 0.0416], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 03:45:44,900 INFO [train.py:901] (3/4) Epoch 30, batch 3650, loss[loss=0.1907, simple_loss=0.2668, pruned_loss=0.05734, over 7909.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2815, pruned_loss=0.05669, over 1617948.86 frames. ], batch size: 20, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:46:05,849 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=238082.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:46:12,776 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=238092.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:46:20,737 INFO [train.py:901] (3/4) Epoch 30, batch 3700, loss[loss=0.2354, simple_loss=0.3268, pruned_loss=0.07203, over 8540.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2812, pruned_loss=0.0567, over 1613229.84 frames. ], batch size: 39, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:46:29,083 WARNING [train.py:1067] (3/4) Exclude cut with ID 2411-132532-0017-25057_sp1.1 from training. Duration: 0.9681875 +2023-02-09 03:46:33,263 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.392e+02 2.352e+02 3.001e+02 3.686e+02 7.575e+02, threshold=6.003e+02, percent-clipped=3.0 +2023-02-09 03:46:50,315 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=238144.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:46:56,063 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=238152.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:46:57,363 INFO [train.py:901] (3/4) Epoch 30, batch 3750, loss[loss=0.1875, simple_loss=0.2727, pruned_loss=0.05119, over 8097.00 frames. ], tot_loss[loss=0.1978, simple_loss=0.2815, pruned_loss=0.05701, over 1614748.30 frames. ], batch size: 23, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:46:59,943 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.16 vs. limit=2.0 +2023-02-09 03:47:18,051 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.30 vs. limit=5.0 +2023-02-09 03:47:33,674 INFO [train.py:901] (3/4) Epoch 30, batch 3800, loss[loss=0.1975, simple_loss=0.2689, pruned_loss=0.06307, over 7646.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2815, pruned_loss=0.05699, over 1614532.54 frames. ], batch size: 19, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:47:46,040 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.710e+02 2.427e+02 2.911e+02 3.474e+02 7.215e+02, threshold=5.821e+02, percent-clipped=2.0 +2023-02-09 03:47:47,588 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=238224.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:48:05,484 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=238249.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:48:09,469 INFO [train.py:901] (3/4) Epoch 30, batch 3850, loss[loss=0.1777, simple_loss=0.2556, pruned_loss=0.04995, over 7926.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2814, pruned_loss=0.05666, over 1618307.78 frames. ], batch size: 20, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:48:17,656 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-02-09 03:48:18,889 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=238267.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:48:23,698 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=238274.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:48:37,703 WARNING [train.py:1067] (3/4) Exclude cut with ID 3557-8342-0013-71585_sp1.1 from training. Duration: 0.836375 +2023-02-09 03:48:39,921 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=238296.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:48:45,174 INFO [train.py:901] (3/4) Epoch 30, batch 3900, loss[loss=0.2006, simple_loss=0.2949, pruned_loss=0.05309, over 8462.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2811, pruned_loss=0.05636, over 1616995.13 frames. ], batch size: 25, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:48:57,808 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=238321.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:48:58,292 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.463e+02 2.361e+02 2.887e+02 3.538e+02 6.169e+02, threshold=5.773e+02, percent-clipped=2.0 +2023-02-09 03:49:20,476 INFO [train.py:901] (3/4) Epoch 30, batch 3950, loss[loss=0.1714, simple_loss=0.2442, pruned_loss=0.04927, over 7543.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2815, pruned_loss=0.05665, over 1618094.91 frames. ], batch size: 18, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:49:46,427 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=238389.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:49:56,966 INFO [train.py:901] (3/4) Epoch 30, batch 4000, loss[loss=0.1498, simple_loss=0.2297, pruned_loss=0.03493, over 7667.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.281, pruned_loss=0.05632, over 1614658.99 frames. ], batch size: 19, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:50:01,963 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.0221, 1.4734, 1.7144, 1.3473, 1.1054, 1.4340, 1.8276, 1.7446], + device='cuda:3'), covar=tensor([0.0539, 0.1323, 0.1706, 0.1514, 0.0620, 0.1545, 0.0746, 0.0631], + device='cuda:3'), in_proj_covar=tensor([0.0101, 0.0155, 0.0192, 0.0162, 0.0102, 0.0165, 0.0114, 0.0148], + device='cuda:3'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0008, 0.0008], + device='cuda:3') +2023-02-09 03:50:09,960 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.496e+02 2.354e+02 2.920e+02 3.674e+02 8.815e+02, threshold=5.839e+02, percent-clipped=5.0 +2023-02-09 03:50:12,702 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=238426.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:50:20,360 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=238436.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:50:21,793 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=238438.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:50:28,636 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.0933, 2.4742, 3.8120, 2.0411, 2.1761, 3.7848, 0.9660, 2.4483], + device='cuda:3'), covar=tensor([0.1388, 0.1077, 0.0249, 0.1606, 0.2150, 0.0286, 0.1884, 0.1208], + device='cuda:3'), in_proj_covar=tensor([0.0204, 0.0206, 0.0139, 0.0225, 0.0279, 0.0149, 0.0175, 0.0201], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 03:50:32,682 INFO [train.py:901] (3/4) Epoch 30, batch 4050, loss[loss=0.1997, simple_loss=0.288, pruned_loss=0.05567, over 8594.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2803, pruned_loss=0.05583, over 1615852.50 frames. ], batch size: 34, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:50:38,563 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.0723, 2.2001, 1.8788, 2.5490, 1.4731, 1.7389, 2.0658, 2.1594], + device='cuda:3'), covar=tensor([0.0606, 0.0629, 0.0746, 0.0416, 0.0941, 0.1106, 0.0699, 0.0696], + device='cuda:3'), in_proj_covar=tensor([0.0230, 0.0193, 0.0243, 0.0214, 0.0201, 0.0245, 0.0248, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0005, 0.0006, 0.0005], + device='cuda:3') +2023-02-09 03:50:57,403 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=238488.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:51:09,333 INFO [train.py:901] (3/4) Epoch 30, batch 4100, loss[loss=0.1947, simple_loss=0.2872, pruned_loss=0.05108, over 8252.00 frames. ], tot_loss[loss=0.1957, simple_loss=0.2804, pruned_loss=0.05549, over 1614386.33 frames. ], batch size: 24, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:51:21,789 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.512e+02 2.287e+02 2.925e+02 3.934e+02 1.031e+03, threshold=5.850e+02, percent-clipped=7.0 +2023-02-09 03:51:22,761 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=238523.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:51:36,004 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=238541.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:51:41,367 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=238548.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:51:43,289 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=238551.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:51:45,010 INFO [train.py:901] (3/4) Epoch 30, batch 4150, loss[loss=0.2131, simple_loss=0.2943, pruned_loss=0.06593, over 8401.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2809, pruned_loss=0.05606, over 1616338.14 frames. ], batch size: 49, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:51:57,586 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6628, 2.0041, 3.0467, 1.5212, 2.2881, 2.1317, 1.7125, 2.4540], + device='cuda:3'), covar=tensor([0.1953, 0.2805, 0.1012, 0.4897, 0.2122, 0.3437, 0.2597, 0.2479], + device='cuda:3'), in_proj_covar=tensor([0.0548, 0.0646, 0.0571, 0.0682, 0.0675, 0.0622, 0.0574, 0.0653], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 03:52:11,579 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.35 vs. limit=2.0 +2023-02-09 03:52:19,916 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=238603.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:52:20,448 INFO [train.py:901] (3/4) Epoch 30, batch 4200, loss[loss=0.2207, simple_loss=0.2963, pruned_loss=0.07259, over 6900.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2811, pruned_loss=0.05671, over 1611502.93 frames. ], batch size: 71, lr: 2.51e-03, grad_scale: 16.0 +2023-02-09 03:52:30,425 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.7735, 1.3470, 2.8961, 1.5033, 2.2293, 3.1029, 3.2584, 2.6711], + device='cuda:3'), covar=tensor([0.1080, 0.1683, 0.0285, 0.1951, 0.0832, 0.0271, 0.0560, 0.0500], + device='cuda:3'), in_proj_covar=tensor([0.0309, 0.0329, 0.0296, 0.0328, 0.0329, 0.0282, 0.0449, 0.0309], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:3') +2023-02-09 03:52:33,714 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.657e+02 2.494e+02 3.256e+02 4.447e+02 1.288e+03, threshold=6.511e+02, percent-clipped=8.0 +2023-02-09 03:52:41,527 WARNING [train.py:1067] (3/4) Exclude cut with ID 7255-291500-0005-9467 from training. Duration: 25.035 +2023-02-09 03:52:50,648 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=238645.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:52:56,592 INFO [train.py:901] (3/4) Epoch 30, batch 4250, loss[loss=0.1757, simple_loss=0.2575, pruned_loss=0.04696, over 7923.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2818, pruned_loss=0.05701, over 1612170.31 frames. ], batch size: 20, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:53:02,698 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.8194, 1.6717, 2.3427, 1.5118, 1.4102, 2.3418, 0.4169, 1.4686], + device='cuda:3'), covar=tensor([0.1401, 0.1110, 0.0321, 0.1055, 0.2054, 0.0335, 0.1744, 0.1057], + device='cuda:3'), in_proj_covar=tensor([0.0204, 0.0206, 0.0139, 0.0224, 0.0279, 0.0149, 0.0176, 0.0201], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 03:53:05,080 WARNING [train.py:1067] (3/4) Exclude cut with ID 1914-133440-0024-53073_sp0.9 from training. Duration: 25.2444375 +2023-02-09 03:53:07,994 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=238670.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:53:27,925 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=238699.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:53:31,133 INFO [train.py:901] (3/4) Epoch 30, batch 4300, loss[loss=0.2145, simple_loss=0.2895, pruned_loss=0.06973, over 7927.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2817, pruned_loss=0.05713, over 1608924.40 frames. ], batch size: 20, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:53:44,805 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.604e+02 2.303e+02 2.743e+02 3.342e+02 6.438e+02, threshold=5.486e+02, percent-clipped=0.0 +2023-02-09 03:54:04,298 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.3201, 2.1266, 1.7642, 2.0517, 1.8095, 1.5080, 1.7275, 1.7307], + device='cuda:3'), covar=tensor([0.1240, 0.0432, 0.1255, 0.0512, 0.0697, 0.1632, 0.0907, 0.0843], + device='cuda:3'), in_proj_covar=tensor([0.0356, 0.0243, 0.0345, 0.0314, 0.0301, 0.0348, 0.0349, 0.0318], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-02-09 03:54:06,886 INFO [train.py:901] (3/4) Epoch 30, batch 4350, loss[loss=0.2235, simple_loss=0.2952, pruned_loss=0.07594, over 8734.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2812, pruned_loss=0.05689, over 1612799.61 frames. ], batch size: 39, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:54:27,361 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=238782.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:54:36,397 WARNING [train.py:1067] (3/4) Exclude cut with ID 5622-44585-0006-50425_sp0.9 from training. Duration: 28.638875 +2023-02-09 03:54:37,982 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=238797.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:54:42,561 INFO [train.py:901] (3/4) Epoch 30, batch 4400, loss[loss=0.2057, simple_loss=0.2958, pruned_loss=0.05774, over 8099.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2814, pruned_loss=0.05691, over 1607557.57 frames. ], batch size: 23, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:54:44,441 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=4.22 vs. limit=5.0 +2023-02-09 03:54:44,937 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=238807.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:54:53,640 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.72 vs. limit=2.0 +2023-02-09 03:54:55,878 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.538e+02 2.573e+02 3.023e+02 3.983e+02 6.680e+02, threshold=6.046e+02, percent-clipped=2.0 +2023-02-09 03:54:56,114 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=238822.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:55:03,783 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=238832.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:55:15,274 WARNING [train.py:1067] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp0.9 from training. Duration: 33.038875 +2023-02-09 03:55:18,608 INFO [train.py:901] (3/4) Epoch 30, batch 4450, loss[loss=0.1672, simple_loss=0.2429, pruned_loss=0.04578, over 7422.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.282, pruned_loss=0.0574, over 1606355.92 frames. ], batch size: 17, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:55:22,533 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=238859.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:55:40,874 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=238884.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:55:50,474 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=238897.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:55:55,016 INFO [train.py:901] (3/4) Epoch 30, batch 4500, loss[loss=0.213, simple_loss=0.2892, pruned_loss=0.06839, over 8530.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.282, pruned_loss=0.05747, over 1608243.53 frames. ], batch size: 26, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:56:07,437 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.647e+02 2.335e+02 2.828e+02 3.474e+02 8.376e+02, threshold=5.656e+02, percent-clipped=3.0 +2023-02-09 03:56:08,191 WARNING [train.py:1067] (3/4) Exclude cut with ID 3033-130750-0096-107983 from training. Duration: 0.83 +2023-02-09 03:56:12,490 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.6819, 1.7116, 2.0929, 1.8296, 1.1391, 1.8037, 2.3113, 2.0655], + device='cuda:3'), covar=tensor([0.0481, 0.1151, 0.1557, 0.1323, 0.0585, 0.1364, 0.0599, 0.0607], + device='cuda:3'), in_proj_covar=tensor([0.0101, 0.0155, 0.0192, 0.0163, 0.0102, 0.0165, 0.0114, 0.0148], + device='cuda:3'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0008, 0.0008], + device='cuda:3') +2023-02-09 03:56:29,264 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.47 vs. limit=2.0 +2023-02-09 03:56:30,975 INFO [train.py:901] (3/4) Epoch 30, batch 4550, loss[loss=0.1819, simple_loss=0.2587, pruned_loss=0.05252, over 7711.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.2805, pruned_loss=0.05644, over 1605780.77 frames. ], batch size: 18, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:56:31,138 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=238954.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 03:57:06,048 INFO [train.py:901] (3/4) Epoch 30, batch 4600, loss[loss=0.2648, simple_loss=0.3297, pruned_loss=0.09996, over 8563.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.281, pruned_loss=0.05696, over 1607475.60 frames. ], batch size: 31, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:57:19,157 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.552e+02 2.347e+02 2.832e+02 3.443e+02 5.144e+02, threshold=5.665e+02, percent-clipped=0.0 +2023-02-09 03:57:26,213 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.2899, 3.2121, 2.9970, 1.6065, 2.9188, 2.9501, 2.8796, 2.8342], + device='cuda:3'), covar=tensor([0.1071, 0.0779, 0.1208, 0.4463, 0.1113, 0.1281, 0.1634, 0.1097], + device='cuda:3'), in_proj_covar=tensor([0.0556, 0.0466, 0.0458, 0.0571, 0.0451, 0.0481, 0.0455, 0.0420], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 03:57:34,217 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=239043.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:57:40,145 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.73 vs. limit=5.0 +2023-02-09 03:57:41,730 INFO [train.py:901] (3/4) Epoch 30, batch 4650, loss[loss=0.1869, simple_loss=0.2687, pruned_loss=0.05248, over 8468.00 frames. ], tot_loss[loss=0.1982, simple_loss=0.2817, pruned_loss=0.05736, over 1609986.84 frames. ], batch size: 25, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:58:17,764 INFO [train.py:901] (3/4) Epoch 30, batch 4700, loss[loss=0.2333, simple_loss=0.3169, pruned_loss=0.07485, over 8555.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2805, pruned_loss=0.05668, over 1610034.41 frames. ], batch size: 39, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:58:25,197 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.0500, 1.2840, 1.1920, 0.6780, 1.2302, 1.0912, 0.0786, 1.2291], + device='cuda:3'), covar=tensor([0.0522, 0.0429, 0.0396, 0.0670, 0.0464, 0.0978, 0.0957, 0.0359], + device='cuda:3'), in_proj_covar=tensor([0.0481, 0.0420, 0.0378, 0.0467, 0.0402, 0.0560, 0.0407, 0.0449], + device='cuda:3'), out_proj_covar=tensor([1.2733e-04, 1.0839e-04, 9.8362e-05, 1.2200e-04, 1.0492e-04, 1.5570e-04, + 1.0844e-04, 1.1746e-04], device='cuda:3') +2023-02-09 03:58:30,952 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.697e+02 2.353e+02 2.866e+02 3.941e+02 8.957e+02, threshold=5.733e+02, percent-clipped=8.0 +2023-02-09 03:58:52,482 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=239153.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:58:52,925 INFO [train.py:901] (3/4) Epoch 30, batch 4750, loss[loss=0.2127, simple_loss=0.2901, pruned_loss=0.06768, over 8300.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2809, pruned_loss=0.05695, over 1609275.87 frames. ], batch size: 23, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:58:55,953 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=239158.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 03:59:05,580 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.32 vs. limit=5.0 +2023-02-09 03:59:10,889 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=239178.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 03:59:12,056 WARNING [train.py:1067] (3/4) Exclude cut with ID 7255-291500-0003-9465 from training. Duration: 26.8349375 +2023-02-09 03:59:14,172 WARNING [train.py:1067] (3/4) Exclude cut with ID 5622-44585-0006-50425 from training. Duration: 25.775 +2023-02-09 03:59:21,387 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.55 vs. limit=5.0 +2023-02-09 03:59:28,648 INFO [train.py:901] (3/4) Epoch 30, batch 4800, loss[loss=0.2319, simple_loss=0.3098, pruned_loss=0.07704, over 8554.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.281, pruned_loss=0.05688, over 1609459.33 frames. ], batch size: 31, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 03:59:35,684 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.5276, 2.8648, 2.3265, 3.9606, 1.5508, 2.0238, 2.6873, 2.8244], + device='cuda:3'), covar=tensor([0.0690, 0.0696, 0.0767, 0.0227, 0.1095, 0.1221, 0.0822, 0.0740], + device='cuda:3'), in_proj_covar=tensor([0.0233, 0.0195, 0.0245, 0.0215, 0.0203, 0.0247, 0.0250, 0.0205], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:3') +2023-02-09 03:59:41,696 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.637e+02 2.393e+02 3.010e+02 3.751e+02 7.640e+02, threshold=6.020e+02, percent-clipped=2.0 +2023-02-09 03:59:46,063 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.8409, 2.1829, 3.6148, 2.0244, 1.9040, 3.5666, 0.8121, 2.2567], + device='cuda:3'), covar=tensor([0.1097, 0.1211, 0.0227, 0.1439, 0.2103, 0.0243, 0.1834, 0.1079], + device='cuda:3'), in_proj_covar=tensor([0.0203, 0.0205, 0.0138, 0.0223, 0.0276, 0.0148, 0.0174, 0.0200], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 04:00:04,582 INFO [train.py:901] (3/4) Epoch 30, batch 4850, loss[loss=0.1966, simple_loss=0.2821, pruned_loss=0.05551, over 8244.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2805, pruned_loss=0.05655, over 1610562.66 frames. ], batch size: 22, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:00:06,726 WARNING [train.py:1067] (3/4) Exclude cut with ID 3972-170212-0014-103914 from training. Duration: 26.205 +2023-02-09 04:00:36,498 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=239298.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 04:00:40,437 INFO [train.py:901] (3/4) Epoch 30, batch 4900, loss[loss=0.1498, simple_loss=0.2333, pruned_loss=0.03318, over 7440.00 frames. ], tot_loss[loss=0.1955, simple_loss=0.2793, pruned_loss=0.05583, over 1611117.55 frames. ], batch size: 17, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:00:53,056 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.481e+02 2.412e+02 2.818e+02 3.519e+02 1.028e+03, threshold=5.635e+02, percent-clipped=4.0 +2023-02-09 04:01:15,931 INFO [train.py:901] (3/4) Epoch 30, batch 4950, loss[loss=0.2064, simple_loss=0.2897, pruned_loss=0.06157, over 7979.00 frames. ], tot_loss[loss=0.195, simple_loss=0.2787, pruned_loss=0.05571, over 1608818.77 frames. ], batch size: 21, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:01:43,683 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([3.9099, 3.8412, 3.6047, 1.7836, 3.4760, 3.5349, 3.4910, 3.4156], + device='cuda:3'), covar=tensor([0.0825, 0.0569, 0.0919, 0.4236, 0.1004, 0.1064, 0.1311, 0.0789], + device='cuda:3'), in_proj_covar=tensor([0.0556, 0.0465, 0.0455, 0.0569, 0.0451, 0.0480, 0.0453, 0.0418], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 04:01:51,711 INFO [train.py:901] (3/4) Epoch 30, batch 5000, loss[loss=0.1895, simple_loss=0.2587, pruned_loss=0.06013, over 7291.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2801, pruned_loss=0.05628, over 1610577.08 frames. ], batch size: 16, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:01:58,093 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=239413.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 04:01:58,914 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=239414.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:02:05,004 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.597e+02 2.513e+02 3.095e+02 3.810e+02 1.179e+03, threshold=6.190e+02, percent-clipped=9.0 +2023-02-09 04:02:17,683 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=239439.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:02:29,040 INFO [train.py:901] (3/4) Epoch 30, batch 5050, loss[loss=0.1778, simple_loss=0.2616, pruned_loss=0.04701, over 7974.00 frames. ], tot_loss[loss=0.1952, simple_loss=0.2791, pruned_loss=0.05562, over 1612589.28 frames. ], batch size: 21, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:02:52,720 WARNING [train.py:1067] (3/4) Exclude cut with ID 5239-32139-0047-92994 from training. Duration: 27.14 +2023-02-09 04:03:05,849 INFO [train.py:901] (3/4) Epoch 30, batch 5100, loss[loss=0.2182, simple_loss=0.3119, pruned_loss=0.06222, over 8687.00 frames. ], tot_loss[loss=0.1944, simple_loss=0.2787, pruned_loss=0.05499, over 1613317.09 frames. ], batch size: 39, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:03:08,880 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.2337, 1.1078, 1.3271, 0.9971, 1.0071, 1.3086, 0.0820, 0.9196], + device='cuda:3'), covar=tensor([0.1273, 0.1195, 0.0450, 0.0643, 0.2266, 0.0535, 0.1791, 0.1068], + device='cuda:3'), in_proj_covar=tensor([0.0203, 0.0207, 0.0138, 0.0225, 0.0279, 0.0149, 0.0175, 0.0201], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 04:03:20,035 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.557e+02 2.525e+02 3.230e+02 3.994e+02 1.175e+03, threshold=6.461e+02, percent-clipped=6.0 +2023-02-09 04:03:42,220 INFO [train.py:901] (3/4) Epoch 30, batch 5150, loss[loss=0.1954, simple_loss=0.2762, pruned_loss=0.05732, over 8229.00 frames. ], tot_loss[loss=0.1954, simple_loss=0.2798, pruned_loss=0.05554, over 1616218.22 frames. ], batch size: 22, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:03:56,229 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([5.9479, 1.6936, 6.0840, 2.2315, 5.5206, 5.1575, 5.5916, 5.5159], + device='cuda:3'), covar=tensor([0.0446, 0.4994, 0.0371, 0.4046, 0.0966, 0.0916, 0.0489, 0.0505], + device='cuda:3'), in_proj_covar=tensor([0.0699, 0.0671, 0.0755, 0.0671, 0.0757, 0.0648, 0.0660, 0.0733], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 04:04:18,741 INFO [train.py:901] (3/4) Epoch 30, batch 5200, loss[loss=0.1873, simple_loss=0.2701, pruned_loss=0.05223, over 8234.00 frames. ], tot_loss[loss=0.1944, simple_loss=0.2791, pruned_loss=0.05486, over 1610889.76 frames. ], batch size: 22, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:04:30,968 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-02-09 04:04:31,919 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.498e+02 2.360e+02 2.794e+02 3.430e+02 1.458e+03, threshold=5.587e+02, percent-clipped=2.0 +2023-02-09 04:04:37,738 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.7348, 1.4720, 4.9185, 1.8418, 4.4007, 4.0556, 4.3919, 4.3264], + device='cuda:3'), covar=tensor([0.0505, 0.5171, 0.0416, 0.4441, 0.0959, 0.0951, 0.0560, 0.0612], + device='cuda:3'), in_proj_covar=tensor([0.0699, 0.0671, 0.0756, 0.0671, 0.0756, 0.0646, 0.0660, 0.0733], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 04:04:44,695 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=239640.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:04:49,567 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.7668, 1.7173, 2.3495, 1.4304, 1.4590, 2.2566, 0.4702, 1.4235], + device='cuda:3'), covar=tensor([0.1544, 0.1171, 0.0325, 0.1033, 0.2045, 0.0448, 0.1719, 0.1168], + device='cuda:3'), in_proj_covar=tensor([0.0203, 0.0207, 0.0138, 0.0224, 0.0278, 0.0148, 0.0174, 0.0201], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 04:04:54,716 INFO [train.py:901] (3/4) Epoch 30, batch 5250, loss[loss=0.1813, simple_loss=0.2767, pruned_loss=0.04295, over 8452.00 frames. ], tot_loss[loss=0.1946, simple_loss=0.2792, pruned_loss=0.05502, over 1614293.71 frames. ], batch size: 25, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:04:56,150 WARNING [train.py:1067] (3/4) Exclude cut with ID 7859-102521-0017-21930_sp0.9 from training. Duration: 27.25 +2023-02-09 04:05:05,224 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=239669.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 04:05:05,366 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.89 vs. limit=5.0 +2023-02-09 04:05:23,283 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=239694.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 04:05:29,842 INFO [train.py:901] (3/4) Epoch 30, batch 5300, loss[loss=0.224, simple_loss=0.327, pruned_loss=0.06049, over 8488.00 frames. ], tot_loss[loss=0.1939, simple_loss=0.2781, pruned_loss=0.05486, over 1606032.99 frames. ], batch size: 26, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:05:43,734 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.572e+02 2.433e+02 2.937e+02 3.850e+02 7.663e+02, threshold=5.875e+02, percent-clipped=5.0 +2023-02-09 04:06:04,873 INFO [train.py:901] (3/4) Epoch 30, batch 5350, loss[loss=0.1898, simple_loss=0.2659, pruned_loss=0.0568, over 7702.00 frames. ], tot_loss[loss=0.1949, simple_loss=0.279, pruned_loss=0.05545, over 1608677.97 frames. ], batch size: 18, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:06:06,436 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.0999, 1.4409, 1.7798, 1.3752, 1.0442, 1.5453, 1.9711, 1.6329], + device='cuda:3'), covar=tensor([0.0583, 0.1242, 0.1677, 0.1526, 0.0643, 0.1486, 0.0678, 0.0660], + device='cuda:3'), in_proj_covar=tensor([0.0101, 0.0155, 0.0192, 0.0163, 0.0102, 0.0165, 0.0114, 0.0149], + device='cuda:3'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0008, 0.0008], + device='cuda:3') +2023-02-09 04:06:41,542 INFO [train.py:901] (3/4) Epoch 30, batch 5400, loss[loss=0.2004, simple_loss=0.2917, pruned_loss=0.05458, over 8293.00 frames. ], tot_loss[loss=0.1957, simple_loss=0.28, pruned_loss=0.0557, over 1611625.92 frames. ], batch size: 23, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:06:55,042 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.816e+02 2.371e+02 2.881e+02 3.522e+02 8.420e+02, threshold=5.763e+02, percent-clipped=7.0 +2023-02-09 04:06:57,402 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.5800, 1.7331, 4.8354, 1.9383, 4.3347, 4.0799, 4.3247, 4.2801], + device='cuda:3'), covar=tensor([0.0620, 0.4092, 0.0405, 0.3867, 0.0872, 0.0829, 0.0515, 0.0575], + device='cuda:3'), in_proj_covar=tensor([0.0701, 0.0674, 0.0759, 0.0675, 0.0757, 0.0649, 0.0663, 0.0734], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 04:07:07,694 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-02-09 04:07:17,655 INFO [train.py:901] (3/4) Epoch 30, batch 5450, loss[loss=0.1778, simple_loss=0.2586, pruned_loss=0.04854, over 7807.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.2805, pruned_loss=0.05631, over 1610895.67 frames. ], batch size: 20, lr: 2.50e-03, grad_scale: 16.0 +2023-02-09 04:07:49,874 WARNING [train.py:1067] (3/4) Exclude cut with ID 7357-94126-0014-44390_sp0.9 from training. Duration: 31.02225 +2023-02-09 04:07:53,971 INFO [train.py:901] (3/4) Epoch 30, batch 5500, loss[loss=0.2012, simple_loss=0.2955, pruned_loss=0.05344, over 8534.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2802, pruned_loss=0.0564, over 1611243.74 frames. ], batch size: 49, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:08:08,783 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.536e+02 2.378e+02 3.012e+02 4.037e+02 9.246e+02, threshold=6.023e+02, percent-clipped=5.0 +2023-02-09 04:08:30,323 INFO [train.py:901] (3/4) Epoch 30, batch 5550, loss[loss=0.1748, simple_loss=0.2641, pruned_loss=0.04269, over 8589.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2802, pruned_loss=0.05595, over 1613991.51 frames. ], batch size: 34, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:08:50,822 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=239984.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:09:07,149 INFO [train.py:901] (3/4) Epoch 30, batch 5600, loss[loss=0.1984, simple_loss=0.2867, pruned_loss=0.05508, over 8041.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2813, pruned_loss=0.05656, over 1615989.03 frames. ], batch size: 22, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:09:21,043 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.553e+02 2.479e+02 2.939e+02 3.472e+02 8.474e+02, threshold=5.878e+02, percent-clipped=2.0 +2023-02-09 04:09:42,584 INFO [train.py:901] (3/4) Epoch 30, batch 5650, loss[loss=0.1981, simple_loss=0.2802, pruned_loss=0.05804, over 8293.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2813, pruned_loss=0.05663, over 1613754.90 frames. ], batch size: 23, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:09:59,637 WARNING [train.py:1067] (3/4) Exclude cut with ID 6758-72288-0033-148662_sp0.9 from training. Duration: 25.988875 +2023-02-09 04:10:14,057 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=240099.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:10:16,491 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=3.91 vs. limit=5.0 +2023-02-09 04:10:17,324 INFO [train.py:901] (3/4) Epoch 30, batch 5700, loss[loss=0.2316, simple_loss=0.3121, pruned_loss=0.07555, over 8360.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.2806, pruned_loss=0.0564, over 1614751.84 frames. ], batch size: 26, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:10:31,996 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.343e+02 2.506e+02 3.162e+02 4.194e+02 1.225e+03, threshold=6.325e+02, percent-clipped=8.0 +2023-02-09 04:10:53,032 INFO [train.py:901] (3/4) Epoch 30, batch 5750, loss[loss=0.2165, simple_loss=0.2919, pruned_loss=0.07057, over 8081.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2799, pruned_loss=0.05642, over 1609641.98 frames. ], batch size: 21, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:11:03,865 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.87 vs. limit=2.0 +2023-02-09 04:11:04,186 WARNING [train.py:1067] (3/4) Exclude cut with ID 3972-170212-0014-103914_sp0.9 from training. Duration: 29.1166875 +2023-02-09 04:11:28,581 INFO [train.py:901] (3/4) Epoch 30, batch 5800, loss[loss=0.1553, simple_loss=0.2411, pruned_loss=0.03479, over 6763.00 frames. ], tot_loss[loss=0.1966, simple_loss=0.28, pruned_loss=0.0566, over 1607322.08 frames. ], batch size: 15, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:11:31,036 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-02-09 04:11:42,443 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.461e+02 2.400e+02 2.667e+02 3.487e+02 8.848e+02, threshold=5.334e+02, percent-clipped=2.0 +2023-02-09 04:12:04,243 INFO [train.py:901] (3/4) Epoch 30, batch 5850, loss[loss=0.1821, simple_loss=0.2662, pruned_loss=0.04894, over 8341.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2807, pruned_loss=0.05686, over 1612267.36 frames. ], batch size: 49, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:12:39,817 INFO [train.py:901] (3/4) Epoch 30, batch 5900, loss[loss=0.2209, simple_loss=0.3087, pruned_loss=0.06661, over 8511.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2812, pruned_loss=0.0567, over 1615486.11 frames. ], batch size: 49, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:12:53,707 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.373e+02 2.330e+02 2.970e+02 3.920e+02 1.059e+03, threshold=5.939e+02, percent-clipped=6.0 +2023-02-09 04:13:08,088 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.4218, 2.9517, 2.3318, 4.1085, 1.7891, 2.2180, 2.7008, 2.8950], + device='cuda:3'), covar=tensor([0.0722, 0.0804, 0.0759, 0.0261, 0.1065, 0.1141, 0.0858, 0.0715], + device='cuda:3'), in_proj_covar=tensor([0.0230, 0.0193, 0.0244, 0.0213, 0.0201, 0.0245, 0.0247, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0005, 0.0006, 0.0005], + device='cuda:3') +2023-02-09 04:13:15,497 INFO [train.py:901] (3/4) Epoch 30, batch 5950, loss[loss=0.2093, simple_loss=0.3073, pruned_loss=0.05558, over 8238.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2809, pruned_loss=0.05633, over 1617643.16 frames. ], batch size: 24, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:13:16,425 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=240355.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:13:18,464 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=240358.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:13:33,802 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=240380.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:13:50,584 INFO [train.py:901] (3/4) Epoch 30, batch 6000, loss[loss=0.2028, simple_loss=0.2881, pruned_loss=0.05872, over 8340.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2813, pruned_loss=0.05641, over 1618324.72 frames. ], batch size: 26, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:13:50,585 INFO [train.py:926] (3/4) Computing validation loss +2023-02-09 04:14:04,295 INFO [train.py:935] (3/4) Epoch 30, validation: loss=0.1701, simple_loss=0.2695, pruned_loss=0.03536, over 944034.00 frames. +2023-02-09 04:14:04,296 INFO [train.py:936] (3/4) Maximum memory allocated so far is 6543MB +2023-02-09 04:14:17,955 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.732e+02 2.377e+02 3.122e+02 3.554e+02 6.850e+02, threshold=6.243e+02, percent-clipped=2.0 +2023-02-09 04:14:39,390 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6097, 1.3187, 2.4081, 1.4172, 2.2309, 2.5580, 2.7340, 2.1984], + device='cuda:3'), covar=tensor([0.1001, 0.1446, 0.0371, 0.1986, 0.0707, 0.0357, 0.0589, 0.0601], + device='cuda:3'), in_proj_covar=tensor([0.0310, 0.0331, 0.0297, 0.0330, 0.0331, 0.0285, 0.0452, 0.0310], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:3') +2023-02-09 04:14:39,922 INFO [train.py:901] (3/4) Epoch 30, batch 6050, loss[loss=0.1867, simple_loss=0.273, pruned_loss=0.05024, over 7976.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.2816, pruned_loss=0.05647, over 1622314.49 frames. ], batch size: 21, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:14:55,006 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.5194, 1.8435, 4.7097, 1.9647, 4.2193, 4.0085, 4.2543, 4.1698], + device='cuda:3'), covar=tensor([0.0611, 0.4403, 0.0501, 0.4432, 0.1002, 0.0896, 0.0656, 0.0669], + device='cuda:3'), in_proj_covar=tensor([0.0704, 0.0674, 0.0759, 0.0677, 0.0761, 0.0650, 0.0662, 0.0738], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 04:15:16,415 INFO [train.py:901] (3/4) Epoch 30, batch 6100, loss[loss=0.1723, simple_loss=0.2456, pruned_loss=0.04946, over 7692.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2818, pruned_loss=0.0567, over 1623240.42 frames. ], batch size: 18, lr: 2.50e-03, grad_scale: 8.0 +2023-02-09 04:15:30,263 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.619e+02 2.423e+02 2.992e+02 3.767e+02 7.583e+02, threshold=5.983e+02, percent-clipped=4.0 +2023-02-09 04:15:36,682 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-02-09 04:15:40,593 WARNING [train.py:1067] (3/4) Exclude cut with ID 3033-130750-0096-107983_sp0.9 from training. Duration: 0.92225 +2023-02-09 04:15:51,549 INFO [train.py:901] (3/4) Epoch 30, batch 6150, loss[loss=0.1648, simple_loss=0.2393, pruned_loss=0.04516, over 7789.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2812, pruned_loss=0.05675, over 1616824.69 frames. ], batch size: 19, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:16:22,933 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.5693, 1.7953, 1.8534, 1.2473, 1.8908, 1.4908, 0.4553, 1.7912], + device='cuda:3'), covar=tensor([0.0703, 0.0497, 0.0392, 0.0681, 0.0609, 0.1140, 0.1130, 0.0390], + device='cuda:3'), in_proj_covar=tensor([0.0482, 0.0417, 0.0376, 0.0467, 0.0402, 0.0558, 0.0408, 0.0446], + device='cuda:3'), out_proj_covar=tensor([1.2749e-04, 1.0757e-04, 9.7819e-05, 1.2187e-04, 1.0515e-04, 1.5508e-04, + 1.0873e-04, 1.1661e-04], device='cuda:3') +2023-02-09 04:16:28,466 INFO [train.py:901] (3/4) Epoch 30, batch 6200, loss[loss=0.1938, simple_loss=0.2876, pruned_loss=0.04993, over 8254.00 frames. ], tot_loss[loss=0.1968, simple_loss=0.2808, pruned_loss=0.05641, over 1617963.26 frames. ], batch size: 24, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:16:44,278 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.619e+02 2.533e+02 2.968e+02 3.901e+02 6.917e+02, threshold=5.935e+02, percent-clipped=4.0 +2023-02-09 04:17:05,855 INFO [train.py:901] (3/4) Epoch 30, batch 6250, loss[loss=0.2238, simple_loss=0.3011, pruned_loss=0.07324, over 8367.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2806, pruned_loss=0.05657, over 1613208.34 frames. ], batch size: 24, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:17:39,922 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=240702.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:17:41,233 INFO [train.py:901] (3/4) Epoch 30, batch 6300, loss[loss=0.2051, simple_loss=0.2825, pruned_loss=0.06382, over 8545.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2817, pruned_loss=0.05711, over 1613008.76 frames. ], batch size: 49, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:17:54,932 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.808e+02 2.569e+02 3.101e+02 4.376e+02 1.063e+03, threshold=6.203e+02, percent-clipped=9.0 +2023-02-09 04:18:10,185 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.3830, 1.5996, 4.5997, 1.6649, 4.1217, 3.8609, 4.1487, 4.0609], + device='cuda:3'), covar=tensor([0.0579, 0.4465, 0.0518, 0.4657, 0.1040, 0.0949, 0.0550, 0.0649], + device='cuda:3'), in_proj_covar=tensor([0.0700, 0.0669, 0.0754, 0.0673, 0.0755, 0.0645, 0.0657, 0.0732], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 04:18:17,093 INFO [train.py:901] (3/4) Epoch 30, batch 6350, loss[loss=0.1629, simple_loss=0.2487, pruned_loss=0.03852, over 7700.00 frames. ], tot_loss[loss=0.198, simple_loss=0.2819, pruned_loss=0.05707, over 1617349.98 frames. ], batch size: 18, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:18:53,327 INFO [train.py:901] (3/4) Epoch 30, batch 6400, loss[loss=0.1855, simple_loss=0.2659, pruned_loss=0.05251, over 7442.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2809, pruned_loss=0.05654, over 1616043.94 frames. ], batch size: 17, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:19:02,608 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=240817.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:19:07,074 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.453e+02 2.422e+02 2.800e+02 3.642e+02 5.918e+02, threshold=5.600e+02, percent-clipped=0.0 +2023-02-09 04:19:28,698 INFO [train.py:901] (3/4) Epoch 30, batch 6450, loss[loss=0.193, simple_loss=0.2832, pruned_loss=0.05137, over 8470.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2815, pruned_loss=0.05719, over 1616119.38 frames. ], batch size: 25, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:20:03,785 INFO [train.py:901] (3/4) Epoch 30, batch 6500, loss[loss=0.1879, simple_loss=0.269, pruned_loss=0.0534, over 8470.00 frames. ], tot_loss[loss=0.1973, simple_loss=0.281, pruned_loss=0.0568, over 1618683.43 frames. ], batch size: 29, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:20:05,406 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.7019, 1.6841, 4.8978, 1.8006, 4.3765, 4.0826, 4.4088, 4.3250], + device='cuda:3'), covar=tensor([0.0527, 0.4684, 0.0497, 0.4605, 0.0989, 0.0980, 0.0544, 0.0608], + device='cuda:3'), in_proj_covar=tensor([0.0704, 0.0673, 0.0759, 0.0678, 0.0760, 0.0650, 0.0660, 0.0737], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 04:20:17,923 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.512e+02 2.582e+02 3.161e+02 3.840e+02 1.025e+03, threshold=6.322e+02, percent-clipped=7.0 +2023-02-09 04:20:36,286 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=240950.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 04:20:38,870 INFO [train.py:901] (3/4) Epoch 30, batch 6550, loss[loss=0.1788, simple_loss=0.2629, pruned_loss=0.04735, over 7970.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2799, pruned_loss=0.05634, over 1615316.61 frames. ], batch size: 21, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:21:00,574 WARNING [train.py:1067] (3/4) Exclude cut with ID 3557-8342-0013-71585 from training. Duration: 0.92 +2023-02-09 04:21:15,887 INFO [train.py:901] (3/4) Epoch 30, batch 6600, loss[loss=0.1981, simple_loss=0.2916, pruned_loss=0.05235, over 8789.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2813, pruned_loss=0.05684, over 1614215.48 frames. ], batch size: 30, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:21:16,033 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=241004.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:21:20,066 WARNING [train.py:1067] (3/4) Exclude cut with ID 4133-6541-0027-26893_sp1.1 from training. Duration: 0.9681875 +2023-02-09 04:21:29,839 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.473e+02 2.233e+02 3.026e+02 3.930e+02 1.368e+03, threshold=6.053e+02, percent-clipped=4.0 +2023-02-09 04:21:51,499 INFO [train.py:901] (3/4) Epoch 30, batch 6650, loss[loss=0.1892, simple_loss=0.2668, pruned_loss=0.05576, over 7932.00 frames. ], tot_loss[loss=0.1964, simple_loss=0.2804, pruned_loss=0.05615, over 1614123.54 frames. ], batch size: 20, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:22:04,820 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=241073.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:22:23,586 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=241098.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:22:27,625 INFO [train.py:901] (3/4) Epoch 30, batch 6700, loss[loss=0.2091, simple_loss=0.3011, pruned_loss=0.05851, over 8348.00 frames. ], tot_loss[loss=0.1952, simple_loss=0.2795, pruned_loss=0.05545, over 1616632.31 frames. ], batch size: 26, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:22:42,112 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.506e+02 2.246e+02 2.834e+02 3.422e+02 9.903e+02, threshold=5.667e+02, percent-clipped=4.0 +2023-02-09 04:22:43,744 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.7842, 2.5981, 1.8695, 2.4797, 2.3116, 1.6641, 2.1217, 2.2750], + device='cuda:3'), covar=tensor([0.1681, 0.0461, 0.1365, 0.0704, 0.0848, 0.1703, 0.1208, 0.1223], + device='cuda:3'), in_proj_covar=tensor([0.0363, 0.0246, 0.0347, 0.0317, 0.0304, 0.0350, 0.0351, 0.0322], + device='cuda:3'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004, 0.0003], + device='cuda:3') +2023-02-09 04:23:04,036 INFO [train.py:901] (3/4) Epoch 30, batch 6750, loss[loss=0.181, simple_loss=0.2741, pruned_loss=0.04398, over 8337.00 frames. ], tot_loss[loss=0.1944, simple_loss=0.2788, pruned_loss=0.05499, over 1614621.61 frames. ], batch size: 25, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:23:29,825 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.5993, 2.4260, 3.1619, 2.6200, 3.1483, 2.6284, 2.5003, 1.9693], + device='cuda:3'), covar=tensor([0.5617, 0.5520, 0.2086, 0.4077, 0.2887, 0.3477, 0.1980, 0.5929], + device='cuda:3'), in_proj_covar=tensor([0.0967, 0.1038, 0.0852, 0.1011, 0.1031, 0.0948, 0.0778, 0.0859], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 04:23:39,179 INFO [train.py:901] (3/4) Epoch 30, batch 6800, loss[loss=0.1951, simple_loss=0.2864, pruned_loss=0.05192, over 8481.00 frames. ], tot_loss[loss=0.1946, simple_loss=0.2788, pruned_loss=0.05525, over 1613868.42 frames. ], batch size: 27, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:23:42,683 WARNING [train.py:1067] (3/4) Exclude cut with ID 8291-282929-0024-9607_sp0.9 from training. Duration: 26.438875 +2023-02-09 04:23:53,776 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.574e+02 2.178e+02 2.682e+02 3.526e+02 7.087e+02, threshold=5.364e+02, percent-clipped=2.0 +2023-02-09 04:24:15,461 INFO [train.py:901] (3/4) Epoch 30, batch 6850, loss[loss=0.2081, simple_loss=0.2963, pruned_loss=0.05991, over 8363.00 frames. ], tot_loss[loss=0.1952, simple_loss=0.2797, pruned_loss=0.05535, over 1611120.61 frames. ], batch size: 48, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:24:34,764 WARNING [train.py:1067] (3/4) Exclude cut with ID 2929-85685-0079-61403_sp1.1 from training. Duration: 27.0318125 +2023-02-09 04:24:43,877 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=241294.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 04:24:50,576 INFO [train.py:901] (3/4) Epoch 30, batch 6900, loss[loss=0.1959, simple_loss=0.2739, pruned_loss=0.05897, over 8134.00 frames. ], tot_loss[loss=0.1954, simple_loss=0.28, pruned_loss=0.05536, over 1612849.45 frames. ], batch size: 22, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:24:54,881 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.9795, 2.1970, 1.8093, 2.7644, 1.2235, 1.6334, 1.9916, 2.1330], + device='cuda:3'), covar=tensor([0.0745, 0.0712, 0.0924, 0.0346, 0.1051, 0.1303, 0.0820, 0.0785], + device='cuda:3'), in_proj_covar=tensor([0.0231, 0.0194, 0.0245, 0.0214, 0.0202, 0.0246, 0.0248, 0.0204], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:3') +2023-02-09 04:25:05,718 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.702e+02 2.527e+02 3.094e+02 3.969e+02 8.004e+02, threshold=6.188e+02, percent-clipped=9.0 +2023-02-09 04:25:12,695 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.7929, 1.9086, 1.6915, 2.3521, 0.9728, 1.5207, 1.7446, 1.8932], + device='cuda:3'), covar=tensor([0.0781, 0.0778, 0.0978, 0.0396, 0.1133, 0.1416, 0.0781, 0.0749], + device='cuda:3'), in_proj_covar=tensor([0.0231, 0.0193, 0.0244, 0.0213, 0.0202, 0.0246, 0.0247, 0.0203], + device='cuda:3'), out_proj_covar=tensor([0.0005, 0.0005, 0.0006, 0.0005, 0.0005, 0.0006, 0.0006, 0.0005], + device='cuda:3') +2023-02-09 04:25:22,110 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=241348.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:25:26,923 INFO [train.py:901] (3/4) Epoch 30, batch 6950, loss[loss=0.1946, simple_loss=0.2856, pruned_loss=0.05183, over 8486.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2802, pruned_loss=0.0557, over 1614462.05 frames. ], batch size: 28, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:25:46,733 WARNING [train.py:1067] (3/4) Exclude cut with ID 7255-291500-0009-9471_sp0.9 from training. Duration: 26.62775 +2023-02-09 04:25:59,228 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.5338, 1.7780, 2.5738, 1.4354, 1.9703, 1.8582, 1.5494, 1.9953], + device='cuda:3'), covar=tensor([0.2179, 0.2900, 0.1069, 0.4997, 0.2171, 0.3591, 0.2768, 0.2420], + device='cuda:3'), in_proj_covar=tensor([0.0547, 0.0649, 0.0567, 0.0677, 0.0670, 0.0621, 0.0574, 0.0650], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 04:26:04,048 INFO [train.py:901] (3/4) Epoch 30, batch 7000, loss[loss=0.2576, simple_loss=0.3209, pruned_loss=0.09716, over 7280.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.2809, pruned_loss=0.05623, over 1619112.14 frames. ], batch size: 72, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:26:07,753 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=241409.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 04:26:17,959 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.728e+02 2.441e+02 2.932e+02 3.651e+02 7.920e+02, threshold=5.865e+02, percent-clipped=3.0 +2023-02-09 04:26:18,115 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=241424.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:26:40,277 INFO [train.py:901] (3/4) Epoch 30, batch 7050, loss[loss=0.1984, simple_loss=0.2855, pruned_loss=0.05569, over 8359.00 frames. ], tot_loss[loss=0.1961, simple_loss=0.2802, pruned_loss=0.05601, over 1617923.21 frames. ], batch size: 24, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:26:46,795 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=241463.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:27:16,741 INFO [train.py:901] (3/4) Epoch 30, batch 7100, loss[loss=0.1723, simple_loss=0.2633, pruned_loss=0.04067, over 8468.00 frames. ], tot_loss[loss=0.1955, simple_loss=0.2798, pruned_loss=0.05558, over 1616549.83 frames. ], batch size: 25, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:27:30,715 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.720e+02 2.381e+02 2.857e+02 3.660e+02 8.579e+02, threshold=5.714e+02, percent-clipped=3.0 +2023-02-09 04:27:51,587 INFO [train.py:901] (3/4) Epoch 30, batch 7150, loss[loss=0.1888, simple_loss=0.266, pruned_loss=0.05575, over 7697.00 frames. ], tot_loss[loss=0.195, simple_loss=0.2796, pruned_loss=0.05523, over 1615767.13 frames. ], batch size: 18, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:28:28,712 INFO [train.py:901] (3/4) Epoch 30, batch 7200, loss[loss=0.2021, simple_loss=0.2883, pruned_loss=0.05796, over 8336.00 frames. ], tot_loss[loss=0.1954, simple_loss=0.2798, pruned_loss=0.0555, over 1616613.19 frames. ], batch size: 26, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:28:43,479 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.677e+02 2.259e+02 2.765e+02 3.853e+02 1.030e+03, threshold=5.530e+02, percent-clipped=3.0 +2023-02-09 04:28:53,986 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=241639.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:29:03,954 INFO [train.py:901] (3/4) Epoch 30, batch 7250, loss[loss=0.2116, simple_loss=0.2891, pruned_loss=0.06708, over 7922.00 frames. ], tot_loss[loss=0.1951, simple_loss=0.2793, pruned_loss=0.05545, over 1615490.98 frames. ], batch size: 20, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:29:11,981 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=241665.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 04:29:30,390 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=241690.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 04:29:40,026 INFO [train.py:901] (3/4) Epoch 30, batch 7300, loss[loss=0.2152, simple_loss=0.3036, pruned_loss=0.06339, over 8500.00 frames. ], tot_loss[loss=0.1955, simple_loss=0.2795, pruned_loss=0.05575, over 1616377.60 frames. ], batch size: 28, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:29:50,469 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=241719.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:29:53,729 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.652e+02 2.349e+02 2.997e+02 3.899e+02 6.597e+02, threshold=5.994e+02, percent-clipped=5.0 +2023-02-09 04:29:59,160 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.9419, 1.4699, 1.8230, 1.4510, 0.9509, 1.5836, 1.7312, 1.6752], + device='cuda:3'), covar=tensor([0.0573, 0.1211, 0.1617, 0.1432, 0.0594, 0.1400, 0.0694, 0.0658], + device='cuda:3'), in_proj_covar=tensor([0.0100, 0.0154, 0.0191, 0.0162, 0.0102, 0.0163, 0.0113, 0.0148], + device='cuda:3'), out_proj_covar=tensor([0.0007, 0.0009, 0.0010, 0.0010, 0.0006, 0.0010, 0.0007, 0.0008], + device='cuda:3') +2023-02-09 04:30:08,890 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=241744.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:30:15,563 INFO [train.py:901] (3/4) Epoch 30, batch 7350, loss[loss=0.1908, simple_loss=0.2868, pruned_loss=0.0474, over 8497.00 frames. ], tot_loss[loss=0.1958, simple_loss=0.2798, pruned_loss=0.05595, over 1611553.46 frames. ], batch size: 26, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:30:25,405 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=241768.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:30:39,735 WARNING [train.py:1067] (3/4) Exclude cut with ID 7357-94126-0026-44402_sp0.9 from training. Duration: 25.061125 +2023-02-09 04:30:51,553 INFO [train.py:901] (3/4) Epoch 30, batch 7400, loss[loss=0.1923, simple_loss=0.2726, pruned_loss=0.05601, over 8345.00 frames. ], tot_loss[loss=0.1957, simple_loss=0.2795, pruned_loss=0.05598, over 1610774.10 frames. ], batch size: 26, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:30:59,775 WARNING [train.py:1067] (3/4) Exclude cut with ID 774-127930-0014-48411_sp1.1 from training. Duration: 0.95 +2023-02-09 04:31:00,900 INFO [scaling.py:679] (3/4) Whitening: num_groups=1, num_channels=256, metric=2.70 vs. limit=5.0 +2023-02-09 04:31:04,083 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=241821.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:31:05,934 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.776e+02 2.497e+02 3.037e+02 3.880e+02 5.984e+02, threshold=6.074e+02, percent-clipped=0.0 +2023-02-09 04:31:24,332 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=241849.0, num_to_drop=1, layers_to_drop={1} +2023-02-09 04:31:27,625 INFO [train.py:901] (3/4) Epoch 30, batch 7450, loss[loss=0.1857, simple_loss=0.2856, pruned_loss=0.04294, over 8547.00 frames. ], tot_loss[loss=0.1952, simple_loss=0.2792, pruned_loss=0.05555, over 1607679.47 frames. ], batch size: 31, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:31:28,705 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([2.6416, 2.4641, 3.1727, 2.6315, 3.2009, 2.7010, 2.6026, 2.1064], + device='cuda:3'), covar=tensor([0.5553, 0.5307, 0.2126, 0.4183, 0.2608, 0.3124, 0.1816, 0.5604], + device='cuda:3'), in_proj_covar=tensor([0.0967, 0.1033, 0.0850, 0.1009, 0.1030, 0.0944, 0.0778, 0.0858], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:3') +2023-02-09 04:31:40,211 WARNING [train.py:1067] (3/4) Exclude cut with ID 7699-105389-0094-102071_sp0.9 from training. Duration: 26.6166875 +2023-02-09 04:31:48,184 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=241883.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:32:02,786 INFO [train.py:901] (3/4) Epoch 30, batch 7500, loss[loss=0.1625, simple_loss=0.2515, pruned_loss=0.03672, over 8252.00 frames. ], tot_loss[loss=0.195, simple_loss=0.2787, pruned_loss=0.05562, over 1605900.63 frames. ], batch size: 24, lr: 2.49e-03, grad_scale: 16.0 +2023-02-09 04:32:18,960 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.524e+02 2.491e+02 2.852e+02 3.531e+02 9.058e+02, threshold=5.704e+02, percent-clipped=2.0 +2023-02-09 04:32:37,298 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.1446, 1.4802, 4.3294, 1.7249, 3.8563, 3.6588, 3.9403, 3.8450], + device='cuda:3'), covar=tensor([0.0708, 0.4659, 0.0598, 0.4390, 0.1076, 0.0950, 0.0656, 0.0727], + device='cuda:3'), in_proj_covar=tensor([0.0700, 0.0673, 0.0759, 0.0677, 0.0761, 0.0648, 0.0660, 0.0736], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 04:32:39,981 INFO [train.py:901] (3/4) Epoch 30, batch 7550, loss[loss=0.2197, simple_loss=0.3018, pruned_loss=0.06878, over 6819.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.2801, pruned_loss=0.05649, over 1606327.55 frames. ], batch size: 15, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:32:43,672 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([4.2123, 4.1826, 3.8310, 1.9511, 3.7412, 3.8329, 3.6303, 3.7378], + device='cuda:3'), covar=tensor([0.0757, 0.0587, 0.1094, 0.4385, 0.0939, 0.1092, 0.1383, 0.0746], + device='cuda:3'), in_proj_covar=tensor([0.0552, 0.0460, 0.0453, 0.0563, 0.0449, 0.0477, 0.0450, 0.0416], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 04:33:01,519 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=241983.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:33:17,233 INFO [train.py:901] (3/4) Epoch 30, batch 7600, loss[loss=0.2123, simple_loss=0.2965, pruned_loss=0.06403, over 8197.00 frames. ], tot_loss[loss=0.1961, simple_loss=0.2798, pruned_loss=0.05622, over 1605545.95 frames. ], batch size: 23, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:33:32,889 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.614e+02 2.414e+02 3.070e+02 3.745e+02 6.631e+02, threshold=6.140e+02, percent-clipped=3.0 +2023-02-09 04:33:45,823 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.7400, 2.0279, 2.8272, 1.5646, 2.2445, 2.0503, 1.8047, 2.1702], + device='cuda:3'), covar=tensor([0.1925, 0.2627, 0.0953, 0.4581, 0.2026, 0.3278, 0.2556, 0.2517], + device='cuda:3'), in_proj_covar=tensor([0.0546, 0.0649, 0.0566, 0.0678, 0.0670, 0.0620, 0.0573, 0.0648], + device='cuda:3'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:3') +2023-02-09 04:33:54,462 INFO [train.py:901] (3/4) Epoch 30, batch 7650, loss[loss=0.2373, simple_loss=0.3154, pruned_loss=0.07961, over 8495.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2795, pruned_loss=0.05625, over 1607544.06 frames. ], batch size: 26, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:34:25,726 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=242098.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:34:29,809 INFO [train.py:901] (3/4) Epoch 30, batch 7700, loss[loss=0.167, simple_loss=0.2596, pruned_loss=0.03722, over 7980.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2803, pruned_loss=0.05701, over 1610601.01 frames. ], batch size: 21, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:34:44,312 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.689e+02 2.431e+02 3.028e+02 3.722e+02 6.918e+02, threshold=6.057e+02, percent-clipped=1.0 +2023-02-09 04:34:54,343 WARNING [train.py:1067] (3/4) Exclude cut with ID 7357-94126-0009-44385_sp0.9 from training. Duration: 27.02225 +2023-02-09 04:34:55,277 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=242139.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:35:05,988 INFO [train.py:901] (3/4) Epoch 30, batch 7750, loss[loss=0.2255, simple_loss=0.3071, pruned_loss=0.07194, over 8533.00 frames. ], tot_loss[loss=0.1965, simple_loss=0.28, pruned_loss=0.05647, over 1613582.10 frames. ], batch size: 49, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:35:13,195 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=242164.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:35:13,204 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.4295, 1.3946, 1.7398, 1.1181, 1.0679, 1.6957, 0.2922, 1.1704], + device='cuda:3'), covar=tensor([0.1343, 0.1040, 0.0400, 0.0851, 0.2380, 0.0502, 0.1676, 0.1192], + device='cuda:3'), in_proj_covar=tensor([0.0203, 0.0206, 0.0138, 0.0223, 0.0277, 0.0149, 0.0175, 0.0200], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 04:35:13,774 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=242165.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:35:18,596 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=242171.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:35:20,781 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6687, 1.2632, 3.1282, 1.4335, 2.3241, 3.3579, 3.5807, 2.8970], + device='cuda:3'), covar=tensor([0.1380, 0.2105, 0.0349, 0.2338, 0.1023, 0.0280, 0.0516, 0.0537], + device='cuda:3'), in_proj_covar=tensor([0.0311, 0.0331, 0.0298, 0.0330, 0.0332, 0.0287, 0.0455, 0.0311], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0004, 0.0002], + device='cuda:3') +2023-02-09 04:35:34,906 INFO [zipformer.py:1185] (3/4) warmup_begin=666.7, warmup_end=1333.3, batch_count=242193.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 04:35:42,648 INFO [train.py:901] (3/4) Epoch 30, batch 7800, loss[loss=0.1769, simple_loss=0.2656, pruned_loss=0.04412, over 7790.00 frames. ], tot_loss[loss=0.1977, simple_loss=0.2815, pruned_loss=0.0569, over 1617293.46 frames. ], batch size: 19, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:35:57,866 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.684e+02 2.396e+02 3.014e+02 3.960e+02 8.063e+02, threshold=6.029e+02, percent-clipped=4.0 +2023-02-09 04:36:11,177 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=242244.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:36:18,001 INFO [train.py:901] (3/4) Epoch 30, batch 7850, loss[loss=0.1688, simple_loss=0.2436, pruned_loss=0.04695, over 7803.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2807, pruned_loss=0.05683, over 1615410.31 frames. ], batch size: 19, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:36:36,021 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=242280.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:36:52,298 INFO [train.py:901] (3/4) Epoch 30, batch 7900, loss[loss=0.1495, simple_loss=0.2413, pruned_loss=0.02886, over 7809.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2808, pruned_loss=0.05677, over 1619948.01 frames. ], batch size: 20, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:36:54,985 INFO [zipformer.py:1185] (3/4) warmup_begin=2000.0, warmup_end=2666.7, batch_count=242308.0, num_to_drop=1, layers_to_drop={0} +2023-02-09 04:37:06,486 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.670e+02 2.359e+02 2.894e+02 3.889e+02 1.272e+03, threshold=5.788e+02, percent-clipped=10.0 +2023-02-09 04:37:08,044 INFO [zipformer.py:1185] (3/4) warmup_begin=1333.3, warmup_end=2000.0, batch_count=242327.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:37:26,030 INFO [scaling.py:679] (3/4) Whitening: num_groups=8, num_channels=192, metric=1.69 vs. limit=2.0 +2023-02-09 04:37:26,303 INFO [train.py:901] (3/4) Epoch 30, batch 7950, loss[loss=0.1834, simple_loss=0.2704, pruned_loss=0.04819, over 7966.00 frames. ], tot_loss[loss=0.197, simple_loss=0.2807, pruned_loss=0.05667, over 1616912.76 frames. ], batch size: 21, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:37:26,516 INFO [zipformer.py:1185] (3/4) warmup_begin=3333.3, warmup_end=4000.0, batch_count=242354.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:37:43,577 INFO [zipformer.py:1185] (3/4) warmup_begin=2666.7, warmup_end=3333.3, batch_count=242379.0, num_to_drop=0, layers_to_drop=set() +2023-02-09 04:37:53,732 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.6552, 1.8466, 1.8871, 1.3756, 1.9712, 1.4749, 0.4562, 1.8274], + device='cuda:3'), covar=tensor([0.0534, 0.0378, 0.0325, 0.0565, 0.0477, 0.0981, 0.0994, 0.0301], + device='cuda:3'), in_proj_covar=tensor([0.0484, 0.0422, 0.0378, 0.0471, 0.0407, 0.0564, 0.0409, 0.0449], + device='cuda:3'), out_proj_covar=tensor([1.2803e-04, 1.0876e-04, 9.8259e-05, 1.2281e-04, 1.0642e-04, 1.5677e-04, + 1.0904e-04, 1.1734e-04], device='cuda:3') +2023-02-09 04:38:00,610 INFO [train.py:901] (3/4) Epoch 30, batch 8000, loss[loss=0.1622, simple_loss=0.2527, pruned_loss=0.03586, over 8136.00 frames. ], tot_loss[loss=0.1956, simple_loss=0.2795, pruned_loss=0.05588, over 1613207.57 frames. ], batch size: 22, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:38:14,884 INFO [optim.py:369] (3/4) Clipping_scale=2.0, grad-norm quartiles 1.710e+02 2.478e+02 2.968e+02 3.707e+02 1.083e+03, threshold=5.936e+02, percent-clipped=5.0 +2023-02-09 04:38:35,242 INFO [train.py:901] (3/4) Epoch 30, batch 8050, loss[loss=0.1759, simple_loss=0.2512, pruned_loss=0.05032, over 7538.00 frames. ], tot_loss[loss=0.196, simple_loss=0.2795, pruned_loss=0.05624, over 1602405.21 frames. ], batch size: 18, lr: 2.49e-03, grad_scale: 8.0 +2023-02-09 04:38:53,666 INFO [zipformer.py:2431] (3/4) attn_weights_entropy = tensor([1.2339, 1.0619, 1.3155, 1.0268, 1.0167, 1.3212, 0.1414, 0.9189], + device='cuda:3'), covar=tensor([0.1365, 0.1272, 0.0495, 0.0674, 0.2424, 0.0570, 0.1781, 0.1134], + device='cuda:3'), in_proj_covar=tensor([0.0204, 0.0207, 0.0139, 0.0224, 0.0278, 0.0150, 0.0175, 0.0200], + device='cuda:3'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0004, 0.0002, 0.0003, 0.0003], + device='cuda:3') +2023-02-09 04:38:58,180 INFO [train.py:1165] (3/4) Done!